diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9763 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.998615597600369,
+  "global_step": 1623,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 1.474,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 1.606,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 1.3749,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 1.1853,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 1.0875,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 1.2057,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.1873,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 1.2746,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 1.1772,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 1.1983,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.1898,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 1.1077,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 1.1141,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.1189,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 1.1466,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 1.1978,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.1208,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 1.0906,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 0.9878,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 1.0768,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.0951,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 1.1037,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 1.203,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.0092,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 1.0867,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 1.1116,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 1.1422,
+      "step": 27
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.0838,
+      "step": 28
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 1.0814,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 1.0979,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 1.0702,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 1.1343,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 1.1528,
+      "step": 33
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 1.1043,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.0949,
+      "step": 35
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 1.2111,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 1.1106,
+      "step": 37
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 1.083,
+      "step": 38
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 1.0464,
+      "step": 39
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 1.0985,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 1.1269,
+      "step": 41
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 1.2304,
+      "step": 42
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 1.0777,
+      "step": 43
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 1.0537,
+      "step": 44
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 1.0442,
+      "step": 45
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 1.0942,
+      "step": 46
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 1.0321,
+      "step": 47
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 1.0969,
+      "step": 48
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.0509,
+      "step": 49
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999998008133858e-05,
+      "loss": 1.1962,
+      "step": 50
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999992032543367e-05,
+      "loss": 1.1959,
+      "step": 51
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999820732523318e-05,
+      "loss": 1.1079,
+      "step": 52
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999681303004274e-05,
+      "loss": 1.1107,
+      "step": 53
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999950203743199e-05,
+      "loss": 1.0374,
+      "step": 54
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999928293652062e-05,
+      "loss": 1.1153,
+      "step": 55
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999902400114299e-05,
+      "loss": 1.0821,
+      "step": 56
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999872523233064e-05,
+      "loss": 1.1063,
+      "step": 57
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998386631273784e-05,
+      "loss": 1.0502,
+      "step": 58
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999800819932131e-05,
+      "loss": 1.0651,
+      "step": 59
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997589937980796e-05,
+      "loss": 1.029,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999713184891848e-05,
+      "loss": 1.0361,
+      "step": 61
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999663393395927e-05,
+      "loss": 1.1057,
+      "step": 62
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999609619508672e-05,
+      "loss": 1.1235,
+      "step": 63
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995518634443044e-05,
+      "loss": 1.0977,
+      "step": 64
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999490125432909e-05,
+      "loss": 1.1004,
+      "step": 65
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994244057204325e-05,
+      "loss": 1.0567,
+      "step": 66
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993547045686856e-05,
+      "loss": 1.0747,
+      "step": 67
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992810222553387e-05,
+      "loss": 1.0127,
+      "step": 68
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992033590739224e-05,
+      "loss": 1.0685,
+      "step": 69
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991217153338262e-05,
+      "loss": 1.0852,
+      "step": 70
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990360913602966e-05,
+      "loss": 1.1406,
+      "step": 71
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989464874944367e-05,
+      "loss": 1.0662,
+      "step": 72
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9988529040932043e-05,
+      "loss": 1.1209,
+      "step": 73
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987553415294108e-05,
+      "loss": 1.0658,
+      "step": 74
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998653800191719e-05,
+      "loss": 1.0456,
+      "step": 75
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998548280484643e-05,
+      "loss": 1.0471,
+      "step": 76
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9984387828285444e-05,
+      "loss": 1.0482,
+      "step": 77
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9983253076596326e-05,
+      "loss": 1.1459,
+      "step": 78
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982078554299627e-05,
+      "loss": 1.1808,
+      "step": 79
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998086426607433e-05,
+      "loss": 1.1517,
+      "step": 80
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979610216757833e-05,
+      "loss": 1.0436,
+      "step": 81
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978316411345927e-05,
+      "loss": 1.0514,
+      "step": 82
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976982854992793e-05,
+      "loss": 1.1283,
+      "step": 83
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997560955301096e-05,
+      "loss": 1.0385,
+      "step": 84
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9974196510871295e-05,
+      "loss": 1.0939,
+      "step": 85
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997274373420298e-05,
+      "loss": 1.1783,
+      "step": 86
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997125122879349e-05,
+      "loss": 1.1195,
+      "step": 87
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9969719000588564e-05,
+      "loss": 1.124,
+      "step": 88
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9968147055692194e-05,
+      "loss": 1.1471,
+      "step": 89
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9966535400366584e-05,
+      "loss": 1.0936,
+      "step": 90
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9964884041032135e-05,
+      "loss": 1.1899,
+      "step": 91
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9963192984267423e-05,
+      "loss": 1.1015,
+      "step": 92
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9961462236809166e-05,
+      "loss": 1.0122,
+      "step": 93
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.99596918055522e-05,
+      "loss": 1.1511,
+      "step": 94
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9957881697549446e-05,
+      "loss": 1.1892,
+      "step": 95
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995603192001189e-05,
+      "loss": 1.1046,
+      "step": 96
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9954142480308553e-05,
+      "loss": 1.0578,
+      "step": 97
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9952213385966453e-05,
+      "loss": 1.0707,
+      "step": 98
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9950244644670594e-05,
+      "loss": 1.0442,
+      "step": 99
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9948236264263903e-05,
+      "loss": 1.1229,
+      "step": 100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946188252747235e-05,
+      "loss": 1.1009,
+      "step": 101
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994410061827932e-05,
+      "loss": 1.1013,
+      "step": 102
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941973369176738e-05,
+      "loss": 1.0636,
+      "step": 103
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993980651391387e-05,
+      "loss": 1.0701,
+      "step": 104
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9937600061122897e-05,
+      "loss": 1.0738,
+      "step": 105
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9935354019593732e-05,
+      "loss": 1.0734,
+      "step": 106
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9933068398274004e-05,
+      "loss": 1.1487,
+      "step": 107
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9930743206269016e-05,
+      "loss": 1.1047,
+      "step": 108
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992837845284171e-05,
+      "loss": 1.1385,
+      "step": 109
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9925974147412634e-05,
+      "loss": 1.0883,
+      "step": 110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9923530299559895e-05,
+      "loss": 1.0845,
+      "step": 111
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9921046919019123e-05,
+      "loss": 1.1773,
+      "step": 112
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991852401568345e-05,
+      "loss": 1.0363,
+      "step": 113
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9915961599603446e-05,
+      "loss": 1.1372,
+      "step": 114
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9913359680987086e-05,
+      "loss": 1.0842,
+      "step": 115
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9910718270199717e-05,
+      "loss": 1.0717,
+      "step": 116
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990803737776402e-05,
+      "loss": 1.1578,
+      "step": 117
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9905317014359947e-05,
+      "loss": 1.1626,
+      "step": 118
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9902557190824696e-05,
+      "loss": 1.1503,
+      "step": 119
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9899757918152667e-05,
+      "loss": 1.1666,
+      "step": 120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9896919207495416e-05,
+      "loss": 1.1354,
+      "step": 121
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9894041070161604e-05,
+      "loss": 1.1205,
+      "step": 122
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9891123517616955e-05,
+      "loss": 1.0502,
+      "step": 123
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888166561484226e-05,
+      "loss": 1.0906,
+      "step": 124
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9885170213543134e-05,
+      "loss": 1.2075,
+      "step": 125
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9882134485730326e-05,
+      "loss": 1.1792,
+      "step": 126
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.987905939013933e-05,
+      "loss": 1.1856,
+      "step": 127
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9875944939020503e-05,
+      "loss": 1.089,
+      "step": 128
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9872791144780986e-05,
+      "loss": 1.166,
+      "step": 129
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986959801998465e-05,
+      "loss": 1.045,
+      "step": 130
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986636557735205e-05,
+      "loss": 1.1216,
+      "step": 131
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986309382976037e-05,
+      "loss": 1.1878,
+      "step": 132
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985978279024338e-05,
+      "loss": 1.0644,
+      "step": 133
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985643247199137e-05,
+      "loss": 1.1067,
+      "step": 134
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985304288835112e-05,
+      "loss": 1.1286,
+      "step": 135
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984961405282581e-05,
+      "loss": 1.1075,
+      "step": 136
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9846145979075015e-05,
+      "loss": 1.1135,
+      "step": 137
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9842638680914607e-05,
+      "loss": 1.1348,
+      "step": 138
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9839092172316726e-05,
+      "loss": 1.0342,
+      "step": 139
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9835506467409707e-05,
+      "loss": 1.1042,
+      "step": 140
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9831881580478045e-05,
+      "loss": 1.071,
+      "step": 141
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9828217525962314e-05,
+      "loss": 1.1098,
+      "step": 142
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982451431845913e-05,
+      "loss": 1.0992,
+      "step": 143
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.982077197272108e-05,
+      "loss": 1.0778,
+      "step": 144
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9816990503656662e-05,
+      "loss": 1.014,
+      "step": 145
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9813169926330244e-05,
+      "loss": 1.0123,
+      "step": 146
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9809310255961977e-05,
+      "loss": 1.1205,
+      "step": 147
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9805411507927757e-05,
+      "loss": 1.1662,
+      "step": 148
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9801473697759158e-05,
+      "loss": 1.1896,
+      "step": 149
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9797496841143355e-05,
+      "loss": 1.0423,
+      "step": 150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979348095392308e-05,
+      "loss": 0.9753,
+      "step": 151
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978942605209655e-05,
+      "loss": 1.1847,
+      "step": 152
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9785332151817414e-05,
+      "loss": 1.0538,
+      "step": 153
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9781199269394675e-05,
+      "loss": 1.1153,
+      "step": 154
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977702742129263e-05,
+      "loss": 1.0966,
+      "step": 155
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9772816624130798e-05,
+      "loss": 1.0638,
+      "step": 156
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9768566894683876e-05,
+      "loss": 1.0693,
+      "step": 157
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9764278249881644e-05,
+      "loss": 1.0461,
+      "step": 158
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9759950706808917e-05,
+      "loss": 1.0819,
+      "step": 159
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9755584282705466e-05,
+      "loss": 1.1481,
+      "step": 160
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9751178994965955e-05,
+      "loss": 1.1925,
+      "step": 161
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9746734861139873e-05,
+      "loss": 1.0585,
+      "step": 162
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974225189893146e-05,
+      "loss": 1.1063,
+      "step": 163
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9737730126199633e-05,
+      "loss": 1.1235,
+      "step": 164
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.973316956095793e-05,
+      "loss": 1.0368,
+      "step": 165
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.972857022137442e-05,
+      "loss": 1.0718,
+      "step": 166
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.972393212577164e-05,
+      "loss": 1.0995,
+      "step": 167
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9719255292626522e-05,
+      "loss": 1.0803,
+      "step": 168
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971453974057031e-05,
+      "loss": 1.168,
+      "step": 169
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9709785488388512e-05,
+      "loss": 1.1011,
+      "step": 170
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.970499255502079e-05,
+      "loss": 1.122,
+      "step": 171
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9700160959560907e-05,
+      "loss": 1.1292,
+      "step": 172
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.969529072125665e-05,
+      "loss": 1.0836,
+      "step": 173
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.969038185950974e-05,
+      "loss": 1.1087,
+      "step": 174
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9685434393875768e-05,
+      "loss": 1.1094,
+      "step": 175
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9680448344064116e-05,
+      "loss": 1.1205,
+      "step": 176
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967542372993787e-05,
+      "loss": 1.022,
+      "step": 177
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9670360571513747e-05,
+      "loss": 1.1467,
+      "step": 178
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9665258888962018e-05,
+      "loss": 1.0075,
+      "step": 179
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9660118702606414e-05,
+      "loss": 1.1371,
+      "step": 180
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9654940032924066e-05,
+      "loss": 1.054,
+      "step": 181
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964972290054541e-05,
+      "loss": 1.1512,
+      "step": 182
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9644467326254094e-05,
+      "loss": 1.0107,
+      "step": 183
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9639173330986933e-05,
+      "loss": 1.1752,
+      "step": 184
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9633840935833775e-05,
+      "loss": 1.1311,
+      "step": 185
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9628470162037465e-05,
+      "loss": 1.0723,
+      "step": 186
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9623061030993717e-05,
+      "loss": 1.0136,
+      "step": 187
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961761356425107e-05,
+      "loss": 1.0894,
+      "step": 188
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961212778351077e-05,
+      "loss": 1.0824,
+      "step": 189
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9606603710626698e-05,
+      "loss": 1.0663,
+      "step": 190
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9601041367605282e-05,
+      "loss": 1.1014,
+      "step": 191
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.959544077660541e-05,
+      "loss": 1.1258,
+      "step": 192
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9589801959938332e-05,
+      "loss": 1.0874,
+      "step": 193
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584124940067587e-05,
+      "loss": 1.0975,
+      "step": 194
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9578409739608902e-05,
+      "loss": 1.1002,
+      "step": 195
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.957265638133011e-05,
+      "loss": 1.0525,
+      "step": 196
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.956686488815104e-05,
+      "loss": 1.1092,
+      "step": 197
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.956103528314346e-05,
+      "loss": 1.1902,
+      "step": 198
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955516758953095e-05,
+      "loss": 1.1797,
+      "step": 199
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9549261830688833e-05,
+      "loss": 1.0531,
+      "step": 200
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9543318030144072e-05,
+      "loss": 1.0864,
+      "step": 201
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9537336211575175e-05,
+      "loss": 0.9924,
+      "step": 202
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9531316398812105e-05,
+      "loss": 1.1942,
+      "step": 203
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9525258615836185e-05,
+      "loss": 1.1197,
+      "step": 204
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9519162886780006e-05,
+      "loss": 1.0611,
+      "step": 205
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.951302923592731e-05,
+      "loss": 1.1128,
+      "step": 206
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.950685768771293e-05,
+      "loss": 1.0707,
+      "step": 207
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9500648266722658e-05,
+      "loss": 1.0915,
+      "step": 208
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.949440099769316e-05,
+      "loss": 1.0628,
+      "step": 209
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9488115905511894e-05,
+      "loss": 1.0866,
+      "step": 210
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9481793015216975e-05,
+      "loss": 1.0386,
+      "step": 211
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9475432351997115e-05,
+      "loss": 1.1326,
+      "step": 212
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.946903394119148e-05,
+      "loss": 1.0705,
+      "step": 213
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9462597808289637e-05,
+      "loss": 1.0917,
+      "step": 214
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9456123978931408e-05,
+      "loss": 1.091,
+      "step": 215
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9449612478906808e-05,
+      "loss": 1.1072,
+      "step": 216
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9443063334155896e-05,
+      "loss": 1.1336,
+      "step": 217
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.943647657076872e-05,
+      "loss": 1.1053,
+      "step": 218
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9429852214985182e-05,
+      "loss": 1.2072,
+      "step": 219
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9423190293194935e-05,
+      "loss": 1.0611,
+      "step": 220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9416490831937298e-05,
+      "loss": 1.145,
+      "step": 221
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9409753857901132e-05,
+      "loss": 1.1795,
