diff --git "a/checkpoint-1484/trainer_state.json" "b/checkpoint-1484/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1484/trainer_state.json"
@@ -0,0 +1,8923 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1484,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.462686567164179e-08,
+      "loss": 0.6423,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 0.6177,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2388059701492537e-07,
+      "loss": 0.5907,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 0.6307,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7313432835820895e-07,
+      "loss": 0.721,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 0.6646,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.223880597014925e-07,
+      "loss": 0.66,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 0.6376,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.716417910447762e-07,
+      "loss": 0.6132,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 0.633,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.208955223880598e-07,
+      "loss": 0.6168,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.5818,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.701492537313434e-07,
+      "loss": 0.724,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 0.5968,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.119402985074627e-06,
+      "loss": 0.5895,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 0.5987,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2686567164179105e-06,
+      "loss": 0.6585,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 0.5707,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.417910447761194e-06,
+      "loss": 0.5809,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 0.6062,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5671641791044779e-06,
+      "loss": 0.6142,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 0.581,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7164179104477613e-06,
+      "loss": 0.6735,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 0.6011,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.865671641791045e-06,
+      "loss": 0.5916,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 0.6114,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0149253731343284e-06,
+      "loss": 0.5439,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.6002,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.1641791044776118e-06,
+      "loss": 0.6543,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 0.668,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.3134328358208956e-06,
+      "loss": 0.5723,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.6427,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.4626865671641794e-06,
+      "loss": 0.6142,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 0.5646,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.6119402985074627e-06,
+      "loss": 0.5973,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.5616,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.7611940298507465e-06,
+      "loss": 0.5881,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 0.576,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9104477611940303e-06,
+      "loss": 0.679,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.5761,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.0597014925373137e-06,
+      "loss": 0.6966,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.58,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.208955223880597e-06,
+      "loss": 0.5924,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 0.5976,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.3582089552238813e-06,
+      "loss": 0.6475,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 0.56,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.5074626865671646e-06,
+      "loss": 0.6735,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.5665,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.656716417910448e-06,
+      "loss": 0.6151,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.5687,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.8059701492537314e-06,
+      "loss": 0.5896,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 0.5482,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.955223880597015e-06,
+      "loss": 0.5854,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.543,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.104477611940299e-06,
+      "loss": 0.5193,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.5576,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.253731343283583e-06,
+      "loss": 0.6328,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.543,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.402985074626866e-06,
+      "loss": 0.5874,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 0.6383,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.5522388059701495e-06,
+      "loss": 0.5593,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 0.6524,
+      "step": 62
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.701492537313434e-06,
+      "loss": 0.614,
+      "step": 63
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.5952,
+      "step": 64
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.850746268656717e-06,
+      "loss": 0.5631,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.6144,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-06,
+      "loss": 0.634,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.999997353301952e-06,
+      "loss": 0.5947,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.999989413213412e-06,
+      "loss": 0.6104,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.999976179751192e-06,
+      "loss": 0.5968,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999957652943311e-06,
+      "loss": 0.6113,
+      "step": 71
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999933832828997e-06,
+      "loss": 0.6135,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999904719458687e-06,
+      "loss": 0.6171,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999870312894024e-06,
+      "loss": 0.5515,
+      "step": 74
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9998306132078584e-06,
+      "loss": 0.6273,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999785620484249e-06,
+      "loss": 0.6061,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.999735334818462e-06,
+      "loss": 0.5914,
+      "step": 77
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9996797563169695e-06,
+      "loss": 0.606,
+      "step": 78
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999618885097451e-06,
+      "loss": 0.5641,
+      "step": 79
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999552721288794e-06,
+      "loss": 0.5109,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.99948126503109e-06,
+      "loss": 0.5312,
+      "step": 81
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999404516475637e-06,
+      "loss": 0.6127,
+      "step": 82
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.99932247578494e-06,
+      "loss": 0.5973,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999235143132708e-06,
+      "loss": 0.6037,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.999142518703857e-06,
+      "loss": 0.576,
+      "step": 85
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.999044602694504e-06,
+      "loss": 0.5563,
+      "step": 86
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.998941395311973e-06,
+      "loss": 0.6013,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.998832896774791e-06,
+      "loss": 0.5999,
+      "step": 88
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.99871910731269e-06,
+      "loss": 0.5908,
+      "step": 89
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.998600027166601e-06,
+      "loss": 0.6267,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.99847565658866e-06,
+      "loss": 0.5976,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.998345995842204e-06,
+      "loss": 0.5706,
+      "step": 92
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.998211045201771e-06,
+      "loss": 0.5906,
+      "step": 93
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.998070804953101e-06,
+      "loss": 0.5676,
+      "step": 94
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997925275393132e-06,
+      "loss": 0.5702,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997774456830002e-06,
+      "loss": 0.5529,
+      "step": 96
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997618349583049e-06,
+      "loss": 0.5694,
+      "step": 97
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9974569539828064e-06,
+      "loss": 0.5443,
+      "step": 98
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997290270371008e-06,
+      "loss": 0.5213,
+      "step": 99
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.997118299100582e-06,
+      "loss": 0.5883,
+      "step": 100
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.996941040535654e-06,
+      "loss": 0.5835,
+      "step": 101
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.996758495051542e-06,
+      "loss": 0.5756,
+      "step": 102
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.996570663034764e-06,
+      "loss": 0.5719,
+      "step": 103
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9963775448830225e-06,
+      "loss": 0.6012,
+      "step": 104
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.996179141005222e-06,
+      "loss": 0.6051,
+      "step": 105
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9959754518214535e-06,
+      "loss": 0.5905,
+      "step": 106
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.995766477762999e-06,
+      "loss": 0.5492,
+      "step": 107
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.995552219272333e-06,
+      "loss": 0.6055,
+      "step": 108
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.995332676803117e-06,
