diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,6016 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3071393563143276,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8985507246376816e-07,
+      "loss": 2.3988,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.797101449275363e-07,
+      "loss": 2.2096,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.695652173913044e-07,
+      "loss": 1.8346,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1594202898550726e-06,
+      "loss": 1.8346,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4492753623188408e-06,
+      "loss": 1.9785,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7391304347826088e-06,
+      "loss": 1.8236,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.028985507246377e-06,
+      "loss": 1.8187,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3188405797101453e-06,
+      "loss": 1.8793,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6086956521739132e-06,
+      "loss": 1.928,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8985507246376816e-06,
+      "loss": 1.76,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188405797101449e-06,
+      "loss": 1.7108,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 1.7712,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.768115942028986e-06,
+      "loss": 1.8288,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 1.8112,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.7858,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 1.7692,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.927536231884059e-06,
+      "loss": 1.6936,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 1.8671,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.507246376811595e-06,
+      "loss": 1.7851,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.797101449275363e-06,
+      "loss": 1.8736,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.086956521739132e-06,
+      "loss": 1.828,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.376811594202898e-06,
+      "loss": 1.6725,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.6799,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.956521739130435e-06,
+      "loss": 1.9097,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 1.8499,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.536231884057972e-06,
+      "loss": 1.7507,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.82608695652174e-06,
+      "loss": 1.6498,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.115942028985508e-06,
+      "loss": 1.6504,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.405797101449275e-06,
+      "loss": 1.5428,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 1.7138,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.985507246376812e-06,
+      "loss": 1.6575,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.275362318840581e-06,
+      "loss": 1.8043,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.565217391304349e-06,
+      "loss": 1.7853,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.855072463768118e-06,
+      "loss": 1.6425,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 1.7267,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0434782608695653e-05,
+      "loss": 1.6063,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0724637681159422e-05,
+      "loss": 1.7126,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.101449275362319e-05,
+      "loss": 1.6379,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1304347826086957e-05,
+      "loss": 1.7932,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1594202898550726e-05,
+      "loss": 1.5776,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1884057971014494e-05,
+      "loss": 1.6716,
+      "step": 41
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.2173913043478263e-05,
+      "loss": 1.5683,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2463768115942029e-05,
+      "loss": 1.7941,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2753623188405797e-05,
+      "loss": 1.6677,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 1.6677,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.8042,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3623188405797103e-05,
+      "loss": 1.6045,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.391304347826087e-05,
+      "loss": 1.6353,
+      "step": 48
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.420289855072464e-05,
+      "loss": 1.6586,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 1.626,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4782608695652174e-05,
+      "loss": 1.6165,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5072463768115944e-05,
+      "loss": 1.6579,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.536231884057971e-05,
+      "loss": 1.585,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.565217391304348e-05,
+      "loss": 1.585,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 1.594,
+      "step": 55
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6231884057971015e-05,
+      "loss": 1.6924,
+      "step": 56
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6521739130434785e-05,
+      "loss": 1.8485,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.681159420289855e-05,
+      "loss": 1.6987,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.710144927536232e-05,
+      "loss": 1.5926,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 1.6222,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7681159420289858e-05,
+      "loss": 1.6906,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7971014492753624e-05,
+      "loss": 1.6466,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8260869565217393e-05,
+      "loss": 1.6629,
+      "step": 63
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8550724637681162e-05,
+      "loss": 1.7639,
+      "step": 64
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 1.5877,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 1.6888,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9420289855072467e-05,
+      "loss": 1.6415,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9710144927536236e-05,
+      "loss": 1.6959,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.6053,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999999004091687e-05,
+      "loss": 1.5844,
+      "step": 70
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999960163687307e-05,
+      "loss": 1.7203,
+      "step": 71
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999910368370826e-05,
+      "loss": 1.6762,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999840655066608e-05,
+      "loss": 1.7704,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999751023913506e-05,
+      "loss": 1.8186,
+      "step": 74
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999964147509006e-05,
+      "loss": 1.816,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999951200881446e-05,
+      "loss": 1.6055,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999362625344584e-05,
+      "loss": 1.7048,
+      "step": 77
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999193324977974e-05,
+      "loss": 1.6125,
+      "step": 78
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999004108051846e-05,
+      "loss": 1.7239,
+      "step": 79
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998794974943087e-05,
+      "loss": 1.7377,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998565926068253e-05,
+      "loss": 1.6942,
+      "step": 81
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998316961883563e-05,
+      "loss": 1.6398,
+      "step": 82
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999804808288491e-05,
+      "loss": 1.5898,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997759289607854e-05,
+      "loss": 1.6926,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997450582627614e-05,
+      "loss": 1.5962,
+      "step": 85
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999712196255908e-05,
+      "loss": 1.7171,
+      "step": 86
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996773430056806e-05,
+      "loss": 1.6443,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996404985814998e-05,
+      "loss": 1.7252,
+      "step": 88
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996016630567535e-05,
+      "loss": 1.6777,
+      "step": 89
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995608365087945e-05,
+      "loss": 1.6758,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995180190189424e-05,
+      "loss": 1.6416,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999473210672481e-05,
+      "loss": 1.6161,
+      "step": 92
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999426411558661e-05,
+      "loss": 1.8361,
+      "step": 93
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993776217706972e-05,
+      "loss": 1.6755,
+      "step": 94
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993268414057704e-05,
+      "loss": 1.5446,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992740705650252e-05,
+      "loss": 1.6776,
+      "step": 96
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999219309353572e-05,
+      "loss": 1.5742,
+      "step": 97
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999162557880485e-05,
+      "loss": 1.7124,
+      "step": 98
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991038162588018e-05,
+      "loss": 1.6711,
+      "step": 99
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999043084605526e-05,
+      "loss": 1.5564,
+      "step": 100
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.998980363041624e-05,
+      "loss": 1.666,
+      "step": 101
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989156516920248e-05,
+      "loss": 1.7261,
+      "step": 102
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9988489506856218e-05,
+      "loss": 1.5975,
+      "step": 103
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987802601552717e-05,
+      "loss": 1.5775,
+      "step": 104
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987095802377933e-05,
+      "loss": 1.6664,
+      "step": 105
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986369110739675e-05,
+      "loss": 1.6425,
+      "step": 106
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9985622528085382e-05,
+      "loss": 1.7126,
+      "step": 107
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998485605590211e-05,
+      "loss": 1.5268,
+      "step": 108
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9984069695716534e-05,
+      "loss": 1.7186,
+      "step": 109
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998326344909494e-05,
+      "loss": 1.5536,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982437317643218e-05,
+      "loss": 1.743,
+      "step": 111
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998159130300687e-05,
+      "loss": 1.7855,
+      "step": 112
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9980725406871007e-05,
+      "loss": 1.6741,
+      "step": 113
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979839630960333e-05,
+      "loss": 1.6982,
+      "step": 114
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997893397703915e-05,
+      "loss": 1.9001,
+      "step": 115
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978008446911354e-05,
+      "loss": 1.8314,
+      "step": 116
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9977063042420438e-05,
+      "loss": 1.6957,
+      "step": 117
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976097765449463e-05,
+      "loss": 1.6764,
+      "step": 118
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9975112617921097e-05,
+      "loss": 1.6958,
+      "step": 119
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997410760179756e-05,
+      "loss": 1.6011,
+      "step": 120
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9973082719080673e-05,
+      "loss": 1.665,
+      "step": 121
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9972037971811802e-05,
+      "loss": 1.6208,
+      "step": 122
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.99709733620719e-05,
+      "loss": 1.6691,
+      "step": 123
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9969888891981473e-05,
+      "loss": 1.8241,
+      "step": 124
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9968784563700586e-05,
+      "loss": 1.7574,
+      "step": 125
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9967660379428855e-05,
+      "loss": 1.6484,
+      "step": 126
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9966516341405452e-05,
+      "loss": 1.5869,
+      "step": 127
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996535245190909e-05,
+      "loss": 1.7188,
+      "step": 128
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996416871325803e-05,
+      "loss": 1.6428,
+      "step": 129
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996296512781005e-05,
+      "loss": 1.7432,
+      "step": 130
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996174169796248e-05,
+      "loss": 1.6332,
+      "step": 131
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996049842615217e-05,
+      "loss": 1.7188,
+      "step": 132
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9959235314855485e-05,
+      "loss": 1.7858,
+      "step": 133
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9957952366588307e-05,
+      "loss": 1.5635,
+      "step": 134
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995664958390604e-05,
+      "loss": 1.7286,
+      "step": 135
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9955326969403587e-05,
+      "loss": 1.5419,
+      "step": 136
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9953984525715354e-05,
+      "loss": 1.6727,
+      "step": 137
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995262225551524e-05,
+      "loss": 1.6839,
+      "step": 138
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9951240161516643e-05,
+      "loss": 1.5626,
+      "step": 139
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9949838246472436e-05,
+      "loss": 1.6662,
+      "step": 140