+      "step": 222
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.940297939792473e-05,
+      "loss": 1.1826,
+      "step": 223
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.939616747899574e-05,
+      "loss": 1.0747,
+      "step": 224
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9389318128251014e-05,
+      "loss": 1.0709,
+      "step": 225
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.938243137297653e-05,
+      "loss": 1.0236,
+      "step": 226
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.937550724060728e-05,
+      "loss": 0.9554,
+      "step": 227
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9368545758727163e-05,
+      "loss": 1.1775,
+      "step": 228
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9361546955068845e-05,
+      "loss": 1.1175,
+      "step": 229
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9354510857513692e-05,
+      "loss": 1.0056,
+      "step": 230
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934743749409163e-05,
+      "loss": 1.1486,
+      "step": 231
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9340326892981052e-05,
+      "loss": 1.0826,
+      "step": 232
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9333179082508686e-05,
+      "loss": 1.0676,
+      "step": 233
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9325994091149488e-05,
+      "loss": 1.0855,
+      "step": 234
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9318771947526555e-05,
+      "loss": 1.0594,
+      "step": 235
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.931151268041096e-05,
+      "loss": 1.1298,
+      "step": 236
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9304216318721687e-05,
+      "loss": 1.1177,
+      "step": 237
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9296882891525483e-05,
+      "loss": 1.0892,
+      "step": 238
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9289512428036765e-05,
+      "loss": 1.1096,
+      "step": 239
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9282104957617483e-05,
+      "loss": 1.0981,
+      "step": 240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9274660509777012e-05,
+      "loss": 1.2161,
+      "step": 241
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9267179114172045e-05,
+      "loss": 1.0874,
+      "step": 242
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9259660800606457e-05,
+      "loss": 1.164,
+      "step": 243
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9252105599031196e-05,
+      "loss": 1.1774,
+      "step": 244
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9244513539544166e-05,
+      "loss": 1.0881,
+      "step": 245
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.923688465239009e-05,
+      "loss": 1.0809,
+      "step": 246
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9229218967960427e-05,
+      "loss": 1.1099,
+      "step": 247
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9221516516793198e-05,
+      "loss": 1.0855,
+      "step": 248
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9213777329572913e-05,
+      "loss": 1.143,
+      "step": 249
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.920600143713042e-05,
+      "loss": 1.1826,
+      "step": 250
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9198188870442792e-05,
+      "loss": 1.0915,
+      "step": 251
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9190339660633205e-05,
+      "loss": 1.1156,
+      "step": 252
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.918245383897081e-05,
+      "loss": 1.0454,
+      "step": 253
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9174531436870608e-05,
+      "loss": 1.0303,
+      "step": 254
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9166572485893325e-05,
+      "loss": 1.1221,
+      "step": 255
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.91585770177453e-05,
+      "loss": 1.136,
+      "step": 256
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9150545064278327e-05,
+      "loss": 1.1156,
+      "step": 257
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9142476657489565e-05,
+      "loss": 1.0963,
+      "step": 258
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9134371829521386e-05,
+      "loss": 1.0925,
+      "step": 259
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912623061266125e-05,
+      "loss": 1.0312,
+      "step": 260
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9118053039341594e-05,
+      "loss": 1.0668,
+      "step": 261
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9109839142139672e-05,
+      "loss": 1.047,
+      "step": 262
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910158895377746e-05,
+      "loss": 1.0092,
+      "step": 263
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9093302507121493e-05,
+      "loss": 1.0627,
+      "step": 264
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9084979835182757e-05,
+      "loss": 1.1508,
+      "step": 265
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9076620971116546e-05,
+      "loss": 1.0863,
+      "step": 266
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9068225948222346e-05,
+      "loss": 1.0993,
+      "step": 267
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9059794799943676e-05,
+      "loss": 1.1685,
+      "step": 268
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905132755986797e-05,
+      "loss": 1.0988,
+      "step": 269
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9042824261726447e-05,
+      "loss": 1.1726,
+      "step": 270
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9034284939393977e-05,
+      "loss": 1.1207,
+      "step": 271
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9025709626888926e-05,
+      "loss": 1.148,
+      "step": 272
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9017098358373046e-05,
+      "loss": 1.1281,
+      "step": 273
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9008451168151326e-05,
+      "loss": 1.0742,
+      "step": 274
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8999768090671858e-05,
+      "loss": 1.1112,
+      "step": 275
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8991049160525695e-05,
+      "loss": 1.2168,
+      "step": 276
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8982294412446722e-05,
+      "loss": 1.0861,
+      "step": 277
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8973503881311516e-05,
+      "loss": 1.0164,
+      "step": 278
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.896467760213919e-05,
+      "loss": 1.108,
+      "step": 279
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8955815610091286e-05,
+      "loss": 1.1164,
+      "step": 280
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.89469179404716e-05,
+      "loss": 1.1062,
+      "step": 281
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8937984628726075e-05,
+      "loss": 1.1096,
+      "step": 282
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.892901571044263e-05,
+      "loss": 1.075,
+      "step": 283
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8920011221351026e-05,
+      "loss": 1.1279,
+      "step": 284
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.891097119732275e-05,
+      "loss": 1.0898,
+      "step": 285
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8901895674370826e-05,
+      "loss": 1.1007,
+      "step": 286
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889278468864972e-05,
+      "loss": 1.0872,
+      "step": 287
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8883638276455146e-05,
+      "loss": 1.1086,
+      "step": 288
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8874456474223972e-05,
+      "loss": 1.1765,
+      "step": 289
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8865239318534038e-05,
+      "loss": 1.0242,
+      "step": 290
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8855986846104022e-05,
+      "loss": 1.0309,
+      "step": 291
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.88466990937933e-05,
+      "loss": 1.083,
+      "step": 292
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.883737609860179e-05,
+      "loss": 1.1214,
+      "step": 293
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.88280178976698e-05,
+      "loss": 1.1677,
+      "step": 294
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8818624528277914e-05,
+      "loss": 1.1236,
+      "step": 295
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.880919602784679e-05,
+      "loss": 1.1768,
+      "step": 296
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.879973243393705e-05,
+      "loss": 1.0618,
+      "step": 297
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8790233784249122e-05,
+      "loss": 1.0983,
+      "step": 298
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8780700116623082e-05,
+      "loss": 1.1921,
+      "step": 299
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.877113146903851e-05,
+      "loss": 1.0294,
+      "step": 300
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.876152787961433e-05,
+      "loss": 1.1448,
+      "step": 301
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8751889386608682e-05,
+      "loss": 1.1545,
+      "step": 302
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8742216028418737e-05,
+      "loss": 1.1469,
+      "step": 303
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8732507843580563e-05,
+      "loss": 1.1174,
+      "step": 304
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.872276487076897e-05,
+      "loss": 1.0296,
+      "step": 305
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8712987148797353e-05,
+      "loss": 1.0762,
+      "step": 306
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8703174716617537e-05,
+      "loss": 1.0594,
+      "step": 307
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8693327613319633e-05,
+      "loss": 1.0863,
+      "step": 308
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8683445878131855e-05,
+      "loss": 1.1403,
+      "step": 309
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8673529550420392e-05,
+      "loss": 1.0803,
+      "step": 310
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8663578669689244e-05,
+      "loss": 1.0918,
+      "step": 311
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.865359327558005e-05,
+      "loss": 1.1026,
+      "step": 312
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.864357340787195e-05,
+      "loss": 1.1754,
+      "step": 313
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8633519106481415e-05,
+      "loss": 1.1284,
+      "step": 314
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862343041146209e-05,
+      "loss": 1.0609,
+      "step": 315
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8613307363004636e-05,
+      "loss": 1.1071,
+      "step": 316
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8603150001436563e-05,
+      "loss": 1.0698,
+      "step": 317
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.859295836722208e-05,
+      "loss": 1.103,
+      "step": 318
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8582732500961938e-05,
+      "loss": 1.056,
+      "step": 319
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8572472443393242e-05,
+      "loss": 0.9948,
+      "step": 320
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8562178235389317e-05,
+      "loss": 1.1106,
+      "step": 321
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8551849917959534e-05,
+      "loss": 1.106,
+      "step": 322
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8541487532249144e-05,
+      "loss": 1.1285,
+      "step": 323
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8531091119539115e-05,
+      "loss": 1.0891,
+      "step": 324
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8520660721245974e-05,
+      "loss": 1.0534,
+      "step": 325
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8510196378921636e-05,
+      "loss": 1.0403,
+      "step": 326
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8499698134253234e-05,
+      "loss": 0.9714,
+      "step": 327
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.848916602906297e-05,
+      "loss": 1.0325,
+      "step": 328
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.847860010530793e-05,
+      "loss": 1.1395,
+      "step": 329
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8468000405079927e-05,
+      "loss": 1.1007,
+      "step": 330
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8457366970605328e-05,
+      "loss": 1.1731,
+      "step": 331
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8446699844244885e-05,
+      "loss": 1.1075,
+      "step": 332
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8435999068493583e-05,
+      "loss": 1.2023,
+      "step": 333
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8425264685980437e-05,
+      "loss": 1.0611,
+      "step": 334
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8414496739468362e-05,
+      "loss": 1.086,
+      "step": 335
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.840369527185397e-05,
+      "loss": 1.0907,
+      "step": 336
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8392860326167413e-05,
+      "loss": 1.1008,
+      "step": 337
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.838199194557222e-05,
+      "loss": 1.0151,
+      "step": 338
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8371090173365103e-05,
+      "loss": 1.1411,
+      "step": 339
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8360155052975813e-05,
+      "loss": 1.065,
+      "step": 340
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8349186627966936e-05,
+      "loss": 1.0416,
+      "step": 341
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8338184942033743e-05,
+      "loss": 1.1179,
+      "step": 342
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8327150039004003e-05,
+      "loss": 1.0832,
+      "step": 343
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8316081962837822e-05,
+      "loss": 1.0412,
+      "step": 344
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.830498075762744e-05,
+      "loss": 1.1223,
+      "step": 345
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.82938464675971e-05,
+      "loss": 1.0097,
+      "step": 346
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8282679137102826e-05,
+      "loss": 1.0984,
+      "step": 347
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.827147881063227e-05,
+      "loss": 1.1289,
+      "step": 348
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.826024553280454e-05,
+      "loss": 0.9915,
+      "step": 349
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8248979348370004e-05,
+      "loss": 1.11,
+      "step": 350
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.823768030221013e-05,
+      "loss": 1.0187,
+      "step": 351
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8226348439337284e-05,
+      "loss": 1.088,
+      "step": 352
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.821498380489458e-05,
+      "loss": 1.156,
+      "step": 353
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.820358644415568e-05,
+      "loss": 1.1527,
+      "step": 354
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8192156402524613e-05,
+      "loss": 1.1342,
+      "step": 355
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.818069372553561e-05,
+      "loss": 1.1564,
+      "step": 356
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8169198458852903e-05,
+      "loss": 1.1756,
+      "step": 357
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.815767064827056e-05,
+      "loss": 1.1368,
+      "step": 358
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8146110339712294e-05,
+      "loss": 1.07,
+      "step": 359
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8134517579231274e-05,
+      "loss": 1.0663,
+      "step": 360
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.812289241300996e-05,
+      "loss": 1.1774,
+      "step": 361
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8111234887359895e-05,
+      "loss": 1.0633,
+      "step": 362
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8099545048721547e-05,
+      "loss": 1.1892,
+      "step": 363
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.80878229436641e-05,
+      "loss": 1.0572,
+      "step": 364
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8076068618885284e-05,
+      "loss": 1.1622,
+      "step": 365
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.806428212121118e-05,
+      "loss": 1.0936,
+      "step": 366
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.805246349759604e-05,
+      "loss": 1.0063,
+      "step": 367
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.80406127951221e-05,
+      "loss": 1.1821,
+      "step": 368
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.802873006099938e-05,
+      "loss": 1.0972,
+      "step": 369
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8016815342565516e-05,
+      "loss": 1.12,
+      "step": 370
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8004868687285556e-05,
+      "loss": 1.0554,
+      "step": 371
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7992890142751775e-05,
+      "loss": 1.1383,
+      "step": 372
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.798087975668349e-05,
+      "loss": 1.112,
+      "step": 373
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7968837576926857e-05,
+      "loss": 1.106,
+      "step": 374
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7956763651454703e-05,
+      "loss": 1.1322,
+      "step": 375
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7944658028366313e-05,
+      "loss": 1.1646,
+      "step": 376
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7932520755887248e-05,
+      "loss": 1.1505,
+      "step": 377
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7920351882369154e-05,
+      "loss": 1.1017,
+      "step": 378
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7908151456289562e-05,
+      "loss": 1.1425,
+      "step": 379
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7895919526251705e-05,
+      "loss": 1.0432,
+      "step": 380
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.788365614098432e-05,
+      "loss": 1.0733,
+      "step": 381
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7871361349341446e-05,
+      "loss": 1.0968,
+      "step": 382
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7859035200302244e-05,
+      "loss": 1.0904,
+      "step": 383
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7846677742970793e-05,
+      "loss": 1.0582,
+      "step": 384
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7834289026575892e-05,
+      "loss": 1.0443,
+      "step": 385
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7821869100470872e-05,
+      "loss": 1.0643,
+      "step": 386
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7809418014133394e-05,
+      "loss": 1.1623,
+      "step": 387
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7796935817165252e-05,
+      "loss": 1.0644,
+      "step": 388
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7784422559292174e-05,
+      "loss": 1.0367,
+      "step": 389
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7771878290363638e-05,
+      "loss": 1.1753,
+      "step": 390
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7759303060352645e-05,
+      "loss": 1.0757,
+      "step": 391
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7746696919355547e-05,
+      "loss": 1.0631,
+      "step": 392
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7734059917591837e-05,
+      "loss": 1.1011,
+      "step": 393
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7721392105403944e-05,
+      "loss": 1.0844,
+      "step": 394
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.770869353325704e-05,
+      "loss": 1.0877,
+      "step": 395
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.769596425173884e-05,
+      "loss": 1.137,
+      "step": 396
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7683204311559394e-05,
+      "loss": 0.9882,
+      "step": 397
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.767041376355088e-05,
+      "loss": 1.007,
+      "step": 398
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7657592658667426e-05,
+      "loss": 1.0548,
+      "step": 399
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.764474104798488e-05,
+      "loss": 1.068,
+      "step": 400
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7631858982700613e-05,
+      "loss": 1.1061,
+      "step": 401
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.761894651413333e-05,
+      "loss": 1.0331,
+      "step": 402
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7606003693722845e-05,
+      "loss": 1.0606,
+      "step": 403
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7593030573029896e-05,
+      "loss": 1.072,
+      "step": 404
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7580027203735915e-05,
+      "loss": 1.0993,
+      "step": 405
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.756699363764285e-05,
+      "loss": 1.1815,
+      "step": 406
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.755392992667293e-05,
+      "loss": 1.0083,
+      "step": 407
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7540836122868494e-05,
+      "loss": 1.0869,
+      "step": 408
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.752771227839174e-05,
+      "loss": 1.0856,
+      "step": 409
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.751455844552456e-05,
+      "loss": 1.0129,
+      "step": 410
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.75013746766683e-05,
+      "loss": 1.2001,
+      "step": 411
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7488161024343564e-05,
+      "loss": 1.0917,
+      "step": 412
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7474917541190005e-05,
+      "loss": 1.1171,
+      "step": 413
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7461644279966113e-05,
+      "loss": 1.0736,
+      "step": 414
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.744834129354901e-05,
+      "loss": 1.1439,
+      "step": 415
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7435008634934234e-05,
+      "loss": 1.0607,
+      "step": 416
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7421646357235524e-05,
+      "loss": 1.1784,
+      "step": 417
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.740825451368462e-05,
+      "loss": 1.1382,
+      "step": 418
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7394833157631038e-05,
+      "loss": 1.1742,
+      "step": 419
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.738138234254187e-05,
+      "loss": 1.1069,
+      "step": 420
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7367902122001563e-05,
+      "loss": 1.0965,
+      "step": 421
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7354392549711704e-05,
+      "loss": 1.0791,
+      "step": 422
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7340853679490813e-05,
+      "loss": 1.1507,
+      "step": 423