+      "loss": 0.5788,
+      "step": 109
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.995107850820199e-06,
+      "loss": 0.5784,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.994877741799619e-06,
+      "loss": 0.6152,
+      "step": 111
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.994642350228601e-06,
+      "loss": 0.6223,
+      "step": 112
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.99440167660555e-06,
+      "loss": 0.478,
+      "step": 113
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.99415572144006e-06,
+      "loss": 0.5487,
+      "step": 114
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.993904485252907e-06,
+      "loss": 0.6073,
+      "step": 115
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.993647968576047e-06,
+      "loss": 0.6092,
+      "step": 116
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9933861719526176e-06,
+      "loss": 0.5703,
+      "step": 117
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.993119095936937e-06,
+      "loss": 0.6088,
+      "step": 118
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9928467410945e-06,
+      "loss": 0.5716,
+      "step": 119
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.992569108001981e-06,
+      "loss": 0.5494,
+      "step": 120
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9922861972472255e-06,
+      "loss": 0.5917,
+      "step": 121
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.991998009429261e-06,
+      "loss": 0.6129,
+      "step": 122
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9917045451582815e-06,
+      "loss": 0.593,
+      "step": 123
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.991405805055658e-06,
+      "loss": 0.5571,
+      "step": 124
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.991101789753928e-06,
+      "loss": 0.5656,
+      "step": 125
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9907924998968035e-06,
+      "loss": 0.5481,
+      "step": 126
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.990477936139159e-06,
+      "loss": 0.6061,
+      "step": 127
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.990158099147043e-06,
+      "loss": 0.5552,
+      "step": 128
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.989832989597661e-06,
+      "loss": 0.5966,
+      "step": 129
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9895026081793884e-06,
+      "loss": 0.5445,
+      "step": 130
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.98916695559176e-06,
+      "loss": 0.5757,
+      "step": 131
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.988826032545474e-06,
+      "loss": 0.5566,
+      "step": 132
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9884798397623856e-06,
+      "loss": 0.5119,
+      "step": 133
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.98812837797551e-06,
+      "loss": 0.5603,
+      "step": 134
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.987771647929017e-06,
+      "loss": 0.5771,
+      "step": 135
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.987409650378232e-06,
+      "loss": 0.6126,
+      "step": 136
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.987042386089633e-06,
+      "loss": 0.5963,
+      "step": 137
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.986669855840852e-06,
+      "loss": 0.5413,
+      "step": 138
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9862920604206675e-06,
+      "loss": 0.5406,
+      "step": 139
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.985909000629009e-06,
+      "loss": 0.6792,
+      "step": 140
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.98552067727695e-06,
+      "loss": 0.5665,
+      "step": 141
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.98512709118671e-06,
+      "loss": 0.5529,
+      "step": 142
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.984728243191653e-06,
+      "loss": 0.5199,
+      "step": 143
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.984324134136284e-06,
+      "loss": 0.5718,
+      "step": 144
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.983914764876245e-06,
+      "loss": 0.5387,
+      "step": 145
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9835001362783175e-06,
+      "loss": 0.6141,
+      "step": 146
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.98308024922042e-06,
+      "loss": 0.5685,
+      "step": 147
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.982655104591604e-06,
+      "loss": 0.5851,
+      "step": 148
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.982224703292051e-06,
+      "loss": 0.5623,
+      "step": 149
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.981789046233076e-06,
+      "loss": 0.5067,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.981348134337123e-06,
+      "loss": 0.6005,
+      "step": 151
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.980901968537758e-06,
+      "loss": 0.6434,
+      "step": 152
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.980450549779675e-06,
+      "loss": 0.659,
+      "step": 153
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.979993879018688e-06,
+      "loss": 0.5762,
+      "step": 154
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9795319572217356e-06,
+      "loss": 0.5527,
+      "step": 155
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.9790647853668686e-06,
+      "loss": 0.5685,
+      "step": 156
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.978592364443259e-06,
+      "loss": 0.6348,
+      "step": 157
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.978114695451192e-06,
+      "loss": 0.6047,
+      "step": 158
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.977631779402061e-06,
+      "loss": 0.6143,
+      "step": 159
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.977143617318376e-06,
+      "loss": 0.6329,
+      "step": 160
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.976650210233749e-06,
+      "loss": 0.581,
+      "step": 161
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9761515591929e-06,
+      "loss": 0.5456,
+      "step": 162
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.975647665251652e-06,
+      "loss": 0.518,
+      "step": 163
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.975138529476929e-06,
+      "loss": 0.5648,
+      "step": 164
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.974624152946755e-06,
+      "loss": 0.5921,
+      "step": 165
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9741045367502475e-06,
+      "loss": 0.5568,
+      "step": 166
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.973579681987622e-06,
+      "loss": 0.4872,
+      "step": 167
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.973049589770183e-06,
+      "loss": 0.6711,
+      "step": 168
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.972514261220327e-06,
+      "loss": 0.6142,
+      "step": 169
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.971973697471535e-06,
+      "loss": 0.6344,
+      "step": 170
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.971427899668376e-06,
+      "loss": 0.6002,
+      "step": 171
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.970876868966497e-06,
+      "loss": 0.5672,
+      "step": 172
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.97032060653263e-06,
+      "loss": 0.6347,
+      "step": 173
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.969759113544581e-06,
+      "loss": 0.5874,
+      "step": 174
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.969192391191231e-06,
+      "loss": 0.501,
+      "step": 175
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9686204406725355e-06,
+      "loss": 0.5753,
+      "step": 176
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.968043263199518e-06,
+      "loss": 0.5443,
+      "step": 177
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.967460859994272e-06,
+      "loss": 0.5693,
+      "step": 178
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.966873232289951e-06,
+      "loss": 0.5469,
+      "step": 179
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.966280381330775e-06,
+      "loss": 0.5876,
+      "step": 180
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.965682308372022e-06,
+      "loss": 0.5709,
+      "step": 181
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.965079014680027e-06,
+      "loss": 0.5215,
+      "step": 182
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.964470501532179e-06,
+      "loss": 0.5841,
+      "step": 183
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.963856770216917e-06,
+      "loss": 0.6485,
+      "step": 184
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.963237822033733e-06,
+      "loss": 0.6201,
+      "step": 185
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.962613658293158e-06,
+      "loss": 0.5295,
+      "step": 186
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.961984280316775e-06,
+      "loss": 0.6078,
+      "step": 187
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.961349689437199e-06,
+      "loss": 0.6665,
+      "step": 188
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.96070988699809e-06,
+      "loss": 0.5529,
+      "step": 189
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.960064874354136e-06,
+      "loss": 0.5783,
+      "step": 190
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9594146528710596e-06,
+      "loss": 0.5181,
+      "step": 191
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9587592239256154e-06,
+      "loss": 0.5941,
+      "step": 192
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.95809858890558e-06,
+      "loss": 0.5932,
+      "step": 193
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.957432749209755e-06,
+      "loss": 0.601,
+      "step": 194
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.956761706247961e-06,
+      "loss": 0.6308,
+      "step": 195