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9948416513174976e-05,
+      "loss": 1.6306,
+      "step": 141
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946974964456094e-05,
+      "loss": 1.6706,
+      "step": 142
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9945513603187096e-05,
+      "loss": 1.8325,
+      "step": 143
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9944032432278743e-05,
+      "loss": 1.7128,
+      "step": 144
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9942531454681254e-05,
+      "loss": 1.6979,
+      "step": 145
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941010673384307e-05,
+      "loss": 1.6811,
+      "step": 146
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9939470091417012e-05,
+      "loss": 1.5436,
+      "step": 147
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993790971184793e-05,
+      "loss": 1.705,
+      "step": 148
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9936329537785054e-05,
+      "loss": 1.5621,
+      "step": 149
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9934729572375792e-05,
+      "loss": 1.7562,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.993310981880699e-05,
+      "loss": 1.6263,
+      "step": 151
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931470280304895e-05,
+      "loss": 1.6247,
+      "step": 152
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992981096013517e-05,
+      "loss": 1.6521,
+      "step": 153
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9928131861602878e-05,
+      "loss": 1.5587,
+      "step": 154
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992643298805247e-05,
+      "loss": 1.6085,
+      "step": 155
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992471434286779e-05,
+      "loss": 1.5774,
+      "step": 156
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9922975929472076e-05,
+      "loss": 1.731,
+      "step": 157
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9921217751327916e-05,
+      "loss": 1.5823,
+      "step": 158
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9919439811937283e-05,
+      "loss": 1.5621,
+      "step": 159
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9917642114841505e-05,
+      "loss": 1.6566,
+      "step": 160
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9915824663621267e-05,
+      "loss": 1.5214,
+      "step": 161
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9913987461896597e-05,
+      "loss": 1.7257,
+      "step": 162
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9912130513326863e-05,
+      "loss": 1.7401,
+      "step": 163
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991025382161077e-05,
+      "loss": 1.5845,
+      "step": 164
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9908357390486342e-05,
+      "loss": 1.6235,
+      "step": 165
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990644122373092e-05,
+      "loss": 1.571,
+      "step": 166
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990450532516116e-05,
+      "loss": 1.8406,
+      "step": 167
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990254969863301e-05,
+      "loss": 1.8326,
+      "step": 168
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9900574348041728e-05,
+      "loss": 1.6099,
+      "step": 169
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9898579277321845e-05,
+      "loss": 1.4999,
+      "step": 170
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.989656449044718e-05,
+      "loss": 1.6982,
+      "step": 171
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9894529991430814e-05,
+      "loss": 1.7241,
+      "step": 172
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.98924757843251e-05,
+      "loss": 1.6838,
+      "step": 173
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9890401873221642e-05,
+      "loss": 1.5703,
+      "step": 174
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888308262251286e-05,
+      "loss": 1.6669,
+      "step": 175
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9886194955584126e-05,
+      "loss": 1.682,
+      "step": 176
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988406195742948e-05,
+      "loss": 1.7266,
+      "step": 177
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9881909272035888e-05,
+      "loss": 1.7065,
+      "step": 178
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9879736903691107e-05,
+      "loss": 1.6784,
+      "step": 179
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9877544856722094e-05,
+      "loss": 1.8589,
+      "step": 180
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9875333135495e-05,
+      "loss": 1.7492,
+      "step": 181
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987310174441518e-05,
+      "loss": 1.6953,
+      "step": 182
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987085068792715e-05,
+      "loss": 1.6772,
+      "step": 183
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.98685799705146e-05,
+      "loss": 1.6507,
+      "step": 184
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9866289596700383e-05,
+      "loss": 1.6677,
+      "step": 185
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.98639795710465e-05,
+      "loss": 1.5531,
+      "step": 186
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9861649898154107e-05,
+      "loss": 1.692,
+      "step": 187
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985930058266348e-05,
+      "loss": 1.5314,
+      "step": 188
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9856931629254032e-05,
+      "loss": 1.5806,
+      "step": 189
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985454304264427e-05,
+      "loss": 1.6998,
+      "step": 190
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985213482759183e-05,
+      "loss": 1.532,
+      "step": 191
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9849706988893433e-05,
+      "loss": 1.4833,
+      "step": 192
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984725953138489e-05,
+      "loss": 1.6384,
+      "step": 193
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984479245994108e-05,
+      "loss": 1.602,
+      "step": 194
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984230577947597e-05,
+      "loss": 1.6837,
+      "step": 195
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983979949494256e-05,
+      "loss": 1.6221,
+      "step": 196
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9837273611332918e-05,
+      "loss": 1.5824,
+      "step": 197
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983472813367813e-05,
+      "loss": 1.5726,
+      "step": 198
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9832163067048335e-05,
+      "loss": 1.8058,
+      "step": 199
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982957841655266e-05,
+      "loss": 1.6665,
+      "step": 200
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9826974187339267e-05,
+      "loss": 1.6163,
+      "step": 201
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9824350384595295e-05,
+      "loss": 1.6621,
+      "step": 202
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9821707013546885e-05,
+      "loss": 1.6097,
+      "step": 203
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9819044079459142e-05,
+      "loss": 1.6398,
+      "step": 204
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9816361587636143e-05,
+      "loss": 1.5231,
+      "step": 205
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9813659543420924e-05,
+      "loss": 1.5495,
+      "step": 206
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981093795219546e-05,
+      "loss": 1.6761,
+      "step": 207
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9808196819380656e-05,
+      "loss": 1.5347,
+      "step": 208
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9805436150436352e-05,
+      "loss": 1.5175,
+      "step": 209
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980265595086129e-05,
+      "loss": 1.6109,
+      "step": 210
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9799856226193125e-05,
+      "loss": 1.5862,
+      "step": 211
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9797036982008385e-05,
+      "loss": 1.6155,
+      "step": 212
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9794198223922496e-05,
+      "loss": 1.6225,
+      "step": 213
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979133995758974e-05,
+      "loss": 1.7578,
+      "step": 214
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978846218870326e-05,
+      "loss": 1.6944,
+      "step": 215
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9785564922995042e-05,
+      "loss": 1.5978,
+      "step": 216
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978264816623591e-05,
+      "loss": 1.4836,
+      "step": 217
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.977971192423551e-05,
+      "loss": 1.7529,
+      "step": 218
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9776756202842297e-05,
+      "loss": 1.5821,
+      "step": 219
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977378100794352e-05,
+      "loss": 1.5954,
+      "step": 220
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977078634546523e-05,
+      "loss": 1.5528,
+      "step": 221
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976777222137224e-05,
+      "loss": 1.7068,
+      "step": 222
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9764738641668137e-05,
+      "loss": 1.6066,
+      "step": 223
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9761685612395253e-05,
+      "loss": 1.639,
+      "step": 224
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9758613139634662e-05,
+      "loss": 1.5999,
+      "step": 225
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9755521229506164e-05,
+      "loss": 1.6008,
+      "step": 226
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9752409888168285e-05,
+      "loss": 1.7106,
+      "step": 227
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 1.6942,
+      "step": 228
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974612893669194e-05,
+      "loss": 1.7407,
+      "step": 229
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9742959339063977e-05,
+      "loss": 1.6449,
+      "step": 230
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9739770335247616e-05,
+      "loss": 1.5437,
+      "step": 231
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.973656193159476e-05,
+      "loss": 1.5952,
+      "step": 232
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9733334134495963e-05,
+      "loss": 1.7003,
+      "step": 233
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9730086950380404e-05,
+      "loss": 1.5441,
+      "step": 234
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9726820385715877e-05,
+      "loss": 1.5202,
+      "step": 235
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9723534447008785e-05,
+      "loss": 1.6691,
+      "step": 236
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.972022914080411e-05,
+      "loss": 1.6877,
+      "step": 237
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9716904473685417e-05,
+      "loss": 1.5395,
+      "step": 238
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971356045227484e-05,
+      "loss": 1.6574,
+      "step": 239
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9710197083233044e-05,
+      "loss": 1.5122,
+      "step": 240
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.970681437325925e-05,
+      "loss": 1.4915,
+      "step": 241
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.97034123290912e-05,
+      "loss": 1.5517,
+      "step": 242
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9699990957505136e-05,
+      "loss": 1.7456,
+      "step": 243
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9696550265315805e-05,
+      "loss": 1.4699,
+      "step": 244
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9693090259376436e-05,
+      "loss": 1.5823,
+      "step": 245
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9689610946578725e-05,
+      "loss": 1.7989,
+      "step": 246
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9686112333852826e-05,
+      "loss": 1.4173,
+      "step": 247
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.968259442816733e-05,
+      "loss": 1.605,
+      "step": 248
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9679057236529266e-05,
+      "loss": 1.6035,
+      "step": 249
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9675500765984066e-05,
+      "loss": 1.5451,
+      "step": 250
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9671925023615572e-05,
+      "loss": 1.8563,
+      "step": 251
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9668330016546004e-05,
+      "loss": 1.464,
+      "step": 252
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9664715751935958e-05,
+      "loss": 1.5198,
+      "step": 253
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9661082236984387e-05,
+      "loss": 1.4562,
+      "step": 254
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965742947892858e-05,
+      "loss": 1.6719,
+      "step": 255
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965375748504417e-05,
+      "loss": 1.629,
+      "step": 256
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9650066262645097e-05,
+      "loss": 1.5683,
+      "step": 257
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964635581908359e-05,
+      "loss": 1.6081,
+      "step": 258
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9642626161750176e-05,
+      "loss": 1.5387,
+      "step": 259
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9638877298073645e-05,
+      "loss": 1.5545,
+      "step": 260
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9635109235521057e-05,
+      "loss": 1.5341,
+      "step": 261
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963132198159769e-05,
+      "loss": 1.5831,
+      "step": 262
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9627515543847068e-05,
+      "loss": 1.6099,
+      "step": 263
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962368992985091e-05,