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.732728556527413e-05,
+      "loss": 1.0708,
+      "step": 424
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7313688261113383e-05,
+      "loss": 1.0396,
+      "step": 425
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7300061821176593e-05,
+      "loss": 1.1097,
+      "step": 426
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7286406299747847e-05,
+      "loss": 1.1252,
+      "step": 427
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.727272175122709e-05,
+      "loss": 1.0691,
+      "step": 428
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7259008230129897e-05,
+      "loss": 1.0967,
+      "step": 429
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7245265791087268e-05,
+      "loss": 1.1096,
+      "step": 430
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.72314944888454e-05,
+      "loss": 1.1743,
+      "step": 431
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7217694378265473e-05,
+      "loss": 1.1326,
+      "step": 432
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.720386551432343e-05,
+      "loss": 1.0186,
+      "step": 433
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7190007952109772e-05,
+      "loss": 1.0076,
+      "step": 434
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.717612174682931e-05,
+      "loss": 1.1326,
+      "step": 435
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7162206953800965e-05,
+      "loss": 1.0971,
+      "step": 436
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7148263628457557e-05,
+      "loss": 1.1199,
+      "step": 437
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7134291826345555e-05,
+      "loss": 1.0049,
+      "step": 438
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.712029160312488e-05,
+      "loss": 1.1472,
+      "step": 439
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.710626301456867e-05,
+      "loss": 1.1984,
+      "step": 440
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.709220611656307e-05,
+      "loss": 0.9672,
+      "step": 441
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7078120965106995e-05,
+      "loss": 1.119,
+      "step": 442
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7064007616311923e-05,
+      "loss": 1.1132,
+      "step": 443
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7049866126401654e-05,
+      "loss": 0.9424,
+      "step": 444
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7035696551712094e-05,
+      "loss": 1.1978,
+      "step": 445
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.702149894869104e-05,
+      "loss": 1.0183,
+      "step": 446
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7007273373897943e-05,
+      "loss": 1.1084,
+      "step": 447
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6993019884003682e-05,
+      "loss": 1.077,
+      "step": 448
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6978738535790343e-05,
+      "loss": 1.091,
+      "step": 449
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6964429386150993e-05,
+      "loss": 1.129,
+      "step": 450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.695009249208946e-05,
+      "loss": 1.0977,
+      "step": 451
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6935727910720086e-05,
+      "loss": 1.0456,
+      "step": 452
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6921335699267524e-05,
+      "loss": 1.0746,
+      "step": 453
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.690691591506648e-05,
+      "loss": 1.1152,
+      "step": 454
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6892468615561527e-05,
+      "loss": 1.064,
+      "step": 455
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6877993858306828e-05,
+      "loss": 1.1105,
+      "step": 456
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6863491700965945e-05,
+      "loss": 1.1054,
+      "step": 457
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6848962201311593e-05,
+      "loss": 1.0579,
+      "step": 458
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6834405417225404e-05,
+      "loss": 1.0975,
+      "step": 459
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.681982140669771e-05,
+      "loss": 0.9444,
+      "step": 460
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.680521022782731e-05,
+      "loss": 1.0604,
+      "step": 461
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.679057193882122e-05,
+      "loss": 1.1474,
+      "step": 462
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6775906597994472e-05,
+      "loss": 1.1401,
+      "step": 463
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.676121426376985e-05,
+      "loss": 1.1319,
+      "step": 464
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.674649499467769e-05,
+      "loss": 1.0954,
+      "step": 465
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6731748849355612e-05,
+      "loss": 1.104,
+      "step": 466
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6716975886548316e-05,
+      "loss": 1.1196,
+      "step": 467
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6702176165107323e-05,
+      "loss": 1.0606,
+      "step": 468
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6687349743990772e-05,
+      "loss": 1.1016,
+      "step": 469
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6672496682263147e-05,
+      "loss": 1.1304,
+      "step": 470
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665761703909507e-05,
+      "loss": 1.0418,
+      "step": 471
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.664271087376306e-05,
+      "loss": 1.1369,
+      "step": 472
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6627778245649292e-05,
+      "loss": 1.0737,
+      "step": 473
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.661281921424135e-05,
+      "loss": 1.1249,
+      "step": 474
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6597833839132016e-05,
+      "loss": 1.147,
+      "step": 475
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6582822180019014e-05,
+      "loss": 1.1221,
+      "step": 476
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6567784296704766e-05,
+      "loss": 1.0297,
+      "step": 477
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.655272024909618e-05,
+      "loss": 1.0731,
+      "step": 478
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6537630097204394e-05,
+      "loss": 1.056,
+      "step": 479
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.652251390114452e-05,
+      "loss": 1.1721,
+      "step": 480
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6507371721135446e-05,
+      "loss": 1.0644,
+      "step": 481
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6492203617499564e-05,
+      "loss": 1.0807,
+      "step": 482
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6477009650662534e-05,
+      "loss": 1.035,
+      "step": 483
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6461789881153053e-05,
+      "loss": 0.999,
+      "step": 484
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.644654436960261e-05,
+      "loss": 1.1095,
+      "step": 485
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6431273176745235e-05,
+      "loss": 1.1145,
+      "step": 486
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.641597636341728e-05,
+      "loss": 1.0846,
+      "step": 487
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6400653990557153e-05,
+      "loss": 1.1037,
+      "step": 488
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6385306119205085e-05,
+      "loss": 1.0504,
+      "step": 489
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.636993281050288e-05,
+      "loss": 1.1619,
+      "step": 490
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6354534125693693e-05,
+      "loss": 1.1019,
+      "step": 491
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6339110126121757e-05,
+      "loss": 1.1267,
+      "step": 492
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6323660873232162e-05,
+      "loss": 1.1174,
+      "step": 493
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6308186428570586e-05,
+      "loss": 1.0981,
+      "step": 494
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.629268685378308e-05,
+      "loss": 1.0985,
+      "step": 495
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6277162210615803e-05,
+      "loss": 1.069,
+      "step": 496
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6261612560914775e-05,
+      "loss": 1.076,
+      "step": 497
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6246037966625635e-05,
+      "loss": 1.0856,
+      "step": 498
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.62304384897934e-05,
+      "loss": 1.0915,
+      "step": 499
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6214814192562205e-05,
+      "loss": 1.0854,
+      "step": 500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.619916513717507e-05,
+      "loss": 0.9778,
+      "step": 501
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.618349138597365e-05,
+      "loss": 1.115,
+      "step": 502
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.616779300139796e-05,
+      "loss": 1.0169,
+      "step": 503
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.615207004598617e-05,
+      "loss": 1.0258,
+      "step": 504
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6136322582374325e-05,
+      "loss": 1.0615,
+      "step": 505
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6120550673296098e-05,
+      "loss": 1.122,
+      "step": 506
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6104754381582555e-05,
+      "loss": 1.1374,
+      "step": 507
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6088933770161903e-05,
+      "loss": 1.0954,
+      "step": 508
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6073088902059207e-05,
+      "loss": 1.1041,
+      "step": 509
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.605721984039619e-05,
+      "loss": 1.0609,
+      "step": 510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6041326648390944e-05,
+      "loss": 1.0797,
+      "step": 511
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6025409389357687e-05,
+      "loss": 1.0636,
+      "step": 512
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.600946812670652e-05,
+      "loss": 1.1396,
+      "step": 513
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5993502923943164e-05,
+      "loss": 1.0405,
+      "step": 514
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5977513844668715e-05,
+      "loss": 1.1018,
+      "step": 515
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5961500952579388e-05,
+      "loss": 1.123,
+      "step": 516
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.594546431146625e-05,
+      "loss": 1.1249,
+      "step": 517
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5929403985214987e-05,
+      "loss": 1.1015,
+      "step": 518
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5913320037805646e-05,
+      "loss": 1.1355,
+      "step": 519
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.589721253331236e-05,
+      "loss": 1.0487,
+      "step": 520
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.588108153590312e-05,
+      "loss": 1.0149,
+      "step": 521
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.58649271098395e-05,
+      "loss": 1.0653,
+      "step": 522
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5848749319476406e-05,
+      "loss": 1.0972,
+      "step": 523
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.583254822926183e-05,
+      "loss": 1.1424,
+      "step": 524
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.581632390373657e-05,
+      "loss": 1.0865,
+      "step": 525
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5800076407534e-05,
+      "loss": 1.197,
+      "step": 526
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5783805805379794e-05,
+      "loss": 1.1024,
+      "step": 527
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5767512162091674e-05,
+      "loss": 1.0576,
+      "step": 528
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5751195542579157e-05,
+      "loss": 1.1478,
+      "step": 529
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.573485601184328e-05,
+      "loss": 1.1102,
+      "step": 530
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5718493634976368e-05,
+      "loss": 1.0605,
+      "step": 531
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5702108477161745e-05,
+      "loss": 1.1454,
+      "step": 532
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.568570060367349e-05,
+      "loss": 1.0505,
+      "step": 533
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5669270079876183e-05,
+      "loss": 1.1561,
+      "step": 534
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.565281697122463e-05,
+      "loss": 1.1271,
+      "step": 535
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5636341343263616e-05,
+      "loss": 1.0493,
+      "step": 536
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5619843261627624e-05,
+      "loss": 1.0812,
+      "step": 537
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5603322792040597e-05,
+      "loss": 1.1063,
+      "step": 538
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.558678000031567e-05,
+      "loss": 1.1559,
+      "step": 539
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5570214952354884e-05,
+      "loss": 1.0684,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5553627714148965e-05,
+      "loss": 1.2365,
+      "step": 541
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.553701835177703e-05,
+      "loss": 1.4796,
+      "step": 542
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5520386931406326e-05,
+      "loss": 0.7428,
+      "step": 543
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5503733519291982e-05,
+      "loss": 0.7332,
+      "step": 544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5487058181776734e-05,
+      "loss": 0.7831,
+      "step": 545
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5470360985290663e-05,
+      "loss": 0.7408,
+      "step": 546
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5453641996350923e-05,
+      "loss": 0.7092,
+      "step": 547
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.54369012815615e-05,
+      "loss": 0.7152,
+      "step": 548
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5420138907612915e-05,
+      "loss": 0.69,
+      "step": 549
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5403354941281978e-05,
+      "loss": 0.6491,
+      "step": 550
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.538654944943152e-05,
+      "loss": 0.6676,
+      "step": 551
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5369722499010114e-05,
+      "loss": 0.6893,
+      "step": 552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.535287415705184e-05,
+      "loss": 0.6783,
+      "step": 553
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.533600449067597e-05,
+      "loss": 0.6745,
+      "step": 554
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5319113567086738e-05,
+      "loss": 0.6781,
+      "step": 555
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5302201453573072e-05,
+      "loss": 0.6328,
+      "step": 556
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5285268217508293e-05,
+      "loss": 0.7567,
+      "step": 557
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.526831392634988e-05,
+      "loss": 0.7942,
+      "step": 558
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5251338647639204e-05,
+      "loss": 0.7095,
+      "step": 559
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.523434244900122e-05,
+      "loss": 0.6564,
+      "step": 560
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5217325398144234e-05,
+      "loss": 0.6648,
+      "step": 561
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5200287562859623e-05,
+      "loss": 0.6278,
+      "step": 562
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.518322901102156e-05,
+      "loss": 0.6989,
+      "step": 563
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.516614981058675e-05,
+      "loss": 0.7533,
+      "step": 564
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5149050029594152e-05,
+      "loss": 0.6613,
+      "step": 565
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.513192973616472e-05,
+      "loss": 0.6644,
+      "step": 566
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5114788998501117e-05,
+      "loss": 0.7481,
+      "step": 567
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5097627884887454e-05,
+      "loss": 0.6838,
+      "step": 568
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5080446463689013e-05,
+      "loss": 0.6663,
+      "step": 569
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5063244803351975e-05,
+      "loss": 0.7884,
+      "step": 570
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5046022972403154e-05,
+      "loss": 0.7401,
+      "step": 571
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.502878103944971e-05,
+      "loss": 0.7643,
+      "step": 572
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5011519073178883e-05,
+      "loss": 0.7151,
+      "step": 573
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.499423714235774e-05,
+      "loss": 0.747,
+      "step": 574
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4976935315832853e-05,
+      "loss": 0.6564,
+      "step": 575
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4959613662530073e-05,
+      "loss": 0.7131,
+      "step": 576
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4942272251454229e-05,
+      "loss": 0.7624,
+      "step": 577
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4924911151688858e-05,
+      "loss": 0.7194,
+      "step": 578
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4907530432395937e-05,
+      "loss": 0.7649,
+      "step": 579
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4890130162815593e-05,
+      "loss": 0.7174,
+      "step": 580
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4872710412265848e-05,
+      "loss": 0.6927,
+      "step": 581
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.485527125014232e-05,
+      "loss": 0.6947,
+      "step": 582
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4837812745917966e-05,
+      "loss": 0.6964,
+      "step": 583
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4820334969142788e-05,
+      "loss": 0.6159,
+      "step": 584
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4802837989443576e-05,
+      "loss": 0.7238,
+      "step": 585
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.478532187652361e-05,
+      "loss": 0.7155,
+      "step": 586
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4767786700162392e-05,
+      "loss": 0.6852,
+      "step": 587
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4750232530215373e-05,
+      "loss": 0.6454,
+      "step": 588
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4732659436613666e-05,
+      "loss": 0.7559,
+      "step": 589
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4715067489363772e-05,
+      "loss": 0.7302,
+      "step": 590
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4697456758547297e-05,
+      "loss": 0.7132,
+      "step": 591
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4679827314320681e-05,
+      "loss": 0.7515,
+      "step": 592
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4662179226914903e-05,
+      "loss": 0.713,
+      "step": 593
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.464451256663523e-05,
+      "loss": 0.7046,
+      "step": 594
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4626827403860899e-05,
+      "loss": 0.7507,
+      "step": 595
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4609123809044863e-05,
+      "loss": 0.6705,
+      "step": 596
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4591401852713511e-05,
+      "loss": 0.6182,
+      "step": 597
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4573661605466365e-05,
+      "loss": 0.7243,
+      "step": 598
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4555903137975827e-05,
+      "loss": 0.7366,
+      "step": 599
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4538126520986872e-05,
+      "loss": 0.6806,
+      "step": 600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4520331825316786e-05,
+      "loss": 0.6926,
+      "step": 601
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4502519121854874e-05,
+      "loss": 0.773,
+      "step": 602
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4484688481562175e-05,
+      "loss": 0.6615,
+      "step": 603
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4466839975471186e-05,
+      "loss": 0.7147,
+      "step": 604
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4448973674685581e-05,
+      "loss": 0.7338,
+      "step": 605
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4431089650379913e-05,
+      "loss": 0.7248,
+      "step": 606
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4413187973799352e-05,
+      "loss": 0.6493,
+      "step": 607
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4395268716259383e-05,
+      "loss": 0.753,
+      "step": 608
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4377331949145532e-05,
+      "loss": 0.7592,
+      "step": 609
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4359377743913075e-05,
+      "loss": 0.6832,
+      "step": 610
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4341406172086765e-05,
+      "loss": 0.7126,
+      "step": 611
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.432341730526052e-05,
+      "loss": 0.6899,
+      "step": 612
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4305411215097185e-05,
+      "loss": 0.7361,
+      "step": 613
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4287387973328193e-05,
+      "loss": 0.6775,
+      "step": 614
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4269347651753316e-05,
+      "loss": 0.7284,
+      "step": 615
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4251290322240366e-05,
+      "loss": 0.6924,
+      "step": 616
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4233216056724912e-05,
+      "loss": 0.6872,
+      "step": 617
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4215124927209987e-05,
+      "loss": 0.7455,
+      "step": 618
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4197017005765804e-05,
+      "loss": 0.7145,
+      "step": 619
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4178892364529477e-05,
+      "loss": 0.7634,
+      "step": 620
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4160751075704728e-05,
+      "loss": 0.6267,
+      "step": 621
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4142593211561591e-05,
+      "loss": 0.6794,
+      "step": 622
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4124418844436138e-05,
+      "loss": 0.684,
+      "step": 623
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4106228046730181e-05,
+      "loss": 0.757,
+      "step": 624
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4088020890910988e-05,