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.956085461441037e-06,
+      "loss": 0.5827,
+      "step": 196
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.955404016220835e-06,
+      "loss": 0.5602,
+      "step": 197
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.95471737203022e-06,
+      "loss": 0.5478,
+      "step": 198
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.954025530323063e-06,
+      "loss": 0.5433,
+      "step": 199
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.953328492564241e-06,
+      "loss": 0.5952,
+      "step": 200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.952626260229633e-06,
+      "loss": 0.6151,
+      "step": 201
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.951918834806116e-06,
+      "loss": 0.5552,
+      "step": 202
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.9512062177915634e-06,
+      "loss": 0.5577,
+      "step": 203
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.950488410694842e-06,
+      "loss": 0.6224,
+      "step": 204
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.949765415035805e-06,
+      "loss": 0.5383,
+      "step": 205
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.949037232345294e-06,
+      "loss": 0.6319,
+      "step": 206
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.948303864165133e-06,
+      "loss": 0.6086,
+      "step": 207
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.947565312048125e-06,
+      "loss": 0.5733,
+      "step": 208
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.94682157755805e-06,
+      "loss": 0.5885,
+      "step": 209
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.946072662269661e-06,
+      "loss": 0.5965,
+      "step": 210
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.945318567768677e-06,
+      "loss": 0.5688,
+      "step": 211
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9445592956517905e-06,
+      "loss": 0.6098,
+      "step": 212
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.943794847526651e-06,
+      "loss": 0.6059,
+      "step": 213
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.943025225011869e-06,
+      "loss": 0.5723,
+      "step": 214
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.942250429737011e-06,
+      "loss": 0.6188,
+      "step": 215
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.941470463342596e-06,
+      "loss": 0.5953,
+      "step": 216
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.940685327480094e-06,
+      "loss": 0.5418,
+      "step": 217
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.939895023811918e-06,
+      "loss": 0.5664,
+      "step": 218
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.939099554011424e-06,
+      "loss": 0.6042,
+      "step": 219
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.938298919762907e-06,
+      "loss": 0.5681,
+      "step": 220
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.937493122761596e-06,
+      "loss": 0.5376,
+      "step": 221
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.936682164713654e-06,
+      "loss": 0.5843,
+      "step": 222
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.935866047336168e-06,
+      "loss": 0.6206,
+      "step": 223
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.935044772357152e-06,
+      "loss": 0.5741,
+      "step": 224
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.934218341515538e-06,
+      "loss": 0.6445,
+      "step": 225
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.933386756561178e-06,
+      "loss": 0.5918,
+      "step": 226
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.932550019254835e-06,
+      "loss": 0.5168,
+      "step": 227
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.931708131368181e-06,
+      "loss": 0.5812,
+      "step": 228
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9308610946837955e-06,
+      "loss": 0.6094,
+      "step": 229
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.930008910995158e-06,
+      "loss": 0.5939,
+      "step": 230
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9291515821066475e-06,
+      "loss": 0.5721,
+      "step": 231
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.928289109833536e-06,
+      "loss": 0.5902,
+      "step": 232
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.927421496001988e-06,
+      "loss": 0.545,
+      "step": 233
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.92654874244905e-06,
+      "loss": 0.5317,
+      "step": 234
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.925670851022656e-06,
+      "loss": 0.5704,
+      "step": 235
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.924787823581616e-06,
+      "loss": 0.5337,
+      "step": 236
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.923899661995617e-06,
+      "loss": 0.6223,
+      "step": 237
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.923006368145215e-06,
+      "loss": 0.6146,
+      "step": 238
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.9221079439218314e-06,
+      "loss": 0.6047,
+      "step": 239
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.921204391227754e-06,
+      "loss": 0.5593,
+      "step": 240
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.920295711976129e-06,
+      "loss": 0.5095,
+      "step": 241
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9193819080909534e-06,
+      "loss": 0.582,
+      "step": 242
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.918462981507079e-06,
+      "loss": 0.6257,
+      "step": 243
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.917538934170202e-06,
+      "loss": 0.6386,
+      "step": 244
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9166097680368615e-06,
+      "loss": 0.6566,
+      "step": 245
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.915675485074438e-06,
+      "loss": 0.4939,
+      "step": 246
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.91473608726114e-06,
+      "loss": 0.5267,
+      "step": 247
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.913791576586012e-06,
+      "loss": 0.5606,
+      "step": 248
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.91284195504892e-06,
+      "loss": 0.6032,
+      "step": 249
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.911887224660553e-06,
+      "loss": 0.6338,
+      "step": 250
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.91092738744242e-06,
+      "loss": 0.5635,
+      "step": 251
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.9099624454268365e-06,
+      "loss": 0.6003,
+      "step": 252
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.908992400656934e-06,
+      "loss": 0.5308,
+      "step": 253
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.908017255186644e-06,
+      "loss": 0.5695,
+      "step": 254
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.907037011080697e-06,
+      "loss": 0.5377,
+      "step": 255
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.906051670414623e-06,
+      "loss": 0.4869,
+      "step": 256
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.905061235274743e-06,
+      "loss": 0.6222,
+      "step": 257
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.904065707758159e-06,
+      "loss": 0.545,
+      "step": 258
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.9030650899727636e-06,
+      "loss": 0.6023,
+      "step": 259
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.902059384037221e-06,
+      "loss": 0.6098,
+      "step": 260
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.9010485920809715e-06,
+      "loss": 0.6025,
+      "step": 261
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.900032716244225e-06,
+      "loss": 0.5454,
+      "step": 262
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.899011758677954e-06,
+      "loss": 0.5883,
+      "step": 263
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.897985721543891e-06,
+      "loss": 0.5959,
+      "step": 264
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.896954607014526e-06,
+      "loss": 0.5522,
+      "step": 265
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8959184172730955e-06,
+      "loss": 0.5334,
+      "step": 266
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.894877154513588e-06,
+      "loss": 0.6167,
+      "step": 267
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.893830820940727e-06,
+      "loss": 0.537,
+      "step": 268
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.892779418769977e-06,
+      "loss": 0.6554,
+      "step": 269
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.891722950227534e-06,
+      "loss": 0.5863,
+      "step": 270
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8906614175503196e-06,
+      "loss": 0.57,
+      "step": 271
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.889594822985978e-06,
+      "loss": 0.5661,
+      "step": 272
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8885231687928746e-06,
+      "loss": 0.6108,
+      "step": 273
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.887446457240083e-06,
+      "loss": 0.5673,
+      "step": 274
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8863646906073895e-06,
+      "loss": 0.5654,
+      "step": 275
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.885277871185281e-06,
+      "loss": 0.5476,
+      "step": 276
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.884186001274944e-06,
+      "loss": 0.5836,
+      "step": 277
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.8830890831882575e-06,
+      "loss": 0.5896,
+      "step": 278
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.881987119247791e-06,
+      "loss": 0.5937,
+      "step": 279
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.8808801117867976e-06,
+      "loss": 0.5862,
+      "step": 280
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.879768063149208e-06,
+      "loss": 0.574,
+      "step": 281