+      "loss": 1.5442,
+      "step": 264
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961984514722914e-05,
+      "loss": 1.6946,
+      "step": 265
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961598120363986e-05,
+      "loss": 1.6732,
+      "step": 266
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961209810677934e-05,
+      "loss": 1.6603,
+      "step": 267
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9608195864381994e-05,
+      "loss": 1.4686,
+      "step": 268
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.960427448422037e-05,
+      "loss": 1.7164,
+      "step": 269
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9600333974105145e-05,
+      "loss": 1.6552,
+      "step": 270
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9596374341885093e-05,
+      "loss": 1.4366,
+      "step": 271
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9592395595447064e-05,
+      "loss": 1.6662,
+      "step": 272
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9588397742716004e-05,
+      "loss": 1.6731,
+      "step": 273
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584380791654896e-05,
+      "loss": 1.4749,
+      "step": 274
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.958034475026477e-05,
+      "loss": 1.5257,
+      "step": 275
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9576289626584685e-05,
+      "loss": 1.5007,
+      "step": 276
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.95722154286917e-05,
+      "loss": 1.5079,
+      "step": 277
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9568122164700867e-05,
+      "loss": 1.5105,
+      "step": 278
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9564009842765225e-05,
+      "loss": 1.5288,
+      "step": 279
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9559878471075763e-05,
+      "loss": 1.4527,
+      "step": 280
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955572805786141e-05,
+      "loss": 1.5178,
+      "step": 281
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955155861138903e-05,
+      "loss": 1.5058,
+      "step": 282
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9547370139963406e-05,
+      "loss": 1.4538,
+      "step": 283
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.954316265192719e-05,
+      "loss": 1.6261,
+      "step": 284
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9538936155660934e-05,
+      "loss": 1.578,
+      "step": 285
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9534690659583045e-05,
+      "loss": 1.4569,
+      "step": 286
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.953042617214977e-05,
+      "loss": 1.5795,
+      "step": 287
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9526142701855187e-05,
+      "loss": 1.625,
+      "step": 288
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9521840257231183e-05,
+      "loss": 1.6226,
+      "step": 289
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517518846847437e-05,
+      "loss": 1.5874,
+      "step": 290
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.951317847931141e-05,
+      "loss": 1.5774,
+      "step": 291
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9508819163268315e-05,
+      "loss": 1.525,
+      "step": 292
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9504440907401113e-05,
+      "loss": 1.6432,
+      "step": 293
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9500043720430484e-05,
+      "loss": 1.6038,
+      "step": 294
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9495627611114817e-05,
+      "loss": 1.6018,
+      "step": 295
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9491192588250198e-05,
+      "loss": 1.5559,
+      "step": 296
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9486738660670373e-05,
+      "loss": 1.6093,
+      "step": 297
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.948226583724675e-05,
+      "loss": 1.4772,
+      "step": 298
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.947777412688838e-05,
+      "loss": 1.6215,
+      "step": 299
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9473263538541916e-05,
+      "loss": 1.6031,
+      "step": 300
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9468734081191627e-05,
+      "loss": 1.6468,
+      "step": 301
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9464185763859365e-05,
+      "loss": 1.5391,
+      "step": 302
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.945961859560454e-05,
+      "loss": 1.5916,
+      "step": 303
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9455032585524115e-05,
+      "loss": 1.5238,
+      "step": 304
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9450427742752583e-05,
+      "loss": 1.5679,
+      "step": 305
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9445804076461946e-05,
+      "loss": 1.4694,
+      "step": 306
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.944116159586169e-05,
+      "loss": 1.7224,
+      "step": 307
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.94365003101988e-05,
+      "loss": 1.5224,
+      "step": 308
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.943182022875769e-05,
+      "loss": 1.4027,
+      "step": 309
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9427121360860233e-05,
+      "loss": 1.6428,
+      "step": 310
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9422403715865708e-05,
+      "loss": 1.5872,
+      "step": 311
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9417667303170803e-05,
+      "loss": 1.5616,
+      "step": 312
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9412912132209573e-05,
+      "loss": 1.7097,
+      "step": 313
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9408138212453456e-05,
+      "loss": 1.5574,
+      "step": 314
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.940334555341122e-05,
+      "loss": 1.6133,
+      "step": 315
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9398534164628968e-05,
+      "loss": 1.5629,
+      "step": 316
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.93937040556901e-05,
+      "loss": 1.4919,
+      "step": 317
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.938885523621531e-05,
+      "loss": 1.4721,
+      "step": 318
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9383987715862554e-05,
+      "loss": 1.7848,
+      "step": 319
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9379101504327044e-05,
+      "loss": 1.5623,
+      "step": 320
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9374196611341212e-05,
+      "loss": 1.6131,
+      "step": 321
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9369273046674708e-05,
+      "loss": 1.6132,
+      "step": 322
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.936433082013437e-05,
+      "loss": 1.5333,
+      "step": 323
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.935936994156421e-05,
+      "loss": 1.554,
+      "step": 324
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9354390420845387e-05,
+      "loss": 1.5865,
+      "step": 325
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934939226789619e-05,
+      "loss": 1.5857,
+      "step": 326
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9344375492672024e-05,
+      "loss": 1.5555,
+      "step": 327
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933934010516539e-05,
+      "loss": 1.5558,
+      "step": 328
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933428611540585e-05,
+      "loss": 1.6296,
+      "step": 329
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932921353346003e-05,
+      "loss": 1.5004,
+      "step": 330
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932412236943158e-05,
+      "loss": 1.6247,
+      "step": 331
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9319012633461164e-05,
+      "loss": 1.6277,
+      "step": 332
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9313884335726443e-05,
+      "loss": 1.7737,
+      "step": 333
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9308737486442045e-05,
+      "loss": 1.5115,
+      "step": 334
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9303572095859545e-05,
+      "loss": 1.5231,
+      "step": 335
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9298388174267463e-05,
+      "loss": 1.5685,
+      "step": 336
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9293185731991212e-05,
+      "loss": 1.4307,
+      "step": 337
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9287964779393105e-05,
+      "loss": 1.6424,
+      "step": 338
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9282725326872324e-05,
+      "loss": 1.5317,
+      "step": 339
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.92774673848649e-05,
+      "loss": 1.5713,
+      "step": 340
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.927219096384368e-05,
+      "loss": 1.4465,
+      "step": 341
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9266896074318335e-05,
+      "loss": 1.4761,
+      "step": 342
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9261582726835316e-05,
+      "loss": 1.6569,
+      "step": 343
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.925625093197783e-05,
+      "loss": 1.588,
+      "step": 344
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9250900700365837e-05,
+      "loss": 1.6713,
+      "step": 345
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.924553204265602e-05,
+      "loss": 1.6163,
+      "step": 346
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9240144969541754e-05,
+      "loss": 1.5906,
+      "step": 347
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.923473949175311e-05,
+      "loss": 1.561,
+      "step": 348
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9229315620056805e-05,
+      "loss": 1.4219,
+      "step": 349
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9223873365256192e-05,
+      "loss": 1.5169,
+      "step": 350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.921841273819125e-05,
+      "loss": 1.6726,
+      "step": 351
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9212933749738546e-05,
+      "loss": 1.4714,
+      "step": 352
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.920743641081122e-05,
+      "loss": 1.5211,
+      "step": 353
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.920192073235896e-05,
+      "loss": 1.6119,
+      "step": 354
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.919638672536799e-05,
+      "loss": 1.5366,
+      "step": 355
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9190834400861035e-05,
+      "loss": 1.4784,
+      "step": 356
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.918526376989731e-05,
+      "loss": 1.5361,
+      "step": 357
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9179674843572487e-05,
+      "loss": 1.734,
+      "step": 358
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9174067633018682e-05,
+      "loss": 1.5957,
+      "step": 359
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9168442149404438e-05,
+      "loss": 1.5167,
+      "step": 360
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.916279840393467e-05,
+      "loss": 1.4856,
+      "step": 361
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9157136407850706e-05,
+      "loss": 1.5079,
+      "step": 362
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9151456172430186e-05,
+      "loss": 1.6342,
+      "step": 363
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9145757708987103e-05,
+      "loss": 1.6238,
+      "step": 364
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.914004102887176e-05,
+      "loss": 1.602,
+      "step": 365
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9134306143470722e-05,
+      "loss": 1.6316,
+      "step": 366
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9128553064206835e-05,
+      "loss": 1.3469,
+      "step": 367
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912278180253918e-05,
+      "loss": 1.5529,
+      "step": 368
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.911699236996305e-05,
+      "loss": 1.4857,
+      "step": 369
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9111184778009934e-05,
+      "loss": 1.5868,
+      "step": 370
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9105359038247484e-05,
+      "loss": 1.5045,
+      "step": 371
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9099515162279515e-05,
+      "loss": 1.6578,
+      "step": 372
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.909365316174595e-05,
+      "loss": 1.4851,
+      "step": 373
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.908777304832282e-05,
+      "loss": 1.4573,
+      "step": 374
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9081874833722234e-05,
+      "loss": 1.5313,
+      "step": 375
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9075958529692355e-05,
+      "loss": 1.4284,
+      "step": 376
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9070024148017375e-05,
+      "loss": 1.3889,
+      "step": 377
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.906407170051749e-05,
+      "loss": 1.5237,
+      "step": 378
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905810119904889e-05,
+      "loss": 1.5036,
+      "step": 379
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9052112655503713e-05,
+      "loss": 1.6423,
+      "step": 380
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9046106081810047e-05,
+      "loss": 1.5939,
+      "step": 381
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.904008148993188e-05,
+      "loss": 1.643,
+      "step": 382
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903403889186909e-05,
+      "loss": 1.5539,
+      "step": 383
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9027978299657436e-05,
+      "loss": 1.5466,
+      "step": 384
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9021899725368498e-05,
+      "loss": 1.4804,
+      "step": 385
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9015803181109685e-05,
+      "loss": 1.6738,
+      "step": 386
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 1.6189,