+      "loss": 0.7026,
+      "step": 625
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4069797449510992e-05,
+      "loss": 0.6711,
+      "step": 626
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4051557795127507e-05,
+      "loss": 0.7137,
+      "step": 627
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4033302000422431e-05,
+      "loss": 0.7015,
+      "step": 628
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4015030138121964e-05,
+      "loss": 0.7242,
+      "step": 629
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3996742281016312e-05,
+      "loss": 0.7331,
+      "step": 630
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3978438501959406e-05,
+      "loss": 0.7056,
+      "step": 631
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3960118873868597e-05,
+      "loss": 0.7064,
+      "step": 632
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3941783469724382e-05,
+      "loss": 0.7558,
+      "step": 633
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3923432362570102e-05,
+      "loss": 0.7101,
+      "step": 634
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3905065625511654e-05,
+      "loss": 0.6925,
+      "step": 635
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3886683331717198e-05,
+      "loss": 0.6413,
+      "step": 636
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.386828555441688e-05,
+      "loss": 0.6689,
+      "step": 637
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3849872366902512e-05,
+      "loss": 0.7329,
+      "step": 638
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3831443842527305e-05,
+      "loss": 0.7484,
+      "step": 639
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.381300005470557e-05,
+      "loss": 0.6971,
+      "step": 640
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3794541076912415e-05,
+      "loss": 0.6569,
+      "step": 641
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3776066982683467e-05,
+      "loss": 0.7161,
+      "step": 642
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3757577845614576e-05,
+      "loss": 0.6705,
+      "step": 643
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3739073739361509e-05,
+      "loss": 0.7057,
+      "step": 644
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3720554737639673e-05,
+      "loss": 0.7566,
+      "step": 645
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3702020914223815e-05,
+      "loss": 0.69,
+      "step": 646
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3683472342947722e-05,
+      "loss": 0.7135,
+      "step": 647
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.366490909770394e-05,
+      "loss": 0.6341,
+      "step": 648
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3646331252443465e-05,
+      "loss": 0.7568,
+      "step": 649
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.362773888117546e-05,
+      "loss": 0.64,
+      "step": 650
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3609132057966963e-05,
+      "loss": 0.7737,
+      "step": 651
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.359051085694256e-05,
+      "loss": 0.6731,
+      "step": 652
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3571875352284143e-05,
+      "loss": 0.6799,
+      "step": 653
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3553225618230567e-05,
+      "loss": 0.7575,
+      "step": 654
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3534561729077384e-05,
+      "loss": 0.7428,
+      "step": 655
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.351588375917653e-05,
+      "loss": 0.7076,
+      "step": 656
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3497191782936042e-05,
+      "loss": 0.7425,
+      "step": 657
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3478485874819738e-05,
+      "loss": 0.7202,
+      "step": 658
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3459766109346957e-05,
+      "loss": 0.7411,
+      "step": 659
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.344103256109223e-05,
+      "loss": 0.6791,
+      "step": 660
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3422285304684996e-05,
+      "loss": 0.6761,
+      "step": 661
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.340352441480931e-05,
+      "loss": 0.6458,
+      "step": 662
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3384749966203534e-05,
+      "loss": 0.6818,
+      "step": 663
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.336596203366004e-05,
+      "loss": 0.6874,
+      "step": 664
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3347160692024927e-05,
+      "loss": 0.688,
+      "step": 665
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3328346016197706e-05,
+      "loss": 0.6961,
+      "step": 666
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3309518081131006e-05,
+      "loss": 0.6897,
+      "step": 667
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.329067696183028e-05,
+      "loss": 0.6978,
+      "step": 668
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3271822733353503e-05,
+      "loss": 0.6297,
+      "step": 669
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3252955470810877e-05,
+      "loss": 0.7844,
+      "step": 670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3234075249364524e-05,
+      "loss": 0.7129,
+      "step": 671
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3215182144228189e-05,
+      "loss": 0.7118,
+      "step": 672
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3196276230666945e-05,
+      "loss": 0.7164,
+      "step": 673
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3177357583996897e-05,
+      "loss": 0.6831,
+      "step": 674
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3158426279584858e-05,
+      "loss": 0.7077,
+      "step": 675
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3139482392848081e-05,
+      "loss": 0.6771,
+      "step": 676
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3120525999253944e-05,
+      "loss": 0.76,
+      "step": 677
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3101557174319635e-05,
+      "loss": 0.6509,
+      "step": 678
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3082575993611881e-05,
+      "loss": 0.6796,
+      "step": 679
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3063582532746626e-05,
+      "loss": 0.6152,
+      "step": 680
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3044576867388728e-05,
+      "loss": 0.6556,
+      "step": 681
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3025559073251667e-05,
+      "loss": 0.8051,
+      "step": 682
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3006529226097254e-05,
+      "loss": 0.6732,
+      "step": 683
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.2987487401735298e-05,
+      "loss": 0.6455,
+      "step": 684
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2968433676023329e-05,
+      "loss": 0.7114,
+      "step": 685
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2949368124866295e-05,
+      "loss": 0.7019,
+      "step": 686
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2930290824216241e-05,
+      "loss": 0.7489,
+      "step": 687
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2911201850072029e-05,
+      "loss": 0.7816,
+      "step": 688
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2892101278479021e-05,
+      "loss": 0.7069,
+      "step": 689
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2872989185528781e-05,
+      "loss": 0.6617,
+      "step": 690
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2853865647358772e-05,
+      "loss": 0.7156,
+      "step": 691
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2834730740152044e-05,
+      "loss": 0.6673,
+      "step": 692
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2815584540136952e-05,
+      "loss": 0.7417,
+      "step": 693
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.279642712358683e-05,
+      "loss": 0.6312,
+      "step": 694
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2777258566819695e-05,
+      "loss": 0.6925,
+      "step": 695
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2758078946197943e-05,
+      "loss": 0.6707,
+      "step": 696
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2738888338128053e-05,
+      "loss": 0.7188,
+      "step": 697
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2719686819060266e-05,
+      "loss": 0.6821,
+      "step": 698
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2700474465488295e-05,
+      "loss": 0.7034,
+      "step": 699
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2681251353949015e-05,
+      "loss": 0.6956,
+      "step": 700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2662017561022152e-05,
+      "loss": 0.7367,
+      "step": 701
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2642773163329992e-05,
+      "loss": 0.6863,
+      "step": 702
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.262351823753706e-05,
+      "loss": 0.7096,
+      "step": 703
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2604252860349826e-05,
+      "loss": 0.6663,
+      "step": 704
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2584977108516396e-05,
+      "loss": 0.6889,
+      "step": 705
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2565691058826209e-05,
+      "loss": 0.7079,
+      "step": 706
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2546394788109714e-05,
+      "loss": 0.7301,
+      "step": 707
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2527088373238097e-05,
+      "loss": 0.7303,
+      "step": 708
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2507771891122941e-05,
+      "loss": 0.7453,
+      "step": 709
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.248844541871594e-05,
+      "loss": 0.6898,
+      "step": 710
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2469109033008588e-05,
+      "loss": 0.7573,
+      "step": 711
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2449762811031868e-05,
+      "loss": 0.6876,
+      "step": 712
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2430406829855948e-05,
+      "loss": 0.6897,
+      "step": 713
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2411041166589879e-05,
+      "loss": 0.7643,
+      "step": 714
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2391665898381273e-05,
+      "loss": 0.6308,
+      "step": 715
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2372281102416016e-05,
+      "loss": 0.7062,
+      "step": 716
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2352886855917943e-05,
+      "loss": 0.6843,
+      "step": 717
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.233348323614854e-05,
+      "loss": 0.6932,
+      "step": 718
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2314070320406637e-05,
+      "loss": 0.6494,
+      "step": 719
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.229464818602809e-05,
+      "loss": 0.763,
+      "step": 720
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2275216910385483e-05,
+      "loss": 0.6057,
+      "step": 721
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2255776570887816e-05,
+      "loss": 0.6944,
+      "step": 722
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.22363272449802e-05,
+      "loss": 0.6254,
+      "step": 723
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2216869010143535e-05,
+      "loss": 0.6773,
+      "step": 724
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.219740194389423e-05,
+      "loss": 0.6274,
+      "step": 725
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2177926123783857e-05,
+      "loss": 0.6523,
+      "step": 726
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2158441627398868e-05,
+      "loss": 0.6456,
+      "step": 727
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.213894853236029e-05,
+      "loss": 0.736,
+      "step": 728
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.211944691632339e-05,
+      "loss": 0.6846,
+      "step": 729
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2099936856977381e-05,
+      "loss": 0.6841,
+      "step": 730
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.208041843204512e-05,
+      "loss": 0.7427,
+      "step": 731
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.206089171928279e-05,
+      "loss": 0.6868,
+      "step": 732
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2041356796479582e-05,
+      "loss": 0.6645,
+      "step": 733
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2021813741457398e-05,
+      "loss": 0.6106,
+      "step": 734
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2002262632070542e-05,
+      "loss": 0.6965,
+      "step": 735
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1982703546205396e-05,
+      "loss": 0.7218,
+      "step": 736
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1963136561780127e-05,
+      "loss": 0.6761,
+      "step": 737
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1943561756744353e-05,
+      "loss": 0.7337,
+      "step": 738
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1923979209078864e-05,
+      "loss": 0.6419,
+      "step": 739
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1904388996795287e-05,
+      "loss": 0.6932,
+      "step": 740
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.188479119793578e-05,
+      "loss": 0.5949,
+      "step": 741
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1865185890572728e-05,
+      "loss": 0.7152,
+      "step": 742
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.184557315280843e-05,
+      "loss": 0.6741,
+      "step": 743
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.182595306277478e-05,
+      "loss": 0.6872,
+      "step": 744
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.180632569863296e-05,
+      "loss": 0.6738,
+      "step": 745
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1786691138573142e-05,
+      "loss": 0.7456,
+      "step": 746
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1767049460814155e-05,
+      "loss": 0.6536,
+      "step": 747
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.174740074360318e-05,
+      "loss": 0.7134,
+      "step": 748
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1727745065215453e-05,
+      "loss": 0.7004,
+      "step": 749
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1708082503953925e-05,
+      "loss": 0.7227,
+      "step": 750
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1688413138148989e-05,
+      "loss": 0.7089,
+      "step": 751
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.166873704615812e-05,
+      "loss": 0.6793,
+      "step": 752
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1649054306365607e-05,
+      "loss": 0.6607,
+      "step": 753
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1629364997182218e-05,
+      "loss": 0.642,
+      "step": 754
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1609669197044887e-05,
+      "loss": 0.6734,
+      "step": 755
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1589966984416405e-05,
+      "loss": 0.7343,
+      "step": 756
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1570258437785118e-05,
+      "loss": 0.6887,
+      "step": 757
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.15505436356646e-05,
+      "loss": 0.6809,
+      "step": 758
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1530822656593339e-05,
+      "loss": 0.6954,
+      "step": 759
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1511095579134441e-05,
+      "loss": 0.6589,
+      "step": 760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1491362481875301e-05,
+      "loss": 0.704,
+      "step": 761
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1471623443427293e-05,
+      "loss": 0.6989,
+      "step": 762
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1451878542425465e-05,
+      "loss": 0.6302,
+      "step": 763
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1432127857528213e-05,
+      "loss": 0.7392,
+      "step": 764
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.141237146741698e-05,
+      "loss": 0.7066,
+      "step": 765
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1392609450795937e-05,
+      "loss": 0.7327,
+      "step": 766
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.137284188639166e-05,
+      "loss": 0.7534,
+      "step": 767
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1353068852952844e-05,
+      "loss": 0.6737,
+      "step": 768
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1333290429249956e-05,
+      "loss": 0.7001,
+      "step": 769
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.131350669407494e-05,
+      "loss": 0.702,
+      "step": 770
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1293717726240899e-05,
+      "loss": 0.6883,
+      "step": 771
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1273923604581789e-05,
+      "loss": 0.7174,
+      "step": 772
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1254124407952086e-05,
+      "loss": 0.712,
+      "step": 773
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.123432021522649e-05,
+      "loss": 0.7152,
+      "step": 774
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1214511105299603e-05,
+      "loss": 0.7056,
+      "step": 775
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1194697157085616e-05,
+      "loss": 0.7409,
+      "step": 776
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1174878449517997e-05,
+      "loss": 0.6407,
+      "step": 777
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1155055061549168e-05,
+      "loss": 0.6609,
+      "step": 778
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1135227072150199e-05,
+      "loss": 0.6593,
+      "step": 779
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1115394560310493e-05,
+      "loss": 0.7552,
+      "step": 780
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1095557605037466e-05,
+      "loss": 0.7579,
+      "step": 781
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.107571628535624e-05,
+      "loss": 0.6239,
+      "step": 782
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.105587068030932e-05,
+      "loss": 0.7875,
+      "step": 783
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1036020868956283e-05,
+      "loss": 0.6953,
+      "step": 784
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1016166930373463e-05,
+      "loss": 0.7434,
+      "step": 785
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0996308943653637e-05,
+      "loss": 0.6912,
+      "step": 786
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0976446987905705e-05,
+      "loss": 0.7109,
+      "step": 787
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0956581142254383e-05,
+      "loss": 0.6776,
+      "step": 788
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0936711485839884e-05,
+      "loss": 0.65,
+      "step": 789
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0916838097817597e-05,
+      "loss": 0.729,
+      "step": 790
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.089696105735778e-05,
+      "loss": 0.6041,
+      "step": 791
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0877080443645242e-05,
+      "loss": 0.6902,
+      "step": 792
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0857196335879025e-05,
+      "loss": 0.7099,
+      "step": 793
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.083730881327209e-05,
+      "loss": 0.7203,
+      "step": 794
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0817417955051004e-05,
+      "loss": 0.6735,
+      "step": 795
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0797523840455622e-05,
+      "loss": 0.6997,
+      "step": 796
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0777626548738766e-05,
+      "loss": 0.6938,
+      "step": 797
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0757726159165929e-05,
+      "loss": 0.7082,
+      "step": 798
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0737822751014921e-05,
+      "loss": 0.6935,
+      "step": 799
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0717916403575605e-05,
+      "loss": 0.5979,
+      "step": 800
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0698007196149533e-05,
+      "loss": 0.6648,
+      "step": 801
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0678095208049659e-05,
+      "loss": 0.7072,
+      "step": 802
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0658180518600016e-05,
+      "loss": 0.6375,
+      "step": 803
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0638263207135388e-05,
+      "loss": 0.6449,
+      "step": 804
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0618343353001019e-05,
+      "loss": 0.7007,
+      "step": 805
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.059842103555227e-05,
+      "loss": 0.7223,
+      "step": 806
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0578496334154324e-05,
+      "loss": 0.7434,
+      "step": 807
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0558569328181851e-05,
+      "loss": 0.742,
+      "step": 808
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0538640097018717e-05,
+      "loss": 0.6165,
+      "step": 809
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0518708720057638e-05,
+      "loss": 0.7405,
+      "step": 810
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0498775276699882e-05,
+      "loss": 0.7002,
+      "step": 811
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0478839846354957e-05,
+      "loss": 0.7065,
+      "step": 812
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0458902508440276e-05,
+      "loss": 0.6862,
+      "step": 813
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0438963342380856e-05,
+      "loss": 0.7285,
+      "step": 814
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0419022427608998e-05,
+      "loss": 0.7149,
+      "step": 815
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0399079843563965e-05,
+      "loss": 0.782,
+      "step": 816
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.037913566969168e-05,
+      "loss": 0.7855,
+      "step": 817
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0359189985444384e-05,
+      "loss": 0.711,
+      "step": 818
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0339242870280348e-05,
+      "loss": 0.7289,
+      "step": 819
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0319294403663536e-05,
+      "loss": 0.7479,
+      "step": 820
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0299344665063304e-05,
+      "loss": 0.7089,
+      "step": 821
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0279393733954062e-05,
+      "loss": 0.7515,
+      "step": 822
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0259441689814983e-05,
+      "loss": 0.7231,
+      "step": 823
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0239488612129669e-05,
+      "loss": 0.6972,
+      "step": 824
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0219534580385838e-05,
+      "loss": 0.5787,
+      "step": 825
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0199579674075014e-05,