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.878650975689628e-06,
+      "loss": 0.5378,
+      "step": 282
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.8775288517733335e-06,
+      "loss": 0.5579,
+      "step": 283
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.876401693776261e-06,
+      "loss": 0.5522,
+      "step": 284
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.875269504085009e-06,
+      "loss": 0.5588,
+      "step": 285
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.874132285096828e-06,
+      "loss": 0.5635,
+      "step": 286
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.87299003921962e-06,
+      "loss": 0.571,
+      "step": 287
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.871842768871928e-06,
+      "loss": 0.5763,
+      "step": 288
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.8706904764829345e-06,
+      "loss": 0.5428,
+      "step": 289
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.869533164492456e-06,
+      "loss": 0.6225,
+      "step": 290
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.868370835350935e-06,
+      "loss": 0.6056,
+      "step": 291
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.867203491519441e-06,
+      "loss": 0.6129,
+      "step": 292
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.866031135469657e-06,
+      "loss": 0.5854,
+      "step": 293
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.864853769683885e-06,
+      "loss": 0.556,
+      "step": 294
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.863671396655026e-06,
+      "loss": 0.571,
+      "step": 295
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.862484018886591e-06,
+      "loss": 0.5766,
+      "step": 296
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.861291638892682e-06,
+      "loss": 0.5975,
+      "step": 297
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.8600942591979955e-06,
+      "loss": 0.5632,
+      "step": 298
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.858891882337814e-06,
+      "loss": 0.5562,
+      "step": 299
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.8576845108579995e-06,
+      "loss": 0.5462,
+      "step": 300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.85647214731499e-06,
+      "loss": 0.5241,
+      "step": 301
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.855254794275796e-06,
+      "loss": 0.5706,
+      "step": 302
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.854032454317987e-06,
+      "loss": 0.5534,
+      "step": 303
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.852805130029696e-06,
+      "loss": 0.5394,
+      "step": 304
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.85157282400961e-06,
+      "loss": 0.5586,
+      "step": 305
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.850335538866961e-06,
+      "loss": 0.5726,
+      "step": 306
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.849093277221525e-06,
+      "loss": 0.5541,
+      "step": 307
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.847846041703616e-06,
+      "loss": 0.5826,
+      "step": 308
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.846593834954078e-06,
+      "loss": 0.5864,
+      "step": 309
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.845336659624282e-06,
+      "loss": 0.544,
+      "step": 310
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.844074518376118e-06,
+      "loss": 0.5671,
+      "step": 311
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.842807413881993e-06,
+      "loss": 0.5405,
+      "step": 312
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.84153534882482e-06,
+      "loss": 0.6019,
+      "step": 313
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.840258325898017e-06,
+      "loss": 0.5412,
+      "step": 314
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8389763478055e-06,
+      "loss": 0.5927,
+      "step": 315
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.837689417261674e-06,
+      "loss": 0.6319,
+      "step": 316
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.8363975369914355e-06,
+      "loss": 0.5956,
+      "step": 317
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.835100709730155e-06,
+      "loss": 0.613,
+      "step": 318
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.833798938223682e-06,
+      "loss": 0.6687,
+      "step": 319
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.832492225228334e-06,
+      "loss": 0.6312,
+      "step": 320
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.83118057351089e-06,
+      "loss": 0.5668,
+      "step": 321
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.829863985848587e-06,
+      "loss": 0.5697,
+      "step": 322
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.828542465029112e-06,
+      "loss": 0.5391,
+      "step": 323
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.827216013850601e-06,
+      "loss": 0.6234,
+      "step": 324
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.825884635121624e-06,
+      "loss": 0.5776,
+      "step": 325
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.824548331661187e-06,
+      "loss": 0.6005,
+      "step": 326
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.823207106298725e-06,
+      "loss": 0.5165,
+      "step": 327
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.821860961874093e-06,
+      "loss": 0.5904,
+      "step": 328
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.820509901237559e-06,
+      "loss": 0.6053,
+      "step": 329
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.819153927249804e-06,
+      "loss": 0.6007,
+      "step": 330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.817793042781911e-06,
+      "loss": 0.6506,
+      "step": 331
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.8164272507153605e-06,
+      "loss": 0.5246,
+      "step": 332
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.815056553942024e-06,
+      "loss": 0.5676,
+      "step": 333
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.813680955364156e-06,
+      "loss": 0.5418,
+      "step": 334
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.812300457894395e-06,
+      "loss": 0.531,
+      "step": 335
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.810915064455746e-06,
+      "loss": 0.5959,
+      "step": 336
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.809524777981584e-06,
+      "loss": 0.5781,
+      "step": 337
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.808129601415645e-06,
+      "loss": 0.5337,
+      "step": 338
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.806729537712017e-06,
+      "loss": 0.6285,
+      "step": 339
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.805324589835138e-06,
+      "loss": 0.6457,
+      "step": 340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.8039147607597835e-06,
+      "loss": 0.5889,
+      "step": 341
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.80250005347107e-06,
+      "loss": 0.6054,
+      "step": 342
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.801080470964438e-06,
+      "loss": 0.5838,
+      "step": 343
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.799656016245653e-06,
+      "loss": 0.5877,
+      "step": 344
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.798226692330797e-06,
+      "loss": 0.6596,
+      "step": 345
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.79679250224626e-06,
+      "loss": 0.6457,
+      "step": 346
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.795353449028737e-06,
+      "loss": 0.6714,
+      "step": 347
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.793909535725219e-06,
+      "loss": 0.5839,
+      "step": 348
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.792460765392989e-06,
+      "loss": 0.4988,
+      "step": 349
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.791007141099611e-06,
+      "loss": 0.5687,
+      "step": 350
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.789548665922932e-06,
+      "loss": 0.5333,
+      "step": 351
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.788085342951063e-06,
+      "loss": 0.4989,
+      "step": 352
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.786617175282387e-06,
+      "loss": 0.6209,
+      "step": 353
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.785144166025537e-06,
+      "loss": 0.6196,
+      "step": 354
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.7836663182994054e-06,
+      "loss": 0.5705,
+      "step": 355
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.7821836352331235e-06,
+      "loss": 0.6934,
+      "step": 356
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.780696119966064e-06,
+      "loss": 0.5832,
+      "step": 357
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.779203775647828e-06,
+      "loss": 0.5212,
+      "step": 358
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.777706605438246e-06,
+      "loss": 0.5548,
+      "step": 359
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.776204612507361e-06,
+      "loss": 0.5626,
+      "step": 360
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.774697800035433e-06,
+      "loss": 0.585,
+      "step": 361
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.773186171212922e-06,
+      "loss": 0.6399,
+      "step": 362
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.77166972924049e-06,
+      "loss": 0.5562,
+      "step": 363
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.7701484773289874e-06,
+      "loss": 0.5208,
+      "step": 364
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.768622418699449e-06,
+      "loss": 0.5369,
+      "step": 365
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.767091556583089e-06,
+      "loss": 0.5876,
+      "step": 366
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.765555894221291e-06,
+      "loss": 0.6237,
+      "step": 367
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.764015434865601e-06,