+      "step": 387
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9003556231290988e-05,
+      "loss": 1.3754,
+      "step": 388
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8997405850124786e-05,
+      "loss": 1.4995,
+      "step": 389
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8991237547776014e-05,
+      "loss": 1.7472,
+      "step": 390
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.89850513365308e-05,
+      "loss": 1.5819,
+      "step": 391
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.897884722871094e-05,
+      "loss": 1.4441,
+      "step": 392
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8972625236673887e-05,
+      "loss": 1.4998,
+      "step": 393
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8966385372812703e-05,
+      "loss": 1.7693,
+      "step": 394
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.896012764955605e-05,
+      "loss": 1.5972,
+      "step": 395
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.895385207936817e-05,
+      "loss": 1.4894,
+      "step": 396
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8947558674748844e-05,
+      "loss": 1.6153,
+      "step": 397
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8941247448233386e-05,
+      "loss": 1.5506,
+      "step": 398
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8934918412392596e-05,
+      "loss": 1.4783,
+      "step": 399
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8928571579832756e-05,
+      "loss": 1.6733,
+      "step": 400
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.892220696319559e-05,
+      "loss": 1.4514,
+      "step": 401
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.891582457515825e-05,
+      "loss": 1.6588,
+      "step": 402
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8909424428433278e-05,
+      "loss": 1.5374,
+      "step": 403
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.89030065357686e-05,
+      "loss": 1.5054,
+      "step": 404
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8896570909947477e-05,
+      "loss": 1.5964,
+      "step": 405
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889011756378849e-05,
+      "loss": 1.5827,
+      "step": 406
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.888364651014553e-05,
+      "loss": 1.4867,
+      "step": 407
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.887715776190775e-05,
+      "loss": 1.5016,
+      "step": 408
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8870651331999542e-05,
+      "loss": 1.6249,
+      "step": 409
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.886412723338052e-05,
+      "loss": 1.6036,
+      "step": 410
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8857585479045493e-05,
+      "loss": 1.6172,
+      "step": 411
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.885102608202444e-05,
+      "loss": 1.5305,
+      "step": 412
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8844449055382473e-05,
+      "loss": 1.5793,
+      "step": 413
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8837854412219828e-05,
+      "loss": 1.6235,
+      "step": 414
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8831242165671816e-05,
+      "loss": 1.6813,
+      "step": 415
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8824612328908828e-05,
+      "loss": 1.553,
+      "step": 416
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8817964915136277e-05,
+      "loss": 1.5233,
+      "step": 417
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8811299937594598e-05,
+      "loss": 1.5602,
+      "step": 418
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.88046174095592e-05,
+      "loss": 1.6497,
+      "step": 419
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8797917344340453e-05,
+      "loss": 1.5556,
+      "step": 420
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8791199755283664e-05,
+      "loss": 1.4972,
+      "step": 421
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8784464655769033e-05,
+      "loss": 1.4895,
+      "step": 422
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8777712059211643e-05,
+      "loss": 1.4853,
+      "step": 423
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8770941979061432e-05,
+      "loss": 1.5,
+      "step": 424
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8764154428803155e-05,
+      "loss": 1.5266,
+      "step": 425
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.875734942195637e-05,
+      "loss": 1.6001,
+      "step": 426
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.87505269720754e-05,
+      "loss": 1.6224,
+      "step": 427
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8743687092749318e-05,
+      "loss": 1.8942,
+      "step": 428
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8736829797601903e-05,
+      "loss": 1.597,
+      "step": 429
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8729955100291634e-05,
+      "loss": 1.4492,
+      "step": 430
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.872306301451165e-05,
+      "loss": 1.553,
+      "step": 431
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8716153553989716e-05,
+      "loss": 1.6568,
+      "step": 432
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8709226732488216e-05,
+      "loss": 1.6419,
+      "step": 433
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.87022825638041e-05,
+      "loss": 1.4201,
+      "step": 434
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8695321061768886e-05,
+      "loss": 1.5144,
+      "step": 435
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.868834224024861e-05,
+      "loss": 1.6008,
+      "step": 436
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86813461131438e-05,
+      "loss": 1.5232,
+      "step": 437
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8674332694389464e-05,
+      "loss": 1.5325,
+      "step": 438
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8667301997955038e-05,
+      "loss": 1.5205,
+      "step": 439
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.5523,
+      "step": 440
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8653188828095754e-05,
+      "loss": 1.6117,
+      "step": 441
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8646106382781738e-05,
+      "loss": 1.5324,
+      "step": 442
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8639006716009275e-05,
+      "loss": 1.5437,
+      "step": 443
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8631889841919596e-05,
+      "loss": 1.646,
+      "step": 444
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862475577468821e-05,
+      "loss": 1.4662,
+      "step": 445
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8617604528524876e-05,
+      "loss": 1.6842,
+      "step": 446
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8610436117673557e-05,
+      "loss": 1.6015,
+      "step": 447
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8603250556412418e-05,
+      "loss": 1.4642,
+      "step": 448
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8596047859053776e-05,
+      "loss": 1.5905,
+      "step": 449
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8588828039944086e-05,
+      "loss": 1.3811,
+      "step": 450
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8581591113463903e-05,
+      "loss": 1.477,
+      "step": 451
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8574337094027858e-05,
+      "loss": 1.5653,
+      "step": 452
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8567065996084628e-05,
+      "loss": 1.6238,
+      "step": 453
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8559777834116906e-05,
+      "loss": 1.5427,
+      "step": 454
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8552472622641372e-05,
+      "loss": 1.478,
+      "step": 455
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.854515037620867e-05,
+      "loss": 1.4926,
+      "step": 456
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8537811109403372e-05,
+      "loss": 1.5102,
+      "step": 457
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8530454836843953e-05,
+      "loss": 1.5598,
+      "step": 458
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8523081573182754e-05,
+      "loss": 1.6468,
+      "step": 459
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.851569133310597e-05,
+      "loss": 1.5438,
+      "step": 460
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8508284131333604e-05,
+      "loss": 1.4735,
+      "step": 461
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8500859982619438e-05,
+      "loss": 1.4544,
+      "step": 462
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8493418901751016e-05,
+      "loss": 1.5896,
+      "step": 463
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8485960903549614e-05,
+      "loss": 1.5448,
+      "step": 464
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.847848600287019e-05,
+      "loss": 1.48,
+      "step": 465
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8470994214601378e-05,
+      "loss": 1.5051,
+      "step": 466
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.846348555366544e-05,
+      "loss": 1.4919,
+      "step": 467
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.845596003501826e-05,
+      "loss": 1.4351,
+      "step": 468
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8448417673649292e-05,
+      "loss": 1.5142,
+      "step": 469
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.844085848458153e-05,
+      "loss": 1.5037,
+      "step": 470
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8433282482871497e-05,
+      "loss": 1.4373,
+      "step": 471
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8425689683609198e-05,
+      "loss": 1.3986,
+      "step": 472
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8418080101918095e-05,
+      "loss": 1.5879,
+      "step": 473
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.841045375295508e-05,
+      "loss": 1.5525,
+      "step": 474
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8402810651910444e-05,
+      "loss": 1.4899,
+      "step": 475
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.839515081400784e-05,
+      "loss": 1.5907,
+      "step": 476
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8387474254504265e-05,
+      "loss": 1.4677,
+      "step": 477
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8379780988690014e-05,
+      "loss": 1.451,
+      "step": 478
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.837207103188866e-05,
+      "loss": 1.3821,
+      "step": 479
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8364344399457028e-05,
+      "loss": 1.5024,
+      "step": 480
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8356601106785148e-05,
+      "loss": 1.5263,
+      "step": 481
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8348841169296247e-05,
+      "loss": 1.5388,
+      "step": 482
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8341064602446686e-05,
+      "loss": 1.6173,
+      "step": 483
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8333271421725968e-05,
+      "loss": 1.5454,
+      "step": 484
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8325461642656676e-05,
+      "loss": 1.4468,
+      "step": 485
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8317635280794466e-05,
+      "loss": 1.3626,
+      "step": 486
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8309792351728006e-05,
+      "loss": 1.4551,
+      "step": 487
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8301932871078975e-05,
+      "loss": 1.3878,
+      "step": 488
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.829405685450202e-05,
+      "loss": 1.525,
+      "step": 489
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.828616431768472e-05,
+      "loss": 1.4175,
+      "step": 490
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8278255276347563e-05,
+      "loss": 1.5459,
+      "step": 491
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8270329746243903e-05,
+      "loss": 1.612,
+      "step": 492
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.826238774315995e-05,
+      "loss": 1.4209,
+      "step": 493
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8254429282914715e-05,
+      "loss": 1.5253,
+      "step": 494
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.824645438135999e-05,
+      "loss": 1.4988,
+      "step": 495
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.823846305438032e-05,
+      "loss": 1.4734,
+      "step": 496
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8230455317892957e-05,
+      "loss": 1.5854,
+      "step": 497
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.822243118784785e-05,
+      "loss": 1.3589,
+      "step": 498
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8214390680227588e-05,
+      "loss": 1.5336,
+      "step": 499
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.82063338110474e-05,
+      "loss": 1.4623,
+      "step": 500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8198260596355077e-05,
+      "loss": 1.4971,
+      "step": 501
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.819017105223099e-05,
+      "loss": 1.4198,
+      "step": 502
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8182065194788024e-05,
+      "loss": 1.6746,
+      "step": 503
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8173943040171567e-05,
+      "loss": 1.7106,
+      "step": 504
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8165804604559455e-05,
+      "loss": 1.4187,
+      "step": 505
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8157649904161963e-05,
+      "loss": 1.5615,
+      "step": 506
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.814947895522176e-05,
+      "loss": 1.534,
+      "step": 507
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8141291774013876e-05,
+      "loss": 1.4895,
+      "step": 508
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8133088376845675e-05,
+      "loss": 1.4571,
+      "step": 509
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8124868780056814e-05,