+      "loss": 0.6117,
+      "step": 826
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0179623972692194e-05,
+      "loss": 0.7428,
+      "step": 827
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0159667555735559e-05,
+      "loss": 0.6917,
+      "step": 828
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0139710502706129e-05,
+      "loss": 0.6664,
+      "step": 829
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0119752893107453e-05,
+      "loss": 0.6498,
+      "step": 830
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0099794806445315e-05,
+      "loss": 0.6349,
+      "step": 831
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0079836322227386e-05,
+      "loss": 0.6284,
+      "step": 832
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.005987751996292e-05,
+      "loss": 0.7171,
+      "step": 833
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0039918479162443e-05,
+      "loss": 0.6906,
+      "step": 834
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0019959279337434e-05,
+      "loss": 0.6947,
+      "step": 835
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1e-05,
+      "loss": 0.7237,
+      "step": 836
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.980040720662567e-06,
+      "loss": 0.6798,
+      "step": 837
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.960081520837557e-06,
+      "loss": 0.6723,
+      "step": 838
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.940122480037085e-06,
+      "loss": 0.6957,
+      "step": 839
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.92016367777262e-06,
+      "loss": 0.6518,
+      "step": 840
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.900205193554686e-06,
+      "loss": 0.7994,
+      "step": 841
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.880247106892549e-06,
+      "loss": 0.7032,
+      "step": 842
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.860289497293874e-06,
+      "loss": 0.7631,
+      "step": 843
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.840332444264443e-06,
+      "loss": 0.6326,
+      "step": 844
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.82037602730781e-06,
+      "loss": 0.6988,
+      "step": 845
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.800420325924991e-06,
+      "loss": 0.6312,
+      "step": 846
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.780465419614166e-06,
+      "loss": 0.6844,
+      "step": 847
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.760511387870335e-06,
+      "loss": 0.6914,
+      "step": 848
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.74055831018502e-06,
+      "loss": 0.6757,
+      "step": 849
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.720606266045938e-06,
+      "loss": 0.6702,
+      "step": 850
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.7006553349367e-06,
+      "loss": 0.6242,
+      "step": 851
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.680705596336466e-06,
+      "loss": 0.7484,
+      "step": 852
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.660757129719653e-06,
+      "loss": 0.6729,
+      "step": 853
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.640810014555619e-06,
+      "loss": 0.7173,
+      "step": 854
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.620864330308322e-06,
+      "loss": 0.6985,
+      "step": 855
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.600920156436034e-06,
+      "loss": 0.636,
+      "step": 856
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.580977572391005e-06,
+      "loss": 0.7181,
+      "step": 857
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.561036657619147e-06,
+      "loss": 0.7568,
+      "step": 858
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.541097491559727e-06,
+      "loss": 0.6277,
+      "step": 859
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.521160153645045e-06,
+      "loss": 0.5955,
+      "step": 860
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.501224723300118e-06,
+      "loss": 0.676,
+      "step": 861
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.481291279942367e-06,
+      "loss": 0.6867,
+      "step": 862
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.461359902981286e-06,
+      "loss": 0.7463,
+      "step": 863
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.44143067181815e-06,
+      "loss": 0.6915,
+      "step": 864
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.421503665845679e-06,
+      "loss": 0.7083,
+      "step": 865
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.401578964447733e-06,
+      "loss": 0.6901,
+      "step": 866
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.381656646998983e-06,
+      "loss": 0.6523,
+      "step": 867
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.361736792864617e-06,
+      "loss": 0.6289,
+      "step": 868
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.34181948139999e-06,
+      "loss": 0.6726,
+      "step": 869
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.321904791950343e-06,
+      "loss": 0.7632,
+      "step": 870
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.301992803850469e-06,
+      "loss": 0.6241,
+      "step": 871
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.282083596424396e-06,
+      "loss": 0.691,
+      "step": 872
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.262177248985079e-06,
+      "loss": 0.706,
+      "step": 873
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.242273840834078e-06,
+      "loss": 0.7222,
+      "step": 874
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.222373451261235e-06,
+      "loss": 0.7283,
+      "step": 875
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.202476159544381e-06,
+      "loss": 0.6449,
+      "step": 876
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.182582044949e-06,
+      "loss": 0.6712,
+      "step": 877
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.162691186727912e-06,
+      "loss": 0.6561,
+      "step": 878
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.142803664120978e-06,
+      "loss": 0.7161,
+      "step": 879
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.122919556354763e-06,
+      "loss": 0.6653,
+      "step": 880
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.103038942642221e-06,
+      "loss": 0.7283,
+      "step": 881
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.083161902182407e-06,
+      "loss": 0.6188,
+      "step": 882
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.063288514160118e-06,
+      "loss": 0.7333,
+      "step": 883
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.04341885774562e-06,
+      "loss": 0.6764,
+      "step": 884
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.023553012094302e-06,
+      "loss": 0.768,
+      "step": 885
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.003691056346368e-06,
+      "loss": 0.601,
+      "step": 886
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.983833069626543e-06,
+      "loss": 0.6512,
+      "step": 887
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.96397913104372e-06,
+      "loss": 0.5955,
+      "step": 888
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.944129319690682e-06,
+      "loss": 0.6607,
+      "step": 889
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.92428371464376e-06,
+      "loss": 0.6423,
+      "step": 890
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.904442394962536e-06,
+      "loss": 0.7148,
+      "step": 891
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.884605439689512e-06,
+      "loss": 0.7082,
+      "step": 892
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.864772927849804e-06,
+      "loss": 0.8539,
+      "step": 893
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.844944938450837e-06,
+      "loss": 0.7223,
+      "step": 894
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.825121550482005e-06,
+      "loss": 0.6309,
+      "step": 895
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.805302842914386e-06,
+      "loss": 0.651,
+      "step": 896
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.785488894700402e-06,
+      "loss": 0.5931,
+      "step": 897
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.765679784773512e-06,
+      "loss": 0.7616,
+      "step": 898
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.74587559204792e-06,
+      "loss": 0.752,
+      "step": 899
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.726076395418214e-06,
+      "loss": 0.6068,
+      "step": 900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.706282273759101e-06,
+      "loss": 0.6567,
+      "step": 901
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.686493305925061e-06,
+      "loss": 0.6185,
+      "step": 902
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.666709570750049e-06,
+      "loss": 0.6327,
+      "step": 903
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.646931147047157e-06,
+      "loss": 0.7005,
+      "step": 904
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.627158113608341e-06,
+      "loss": 0.7578,
+      "step": 905
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.607390549204067e-06,
+      "loss": 0.7811,
+      "step": 906
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.587628532583022e-06,
+      "loss": 0.6714,
+      "step": 907
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.567872142471792e-06,
+      "loss": 0.6605,
+      "step": 908
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.548121457574538e-06,
+      "loss": 0.5431,
+      "step": 909
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.52837655657271e-06,
+      "loss": 0.7254,
+      "step": 910
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.5086375181247e-06,
+      "loss": 0.6393,
+      "step": 911
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.48890442086556e-06,
+      "loss": 0.6905,
+      "step": 912
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.469177343406661e-06,
+      "loss": 0.6835,
+      "step": 913
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.449456364335404e-06,
+      "loss": 0.6568,
+      "step": 914
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.429741562214885e-06,
+      "loss": 0.7457,
+      "step": 915
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.410033015583598e-06,
+      "loss": 0.713,
+      "step": 916
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.390330802955118e-06,
+      "loss": 0.742,
+      "step": 917
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.370635002817784e-06,
+      "loss": 0.688,
+      "step": 918
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.350945693634394e-06,
+      "loss": 0.7449,
+      "step": 919
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.331262953841885e-06,
+      "loss": 0.6702,
+      "step": 920
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.311586861851015e-06,
+      "loss": 0.6726,
+      "step": 921
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.291917496046076e-06,
+      "loss": 0.693,
+      "step": 922
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.272254934784549e-06,
+      "loss": 0.7291,
+      "step": 923
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.25259925639682e-06,
+      "loss": 0.631,
+      "step": 924
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.232950539185845e-06,
+      "loss": 0.716,
+      "step": 925
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.21330886142686e-06,
+      "loss": 0.6949,
+      "step": 926
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.193674301367041e-06,
+      "loss": 0.675,
+      "step": 927
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.174046937225225e-06,
+      "loss": 0.7759,
+      "step": 928
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.154426847191573e-06,
+      "loss": 0.6715,
+      "step": 929
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.134814109427272e-06,
+      "loss": 0.6723,
+      "step": 930
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.115208802064226e-06,
+      "loss": 0.6515,
+      "step": 931
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.095611003204718e-06,
+      "loss": 0.6607,
+      "step": 932
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.076020790921138e-06,
+      "loss": 0.6522,
+      "step": 933
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.05643824325565e-06,
+      "loss": 0.7013,
+      "step": 934
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.036863438219876e-06,
+      "loss": 0.6971,
+      "step": 935
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.017296453794604e-06,
+      "loss": 0.757,
+      "step": 936
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.997737367929461e-06,
+      "loss": 0.7161,
+      "step": 937
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.978186258542605e-06,
+      "loss": 0.6116,
+      "step": 938
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.958643203520423e-06,
+      "loss": 0.6992,
+      "step": 939
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.939108280717214e-06,
+      "loss": 0.7158,
+      "step": 940
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.919581567954879e-06,
+      "loss": 0.7667,
+      "step": 941
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.90006314302262e-06,
+      "loss": 0.6089,
+      "step": 942
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.880553083676616e-06,
+      "loss": 0.6422,
+      "step": 943
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.861051467639712e-06,
+      "loss": 0.718,
+      "step": 944
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.841558372601133e-06,
+      "loss": 0.7319,
+      "step": 945
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.822073876216146e-06,
+      "loss": 0.677,
+      "step": 946
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.802598056105773e-06,
+      "loss": 0.6795,
+      "step": 947
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.783130989856465e-06,
+      "loss": 0.6514,
+      "step": 948
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.763672755019806e-06,
+      "loss": 0.7415,
+      "step": 949
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.744223429112187e-06,
+      "loss": 0.6964,
+      "step": 950
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.72478308961452e-06,
+      "loss": 0.6964,
+      "step": 951
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.705351813971913e-06,
+      "loss": 0.7025,
+      "step": 952
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.685929679593363e-06,
+      "loss": 0.7258,
+      "step": 953
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.666516763851461e-06,
+      "loss": 0.6882,
+      "step": 954
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.647113144082062e-06,
+      "loss": 0.7115,
+      "step": 955
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.627718897583987e-06,
+      "loss": 0.6559,
+      "step": 956
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.608334101618729e-06,
+      "loss": 0.7259,
+      "step": 957
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.588958833410123e-06,
+      "loss": 0.6636,
+      "step": 958
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.569593170144053e-06,
+      "loss": 0.7069,
+      "step": 959
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.550237188968136e-06,
+      "loss": 0.717,
+      "step": 960
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.530890966991415e-06,
+      "loss": 0.6547,
+      "step": 961
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.5115545812840615e-06,
+      "loss": 0.7066,
+      "step": 962
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.492228108877061e-06,
+      "loss": 0.6646,
+      "step": 963
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.472911626761905e-06,
+      "loss": 0.6926,
+      "step": 964
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.453605211890285e-06,
+      "loss": 0.7277,
+      "step": 965
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.434308941173798e-06,
+      "loss": 0.72,
+      "step": 966
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.415022891483605e-06,
+      "loss": 0.6588,
+      "step": 967
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.395747139650178e-06,
+      "loss": 0.6625,
+      "step": 968
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.376481762462943e-06,
+      "loss": 0.6292,
+      "step": 969
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.3572268366700115e-06,
+      "loss": 0.7142,
+      "step": 970
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.337982438977853e-06,
+      "loss": 0.6556,
+      "step": 971
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.318748646050989e-06,
+      "loss": 0.648,
+      "step": 972
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.2995255345117085e-06,
+      "loss": 0.7862,
+      "step": 973
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.280313180939737e-06,
+      "loss": 0.6885,
+      "step": 974
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.2611116618719515e-06,
+      "loss": 0.6297,
+      "step": 975
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.241921053802059e-06,
+      "loss": 0.7574,
+      "step": 976
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.2227414331803115e-06,
+      "loss": 0.6622,
+      "step": 977
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.203572876413175e-06,
+      "loss": 0.7607,
+      "step": 978
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.18441545986305e-06,
+      "loss": 0.6963,
+      "step": 979
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.165269259847959e-06,
+      "loss": 0.7073,
+      "step": 980
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.146134352641232e-06,
+      "loss": 0.6794,
+      "step": 981
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.1270108144712215e-06,
+      "loss": 0.7476,
+      "step": 982
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.107898721520984e-06,
+      "loss": 0.6931,
+      "step": 983
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0887981499279735e-06,
+      "loss": 0.6724,
+      "step": 984
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.069709175783763e-06,
+      "loss": 0.6842,
+      "step": 985
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.050631875133708e-06,
+      "loss": 0.675,
+      "step": 986
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.031566323976672e-06,
+      "loss": 0.7298,
+      "step": 987
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.012512598264702e-06,
+      "loss": 0.7302,
+      "step": 988
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.993470773902749e-06,
+      "loss": 0.6803,
+      "step": 989
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.974440926748335e-06,
+      "loss": 0.6088,
+      "step": 990
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9554231326112766e-06,
+      "loss": 0.6285,
+      "step": 991
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.936417467253378e-06,
+      "loss": 0.6888,
+      "step": 992
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.917424006388118e-06,
+      "loss": 0.6566,
+      "step": 993
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.898442825680368e-06,
+      "loss": 0.6899,
+      "step": 994
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.879474000746063e-06,
+      "loss": 0.7107,
+      "step": 995
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.860517607151922e-06,
+      "loss": 0.6549,
+      "step": 996
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.841573720415146e-06,
+      "loss": 0.7016,
+      "step": 997
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.822642416003107e-06,
+      "loss": 0.6476,
+      "step": 998
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.803723769333055e-06,
+      "loss": 0.7161,
+      "step": 999
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.784817855771815e-06,
+      "loss": 0.6959,
+      "step": 1000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.765924750635481e-06,
+      "loss": 0.6791,
+      "step": 1001
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.747044529189124e-06,
+      "loss": 0.6999,
+      "step": 1002
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.7281772666464985e-06,
+      "loss": 0.6133,
+      "step": 1003
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.709323038169722e-06,
+      "loss": 0.7016,
+      "step": 1004
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.690481918868997e-06,
+      "loss": 0.6662,
+      "step": 1005
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6716539838023e-06,
+      "loss": 0.6705,
+      "step": 1006
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.652839307975076e-06,
+      "loss": 0.7864,
+      "step": 1007
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.634037966339964e-06,
+      "loss": 0.7498,
+      "step": 1008
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.61525003379647e-06,
+      "loss": 0.6526,
+      "step": 1009
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.596475585190692e-06,
+      "loss": 0.6405,
+      "step": 1010
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.577714695315004e-06,
+      "loss": 0.7353,
+      "step": 1011
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.5589674389077754e-06,
+      "loss": 0.7205,
+      "step": 1012
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.5402338906530476e-06,
+      "loss": 0.6288,
+      "step": 1013
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.5215141251802635e-06,
+      "loss": 0.6739,
+      "step": 1014
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.502808217063963e-06,
+      "loss": 0.7086,
+      "step": 1015
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.484116240823468e-06,
+      "loss": 0.5953,
+      "step": 1016
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.465438270922617e-06,
+      "loss": 0.6907,
+      "step": 1017
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.446774381769436e-06,
+      "loss": 0.6701,
+      "step": 1018
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.428124647715861e-06,
+      "loss": 0.7424,
+      "step": 1019
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.4094891430574436e-06,
+      "loss": 0.6843,
+      "step": 1020
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.390867942033041e-06,
+      "loss": 0.6154,
+      "step": 1021
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.372261118824539e-06,
+      "loss": 0.6539,
+      "step": 1022
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.353668747556538e-06,
+      "loss": 0.6177,
+      "step": 1023
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.3350909022960625e-06,
+      "loss": 0.7013,
+      "step": 1024
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.316527657052281e-06,
+      "loss": 0.6503,
+      "step": 1025
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.297979085776187e-06,
+      "loss": 0.6751,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.279445262360328e-06,