+      "loss": 0.6593,
+      "step": 368
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.762470181777726e-06,
+      "loss": 0.5541,
+      "step": 369
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.76092013822952e-06,
+      "loss": 0.578,
+      "step": 370
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.759365307502981e-06,
+      "loss": 0.5523,
+      "step": 371
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.757805692890241e-06,
+      "loss": 0.5682,
+      "step": 372
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.7562412976935655e-06,
+      "loss": 0.5599,
+      "step": 373
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.7546721252253395e-06,
+      "loss": 0.5949,
+      "step": 374
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.753098178808063e-06,
+      "loss": 0.4994,
+      "step": 375
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.751519461774344e-06,
+      "loss": 0.6295,
+      "step": 376
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.7499359774668935e-06,
+      "loss": 0.5813,
+      "step": 377
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.748347729238515e-06,
+      "loss": 0.5691,
+      "step": 378
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.746754720452099e-06,
+      "loss": 0.5352,
+      "step": 379
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.745156954480617e-06,
+      "loss": 0.5713,
+      "step": 380
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.743554434707111e-06,
+      "loss": 0.5313,
+      "step": 381
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.741947164524691e-06,
+      "loss": 0.5207,
+      "step": 382
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.7403351473365235e-06,
+      "loss": 0.6032,
+      "step": 383
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.738718386555826e-06,
+      "loss": 0.5859,
+      "step": 384
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.737096885605862e-06,
+      "loss": 0.5978,
+      "step": 385
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.735470647919928e-06,
+      "loss": 0.5586,
+      "step": 386
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.733839676941355e-06,
+      "loss": 0.5008,
+      "step": 387
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.732203976123492e-06,
+      "loss": 0.6217,
+      "step": 388
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.730563548929703e-06,
+      "loss": 0.5872,
+      "step": 389
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.728918398833361e-06,
+      "loss": 0.5123,
+      "step": 390
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.727268529317839e-06,
+      "loss": 0.5139,
+      "step": 391
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.725613943876502e-06,
+      "loss": 0.6775,
+      "step": 392
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.723954646012699e-06,
+      "loss": 0.6362,
+      "step": 393
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.7222906392397596e-06,
+      "loss": 0.6033,
+      "step": 394
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.720621927080982e-06,
+      "loss": 0.574,
+      "step": 395
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.718948513069629e-06,
+      "loss": 0.5384,
+      "step": 396
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.717270400748917e-06,
+      "loss": 0.5976,
+      "step": 397
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.715587593672012e-06,
+      "loss": 0.6397,
+      "step": 398
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.713900095402019e-06,
+      "loss": 0.5382,
+      "step": 399
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.7122079095119765e-06,
+      "loss": 0.6114,
+      "step": 400
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.710511039584849e-06,
+      "loss": 0.5958,
+      "step": 401
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.708809489213519e-06,
+      "loss": 0.6055,
+      "step": 402
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.707103262000778e-06,
+      "loss": 0.5728,
+      "step": 403
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.70539236155932e-06,
+      "loss": 0.5638,
+      "step": 404
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.7036767915117355e-06,
+      "loss": 0.5813,
+      "step": 405
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.7019565554905e-06,
+      "loss": 0.5263,
+      "step": 406
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.700231657137971e-06,
+      "loss": 0.6219,
+      "step": 407
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.698502100106376e-06,
+      "loss": 0.5896,
+      "step": 408
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.6967678880578075e-06,
+      "loss": 0.5986,
+      "step": 409
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.695029024664214e-06,
+      "loss": 0.5573,
+      "step": 410
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.693285513607392e-06,
+      "loss": 0.6516,
+      "step": 411
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.691537358578979e-06,
+      "loss": 0.6078,
+      "step": 412
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.689784563280447e-06,
+      "loss": 0.5442,
+      "step": 413
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.688027131423092e-06,
+      "loss": 0.5692,
+      "step": 414
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.686265066728026e-06,
+      "loss": 0.6515,
+      "step": 415
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.684498372926172e-06,
+      "loss": 0.5343,
+      "step": 416
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.6827270537582544e-06,
+      "loss": 0.5115,
+      "step": 417
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.680951112974791e-06,
+      "loss": 0.643,
+      "step": 418
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.679170554336085e-06,
+      "loss": 0.5451,
+      "step": 419
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.677385381612215e-06,
+      "loss": 0.5823,
+      "step": 420
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.675595598583035e-06,
+      "loss": 0.5244,
+      "step": 421
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.673801209038154e-06,
+      "loss": 0.5801,
+      "step": 422
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.672002216776941e-06,
+      "loss": 0.5131,
+      "step": 423
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.670198625608504e-06,
+      "loss": 0.6185,
+      "step": 424
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.6683904393516955e-06,
+      "loss": 0.5772,
+      "step": 425
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.666577661835092e-06,
+      "loss": 0.5883,
+      "step": 426
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.664760296896993e-06,
+      "loss": 0.528,
+      "step": 427
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.6629383483854115e-06,
+      "loss": 0.6697,
+      "step": 428
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.661111820158068e-06,
+      "loss": 0.4754,
+      "step": 429
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.6592807160823735e-06,
+      "loss": 0.6839,
+      "step": 430
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.657445040035434e-06,
+      "loss": 0.6277,
+      "step": 431
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.655604795904034e-06,
+      "loss": 0.5867,
+      "step": 432
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.65375998758463e-06,
+      "loss": 0.5332,
+      "step": 433
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.6519106189833405e-06,
+      "loss": 0.6095,
+      "step": 434
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.650056694015943e-06,
+      "loss": 0.532,
+      "step": 435
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.6481982166078615e-06,
+      "loss": 0.5194,
+      "step": 436
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.646335190694159e-06,
+      "loss": 0.5159,
+      "step": 437
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.644467620219527e-06,
+      "loss": 0.6093,
+      "step": 438
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.642595509138285e-06,
+      "loss": 0.6159,
+      "step": 439
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.640718861414361e-06,
+      "loss": 0.5613,
+      "step": 440
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.63883768102129e-06,
+      "loss": 0.528,
+      "step": 441
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.6369519719422075e-06,
+      "loss": 0.5926,
+      "step": 442
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.635061738169835e-06,
+      "loss": 0.6028,
+      "step": 443
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.633166983706474e-06,
+      "loss": 0.6421,
+      "step": 444
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.631267712563998e-06,
+      "loss": 0.5765,
+      "step": 445
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.629363928763848e-06,
+      "loss": 0.5149,
+      "step": 446
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.627455636337014e-06,
+      "loss": 0.5851,
+      "step": 447
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.625542839324036e-06,
+      "loss": 0.5438,
+      "step": 448
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.6236255417749906e-06,
+      "loss": 0.5618,
+      "step": 449
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.621703747749484e-06,
+      "loss": 0.6441,
+      "step": 450
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.619777461316643e-06,
+      "loss": 0.6496,
+      "step": 451
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.617846686555107e-06,
+      "loss": 0.598,
+      "step": 452
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.615911427553017e-06,
+      "loss": 0.6012,
+      "step": 453
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.613971688408011e-06,