+      "loss": 1.6314,
+      "step": 510
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8116633000019233e-05,
+      "loss": 1.708,
+      "step": 511
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8108381053137086e-05,
+      "loss": 1.5999,
+      "step": 512
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8100112955846746e-05,
+      "loss": 1.4073,
+      "step": 513
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8091828724616742e-05,
+      "loss": 1.4875,
+      "step": 514
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8083528375947744e-05,
+      "loss": 1.455,
+      "step": 515
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.807521192637253e-05,
+      "loss": 1.462,
+      "step": 516
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8066879392455932e-05,
+      "loss": 1.3701,
+      "step": 517
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8058530790794837e-05,
+      "loss": 1.5939,
+      "step": 518
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.805016613801813e-05,
+      "loss": 1.5063,
+      "step": 519
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8041785450786655e-05,
+      "loss": 1.533,
+      "step": 520
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8033388745793218e-05,
+      "loss": 1.3381,
+      "step": 521
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8024976039762507e-05,
+      "loss": 1.6287,
+      "step": 522
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.801654734945109e-05,
+      "loss": 1.5195,
+      "step": 523
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.800810269164738e-05,
+      "loss": 1.4591,
+      "step": 524
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7999642083171576e-05,
+      "loss": 1.5765,
+      "step": 525
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7991165540875662e-05,
+      "loss": 1.4827,
+      "step": 526
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7982673081643364e-05,
+      "loss": 1.3935,
+      "step": 527
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7974164722390094e-05,
+      "loss": 1.5364,
+      "step": 528
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7965640480062945e-05,
+      "loss": 1.3546,
+      "step": 529
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.795710037164065e-05,
+      "loss": 1.5956,
+      "step": 530
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7948544414133534e-05,
+      "loss": 1.7426,
+      "step": 531
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7939972624583496e-05,
+      "loss": 1.3783,
+      "step": 532
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.793138502006397e-05,
+      "loss": 1.4114,
+      "step": 533
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.792278161767989e-05,
+      "loss": 1.4835,
+      "step": 534
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7914162434567653e-05,
+      "loss": 1.4548,
+      "step": 535
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7905527487895095e-05,
+      "loss": 1.4565,
+      "step": 536
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7896876794861443e-05,
+      "loss": 1.6375,
+      "step": 537
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7888210372697292e-05,
+      "loss": 1.7816,
+      "step": 538
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7879528238664567e-05,
+      "loss": 1.4677,
+      "step": 539
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7870830410056487e-05,
+      "loss": 1.6285,
+      "step": 540
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7862116904197534e-05,
+      "loss": 1.5072,
+      "step": 541
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7853387738443407e-05,
+      "loss": 1.4388,
+      "step": 542
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7844642930181008e-05,
+      "loss": 1.4634,
+      "step": 543
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7835882496828392e-05,
+      "loss": 1.3691,
+      "step": 544
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.782710645583473e-05,
+      "loss": 1.7348,
+      "step": 545
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 1.5476,
+      "step": 546
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7809507620876406e-05,
+      "loss": 1.4226,
+      "step": 547
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7800684861965386e-05,
+      "loss": 1.552,
+      "step": 548
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.779184656552056e-05,
+      "loss": 1.3445,
+      "step": 549
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.778299274914619e-05,
+      "loss": 1.3728,
+      "step": 550
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7774123430477464e-05,
+      "loss": 1.4584,
+      "step": 551
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7765238627180424e-05,
+      "loss": 1.3861,
+      "step": 552
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.775633835695198e-05,
+      "loss": 1.5778,
+      "step": 553
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7747422637519838e-05,
+      "loss": 1.4371,
+      "step": 554
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.773849148664247e-05,
+      "loss": 1.4731,
+      "step": 555
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7729544922109097e-05,
+      "loss": 1.4639,
+      "step": 556
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7720582961739628e-05,
+      "loss": 1.4996,
+      "step": 557
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7711605623384653e-05,
+      "loss": 1.4186,
+      "step": 558
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7702612924925377e-05,
+      "loss": 1.7156,
+      "step": 559
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.769360488427361e-05,
+      "loss": 1.4141,
+      "step": 560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7684581519371714e-05,
+      "loss": 1.4783,
+      "step": 561
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.767554284819258e-05,
+      "loss": 1.4907,
+      "step": 562
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7666488888739587e-05,
+      "loss": 1.4398,
+      "step": 563
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7657419659046555e-05,
+      "loss": 1.5578,
+      "step": 564
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7648335177177725e-05,
+      "loss": 1.2612,
+      "step": 565
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7639235461227727e-05,
+      "loss": 1.4069,
+      "step": 566
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7630120529321518e-05,
+      "loss": 1.4186,
+      "step": 567
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7620990399614385e-05,
+      "loss": 1.6359,
+      "step": 568
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7611845090291858e-05,
+      "loss": 1.5227,
+      "step": 569
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.760268461956972e-05,
+      "loss": 1.5207,
+      "step": 570
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.759350900569395e-05,
+      "loss": 1.3005,
+      "step": 571
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.758431826694069e-05,
+      "loss": 1.5959,
+      "step": 572
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7575112421616203e-05,
+      "loss": 1.4916,
+      "step": 573
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7565891488056846e-05,
+      "loss": 1.562,
+      "step": 574
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7556655484629028e-05,
+      "loss": 1.3952,
+      "step": 575
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7547404429729174e-05,
+      "loss": 1.442,
+      "step": 576
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.753813834178369e-05,
+      "loss": 1.4802,
+      "step": 577
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7528857239248925e-05,
+      "loss": 1.4531,
+      "step": 578
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.751956114061113e-05,
+      "loss": 1.4268,
+      "step": 579
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.751025006438643e-05,
+      "loss": 1.4611,
+      "step": 580
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7500924029120782e-05,
+      "loss": 1.4781,
+      "step": 581
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7491583053389937e-05,
+      "loss": 1.6709,
+      "step": 582
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.748222715579941e-05,
+      "loss": 1.4697,
+      "step": 583
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7472856354984427e-05,
+      "loss": 1.8062,
+      "step": 584
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7463470669609907e-05,
+      "loss": 1.5088,
+      "step": 585
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7454070118370416e-05,
+      "loss": 1.5084,
+      "step": 586
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7444654719990128e-05,
+      "loss": 1.4291,
+      "step": 587
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.743522449322279e-05,
+      "loss": 1.543,
+      "step": 588
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7425779456851683e-05,
+      "loss": 1.4138,
+      "step": 589
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.741631962968959e-05,
+      "loss": 1.4493,
+      "step": 590
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7406845030578747e-05,
+      "loss": 1.4229,
+      "step": 591
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7397355678390823e-05,
+      "loss": 1.6379,
+      "step": 592
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7387851592026868e-05,
+      "loss": 1.6639,
+      "step": 593
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7378332790417275e-05,
+      "loss": 1.3675,
+      "step": 594
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7368799292521754e-05,
+      "loss": 1.5564,
+      "step": 595
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7359251117329284e-05,
+      "loss": 1.4036,
+      "step": 596
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.734968828385808e-05,
+      "loss": 1.5228,
+      "step": 597
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.734011081115555e-05,
+      "loss": 1.4199,
+      "step": 598
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7330518718298263e-05,
+      "loss": 1.3642,
+      "step": 599
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7320912024391915e-05,
+      "loss": 1.6231,
+      "step": 600
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7311290748571273e-05,
+      "loss": 1.4194,
+      "step": 601
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.730165491000016e-05,
+      "loss": 1.5672,
+      "step": 602
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.729200452787139e-05,
+      "loss": 1.4293,
+      "step": 603
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7282339621406764e-05,
+      "loss": 1.4722,
+      "step": 604
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7272660209857e-05,
+      "loss": 1.3389,
+      "step": 605
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.726296631250171e-05,
+      "loss": 1.5645,
+      "step": 606
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7253257948649357e-05,
+      "loss": 1.4608,
+      "step": 607
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7243535137637227e-05,
+      "loss": 1.4738,
+      "step": 608
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7233797898831376e-05,
+      "loss": 1.4042,
+      "step": 609
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7224046251626596e-05,
+      "loss": 1.4588,
+      "step": 610
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.721428021544638e-05,
+      "loss": 1.4518,
+      "step": 611
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.720449980974288e-05,
+      "loss": 1.4475,
+      "step": 612
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7194705053996873e-05,
+      "loss": 1.3607,
+      "step": 613
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7184895967717715e-05,
+      "loss": 1.6416,
+      "step": 614
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.717507257044331e-05,
+      "loss": 1.3898,
+      "step": 615
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7165234881740064e-05,
+      "loss": 1.4316,
+      "step": 616
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7155382921202844e-05,
+      "loss": 1.4831,
+      "step": 617
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7145516708454953e-05,
+      "loss": 1.4551,
+      "step": 618
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.713563626314808e-05,
+      "loss": 1.3642,
+      "step": 619
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7125741604962254e-05,
+      "loss": 1.5155,
+      "step": 620
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.711583275360582e-05,
+      "loss": 1.5094,
+      "step": 621
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7105909728815398e-05,
+      "loss": 1.4877,
+      "step": 622
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.709597255035583e-05,
+      "loss": 1.4399,
+      "step": 623
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7086021238020154e-05,
+      "loss": 1.5668,
+      "step": 624
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7076055811629556e-05,
+      "loss": 1.3735,
+      "step": 625
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7066076291033348e-05,
+      "loss": 1.5926,
+      "step": 626
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7056082696108896e-05,
+      "loss": 1.4911,
+      "step": 627
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7046075046761614e-05,
+      "loss": 1.3751,
+      "step": 628
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7036053362924896e-05,
+      "loss": 1.486,
+      "step": 629
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7026017664560108e-05,
+      "loss": 1.4863,
+      "step": 630
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7015967971656513e-05,
+      "loss": 1.5688,
+      "step": 631
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7005904304231263e-05,
+      "loss": 1.5164,
+      "step": 632
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.699582668232934e-05,