+      "loss": 0.7046,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.260926260638491e-06,
+      "loss": 0.6545,
+      "step": 1028
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.242422154385428e-06,
+      "loss": 0.7078,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.223933017316535e-06,
+      "loss": 0.625,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.205458923087587e-06,
+      "loss": 0.614,
+      "step": 1031
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.186999945294434e-06,
+      "loss": 0.7259,
+      "step": 1032
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.168556157472695e-06,
+      "loss": 0.6501,
+      "step": 1033
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.150127633097489e-06,
+      "loss": 0.7315,
+      "step": 1034
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.131714445583124e-06,
+      "loss": 0.7031,
+      "step": 1035
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.113316668282805e-06,
+      "loss": 0.654,
+      "step": 1036
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.0949343744883495e-06,
+      "loss": 0.6901,
+      "step": 1037
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.076567637429902e-06,
+      "loss": 0.6609,
+      "step": 1038
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.058216530275619e-06,
+      "loss": 0.6729,
+      "step": 1039
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.039881126131406e-06,
+      "loss": 0.7722,
+      "step": 1040
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.021561498040599e-06,
+      "loss": 0.7511,
+      "step": 1041
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.003257718983691e-06,
+      "loss": 0.7145,
+      "step": 1042
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.98496986187804e-06,
+      "loss": 0.6628,
+      "step": 1043
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.966697999577573e-06,
+      "loss": 0.7894,
+      "step": 1044
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.948442204872498e-06,
+      "loss": 0.7522,
+      "step": 1045
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.930202550489014e-06,
+      "loss": 0.7106,
+      "step": 1046
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.911979109089016e-06,
+      "loss": 0.7141,
+      "step": 1047
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.893771953269822e-06,
+      "loss": 0.6893,
+      "step": 1048
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.875581155563863e-06,
+      "loss": 0.6865,
+      "step": 1049
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.857406788438409e-06,
+      "loss": 0.6135,
+      "step": 1050
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.839248924295271e-06,
+      "loss": 0.6774,
+      "step": 1051
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.821107635470524e-06,
+      "loss": 0.7124,
+      "step": 1052
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.8029829942342005e-06,
+      "loss": 0.7102,
+      "step": 1053
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.78487507279002e-06,
+      "loss": 0.7136,
+      "step": 1054
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7667839432750895e-06,
+      "loss": 0.731,
+      "step": 1055
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.748709677759634e-06,
+      "loss": 0.7672,
+      "step": 1056
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.730652348246686e-06,
+      "loss": 0.6498,
+      "step": 1057
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7126120266718085e-06,
+      "loss": 0.6683,
+      "step": 1058
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.694588784902818e-06,
+      "loss": 0.6827,
+      "step": 1059
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.676582694739481e-06,
+      "loss": 0.6814,
+      "step": 1060
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.658593827913241e-06,
+      "loss": 0.6655,
+      "step": 1061
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.640622256086924e-06,
+      "loss": 0.7191,
+      "step": 1062
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.622668050854473e-06,
+      "loss": 0.7353,
+      "step": 1063
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.604731283740621e-06,
+      "loss": 0.716,
+      "step": 1064
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.586812026200654e-06,
+      "loss": 0.6702,
+      "step": 1065
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5689103496200894e-06,
+      "loss": 0.7106,
+      "step": 1066
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.551026325314423e-06,
+      "loss": 0.6428,
+      "step": 1067
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.533160024528817e-06,
+      "loss": 0.6541,
+      "step": 1068
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.515311518437829e-06,
+      "loss": 0.6361,
+      "step": 1069
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.4974808781451295e-06,
+      "loss": 0.6998,
+      "step": 1070
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.479668174683217e-06,
+      "loss": 0.7449,
+      "step": 1071
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.461873479013131e-06,
+      "loss": 0.7642,
+      "step": 1072
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.444096862024174e-06,
+      "loss": 0.7097,
+      "step": 1073
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.426338394533636e-06,
+      "loss": 0.6949,
+      "step": 1074
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.408598147286495e-06,
+      "loss": 0.6348,
+      "step": 1075
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.390876190955138e-06,
+      "loss": 0.5959,
+      "step": 1076
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.373172596139105e-06,
+      "loss": 0.6872,
+      "step": 1077
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.355487433364773e-06,
+      "loss": 0.7053,
+      "step": 1078
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.337820773085099e-06,
+      "loss": 0.6867,
+      "step": 1079
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.3201726856793215e-06,
+      "loss": 0.7495,
+      "step": 1080
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.302543241452707e-06,
+      "loss": 0.7328,
+      "step": 1081
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.284932510636234e-06,
+      "loss": 0.7252,
+      "step": 1082
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.267340563386334e-06,
+      "loss": 0.9129,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.2497674697846275e-06,
+      "loss": 0.3833,
+      "step": 1084
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.232213299837609e-06,
+      "loss": 0.4007,
+      "step": 1085
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.214678123476393e-06,
+      "loss": 0.3596,
+      "step": 1086
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.197162010556426e-06,
+      "loss": 0.4324,
+      "step": 1087
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.179665030857213e-06,
+      "loss": 0.4366,
+      "step": 1088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.1621872540820386e-06,
+      "loss": 0.3468,
+      "step": 1089
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.14472874985768e-06,
+      "loss": 0.4116,
+      "step": 1090
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.127289587734154e-06,
+      "loss": 0.3799,
+      "step": 1091
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.109869837184413e-06,
+      "loss": 0.4633,
+      "step": 1092
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.092469567604068e-06,
+      "loss": 0.4027,
+      "step": 1093
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.075088848311145e-06,
+      "loss": 0.3812,
+      "step": 1094
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.057727748545776e-06,
+      "loss": 0.3522,
+      "step": 1095
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0403863374699304e-06,
+      "loss": 0.3872,
+      "step": 1096
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.023064684167147e-06,
+      "loss": 0.3313,
+      "step": 1097
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.005762857642264e-06,
+      "loss": 0.4181,
+      "step": 1098
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.988480926821119e-06,
+      "loss": 0.4073,
+      "step": 1099
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.971218960550297e-06,
+      "loss": 0.3995,
+      "step": 1100
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.953977027596848e-06,
+      "loss": 0.3836,
+      "step": 1101
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.9367551966480255e-06,
+      "loss": 0.4275,
+      "step": 1102
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.919553536310989e-06,
+      "loss": 0.4039,
+      "step": 1103
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.902372115112549e-06,
+      "loss": 0.4002,
+      "step": 1104
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.885211001498886e-06,
+      "loss": 0.4053,
+      "step": 1105
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.868070263835284e-06,
+      "loss": 0.4154,
+      "step": 1106
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.850949970405852e-06,
+      "loss": 0.4125,
+      "step": 1107
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.833850189413251e-06,
+      "loss": 0.4353,
+      "step": 1108
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.8167709889784454e-06,
+      "loss": 0.3925,
+      "step": 1109
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.799712437140382e-06,
+      "loss": 0.3599,
+      "step": 1110
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.782674601855768e-06,
+      "loss": 0.3203,
+      "step": 1111
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.765657550998783e-06,
+      "loss": 0.3431,
+      "step": 1112
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7486613523607974e-06,
+      "loss": 0.3828,
+      "step": 1113
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7316860736501195e-06,
+      "loss": 0.4199,
+      "step": 1114
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.714731782491713e-06,
+      "loss": 0.4589,
+      "step": 1115
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.697798546426935e-06,
+      "loss": 0.371,
+      "step": 1116
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.680886432913265e-06,
+      "loss": 0.3551,
+      "step": 1117
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.663995509324032e-06,
+      "loss": 0.3459,
+      "step": 1118
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.647125842948161e-06,
+      "loss": 0.4605,
+      "step": 1119
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.630277500989884e-06,
+      "loss": 0.385,
+      "step": 1120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.613450550568482e-06,
+      "loss": 0.3532,
+      "step": 1121
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.596645058718023e-06,
+      "loss": 0.3452,
+      "step": 1122
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.579861092387088e-06,
+      "loss": 0.3406,
+      "step": 1123
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.563098718438502e-06,
+      "loss": 0.4195,
+      "step": 1124
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.546358003649075e-06,
+      "loss": 0.4149,
+      "step": 1125
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.529639014709339e-06,
+      "loss": 0.338,
+      "step": 1126
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.5129418182232706e-06,
+      "loss": 0.3541,
+      "step": 1127
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.49626648070802e-06,
+      "loss": 0.3631,
+      "step": 1128
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.479613068593674e-06,
+      "loss": 0.3259,
+      "step": 1129
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.462981648222972e-06,
+      "loss": 0.4036,
+      "step": 1130
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.446372285851034e-06,
+      "loss": 0.3687,
+      "step": 1131
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.429785047645116e-06,
+      "loss": 0.3716,
+      "step": 1132
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.413219999684334e-06,
+      "loss": 0.3083,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.396677207959404e-06,
+      "loss": 0.3211,
+      "step": 1134
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.38015673837238e-06,
+      "loss": 0.3525,
+      "step": 1135
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.363658656736386e-06,
+      "loss": 0.394,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.347183028775369e-06,
+      "loss": 0.3407,
+      "step": 1137
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.330729920123822e-06,
+      "loss": 0.4269,
+      "step": 1138
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.314299396326511e-06,
+      "loss": 0.3558,
+      "step": 1139
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.297891522838259e-06,
+      "loss": 0.3852,
+      "step": 1140
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2815063650236345e-06,
+      "loss": 0.327,
+      "step": 1141
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2651439881567205e-06,
+      "loss": 0.3901,
+      "step": 1142
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2488044574208435e-06,
+      "loss": 0.303,
+      "step": 1143
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.232487837908329e-06,
+      "loss": 0.2763,
+      "step": 1144
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.216194194620211e-06,
+      "loss": 0.3372,
+      "step": 1145
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.199923592466002e-06,
+      "loss": 0.3183,
+      "step": 1146
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.183676096263432e-06,
+      "loss": 0.4098,
+      "step": 1147
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.167451770738174e-06,
+      "loss": 0.4114,
+      "step": 1148
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.151250680523595e-06,
+      "loss": 0.3756,
+      "step": 1149
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.135072890160502e-06,
+      "loss": 0.4246,
+      "step": 1150
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.1189184640968825e-06,
+      "loss": 0.4246,
+      "step": 1151
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.1027874666876425e-06,
+      "loss": 0.3916,
+      "step": 1152
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.086679962194355e-06,
+      "loss": 0.3029,
+      "step": 1153
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.070596014785013e-06,
+      "loss": 0.4057,
+      "step": 1154
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.0545356885337576e-06,
+      "loss": 0.432,
+      "step": 1155
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.038499047420616e-06,
+      "loss": 0.4509,
+      "step": 1156
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0224861553312855e-06,
+      "loss": 0.3989,
+      "step": 1157
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.006497076056838e-06,
+      "loss": 0.3592,
+      "step": 1158
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.990531873293484e-06,
+      "loss": 0.3361,
+      "step": 1159
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.974590610642314e-06,
+      "loss": 0.3737,
+      "step": 1160
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.95867335160906e-06,
+      "loss": 0.4285,
+      "step": 1161
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.942780159603813e-06,
+      "loss": 0.3335,
+      "step": 1162
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.926911097940792e-06,
+      "loss": 0.387,
+      "step": 1163
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.911066229838098e-06,
+      "loss": 0.4705,
+      "step": 1164
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.895245618417443e-06,
+      "loss": 0.3514,
+      "step": 1165
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.879449326703903e-06,
+      "loss": 0.378,
+      "step": 1166
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.86367741762568e-06,
+      "loss": 0.3788,
+      "step": 1167
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.847929954013833e-06,
+      "loss": 0.3771,
+      "step": 1168
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.832206998602043e-06,
+      "loss": 0.3799,
+      "step": 1169
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.816508614026355e-06,
+      "loss": 0.4331,
+      "step": 1170
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.800834862824928e-06,
+      "loss": 0.3969,
+      "step": 1171
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.785185807437796e-06,
+      "loss": 0.3819,
+      "step": 1172
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7695615102066072e-06,
+      "loss": 0.3915,
+      "step": 1173
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7539620333743676e-06,
+      "loss": 0.3387,
+      "step": 1174
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.738387439085228e-06,
+      "loss": 0.395,
+      "step": 1175
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7228377893841992e-06,
+      "loss": 0.4089,
+      "step": 1176
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.707313146216921e-06,
+      "loss": 0.3822,
+      "step": 1177
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.691813571429418e-06,
+      "loss": 0.4021,
+      "step": 1178
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.676339126767845e-06,
+      "loss": 0.3211,
+      "step": 1179
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.660889873878246e-06,
+      "loss": 0.39,
+      "step": 1180
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6454658743063076e-06,
+      "loss": 0.2979,
+      "step": 1181
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6300671894971207e-06,
+      "loss": 0.3622,
+      "step": 1182
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6146938807949185e-06,
+      "loss": 0.3394,
+      "step": 1183
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.599346009442848e-06,
+      "loss": 0.3531,
+      "step": 1184
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5840236365827206e-06,
+      "loss": 0.4074,
+      "step": 1185
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5687268232547666e-06,
+      "loss": 0.3861,
+      "step": 1186
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5534556303973954e-06,
+      "loss": 0.392,
+      "step": 1187
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5382101188469477e-06,
+      "loss": 0.3473,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.522990349337466e-06,
+      "loss": 0.3114,
+      "step": 1189
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.50779638250044e-06,
+      "loss": 0.3697,
+      "step": 1190
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4926282788645527e-06,
+      "loss": 0.435,
+      "step": 1191
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4774860988554803e-06,
+      "loss": 0.4626,
+      "step": 1192
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4623699027956104e-06,
+      "loss": 0.3332,
+      "step": 1193
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4472797509038215e-06,
+      "loss": 0.3496,
+      "step": 1194
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.432215703295234e-06,
+      "loss": 0.4138,
+      "step": 1195
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4171778199809922e-06,
+      "loss": 0.3486,
+      "step": 1196
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4021661608679867e-06,
+      "loss": 0.433,
+      "step": 1197
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3871807857586527e-06,
+      "loss": 0.34,
+      "step": 1198
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3722217543507085e-06,
+      "loss": 0.342,
+      "step": 1199
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3572891262369377e-06,
+      "loss": 0.3944,
+      "step": 1200
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.342382960904933e-06,
+      "loss": 0.4692,
+      "step": 1201
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3275033177368555e-06,
+      "loss": 0.3618,
+      "step": 1202
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3126502560092323e-06,
+      "loss": 0.3976,
+      "step": 1203
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.297823834892678e-06,
+      "loss": 0.3738,
+      "step": 1204
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2830241134516895e-06,
+      "loss": 0.3182,
+      "step": 1205
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.268251150644388e-06,
+      "loss": 0.4364,
+      "step": 1206
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.253505005322314e-06,
+      "loss": 0.3975,
+      "step": 1207
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.238785736230152e-06,
+      "loss": 0.4279,
+      "step": 1208
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2240934020055305e-06,
+      "loss": 0.3812,
+      "step": 1209
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.209428061178781e-06,
+      "loss": 0.3747,
+      "step": 1210
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1947897721726927e-06,
+      "loss": 0.3212,
+      "step": 1211
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.180178593302291e-06,
+      "loss": 0.4137,
+      "step": 1212
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1655945827745984e-06,
+      "loss": 0.44,
+      "step": 1213
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1510377986884113e-06,
+      "loss": 0.395,
+      "step": 1214
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.136508299034058e-06,
+      "loss": 0.3418,
+      "step": 1215
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1220061416931737e-06,
+      "loss": 0.354,
+      "step": 1216
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1075313844384757e-06,
+      "loss": 0.3282,
+      "step": 1217
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0930840849335196e-06,
+      "loss": 0.3189,
+      "step": 1218
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.07866430073248e-06,
+      "loss": 0.327,
+      "step": 1219
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.064272089279915e-06,
+      "loss": 0.3201,
+      "step": 1220
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0499075079105422e-06,
+      "loss": 0.342,
+      "step": 1221
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0355706138490093e-06,
+      "loss": 0.3873,
+      "step": 1222
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0212614642096595e-06,
+      "loss": 0.369,
+      "step": 1223
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.006980115996324e-06,
+      "loss": 0.3977,
+      "step": 1224
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.9927266261020615e-06,
+      "loss": 0.3488,
+      "step": 1225
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.978501051308961e-06,
+      "loss": 0.3382,
+      "step": 1226
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9643034482879073e-06,
+      "loss": 0.3371,
+      "step": 1227
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.95013387359835e-06,