+      "loss": 0.5478,
+      "step": 454
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.612027473227212e-06,
+      "loss": 0.6098,
+      "step": 455
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.61007878612722e-06,
+      "loss": 0.6157,
+      "step": 456
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.608125631234104e-06,
+      "loss": 0.5677,
+      "step": 457
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.606168012683394e-06,
+      "loss": 0.582,
+      "step": 458
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.604205934620069e-06,
+      "loss": 0.598,
+      "step": 459
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.602239401198552e-06,
+      "loss": 0.5248,
+      "step": 460
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.6002684165827e-06,
+      "loss": 0.5587,
+      "step": 461
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.5982929849457915e-06,
+      "loss": 0.5792,
+      "step": 462
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.596313110470525e-06,
+      "loss": 0.5679,
+      "step": 463
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.594328797349005e-06,
+      "loss": 0.5206,
+      "step": 464
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.592340049782732e-06,
+      "loss": 0.5738,
+      "step": 465
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.590346871982599e-06,
+      "loss": 0.6298,
+      "step": 466
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.588349268168878e-06,
+      "loss": 0.5276,
+      "step": 467
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.58634724257121e-06,
+      "loss": 0.6081,
+      "step": 468
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5843407994286035e-06,
+      "loss": 0.5262,
+      "step": 469
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.5823299429894155e-06,
+      "loss": 0.519,
+      "step": 470
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.580314677511352e-06,
+      "loss": 0.5797,
+      "step": 471
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.57829500726145e-06,
+      "loss": 0.6002,
+      "step": 472
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.576270936516078e-06,
+      "loss": 0.5263,
+      "step": 473
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.574242469560917e-06,
+      "loss": 0.5529,
+      "step": 474
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.5722096106909595e-06,
+      "loss": 0.5836,
+      "step": 475
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.570172364210497e-06,
+      "loss": 0.5576,
+      "step": 476
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.56813073443311e-06,
+      "loss": 0.5774,
+      "step": 477
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.5660847256816605e-06,
+      "loss": 0.6327,
+      "step": 478
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.564034342288282e-06,
+      "loss": 0.5737,
+      "step": 479
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.561979588594372e-06,
+      "loss": 0.5601,
+      "step": 480
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.55992046895058e-06,
+      "loss": 0.5353,
+      "step": 481
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.5578569877168e-06,
+      "loss": 0.6106,
+      "step": 482
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.555789149262162e-06,
+      "loss": 0.619,
+      "step": 483
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.553716957965021e-06,
+      "loss": 0.544,
+      "step": 484
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.551640418212948e-06,
+      "loss": 0.5549,
+      "step": 485
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.549559534402723e-06,
+      "loss": 0.5328,
+      "step": 486
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.547474310940324e-06,
+      "loss": 0.5771,
+      "step": 487
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.545384752240913e-06,
+      "loss": 0.517,
+      "step": 488
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.543290862728838e-06,
+      "loss": 0.6366,
+      "step": 489
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.541192646837613e-06,
+      "loss": 0.6169,
+      "step": 490
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.539090109009912e-06,
+      "loss": 0.5425,
+      "step": 491
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.536983253697561e-06,
+      "loss": 0.61,
+      "step": 492
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.5348720853615304e-06,
+      "loss": 0.6311,
+      "step": 493
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.532756608471917e-06,
+      "loss": 0.5941,
+      "step": 494
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.530636827507946e-06,
+      "loss": 0.4866,
+      "step": 495
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.528512746957953e-06,
+      "loss": 0.5785,
+      "step": 496
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.526384371319377e-06,
+      "loss": 0.5574,
+      "step": 497
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.524251705098752e-06,
+      "loss": 0.5019,
+      "step": 498
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5221147528116986e-06,
+      "loss": 0.5514,
+      "step": 499
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.5199735189829095e-06,
+      "loss": 0.5541,
+      "step": 500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.517828008146145e-06,
+      "loss": 0.5179,
+      "step": 501
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.515678224844219e-06,
+      "loss": 0.5544,
+      "step": 502
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.513524173628994e-06,
+      "loss": 0.5268,
+      "step": 503
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.51136585906137e-06,
+      "loss": 0.5486,
+      "step": 504
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.50920328571127e-06,
+      "loss": 0.599,
+      "step": 505
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.50703645815764e-06,
+      "loss": 0.5492,
+      "step": 506
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.504865380988428e-06,
+      "loss": 0.5157,
+      "step": 507
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.502690058800583e-06,
+      "loss": 0.5522,
+      "step": 508
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.500510496200043e-06,
+      "loss": 0.6046,
+      "step": 509
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.498326697801723e-06,
+      "loss": 0.6006,
+      "step": 510
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4961386682295045e-06,
+      "loss": 0.5056,
+      "step": 511
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.493946412116235e-06,
+      "loss": 0.5967,
+      "step": 512
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.491749934103702e-06,
+      "loss": 0.5706,
+      "step": 513
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.4895492388426394e-06,
+      "loss": 0.5597,
+      "step": 514
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.487344330992706e-06,
+      "loss": 0.5553,
+      "step": 515
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.485135215222485e-06,
+      "loss": 0.6106,
+      "step": 516
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.482921896209463e-06,
+      "loss": 0.5479,
+      "step": 517
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.480704378640031e-06,
+      "loss": 0.5321,
+      "step": 518
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.478482667209471e-06,
+      "loss": 0.5633,
+      "step": 519
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.476256766621937e-06,
+      "loss": 0.5842,
+      "step": 520
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.474026681590464e-06,
+      "loss": 0.5064,
+      "step": 521
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.471792416836938e-06,
+      "loss": 0.6332,
+      "step": 522
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.4695539770920985e-06,
+      "loss": 0.5234,
+      "step": 523
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.467311367095526e-06,
+      "loss": 0.5326,
+      "step": 524
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.465064591595629e-06,
+      "loss": 0.614,
+      "step": 525
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.462813655349637e-06,
+      "loss": 0.5514,
+      "step": 526
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.4605585631235896e-06,
+      "loss": 0.5385,
+      "step": 527
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.458299319692323e-06,
+      "loss": 0.586,
+      "step": 528
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.456035929839468e-06,
+      "loss": 0.5621,
+      "step": 529
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.45376839835743e-06,
+      "loss": 0.6015,
+      "step": 530
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.451496730047387e-06,
+      "loss": 0.5595,
+      "step": 531
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.449220929719274e-06,
+      "loss": 0.5407,
+      "step": 532
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.4469410021917784e-06,
+      "loss": 0.4871,
+      "step": 533
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.444656952292321e-06,
+      "loss": 0.5569,
+      "step": 534
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.442368784857057e-06,
+      "loss": 0.668,
+      "step": 535
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.440076504730855e-06,
+      "loss": 0.5083,
+      "step": 536
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.4377801167672945e-06,
+      "loss": 0.5741,
+      "step": 537
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.435479625828652e-06,
+      "loss": 0.5866,
+      "step": 538
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.4331750367858906e-06,
+      "loss": 0.5299,
+      "step": 539
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.430866354518653e-06,