+      "loss": 1.4478,
+      "step": 633
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6985735126023505e-05,
+      "loss": 1.5325,
+      "step": 634
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6975629655414304e-05,
+      "loss": 1.5206,
+      "step": 635
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6965510290629973e-05,
+      "loss": 1.5853,
+      "step": 636
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.695537705182643e-05,
+      "loss": 1.5414,
+      "step": 637
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6945229959187232e-05,
+      "loss": 1.3947,
+      "step": 638
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6935069032923525e-05,
+      "loss": 1.5173,
+      "step": 639
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6924894293274016e-05,
+      "loss": 1.4498,
+      "step": 640
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6914705760504913e-05,
+      "loss": 1.315,
+      "step": 641
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6904503454909905e-05,
+      "loss": 1.5449,
+      "step": 642
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.689428739681012e-05,
+      "loss": 1.3916,
+      "step": 643
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.688405760655407e-05,
+      "loss": 1.4827,
+      "step": 644
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6873814104517617e-05,
+      "loss": 1.313,
+      "step": 645
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6863556911103948e-05,
+      "loss": 1.3906,
+      "step": 646
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6853286046743505e-05,
+      "loss": 1.4809,
+      "step": 647
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6843001531893967e-05,
+      "loss": 1.4373,
+      "step": 648
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.68327033870402e-05,
+      "loss": 1.3667,
+      "step": 649
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.682239163269422e-05,
+      "loss": 1.6086,
+      "step": 650
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6812066289395157e-05,
+      "loss": 1.6025,
+      "step": 651
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6801727377709195e-05,
+      "loss": 1.3733,
+      "step": 652
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.679137491822955e-05,
+      "loss": 1.3865,
+      "step": 653
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6781008931576433e-05,
+      "loss": 1.4149,
+      "step": 654
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6770629438396973e-05,
+      "loss": 1.4549,
+      "step": 655
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.676023645936523e-05,
+      "loss": 1.443,
+      "step": 656
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6749830015182106e-05,
+      "loss": 1.5147,
+      "step": 657
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.673941012657533e-05,
+      "loss": 1.3656,
+      "step": 658
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6728976814299413e-05,
+      "loss": 1.5812,
+      "step": 659
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6718530099135596e-05,
+      "loss": 1.4519,
+      "step": 660
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.670807000189182e-05,
+      "loss": 1.4256,
+      "step": 661
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6697596543402684e-05,
+      "loss": 1.4598,
+      "step": 662
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6687109744529394e-05,
+      "loss": 1.3108,
+      "step": 663
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.667660962615973e-05,
+      "loss": 1.4414,
+      "step": 664
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6666096209208e-05,
+      "loss": 1.455,
+      "step": 665
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665556951461501e-05,
+      "loss": 1.4531,
+      "step": 666
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6645029563348e-05,
+      "loss": 1.4081,
+      "step": 667
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6634476376400624e-05,
+      "loss": 1.5535,
+      "step": 668
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6623909974792888e-05,
+      "loss": 1.4525,
+      "step": 669
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6613330379571135e-05,
+      "loss": 1.3994,
+      "step": 670
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6602737611807975e-05,
+      "loss": 1.3552,
+      "step": 671
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6592131692602257e-05,
+      "loss": 1.3903,
+      "step": 672
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6581512643079028e-05,
+      "loss": 1.4062,
+      "step": 673
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.657088048438949e-05,
+      "loss": 1.5593,
+      "step": 674
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.656023523771095e-05,
+      "loss": 1.4703,
+      "step": 675
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.654957692424679e-05,
+      "loss": 1.3948,
+      "step": 676
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6538905565226416e-05,
+      "loss": 1.4193,
+      "step": 677
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6528221181905217e-05,
+      "loss": 1.4513,
+      "step": 678
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6517523795564527e-05,
+      "loss": 1.4719,
+      "step": 679
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6506813427511574e-05,
+      "loss": 1.4331,
+      "step": 680
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6496090099079452e-05,
+      "loss": 1.6245,
+      "step": 681
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.648535383162706e-05,
+      "loss": 1.3991,
+      "step": 682
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.647460464653908e-05,
+      "loss": 1.4738,
+      "step": 683
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6463842565225914e-05,
+      "loss": 1.3917,
+      "step": 684
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6453067609123656e-05,
+      "loss": 1.4979,
+      "step": 685
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6442279799694037e-05,
+      "loss": 1.3204,
+      "step": 686
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.64314791584244e-05,
+      "loss": 1.4417,
+      "step": 687
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6420665706827647e-05,
+      "loss": 1.393,
+      "step": 688
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.640983946644218e-05,
+      "loss": 1.5019,
+      "step": 689
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6399000458831893e-05,
+      "loss": 1.4612,
+      "step": 690
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6388148705586097e-05,
+      "loss": 1.4936,
+      "step": 691
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6377284228319496e-05,
+      "loss": 1.4793,
+      "step": 692
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6366407048672135e-05,
+      "loss": 1.5915,
+      "step": 693
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6355517188309366e-05,
+      "loss": 1.531,
+      "step": 694
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6344614668921787e-05,
+      "loss": 1.4645,
+      "step": 695
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6333699512225224e-05,
+      "loss": 1.3448,
+      "step": 696
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6322771739960664e-05,
+      "loss": 1.5339,
+      "step": 697
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6311831373894228e-05,
+      "loss": 1.5129,
+      "step": 698
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6300878435817115e-05,
+      "loss": 1.4603,
+      "step": 699
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.628991294754557e-05,
+      "loss": 1.3988,
+      "step": 700
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6278934930920834e-05,
+      "loss": 1.3577,
+      "step": 701
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.626794440780911e-05,
+      "loss": 1.3563,
+      "step": 702
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.62569414001015e-05,
+      "loss": 1.2694,
+      "step": 703
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6245925929713976e-05,
+      "loss": 1.4932,
+      "step": 704
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 1.5791,
+      "step": 705
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.622385768868716e-05,
+      "loss": 1.336,
+      "step": 706
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6212804962003757e-05,
+      "loss": 1.5013,
+      "step": 707
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.620173986055213e-05,
+      "loss": 1.4369,
+      "step": 708
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6190662406371937e-05,
+      "loss": 1.4798,
+      "step": 709
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6179572621527434e-05,
+      "loss": 1.5673,
+      "step": 710
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.616847052810744e-05,
+      "loss": 1.5329,
+      "step": 711
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6157356148225282e-05,
+      "loss": 1.5054,
+      "step": 712
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6146229504018777e-05,
+      "loss": 1.5076,
+      "step": 713
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.613509061765015e-05,
+      "loss": 1.5683,
+      "step": 714
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6123939511306028e-05,
+      "loss": 1.3936,
+      "step": 715
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6112776207197368e-05,
+      "loss": 1.4506,
+      "step": 716
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6101600727559423e-05,
+      "loss": 1.4543,
+      "step": 717
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.60904130946517e-05,
+      "loss": 1.4286,
+      "step": 718
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.607921333075791e-05,
+      "loss": 1.3896,
+      "step": 719
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6068001458185934e-05,
+      "loss": 1.5686,
+      "step": 720
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6056777499267764e-05,
+      "loss": 1.3608,
+      "step": 721
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6045541476359464e-05,
+      "loss": 1.4546,
+      "step": 722
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.603429341184114e-05,
+      "loss": 1.361,
+      "step": 723
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6023033328116863e-05,
+      "loss": 1.349,
+      "step": 724
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6011761247614664e-05,
+      "loss": 1.6018,
+      "step": 725
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.600047719278645e-05,
+      "loss": 1.3169,
+      "step": 726
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5989181186108003e-05,
+      "loss": 1.3935,
+      "step": 727
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.597787325007889e-05,
+      "loss": 1.4533,
+      "step": 728
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.596655340722244e-05,
+      "loss": 1.3166,
+      "step": 729
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5955221680085715e-05,
+      "loss": 1.5751,
+      "step": 730
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.594387809123943e-05,
+      "loss": 1.4022,
+      "step": 731
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.593252266327794e-05,
+      "loss": 1.4902,
+      "step": 732
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.592115541881917e-05,
+      "loss": 1.4441,
+      "step": 733
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5909776380504583e-05,
+      "loss": 1.4561,
+      "step": 734
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5898385570999146e-05,
+      "loss": 1.5479,
+      "step": 735
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.588698301299126e-05,
+      "loss": 1.4435,
+      "step": 736
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5875568729192728e-05,
+      "loss": 1.4532,
+      "step": 737
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5864142742338707e-05,
+      "loss": 1.4191,
+      "step": 738
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5852705075187674e-05,
+      "loss": 1.5131,
+      "step": 739
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.584125575052136e-05,
+      "loss": 1.6136,
+      "step": 740
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5829794791144723e-05,
+      "loss": 1.6131,
+      "step": 741
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5818322219885888e-05,
+      "loss": 1.3705,
+      "step": 742
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.580683805959612e-05,
+      "loss": 1.3605,
+      "step": 743
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5795342333149757e-05,
+      "loss": 1.4814,
+      "step": 744
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5783835063444176e-05,
+      "loss": 1.6124,
+      "step": 745
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.577231627339975e-05,
+      "loss": 1.4262,
+      "step": 746
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.57607859859598e-05,
+      "loss": 1.3018,
+      "step": 747
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5749244224090537e-05,
+      "loss": 1.4496,
+      "step": 748
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.573769101078104e-05,
+      "loss": 1.4053,
+      "step": 749
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.572612636904319e-05,
+      "loss": 1.4787,
+      "step": 750
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5714550321911636e-05,
+      "loss": 1.4017,
+      "step": 751
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.570296289244373e-05,
+      "loss": 1.4664,
+      "step": 752
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5691364103719515e-05,
+      "loss": 1.3794,
+      "step": 753
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5679753978841655e-05,
+      "loss": 1.5015,
+      "step": 754
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.566813254093538e-05,
+      "loss": 1.4208,
+      "step": 755
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.565649981314847e-05,