+      "loss": 0.3362,
+      "step": 1228
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.93599238368808e-06,
+      "loss": 0.3425,
+      "step": 1229
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9218790348930072e-06,
+      "loss": 0.3593,
+      "step": 1230
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9077938834369333e-06,
+      "loss": 0.3724,
+      "step": 1231
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8937369854313335e-06,
+      "loss": 0.3899,
+      "step": 1232
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.879708396875125e-06,
+      "loss": 0.3526,
+      "step": 1233
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8657081736544454e-06,
+      "loss": 0.3441,
+      "step": 1234
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8517363715424438e-06,
+      "loss": 0.3543,
+      "step": 1235
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8377930461990376e-06,
+      "loss": 0.2987,
+      "step": 1236
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.823878253170693e-06,
+      "loss": 0.3486,
+      "step": 1237
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8099920478902298e-06,
+      "loss": 0.3538,
+      "step": 1238
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7961344856765704e-06,
+      "loss": 0.3806,
+      "step": 1239
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.78230562173453e-06,
+      "loss": 0.3565,
+      "step": 1240
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7685055111546034e-06,
+      "loss": 0.4141,
+      "step": 1241
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.754734208912735e-06,
+      "loss": 0.4008,
+      "step": 1242
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.740991769870106e-06,
+      "loss": 0.331,
+      "step": 1243
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.727278248772911e-06,
+      "loss": 0.3845,
+      "step": 1244
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.713593700252155e-06,
+      "loss": 0.3753,
+      "step": 1245
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.6999381788234104e-06,
+      "loss": 0.3772,
+      "step": 1246
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.68631173888662e-06,
+      "loss": 0.3768,
+      "step": 1247
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.672714434725873e-06,
+      "loss": 0.4167,
+      "step": 1248
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.659146320509187e-06,
+      "loss": 0.3801,
+      "step": 1249
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6456074502882977e-06,
+      "loss": 0.3258,
+      "step": 1250
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6320978779984375e-06,
+      "loss": 0.3361,
+      "step": 1251
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6186176574581303e-06,
+      "loss": 0.4217,
+      "step": 1252
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6051668423689657e-06,
+      "loss": 0.3368,
+      "step": 1253
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5917454863153823e-06,
+      "loss": 0.3655,
+      "step": 1254
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5783536427644772e-06,
+      "loss": 0.3206,
+      "step": 1255
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5649913650657686e-06,
+      "loss": 0.3413,
+      "step": 1256
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5516587064509925e-06,
+      "loss": 0.3991,
+      "step": 1257
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5383557200338873e-06,
+      "loss": 0.3309,
+      "step": 1258
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.525082458810001e-06,
+      "loss": 0.388,
+      "step": 1259
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.511838975656441e-06,
+      "loss": 0.3757,
+      "step": 1260
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4986253233317017e-06,
+      "loss": 0.3033,
+      "step": 1261
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4854415544754397e-06,
+      "loss": 0.4264,
+      "step": 1262
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4722877216082597e-06,
+      "loss": 0.3052,
+      "step": 1263
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.459163877131512e-06,
+      "loss": 0.3198,
+      "step": 1264
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4460700733270703e-06,
+      "loss": 0.317,
+      "step": 1265
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4330063623571542e-06,
+      "loss": 0.3897,
+      "step": 1266
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.419972796264087e-06,
+      "loss": 0.3766,
+      "step": 1267
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.406969426970107e-06,
+      "loss": 0.4255,
+      "step": 1268
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.393996306277152e-06,
+      "loss": 0.36,
+      "step": 1269
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3810534858666724e-06,
+      "loss": 0.3302,
+      "step": 1270
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.36814101729939e-06,
+      "loss": 0.334,
+      "step": 1271
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3552589520151224e-06,
+      "loss": 0.396,
+      "step": 1272
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3424073413325744e-06,
+      "loss": 0.427,
+      "step": 1273
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3295862364491216e-06,
+      "loss": 0.336,
+      "step": 1274
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3167956884406105e-06,
+      "loss": 0.4221,
+      "step": 1275
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.304035748261163e-06,
+      "loss": 0.3729,
+      "step": 1276
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.291306466742963e-06,
+      "loss": 0.3776,
+      "step": 1277
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.278607894596061e-06,
+      "loss": 0.3733,
+      "step": 1278
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2659400824081645e-06,
+      "loss": 0.4081,
+      "step": 1279
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2533030806444534e-06,
+      "loss": 0.4044,
+      "step": 1280
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2406969396473564e-06,
+      "loss": 0.407,
+      "step": 1281
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2281217096363638e-06,
+      "loss": 0.359,
+      "step": 1282
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2155774407078255e-06,
+      "loss": 0.354,
+      "step": 1283
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.203064182834751e-06,
+      "loss": 0.3518,
+      "step": 1284
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1905819858666087e-06,
+      "loss": 0.3925,
+      "step": 1285
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1781308995291283e-06,
+      "loss": 0.368,
+      "step": 1286
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.165710973424112e-06,
+      "loss": 0.4049,
+      "step": 1287
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.153322257029211e-06,
+      "loss": 0.3979,
+      "step": 1288
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.140964799697758e-06,
+      "loss": 0.3133,
+      "step": 1289
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1286386506585567e-06,
+      "loss": 0.4118,
+      "step": 1290
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1163438590156827e-06,
+      "loss": 0.3458,
+      "step": 1291
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1040804737482965e-06,
+      "loss": 0.3559,
+      "step": 1292
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.091848543710441e-06,
+      "loss": 0.4021,
+      "step": 1293
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0796481176308493e-06,
+      "loss": 0.3776,
+      "step": 1294
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0674792441127543e-06,
+      "loss": 0.4148,
+      "step": 1295
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.05534197163369e-06,
+      "loss": 0.3597,
+      "step": 1296
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0432363485452977e-06,
+      "loss": 0.352,
+      "step": 1297
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0311624230731454e-06,
+      "loss": 0.3331,
+      "step": 1298
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.019120243316517e-06,
+      "loss": 0.4426,
+      "step": 1299
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.007109857248227e-06,
+      "loss": 0.3636,
+      "step": 1300
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9951313127144467e-06,
+      "loss": 0.409,
+      "step": 1301
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.983184657434486e-06,
+      "loss": 0.3633,
+      "step": 1302
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9712699390006217e-06,
+      "loss": 0.3264,
+      "step": 1303
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9593872048779005e-06,
+      "loss": 0.3626,
+      "step": 1304
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9475365024039628e-06,
+      "loss": 0.4662,
+      "step": 1305
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.935717878788824e-06,
+      "loss": 0.3928,
+      "step": 1306
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9239313811147176e-06,
+      "loss": 0.3542,
+      "step": 1307
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9121770563359022e-06,
+      "loss": 0.3574,
+      "step": 1308
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9004549512784553e-06,
+      "loss": 0.3174,
+      "step": 1309
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8887651126401063e-06,
+      "loss": 0.4826,
+      "step": 1310
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8771075869900434e-06,
+      "loss": 0.3504,
+      "step": 1311
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8654824207687283e-06,
+      "loss": 0.3626,
+      "step": 1312
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.85388966028771e-06,
+      "loss": 0.358,
+      "step": 1313
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8423293517294393e-06,
+      "loss": 0.3905,
+      "step": 1314
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8308015411470969e-06,
+      "loss": 0.425,
+      "step": 1315
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8193062744643953e-06,
+      "loss": 0.3442,
+      "step": 1316
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8078435974753893e-06,
+      "loss": 0.3709,
+      "step": 1317
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7964135558443241e-06,
+      "loss": 0.382,
+      "step": 1318
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7850161951054235e-06,
+      "loss": 0.3654,
+      "step": 1319
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7736515606627202e-06,
+      "loss": 0.3249,
+      "step": 1320
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7623196977898737e-06,
+      "loss": 0.3248,
+      "step": 1321
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7510206516299989e-06,
+      "loss": 0.3799,
+      "step": 1322
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7397544671954636e-06,
+      "loss": 0.3254,
+      "step": 1323
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7285211893677312e-06,
+      "loss": 0.4236,
+      "step": 1324
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7173208628971772e-06,
+      "loss": 0.4118,
+      "step": 1325
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.706153532402901e-06,
+      "loss": 0.3717,
+      "step": 1326
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6950192423725598e-06,
+      "loss": 0.4089,
+      "step": 1327
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.683918037162181e-06,
+      "loss": 0.3224,
+      "step": 1328
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6728499609959969e-06,
+      "loss": 0.3538,
+      "step": 1329
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6618150579662583e-06,
+      "loss": 0.3222,
+      "step": 1330
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6508133720330654e-06,
+      "loss": 0.3701,
+      "step": 1331
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6398449470241862e-06,
+      "loss": 0.3785,
+      "step": 1332
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.628909826634898e-06,
+      "loss": 0.3711,
+      "step": 1333
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6180080544277853e-06,
+      "loss": 0.3691,
+      "step": 1334
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6071396738325894e-06,
+      "loss": 0.3364,
+      "step": 1335
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5963047281460353e-06,
+      "loss": 0.4223,
+      "step": 1336
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5855032605316422e-06,
+      "loss": 0.3956,
+      "step": 1337
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5747353140195653e-06,
+      "loss": 0.3477,
+      "step": 1338
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.564000931506421e-06,
+      "loss": 0.3998,
+      "step": 1339
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5533001557551164e-06,
+      "loss": 0.4434,
+      "step": 1340
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5426330293946756e-06,
+      "loss": 0.3617,
+      "step": 1341
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5319995949200727e-06,
+      "loss": 0.3527,
+      "step": 1342
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5213998946920693e-06,
+      "loss": 0.4095,
+      "step": 1343
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5108339709370302e-06,
+      "loss": 0.3772,
+      "step": 1344
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5003018657467673e-06,
+      "loss": 0.3081,
+      "step": 1345
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4898036210783684e-06,
+      "loss": 0.3599,
+      "step": 1346
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4793392787540295e-06,
+      "loss": 0.3508,
+      "step": 1347
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4689088804608876e-06,
+      "loss": 0.3706,
+      "step": 1348
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4585124677508567e-06,
+      "loss": 0.348,
+      "step": 1349
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4481500820404659e-06,
+      "loss": 0.3017,
+      "step": 1350
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4378217646106851e-06,
+      "loss": 0.3391,
+      "step": 1351
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4275275566067592e-06,
+      "loss": 0.4018,
+      "step": 1352
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.417267499038064e-06,
+      "loss": 0.3839,
+      "step": 1353
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4070416327779202e-06,
+      "loss": 0.3802,
+      "step": 1354
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.396849998563441e-06,
+      "loss": 0.3643,
+      "step": 1355
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.386692636995368e-06,
+      "loss": 0.4006,
+      "step": 1356
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3765695885379115e-06,
+      "loss": 0.4202,
+      "step": 1357
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3664808935185858e-06,
+      "loss": 0.3526,
+      "step": 1358
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3564265921280485e-06,
+      "loss": 0.4257,
+      "step": 1359
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3464067244199496e-06,
+      "loss": 0.4319,
+      "step": 1360
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3364213303107577e-06,
+      "loss": 0.3326,
+      "step": 1361
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.326470449579611e-06,
+      "loss": 0.3456,
+      "step": 1362
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3165541218681477e-06,
+      "loss": 0.3028,
+      "step": 1363
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.306672386680371e-06,
+      "loss": 0.3757,
+      "step": 1364
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2968252833824635e-06,
+      "loss": 0.3438,
+      "step": 1365
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2870128512026504e-06,
+      "loss": 0.3471,
+      "step": 1366
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2772351292310315e-06,
+      "loss": 0.3252,
+      "step": 1367
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2674921564194408e-06,
+      "loss": 0.3198,
+      "step": 1368
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2577839715812667e-06,
+      "loss": 0.3725,
+      "step": 1369
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2481106133913178e-06,
+      "loss": 0.367,
+      "step": 1370
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2384721203856697e-06,
+      "loss": 0.4296,
+      "step": 1371
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2288685309614933e-06,
+      "loss": 0.5371,
+      "step": 1372
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.219299883376921e-06,
+      "loss": 0.389,
+      "step": 1373
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2097662157508805e-06,
+      "loss": 0.3394,
+      "step": 1374
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2002675660629526e-06,
+      "loss": 0.3395,
+      "step": 1375
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1908039721532138e-06,
+      "loss": 0.4244,
+      "step": 1376
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1813754717220872e-06,
+      "loss": 0.3718,
+      "step": 1377
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.171982102330198e-06,
+      "loss": 0.3667,
+      "step": 1378
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1626239013982165e-06,
+      "loss": 0.3268,
+      "step": 1379
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1533009062067024e-06,
+      "loss": 0.3422,
+      "step": 1380
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1440131538959797e-06,
+      "loss": 0.3874,
+      "step": 1381
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.134760681465964e-06,
+      "loss": 0.3831,
+      "step": 1382
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1255435257760284e-06,
+      "loss": 0.2994,
+      "step": 1383
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.116361723544853e-06,
+      "loss": 0.3896,
+      "step": 1384
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1072153113502849e-06,
+      "loss": 0.3077,
+      "step": 1385
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0981043256291746e-06,
+      "loss": 0.361,
+      "step": 1386
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0890288026772521e-06,
+      "loss": 0.39,
+      "step": 1387
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0799887786489738e-06,
+      "loss": 0.3719,
+      "step": 1388
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.070984289557374e-06,
+      "loss": 0.3584,
+      "step": 1389
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0620153712739268e-06,
+      "loss": 0.329,
+      "step": 1390
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0530820595283996e-06,
+      "loss": 0.416,
+      "step": 1391
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0441843899087178e-06,
+      "loss": 0.3563,
+      "step": 1392
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0353223978608117e-06,
+      "loss": 0.372,
+      "step": 1393
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.026496118688487e-06,
+      "loss": 0.3879,
+      "step": 1394
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0177055875532771e-06,
+      "loss": 0.3698,
+      "step": 1395
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0089508394743053e-06,
+      "loss": 0.3668,
+      "step": 1396
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0002319093281454e-06,
+      "loss": 0.3906,
+      "step": 1397
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.915488318486744e-07,
+      "loss": 0.3522,
+      "step": 1398
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.829016416269554e-07,
+      "loss": 0.4021,
+      "step": 1399
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.742903731110764e-07,
+      "loss": 0.3663,
+      "step": 1400
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.657150606060262e-07,
+      "loss": 0.3567,
+      "step": 1401
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.571757382735548e-07,
+      "loss": 0.3518,
+      "step": 1402
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.486724401320346e-07,
+      "loss": 0.3768,
+      "step": 1403
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.402052000563289e-07,
+      "loss": 0.3886,
+      "step": 1404
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.317740517776552e-07,
+      "loss": 0.4617,
+      "step": 1405
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.233790288834543e-07,
+      "loss": 0.3328,
+      "step": 1406
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.150201648172475e-07,
+      "loss": 0.3563,
+      "step": 1407
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.066974928785121e-07,
+      "loss": 0.417,
+      "step": 1408
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.984110462225437e-07,
+      "loss": 0.3272,
+      "step": 1409
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.901608578603283e-07,
+      "loss": 0.4232,
+      "step": 1410
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.819469606584085e-07,
+      "loss": 0.318,
+      "step": 1411
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.737693873387487e-07,
+      "loss": 0.3853,
+      "step": 1412
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.656281704786152e-07,
+      "loss": 0.392,
+      "step": 1413
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.575233425104368e-07,
+      "loss": 0.3788,
+      "step": 1414
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.49454935721673e-07,
+      "loss": 0.3944,
+      "step": 1415
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.414229822547026e-07,
+      "loss": 0.3906,
+      "step": 1416
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.33427514106675e-07,
+      "loss": 0.4074,
+      "step": 1417
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.254685631293957e-07,
+      "loss": 0.4244,
+      "step": 1418
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.175461610291913e-07,
+      "loss": 0.3299,
+      "step": 1419
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.096603393667978e-07,
+      "loss": 0.352,
+      "step": 1420
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.018111295572118e-07,
+      "loss": 0.377,
+      "step": 1421
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.939985628695834e-07,
+      "loss": 0.3259,
+      "step": 1422
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.862226704270904e-07,
+      "loss": 0.3057,
+      "step": 1423
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.784834832068044e-07,
+      "loss": 0.3481,
+      "step": 1424
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.707810320395782e-07,
+      "loss": 0.3678,
+      "step": 1425
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.631153476099096e-07,
+      "loss": 0.3916,
+      "step": 1426
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.554864604558376e-07,
+      "loss": 0.3691,
+      "step": 1427