+      "loss": 0.525,
+      "step": 540
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.428553583915245e-06,
+      "loss": 0.5077,
+      "step": 541
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.426236729872633e-06,
+      "loss": 0.5841,
+      "step": 542
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.423915797296426e-06,
+      "loss": 0.5103,
+      "step": 543
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.421590791100869e-06,
+      "loss": 0.55,
+      "step": 544
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.419261716208837e-06,
+      "loss": 0.5814,
+      "step": 545
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.416928577551813e-06,
+      "loss": 0.6587,
+      "step": 546
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.414591380069892e-06,
+      "loss": 0.6303,
+      "step": 547
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.412250128711754e-06,
+      "loss": 0.5439,
+      "step": 548
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.409904828434669e-06,
+      "loss": 0.5464,
+      "step": 549
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.4075554842044795e-06,
+      "loss": 0.5362,
+      "step": 550
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.405202100995588e-06,
+      "loss": 0.592,
+      "step": 551
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.402844683790953e-06,
+      "loss": 0.5624,
+      "step": 552
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.400483237582067e-06,
+      "loss": 0.5854,
+      "step": 553
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3981177673689625e-06,
+      "loss": 0.5768,
+      "step": 554
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.395748278160184e-06,
+      "loss": 0.5695,
+      "step": 555
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.393374774972793e-06,
+      "loss": 0.5952,
+      "step": 556
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.390997262832345e-06,
+      "loss": 0.6261,
+      "step": 557
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3886157467728845e-06,
+      "loss": 0.6034,
+      "step": 558
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.386230231836936e-06,
+      "loss": 0.5326,
+      "step": 559
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.383840723075488e-06,
+      "loss": 0.5507,
+      "step": 560
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.38144722554799e-06,
+      "loss": 0.5612,
+      "step": 561
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.3790497443223316e-06,
+      "loss": 0.5735,
+      "step": 562
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.376648284474841e-06,
+      "loss": 0.5491,
+      "step": 563
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.374242851090269e-06,
+      "loss": 0.5162,
+      "step": 564
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.37183344926178e-06,
+      "loss": 0.5451,
+      "step": 565
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.369420084090942e-06,
+      "loss": 0.6291,
+      "step": 566
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.367002760687713e-06,
+      "loss": 0.5922,
+      "step": 567
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.364581484170435e-06,
+      "loss": 0.6281,
+      "step": 568
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.362156259665817e-06,
+      "loss": 0.6049,
+      "step": 569
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3597270923089286e-06,
+      "loss": 0.4983,
+      "step": 570
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.357293987243188e-06,
+      "loss": 0.6074,
+      "step": 571
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.354856949620351e-06,
+      "loss": 0.5515,
+      "step": 572
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3524159846004985e-06,
+      "loss": 0.5663,
+      "step": 573
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.349971097352031e-06,
+      "loss": 0.5298,
+      "step": 574
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.3475222930516484e-06,
+      "loss": 0.5779,
+      "step": 575
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.345069576884348e-06,
+      "loss": 0.6118,
+      "step": 576
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.34261295404341e-06,
+      "loss": 0.5585,
+      "step": 577
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.340152429730384e-06,
+      "loss": 0.6217,
+      "step": 578
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.337688009155084e-06,
+      "loss": 0.5257,
+      "step": 579
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.33521969753557e-06,
+      "loss": 0.5539,
+      "step": 580
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.332747500098143e-06,
+      "loss": 0.5334,
+      "step": 581
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.3302714220773305e-06,
+      "loss": 0.5239,
+      "step": 582
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.327791468715878e-06,
+      "loss": 0.6228,
+      "step": 583
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3253076452647355e-06,
+      "loss": 0.5515,
+      "step": 584
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.322819956983047e-06,
+      "loss": 0.5151,
+      "step": 585
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.320328409138141e-06,
+      "loss": 0.5831,
+      "step": 586
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.317833007005517e-06,
+      "loss": 0.5766,
+      "step": 587
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.315333755868836e-06,
+      "loss": 0.5509,
+      "step": 588
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3128306610199076e-06,
+      "loss": 0.617,
+      "step": 589
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.310323727758682e-06,
+      "loss": 0.521,
+      "step": 590
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.3078129613932354e-06,
+      "loss": 0.5317,
+      "step": 591
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.305298367239758e-06,
+      "loss": 0.5576,
+      "step": 592
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.302779950622551e-06,
+      "loss": 0.579,
+      "step": 593
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.300257716874001e-06,
+      "loss": 0.5904,
+      "step": 594
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.297731671334584e-06,
+      "loss": 0.5276,
+      "step": 595
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2952018193528425e-06,
+      "loss": 0.5281,
+      "step": 596
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2926681662853785e-06,
+      "loss": 0.5217,
+      "step": 597
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.290130717496847e-06,
+      "loss": 0.5252,
+      "step": 598
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2875894783599335e-06,
+      "loss": 0.522,
+      "step": 599
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2850444542553535e-06,
+      "loss": 0.5978,
+      "step": 600
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2824956505718355e-06,
+      "loss": 0.5406,
+      "step": 601
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.279943072706109e-06,
+      "loss": 0.6198,
+      "step": 602
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.2773867260629e-06,
+      "loss": 0.5042,
+      "step": 603
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.274826616054906e-06,
+      "loss": 0.5485,
+      "step": 604
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.2722627481028e-06,
+      "loss": 0.5828,
+      "step": 605
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.269695127635209e-06,
+      "loss": 0.5614,
+      "step": 606
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.267123760088706e-06,
+      "loss": 0.6055,
+      "step": 607
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.264548650907798e-06,
+      "loss": 0.5284,
+      "step": 608
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.261969805544914e-06,
+      "loss": 0.5217,
+      "step": 609
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.259387229460393e-06,
+      "loss": 0.5893,
+      "step": 610
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.2568009281224755e-06,
+      "loss": 0.5702,
+      "step": 611
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.2542109070072884e-06,
+      "loss": 0.5206,
+      "step": 612
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.251617171598834e-06,
+      "loss": 0.486,
+      "step": 613
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.24901972738898e-06,
+      "loss": 0.5843,
+      "step": 614
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.2464185798774485e-06,
+      "loss": 0.5801,
+      "step": 615
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.243813734571798e-06,
+      "loss": 0.5293,
+      "step": 616
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.241205196987422e-06,
+      "loss": 0.5644,
+      "step": 617
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.2385929726475295e-06,
+      "loss": 0.6096,
+      "step": 618
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.235977067083134e-06,
+      "loss": 0.5863,
+      "step": 619
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.233357485833048e-06,
+      "loss": 0.5464,
+      "step": 620
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.230734234443861e-06,
+      "loss": 0.5871,
+      "step": 621
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.228107318469938e-06,
+      "loss": 0.5036,
+      "step": 622
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.225476743473403e-06,
+      "loss": 0.5659,
+      "step": 623
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.222842515024123e-06,
+      "loss": 0.6611,
+      "step": 624
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.220204638699707e-06,
+      "loss": 0.5194,
+      "step": 625
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.217563120085483e-06,
+      "loss": 0.5512,