+      "loss": 1.2914,
+      "step": 756
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5644855818651184e-05,
+      "loss": 1.4495,
+      "step": 757
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.563320058063622e-05,
+      "loss": 1.5148,
+      "step": 758
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5621534122318682e-05,
+      "loss": 1.415,
+      "step": 759
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5609856466936013e-05,
+      "loss": 1.3962,
+      "step": 760
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.559816763774796e-05,
+      "loss": 1.2682,
+      "step": 761
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5586467658036526e-05,
+      "loss": 1.2831,
+      "step": 762
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5574756551105926e-05,
+      "loss": 1.4633,
+      "step": 763
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.556303434028254e-05,
+      "loss": 1.4293,
+      "step": 764
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5551301048914863e-05,
+      "loss": 1.3353,
+      "step": 765
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5539556700373454e-05,
+      "loss": 1.4561,
+      "step": 766
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5527801318050904e-05,
+      "loss": 1.2696,
+      "step": 767
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5516034925361785e-05,
+      "loss": 1.3438,
+      "step": 768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5504257545742585e-05,
+      "loss": 1.3211,
+      "step": 769
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5492469202651687e-05,
+      "loss": 1.2836,
+      "step": 770
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5480669919569313e-05,
+      "loss": 1.4497,
+      "step": 771
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5468859719997468e-05,
+      "loss": 1.3471,
+      "step": 772
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5457038627459905e-05,
+      "loss": 1.4046,
+      "step": 773
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.544520666550207e-05,
+      "loss": 1.3368,
+      "step": 774
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5433363857691067e-05,
+      "loss": 1.3998,
+      "step": 775
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5421510227615587e-05,
+      "loss": 1.4141,
+      "step": 776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.54096457988859e-05,
+      "loss": 1.3368,
+      "step": 777
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5397770595133763e-05,
+      "loss": 1.2981,
+      "step": 778
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.538588464001241e-05,
+      "loss": 1.2526,
+      "step": 779
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.537398795719648e-05,
+      "loss": 1.4553,
+      "step": 780
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.536208057038199e-05,
+      "loss": 1.3564,
+      "step": 781
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5350162503286262e-05,
+      "loss": 1.2815,
+      "step": 782
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.533823377964791e-05,
+      "loss": 1.3393,
+      "step": 783
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5326294423226754e-05,
+      "loss": 1.2078,
+      "step": 784
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5314344457803812e-05,
+      "loss": 1.3704,
+      "step": 785
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5302383907181212e-05,
+      "loss": 1.4243,
+      "step": 786
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5290412795182193e-05,
+      "loss": 1.3263,
+      "step": 787
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5278431145651006e-05,
+      "loss": 1.3138,
+      "step": 788
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.52664389824529e-05,
+      "loss": 1.2868,
+      "step": 789
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5254436329474062e-05,
+      "loss": 1.4212,
+      "step": 790
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5242423210621584e-05,
+      "loss": 1.227,
+      "step": 791
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5230399649823389e-05,
+      "loss": 1.3289,
+      "step": 792
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5218365671028207e-05,
+      "loss": 1.4862,
+      "step": 793
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.520632129820552e-05,
+      "loss": 1.3405,
+      "step": 794
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5194266555345505e-05,
+      "loss": 1.3097,
+      "step": 795
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5182201466459007e-05,
+      "loss": 1.3609,
+      "step": 796
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.517012605557746e-05,
+      "loss": 1.3197,
+      "step": 797
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5158040346752878e-05,
+      "loss": 1.4184,
+      "step": 798
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.514594436405777e-05,
+      "loss": 1.2864,
+      "step": 799
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.513383813158512e-05,
+      "loss": 1.4705,
+      "step": 800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5121721673448319e-05,
+      "loss": 1.2534,
+      "step": 801
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5109595013781136e-05,
+      "loss": 1.2593,
+      "step": 802
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5097458176737647e-05,
+      "loss": 1.602,
+      "step": 803
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5085311186492206e-05,
+      "loss": 1.5528,
+      "step": 804
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5073154067239396e-05,
+      "loss": 1.3256,
+      "step": 805
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.506098684319396e-05,
+      "loss": 1.3318,
+      "step": 806
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5048809538590789e-05,
+      "loss": 1.4107,
+      "step": 807
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5036622177684832e-05,
+      "loss": 1.3083,
+      "step": 808
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5024424784751079e-05,
+      "loss": 1.4553,
+      "step": 809
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5012217384084501e-05,
+      "loss": 1.229,
+      "step": 810
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.4023,
+      "step": 811
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4987772656832369e-05,
+      "loss": 1.3542,
+      "step": 812
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4975535378936228e-05,
+      "loss": 1.3595,
+      "step": 813
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4963288190685993e-05,
+      "loss": 1.4248,
+      "step": 814
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4951031116475819e-05,
+      "loss": 1.4272,
+      "step": 815
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4938764180719548e-05,
+      "loss": 1.3736,
+      "step": 816
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4926487407850667e-05,
+      "loss": 1.324,
+      "step": 817
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4914200822322255e-05,
+      "loss": 1.3547,
+      "step": 818
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.490190444860694e-05,
+      "loss": 1.2825,
+      "step": 819
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.488959831119684e-05,
+      "loss": 1.2481,
+      "step": 820
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4877282434603527e-05,
+      "loss": 1.4295,
+      "step": 821
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4864956843357967e-05,
+      "loss": 1.3544,
+      "step": 822
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4852621562010484e-05,
+      "loss": 1.3211,
+      "step": 823
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4840276615130687e-05,
+      "loss": 1.3247,
+      "step": 824
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.482792202730745e-05,
+      "loss": 1.298,
+      "step": 825
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.481555782314885e-05,
+      "loss": 1.2263,
+      "step": 826
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.480318402728211e-05,
+      "loss": 1.2477,
+      "step": 827
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4790800664353562e-05,
+      "loss": 1.381,
+      "step": 828
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4778407759028599e-05,
+      "loss": 1.4058,
+      "step": 829
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4766005335991608e-05,
+      "loss": 1.2427,
+      "step": 830
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.475359341994595e-05,
+      "loss": 1.2379,
+      "step": 831
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4741172035613884e-05,
+      "loss": 1.3705,
+      "step": 832
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4728741207736525e-05,
+      "loss": 1.2921,
+      "step": 833
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4716300961073808e-05,
+      "loss": 1.4335,
+      "step": 834
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4703851320404416e-05,
+      "loss": 1.3733,
+      "step": 835
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4691392310525761e-05,
+      "loss": 1.205,
+      "step": 836
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4678923956253894e-05,
+      "loss": 1.3927,
+      "step": 837
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.46664462824235e-05,
+      "loss": 1.2401,
+      "step": 838
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4653959313887813e-05,
+      "loss": 1.1496,
+      "step": 839
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4641463075518587e-05,
+      "loss": 1.2139,
+      "step": 840
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.462895759220603e-05,
+      "loss": 1.2934,
+      "step": 841
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4616442888858781e-05,
+      "loss": 1.433,
+      "step": 842
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.460391899040383e-05,
+      "loss": 1.3424,
+      "step": 843
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4591385921786484e-05,
+      "loss": 1.5823,
+      "step": 844
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4578843707970323e-05,
+      "loss": 1.2476,
+      "step": 845
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4566292373937133e-05,
+      "loss": 1.3494,
+      "step": 846
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.455373194468687e-05,
+      "loss": 1.4667,
+      "step": 847
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4541162445237604e-05,
+      "loss": 1.3184,
+      "step": 848
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4528583900625481e-05,
+      "loss": 1.1635,
+      "step": 849
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4515996335904646e-05,
+      "loss": 1.2017,
+      "step": 850
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4503399776147223e-05,
+      "loss": 1.2251,
+      "step": 851
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4490794246443249e-05,
+      "loss": 1.266,
+      "step": 852
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4478179771900634e-05,
+      "loss": 1.2223,
+      "step": 853
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4465556377645086e-05,
+      "loss": 1.3447,
+      "step": 854
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4452924088820101e-05,
+      "loss": 1.3507,
+      "step": 855
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.444028293058688e-05,
+      "loss": 1.3325,
+      "step": 856
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4427632928124288e-05,
+      "loss": 1.2199,
+      "step": 857
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4414974106628816e-05,
+      "loss": 1.3947,
+      "step": 858
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4402306491314508e-05,
+      "loss": 1.2455,
+      "step": 859
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4389630107412942e-05,
+      "loss": 1.2585,
+      "step": 860
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4376944980173138e-05,
+      "loss": 1.1741,
+      "step": 861
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4364251134861552e-05,
+      "loss": 1.4873,
+      "step": 862
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.435154859676199e-05,
+      "loss": 1.2544,
+      "step": 863
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 1.5139,
+      "step": 864
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.432611754342072e-05,
+      "loss": 1.3318,
+      "step": 865
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4313389078833005e-05,
+      "loss": 1.3818,
+      "step": 866
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4300652022765207e-05,
+      "loss": 1.3969,
+      "step": 867
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4287906400587204e-05,
+      "loss": 1.2875,
+      "step": 868
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4275152237685938e-05,
+      "loss": 1.3772,
+      "step": 869
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4262389559465364e-05,
+      "loss": 1.2559,
+      "step": 870
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4249618391346399e-05,
+      "loss": 1.3725,
+      "step": 871
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.423683875876686e-05,
+      "loss": 1.3986,
+      "step": 872
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4224050687181442e-05,
+      "loss": 1.4285,
+      "step": 873
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4211254202061632e-05,
+      "loss": 1.3434,
+      "step": 874
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4198449328895685e-05,
+      "loss": 1.3032,
+      "step": 875
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4185636093188558e-05,
+      "loss": 1.5187,
+      "step": 876
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4172814520461867e-05,
+      "loss": 1.3192,
+      "step": 877
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4159984636253834e-05,
+      "loss": 1.287,
+      "step": 878
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4147146466119235e-05,