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.478944009688061e-07,
+      "loss": 0.4449,
+      "step": 1428
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.40339199393546e-07,
+      "loss": 0.3588,
+      "step": 1429
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.328208858279562e-07,
+      "loss": 0.3652,
+      "step": 1430
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.253394902229905e-07,
+      "loss": 0.3934,
+      "step": 1431
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.178950423825215e-07,
+      "loss": 0.3475,
+      "step": 1432
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.104875719632365e-07,
+      "loss": 0.383,
+      "step": 1433
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.03117108474517e-07,
+      "loss": 0.3837,
+      "step": 1434
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.957836812783158e-07,
+      "loss": 0.3819,
+      "step": 1435
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.884873195890418e-07,
+      "loss": 0.3368,
+      "step": 1436
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.812280524734483e-07,
+      "loss": 0.338,
+      "step": 1437
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.740059088505124e-07,
+      "loss": 0.4005,
+      "step": 1438
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.668209174913187e-07,
+      "loss": 0.4049,
+      "step": 1439
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.596731070189488e-07,
+      "loss": 0.3368,
+      "step": 1440
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.525625059083707e-07,
+      "loss": 0.3619,
+      "step": 1441
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.454891424863107e-07,
+      "loss": 0.4276,
+      "step": 1442
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.384530449311588e-07,
+      "loss": 0.3828,
+      "step": 1443
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.314542412728408e-07,
+      "loss": 0.4158,
+      "step": 1444
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.244927593927186e-07,
+      "loss": 0.3289,
+      "step": 1445
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.175686270234715e-07,
+      "loss": 0.2985,
+      "step": 1446
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.106818717489882e-07,
+      "loss": 0.3786,
+      "step": 1447
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.038325210042629e-07,
+      "loss": 0.3404,
+      "step": 1448
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.9702060207527e-07,
+      "loss": 0.4017,
+      "step": 1449
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.902461420988703e-07,
+      "loss": 0.3857,
+      "step": 1450
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.835091680627025e-07,
+      "loss": 0.3365,
+      "step": 1451
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.768097068050671e-07,
+      "loss": 0.2921,
+      "step": 1452
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.701477850148218e-07,
+      "loss": 0.3199,
+      "step": 1453
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.635234292312819e-07,
+      "loss": 0.3559,
+      "step": 1454
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.569366658441055e-07,
+      "loss": 0.4253,
+      "step": 1455
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.503875210931964e-07,
+      "loss": 0.3356,
+      "step": 1456
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.4387602106859e-07,
+      "loss": 0.3894,
+      "step": 1457
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.374021917103644e-07,
+      "loss": 0.3589,
+      "step": 1458
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.309660588085208e-07,
+      "loss": 0.4043,
+      "step": 1459
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.245676480028905e-07,
+      "loss": 0.3729,
+      "step": 1460
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.18206984783024e-07,
+      "loss": 0.3669,
+      "step": 1461
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.118840944881076e-07,
+      "loss": 0.3923,
+      "step": 1462
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.055990023068391e-07,
+      "loss": 0.3281,
+      "step": 1463
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.993517332773457e-07,
+      "loss": 0.3242,
+      "step": 1464
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.931423122870716e-07,
+      "loss": 0.3832,
+      "step": 1465
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.869707640726917e-07,
+      "loss": 0.2878,
+      "step": 1466
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.808371132199985e-07,
+      "loss": 0.3716,
+      "step": 1467
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.74741384163816e-07,
+      "loss": 0.3555,
+      "step": 1468
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.68683601187897e-07,
+      "loss": 0.4295,
+      "step": 1469
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.626637884248286e-07,
+      "loss": 0.4144,
+      "step": 1470
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.5668196985593106e-07,
+      "loss": 0.364,
+      "step": 1471
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.507381693111679e-07,
+      "loss": 0.3322,
+      "step": 1472
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.448324104690516e-07,
+      "loss": 0.4534,
+      "step": 1473
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.389647168565425e-07,
+      "loss": 0.3736,
+      "step": 1474
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.331351118489602e-07,
+      "loss": 0.3948,
+      "step": 1475
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.2734361866989317e-07,
+      "loss": 0.4038,
+      "step": 1476
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.215902603910993e-07,
+      "loss": 0.3571,
+      "step": 1477
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.158750599324135e-07,
+      "loss": 0.4006,
+      "step": 1478
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1019804006166897e-07,
+      "loss": 0.4315,
+      "step": 1479
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.0455922339459274e-07,
+      "loss": 0.3438,
+      "step": 1480
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9895863239471924e-07,
+      "loss": 0.3494,
+      "step": 1481
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.933962893733023e-07,
+      "loss": 0.3395,
+      "step": 1482
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.8787221648923123e-07,
+      "loss": 0.4443,
+      "step": 1483
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.823864357489315e-07,
+      "loss": 0.3856,
+      "step": 1484
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.7693896900628303e-07,
+      "loss": 0.3504,
+      "step": 1485
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.715298379625376e-07,
+      "loss": 0.3162,
+      "step": 1486
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.661590641662249e-07,
+      "loss": 0.4297,
+      "step": 1487
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6082666901306905e-07,
+      "loss": 0.4059,
+      "step": 1488
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.555326737459064e-07,
+      "loss": 0.3627,
+      "step": 1489
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5027709945459453e-07,
+      "loss": 0.3693,
+      "step": 1490
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.450599670759347e-07,
+      "loss": 0.4464,
+      "step": 1491
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.3988129739358613e-07,
+      "loss": 0.3914,
+      "step": 1492
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.347411110379839e-07,
+      "loss": 0.3807,
+      "step": 1493
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2963942848625363e-07,
+      "loss": 0.3327,
+      "step": 1494
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2457627006213133e-07,
+      "loss": 0.3217,
+      "step": 1495
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.195516559358847e-07,
+      "loss": 0.2933,
+      "step": 1496
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1456560612423304e-07,
+      "loss": 0.3541,
+      "step": 1497
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0961814049026316e-07,
+      "loss": 0.375,
+      "step": 1498
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.047092787433525e-07,
+      "loss": 0.355,
+      "step": 1499
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9983904043909386e-07,
+      "loss": 0.3631,
+      "step": 1500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9500744497921285e-07,
+      "loss": 0.3339,
+      "step": 1501
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.902145116114907e-07,
+      "loss": 0.3562,
+      "step": 1502
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.854602594296896e-07,
+      "loss": 0.4104,
+      "step": 1503
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.807447073734815e-07,
+      "loss": 0.3469,
+      "step": 1504
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.760678742283618e-07,
+      "loss": 0.3024,
+      "step": 1505
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.714297786255804e-07,
+      "loss": 0.3713,
+      "step": 1506
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.668304390420695e-07,
+      "loss": 0.3698,
+      "step": 1507
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.622698738003659e-07,
+      "loss": 0.3458,
+      "step": 1508
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.577481010685423e-07,
+      "loss": 0.4177,
+      "step": 1509
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5326513886012705e-07,
+      "loss": 0.3898,
+      "step": 1510
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.488210050340467e-07,
+      "loss": 0.3943,
+      "step": 1511
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.4441571729453717e-07,
+      "loss": 0.3362,
+      "step": 1512
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.400492931910847e-07,
+      "loss": 0.4642,
+      "step": 1513
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.357217501183573e-07,
+      "loss": 0.3622,
+      "step": 1514
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3143310531612563e-07,
+      "loss": 0.4205,
+      "step": 1515
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2718337586920214e-07,
+      "loss": 0.3596,
+      "step": 1516
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2297257870737333e-07,
+      "loss": 0.4084,
+      "step": 1517
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1880073060532524e-07,
+      "loss": 0.3681,
+      "step": 1518
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.146678481825859e-07,
+      "loss": 0.3971,
+      "step": 1519
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1057394790345075e-07,
+      "loss": 0.368,
+      "step": 1520
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0651904607692396e-07,
+      "loss": 0.3651,
+      "step": 1521
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.025031588566484e-07,
+      "loss": 0.2849,
+      "step": 1522
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.9852630224084458e-07,
+      "loss": 0.3032,
+      "step": 1523
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9458849207224183e-07,
+      "loss": 0.3459,
+      "step": 1524
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.90689744038024e-07,
+      "loss": 0.3957,
+      "step": 1525
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8683007366975926e-07,
+      "loss": 0.335,
+      "step": 1526
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8300949634334043e-07,
+      "loss": 0.3385,
+      "step": 1527
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7922802727892375e-07,
+      "loss": 0.336,
+      "step": 1528
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.754856815408723e-07,
+      "loss": 0.3635,
+      "step": 1529
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7178247403768833e-07,
+      "loss": 0.3704,
+      "step": 1530
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6811841952195762e-07,
+      "loss": 0.3829,
+      "step": 1531
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.644935325902941e-07,
+      "loss": 0.3805,
+      "step": 1532
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.609078276832776e-07,
+      "loss": 0.3746,
+      "step": 1533
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.57361319085394e-07,
+      "loss": 0.3246,
+      "step": 1534
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5385402092498726e-07,
+      "loss": 0.3587,
+      "step": 1535
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5038594717419198e-07,
+      "loss": 0.3741,
+      "step": 1536
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.469571116488855e-07,
+      "loss": 0.4023,
+      "step": 1537
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4356752800863017e-07,
+      "loss": 0.3299,
+      "step": 1538
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4021720975662235e-07,
+      "loss": 0.3417,
+      "step": 1539
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3690617023963237e-07,
+      "loss": 0.3374,
+      "step": 1540
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3363442264795245e-07,
+      "loss": 0.3154,
+      "step": 1541
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.304019800153522e-07,
+      "loss": 0.3652,
+      "step": 1542
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.272088552190165e-07,
+      "loss": 0.4132,
+      "step": 1543
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2405506097949994e-07,
+      "loss": 0.3709,
+      "step": 1544
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2094060986067246e-07,
+      "loss": 0.3014,
+      "step": 1545
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1786551426967719e-07,
+      "loss": 0.3393,
+      "step": 1546
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1482978645686926e-07,
+      "loss": 0.3739,
+      "step": 1547
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.11833438515776e-07,
+      "loss": 0.3939,
+      "step": 1548
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.088764823830457e-07,
+      "loss": 0.3739,
+      "step": 1549
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0595892983839895e-07,
+      "loss": 0.3056,
+      "step": 1550
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0308079250458624e-07,
+      "loss": 0.3051,
+      "step": 1551
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0024208184733374e-07,
+      "loss": 0.3706,
+      "step": 1552
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.744280917530657e-08,
+      "loss": 0.3603,
+      "step": 1553
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.468298564005662e-08,
+      "loss": 0.4334,
+      "step": 1554
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.196262223598152e-08,
+      "loss": 0.4233,
+      "step": 1555
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.928172980028238e-08,
+      "loss": 0.4457,
+      "step": 1556
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.664031901291725e-08,
+      "loss": 0.3682,
+      "step": 1557
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.403840039655775e-08,
+      "loss": 0.3381,
+      "step": 1558
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.147598431655024e-08,
+      "loss": 0.3578,
+      "step": 1559
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.895308098087696e-08,
+      "loss": 0.37,
+      "step": 1560
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.646970044010827e-08,
+      "loss": 0.3783,
+      "step": 1561
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.402585258736716e-08,
+      "loss": 0.3556,
+      "step": 1562
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.16215471582904e-08,
+      "loss": 0.3604,
+      "step": 1563
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.925679373098514e-08,
+      "loss": 0.4096,
+      "step": 1564
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.6931601725998e-08,
+      "loss": 0.3928,
+      "step": 1565
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.464598040626935e-08,
+      "loss": 0.3243,
+      "step": 1566
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.23999388771046e-08,
+      "loss": 0.3489,
+      "step": 1567
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.019348608613195e-08,
+      "loss": 0.4009,
+      "step": 1568
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.802663082326687e-08,
+      "loss": 0.4111,
+      "step": 1569
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.589938172068099e-08,
+      "loss": 0.4222,
+      "step": 1570
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.381174725276661e-08,
+      "loss": 0.3951,
+      "step": 1571
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.176373573609894e-08,
+      "loss": 0.3653,
+      "step": 1572
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.975535532940834e-08,
+      "loss": 0.4385,
+      "step": 1573
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.77866140335459e-08,
+      "loss": 0.4403,
+      "step": 1574
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.585751969144792e-08,
+      "loss": 0.3175,
+      "step": 1575
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.396807998811037e-08,
+      "loss": 0.3328,
+      "step": 1576
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2118302450555596e-08,
+      "loss": 0.3713,
+      "step": 1577
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.030819444780121e-08,
+      "loss": 0.355,
+      "step": 1578
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.8537763190834575e-08,
+      "loss": 0.2843,
+      "step": 1579
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.680701573257839e-08,
+      "loss": 0.3338,
+      "step": 1580
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.511595896786735e-08,
+      "loss": 0.3651,
+      "step": 1581
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.34645996334193e-08,
+      "loss": 0.3487,
+      "step": 1582
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.18529443078075e-08,
+      "loss": 0.4208,
+      "step": 1583
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.028099941143614e-08,
+      "loss": 0.368,
+      "step": 1584
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.8748771206511537e-08,
+      "loss": 0.3403,
+      "step": 1585
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.725626579702101e-08,
+      "loss": 0.3242,
+      "step": 1586
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.5803489128707338e-08,
+      "loss": 0.3641,
+      "step": 1587
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.4390446989043247e-08,
+      "loss": 0.4075,
+      "step": 1588
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.3017145007210305e-08,
+      "loss": 0.3625,
+      "step": 1589
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1683588654074496e-08,
+      "loss": 0.4343,
+      "step": 1590
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.038978324217067e-08,
+      "loss": 0.3669,
+      "step": 1591
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.9135733925671474e-08,
+      "loss": 0.3339,
+      "step": 1592
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7921445700371798e-08,
+      "loss": 0.3553,
+      "step": 1593
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.6746923403674343e-08,
+      "loss": 0.3987,
+      "step": 1594
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.5612171714558532e-08,
+      "loss": 0.3964,
+      "step": 1595
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4517195153572749e-08,
+      "loss": 0.3669,
+      "step": 1596
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3461998082811013e-08,
+      "loss": 0.3538,
+      "step": 1597
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2446584705894104e-08,
+      "loss": 0.3574,
+      "step": 1598
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.1470959067958477e-08,
+      "loss": 0.3754,
+      "step": 1599
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0535125055636252e-08,
+      "loss": 0.3525,
+      "step": 1600
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.639086397037478e-09,
+      "loss": 0.3646,
+      "step": 1601
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.782846661741228e-09,
+      "loss": 0.3331,
+      "step": 1602
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.966409260777851e-09,
+      "loss": 0.3746,
+      "step": 1603
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.189777446614532e-09,
+      "loss": 0.4783,
+      "step": 1604
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.452954313145299e-09,
+      "loss": 0.3464,
+      "step": 1605
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.755942795675484e-09,
+      "loss": 0.4212,
+      "step": 1606
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.098745670913951e-09,
+      "loss": 0.4492,
+      "step": 1607
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.4813655569553304e-09,
+      "loss": 0.4363,
+      "step": 1608
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.903804913280018e-09,
+      "loss": 0.4263,
+      "step": 1609
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.3660660407319745e-09,
+      "loss": 0.3727,
+      "step": 1610
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.8681510815209425e-09,
+      "loss": 0.3278,
+      "step": 1611
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.4100620192057943e-09,
+      "loss": 0.3629,
+      "step": 1612
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.9918006786912024e-09,
+      "loss": 0.3868,
+      "step": 1613
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.613368726218756e-09,
+      "loss": 0.4068,
+      "step": 1614
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2747676693591893e-09,
+      "loss": 0.3861,
+      "step": 1615
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.759988570079426e-10,
+      "loss": 0.4011,
+      "step": 1616
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.170634793818298e-10,
+      "loss": 0.3036,
+      "step": 1617
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.979625680090472e-10,
+      "loss": 0.3637,
+      "step": 1618
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.1869699572806276e-10,
+      "loss": 0.3822,
+      "step": 1619
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.7926747668650656e-10,
+      "loss": 0.3634,
+      "step": 1620
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.967456633339866e-11,
+      "loss": 0.398,
+      "step": 1621
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.991866142136978e-11,
+      "loss": 0.3167,
+      "step": 1622
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.4196,
+      "step": 1623
+    },
+    {
+      "epoch": 3.0,
+      "step": 1623,
+      "total_flos": 2.7187427763499827e+17,
+      "train_loss": 0.723628326029375,
+      "train_runtime": 22133.6544,
+      "train_samples_per_second": 7.048,
+      "train_steps_per_second": 0.073
+    }
+  ],
+  "max_steps": 1623,
+  "num_train_epochs": 3,
+  "total_flos": 2.7187427763499827e+17,
+  "trial_name": null,
+  "trial_params": null
+}