+      "loss": 1.3911,
+      "step": 879
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4134300035629353e-05,
+      "loss": 1.4266,
+      "step": 880
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4121445370371922e-05,
+      "loss": 1.4889,
+      "step": 881
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4108582495951077e-05,
+      "loss": 1.3725,
+      "step": 882
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4095711437987303e-05,
+      "loss": 1.3562,
+      "step": 883
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4082832222117388e-05,
+      "loss": 1.192,
+      "step": 884
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.406994487399437e-05,
+      "loss": 1.2356,
+      "step": 885
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4057049419287486e-05,
+      "loss": 1.354,
+      "step": 886
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4044145883682108e-05,
+      "loss": 1.4592,
+      "step": 887
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4031234292879726e-05,
+      "loss": 1.3197,
+      "step": 888
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4018314672597848e-05,
+      "loss": 1.2761,
+      "step": 889
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4005387048569997e-05,
+      "loss": 1.3186,
+      "step": 890
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3992451446545624e-05,
+      "loss": 1.3058,
+      "step": 891
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3979507892290078e-05,
+      "loss": 1.3268,
+      "step": 892
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3966556411584548e-05,
+      "loss": 1.4442,
+      "step": 893
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3953597030226006e-05,
+      "loss": 1.5029,
+      "step": 894
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.394062977402717e-05,
+      "loss": 1.3462,
+      "step": 895
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3927654668816425e-05,
+      "loss": 1.3557,
+      "step": 896
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3914671740437811e-05,
+      "loss": 1.2279,
+      "step": 897
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3901681014750932e-05,
+      "loss": 1.3755,
+      "step": 898
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.388868251763094e-05,
+      "loss": 1.3507,
+      "step": 899
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3875676274968453e-05,
+      "loss": 1.3406,
+      "step": 900
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3862662312669518e-05,
+      "loss": 1.2504,
+      "step": 901
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3849640656655572e-05,
+      "loss": 1.3773,
+      "step": 902
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3836611332863356e-05,
+      "loss": 1.3935,
+      "step": 903
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3823574367244898e-05,
+      "loss": 1.4741,
+      "step": 904
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3810529785767444e-05,
+      "loss": 1.2404,
+      "step": 905
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3797477614413404e-05,
+      "loss": 1.5165,
+      "step": 906
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3784417879180314e-05,
+      "loss": 1.3026,
+      "step": 907
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.377135060608077e-05,
+      "loss": 1.3751,
+      "step": 908
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3758275821142382e-05,
+      "loss": 1.303,
+      "step": 909
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.374519355040773e-05,
+      "loss": 1.4328,
+      "step": 910
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.373210381993429e-05,
+      "loss": 1.3693,
+      "step": 911
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3719006655794414e-05,
+      "loss": 1.2239,
+      "step": 912
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3705902084075244e-05,
+      "loss": 1.5724,
+      "step": 913
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3692790130878684e-05,
+      "loss": 1.28,
+      "step": 914
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3679670822321347e-05,
+      "loss": 1.2516,
+      "step": 915
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3666544184534483e-05,
+      "loss": 1.2594,
+      "step": 916
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3653410243663953e-05,
+      "loss": 1.3354,
+      "step": 917
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3640269025870152e-05,
+      "loss": 1.335,
+      "step": 918
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3627120557327982e-05,
+      "loss": 1.2795,
+      "step": 919
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3613964864226778e-05,
+      "loss": 1.336,
+      "step": 920
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3600801972770272e-05,
+      "loss": 1.3821,
+      "step": 921
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3587631909176526e-05,
+      "loss": 1.3757,
+      "step": 922
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3574454699677893e-05,
+      "loss": 1.3641,
+      "step": 923
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3561270370520957e-05,
+      "loss": 1.4058,
+      "step": 924
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3548078947966487e-05,
+      "loss": 1.2977,
+      "step": 925
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3534880458289377e-05,
+      "loss": 1.4556,
+      "step": 926
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3521674927778594e-05,
+      "loss": 1.3128,
+      "step": 927
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.350846238273714e-05,
+      "loss": 1.1963,
+      "step": 928
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3495242849481973e-05,
+      "loss": 1.3155,
+      "step": 929
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.348201635434399e-05,
+      "loss": 1.3442,
+      "step": 930
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3468782923667936e-05,
+      "loss": 1.2025,
+      "step": 931
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.345554258381238e-05,
+      "loss": 1.2701,
+      "step": 932
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3442295361149651e-05,
+      "loss": 1.2732,
+      "step": 933
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3429041282065788e-05,
+      "loss": 1.3274,
+      "step": 934
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.341578037296049e-05,
+      "loss": 1.4623,
+      "step": 935
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3402512660247046e-05,
+      "loss": 1.2364,
+      "step": 936
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3389238170352318e-05,
+      "loss": 1.2667,
+      "step": 937
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.337595692971665e-05,
+      "loss": 1.3677,
+      "step": 938
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.336266896479384e-05,
+      "loss": 1.2828,
+      "step": 939
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3349374302051072e-05,
+      "loss": 1.3525,
+      "step": 940
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3336072967968882e-05,
+      "loss": 1.1946,
+      "step": 941
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3322764989041086e-05,
+      "loss": 1.1775,
+      "step": 942
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.330945039177474e-05,
+      "loss": 1.5145,
+      "step": 943
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.329612920269008e-05,
+      "loss": 1.1247,
+      "step": 944
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.328280144832047e-05,
+      "loss": 1.197,
+      "step": 945
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3269467155212352e-05,
+      "loss": 1.4466,
+      "step": 946
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3256126349925195e-05,
+      "loss": 1.309,
+      "step": 947
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3242779059031436e-05,
+      "loss": 1.3885,
+      "step": 948
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.322942530911643e-05,
+      "loss": 1.305,
+      "step": 949
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3216065126778399e-05,
+      "loss": 1.3519,
+      "step": 950
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3202698538628376e-05,
+      "loss": 1.343,
+      "step": 951
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3189325571290158e-05,
+      "loss": 1.3325,
+      "step": 952
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3175946251400234e-05,
+      "loss": 1.1634,
+      "step": 953
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3162560605607763e-05,
+      "loss": 1.202,
+      "step": 954
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3149168660574495e-05,
+      "loss": 1.2865,
+      "step": 955
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.313577044297473e-05,
+      "loss": 1.355,
+      "step": 956
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3122365979495259e-05,
+      "loss": 1.4174,
+      "step": 957
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3108955296835313e-05,
+      "loss": 1.3318,
+      "step": 958
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.309553842170652e-05,
+      "loss": 1.3663,
+      "step": 959
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3082115380832829e-05,
+      "loss": 1.344,
+      "step": 960
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3068686200950475e-05,
+      "loss": 1.3489,
+      "step": 961
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3055250908807926e-05,
+      "loss": 1.3992,
+      "step": 962
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3041809531165819e-05,
+      "loss": 1.3718,
+      "step": 963
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3028362094796908e-05,
+      "loss": 1.3789,
+      "step": 964
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3014908626486032e-05,
+      "loss": 1.3265,
+      "step": 965
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3001449153030024e-05,
+      "loss": 1.2635,
+      "step": 966
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.2987983701237688e-05,
+      "loss": 1.1998,
+      "step": 967
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2974512297929733e-05,
+      "loss": 1.174,
+      "step": 968
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2961034969938732e-05,
+      "loss": 1.2968,
+      "step": 969
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2947551744109044e-05,
+      "loss": 1.2934,
+      "step": 970
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2934062647296783e-05,
+      "loss": 1.1205,
+      "step": 971
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.292056770636976e-05,
+      "loss": 1.4206,
+      "step": 972
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.290706694820742e-05,
+      "loss": 1.2795,
+      "step": 973
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2893560399700797e-05,
+      "loss": 1.3978,
+      "step": 974
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.288004808775246e-05,
+      "loss": 1.3022,
+      "step": 975
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2866530039276455e-05,
+      "loss": 1.1413,
+      "step": 976
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2853006281198257e-05,
+      "loss": 1.3979,
+      "step": 977
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2839476840454714e-05,
+      "loss": 1.337,
+      "step": 978
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.282594174399399e-05,
+      "loss": 1.3242,
+      "step": 979
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2812401018775512e-05,
+      "loss": 1.3954,
+      "step": 980
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2798854691769927e-05,
+      "loss": 1.4023,
+      "step": 981
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.278530278995903e-05,
+      "loss": 1.1767,
+      "step": 982
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2771745340335726e-05,
+      "loss": 1.3525,
+      "step": 983
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2758182369903969e-05,
+      "loss": 1.3312,
+      "step": 984
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2744613905678707e-05,
+      "loss": 1.2137,
+      "step": 985
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2731039974685833e-05,
+      "loss": 1.2834,
+      "step": 986
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2717460603962132e-05,
+      "loss": 1.315,
+      "step": 987
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2703875820555218e-05,
+      "loss": 1.3543,
+      "step": 988
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.269028565152349e-05,
+      "loss": 1.3201,
+      "step": 989
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2676690123936066e-05,
+      "loss": 1.2848,
+      "step": 990
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2663089264872751e-05,
+      "loss": 1.3589,
+      "step": 991
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2649483101423962e-05,
+      "loss": 1.1543,
+      "step": 992
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2635871660690677e-05,
+      "loss": 1.5853,
+      "step": 993
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2622254969784393e-05,
+      "loss": 1.3134,
+      "step": 994
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2608633055827064e-05,
+      "loss": 1.3688,
+      "step": 995
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.259500594595104e-05,
+      "loss": 1.3339,
+      "step": 996
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2581373667299026e-05,
+      "loss": 1.3099,
+      "step": 997
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2567736247024023e-05,
+      "loss": 1.4095,
+      "step": 998
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2554093712289267e-05,
+      "loss": 1.27,
+      "step": 999
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2540446090268193e-05,
+      "loss": 1.2994,
+      "step": 1000
+    }
+  ],
+  "max_steps": 2295,
+  "num_train_epochs": 3,
+  "total_flos": 2.343808717197148e+17,
+  "trial_name": null,
+  "trial_params": null
+}