+      "step": 626
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.214917964774493e-06,
+      "loss": 0.6293,
+      "step": 627
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.2122691783674786e-06,
+      "loss": 0.6099,
+      "step": 628
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.209616766472869e-06,
+      "loss": 0.5465,
+      "step": 629
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.206960734706774e-06,
+      "loss": 0.5462,
+      "step": 630
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.204301088692964e-06,
+      "loss": 0.5083,
+      "step": 631
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.20163783406286e-06,
+      "loss": 0.5827,
+      "step": 632
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.19897097645553e-06,
+      "loss": 0.5706,
+      "step": 633
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.1963005215176655e-06,
+      "loss": 0.5946,
+      "step": 634
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.193626474903578e-06,
+      "loss": 0.5746,
+      "step": 635
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.190948842275182e-06,
+      "loss": 0.6247,
+      "step": 636
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.188267629301986e-06,
+      "loss": 0.6076,
+      "step": 637
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.185582841661079e-06,
+      "loss": 0.528,
+      "step": 638
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.182894485037117e-06,
+      "loss": 0.5771,
+      "step": 639
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.180202565122318e-06,
+      "loss": 0.4491,
+      "step": 640
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.177507087616438e-06,
+      "loss": 0.5962,
+      "step": 641
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.174808058226771e-06,
+      "loss": 0.5363,
+      "step": 642
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.172105482668127e-06,
+      "loss": 0.5633,
+      "step": 643
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.169399366662831e-06,
+      "loss": 0.5211,
+      "step": 644
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.166689715940699e-06,
+      "loss": 0.5,
+      "step": 645
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.163976536239032e-06,
+      "loss": 0.5578,
+      "step": 646
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.161259833302605e-06,
+      "loss": 0.5856,
+      "step": 647
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.15853961288365e-06,
+      "loss": 0.5503,
+      "step": 648
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.155815880741851e-06,
+      "loss": 0.5569,
+      "step": 649
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.153088642644325e-06,
+      "loss": 0.5067,
+      "step": 650
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.150357904365611e-06,
+      "loss": 0.5913,
+      "step": 651
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.147623671687662e-06,
+      "loss": 0.7278,
+      "step": 652
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.1448859503998275e-06,
+      "loss": 0.5854,
+      "step": 653
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.142144746298846e-06,
+      "loss": 0.5993,
+      "step": 654
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.1394000651888285e-06,
+      "loss": 0.5394,
+      "step": 655
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.13665191288125e-06,
+      "loss": 0.6006,
+      "step": 656
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.133900295194931e-06,
+      "loss": 0.6133,
+      "step": 657
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.131145217956036e-06,
+      "loss": 0.5946,
+      "step": 658
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.128386686998048e-06,
+      "loss": 0.5552,
+      "step": 659
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.125624708161768e-06,
+      "loss": 0.5654,
+      "step": 660
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.122859287295294e-06,
+      "loss": 0.568,
+      "step": 661
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.120090430254013e-06,
+      "loss": 0.568,
+      "step": 662
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.1173181429005895e-06,
+      "loss": 0.5896,
+      "step": 663
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.114542431104948e-06,
+      "loss": 0.6089,
+      "step": 664
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.111763300744265e-06,
+      "loss": 0.5207,
+      "step": 665
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.108980757702956e-06,
+      "loss": 0.5555,
+      "step": 666
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.106194807872663e-06,
+      "loss": 0.6308,
+      "step": 667
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.10340545715224e-06,
+      "loss": 0.6148,
+      "step": 668
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.100612711447741e-06,
+      "loss": 0.5744,
+      "step": 669
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.097816576672412e-06,
+      "loss": 0.5493,
+      "step": 670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.09501705874667e-06,
+      "loss": 0.5823,
+      "step": 671
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0922141635981e-06,
+      "loss": 0.5568,
+      "step": 672
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.089407897161434e-06,
+      "loss": 0.5935,
+      "step": 673
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.086598265378545e-06,
+      "loss": 0.5386,
+      "step": 674
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0837852741984305e-06,
+      "loss": 0.5596,
+      "step": 675
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0809689295772e-06,
+      "loss": 0.5187,
+      "step": 676
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.078149237478066e-06,
+      "loss": 0.5811,
+      "step": 677
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.075326203871327e-06,
+      "loss": 0.5738,
+      "step": 678
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.072499834734357e-06,
+      "loss": 0.5527,
+      "step": 679
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.069670136051591e-06,
+      "loss": 0.5897,
+      "step": 680
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.066837113814518e-06,
+      "loss": 0.5307,
+      "step": 681
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.06400077402166e-06,
+      "loss": 0.5536,
+      "step": 682
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.061161122678566e-06,
+      "loss": 0.606,
+      "step": 683
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.058318165797794e-06,
+      "loss": 0.5559,
+      "step": 684
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.055471909398904e-06,
+      "loss": 0.6132,
+      "step": 685
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.052622359508441e-06,
+      "loss": 0.5803,
+      "step": 686
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.049769522159923e-06,
+      "loss": 0.5899,
+      "step": 687
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.04691340339383e-06,
+      "loss": 0.5873,
+      "step": 688
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.044054009257588e-06,
+      "loss": 0.6187,
+      "step": 689
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.04119134580556e-06,
+      "loss": 0.5386,
+      "step": 690
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.038325419099032e-06,
+      "loss": 0.6336,
+      "step": 691
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.035456235206195e-06,
+      "loss": 0.4896,
+      "step": 692
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.0325838002021425e-06,
+      "loss": 0.6069,
+      "step": 693
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.029708120168847e-06,
+      "loss": 0.6055,
+      "step": 694
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.026829201195155e-06,
+      "loss": 0.5544,
+      "step": 695
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.02394704937677e-06,
+      "loss": 0.5978,
+      "step": 696
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.0210616708162395e-06,
+      "loss": 0.5601,
+      "step": 697
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.018173071622946e-06,
+      "loss": 0.5842,
+      "step": 698
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.0152812579130875e-06,
+      "loss": 0.5686,
+      "step": 699
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.012386235809671e-06,
+      "loss": 0.5164,
+      "step": 700
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.009488011442495e-06,
+      "loss": 0.4864,
+      "step": 701
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.006586590948141e-06,
+      "loss": 0.5935,
+      "step": 702
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.003681980469956e-06,
+      "loss": 0.5334,
+      "step": 703
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.0007741861580395e-06,
+      "loss": 0.5446,
+      "step": 704
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.9978632141692355e-06,
+      "loss": 0.518,
+      "step": 705
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.994949070667116e-06,
+      "loss": 0.5973,
+      "step": 706
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.992031761821967e-06,
+      "loss": 0.5407,
+      "step": 707
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.9891112938107755e-06,
+      "loss": 0.5739,
+      "step": 708
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.98618767281722e-06,
+      "loss": 0.5854,
+      "step": 709
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.983260905031656e-06,
+      "loss": 0.6069,
+      "step": 710
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.980330996651097e-06,
+      "loss": 0.6012,
+      "step": 711
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.977397953879211e-06,
+      "loss": 0.5287,
+      "step": 712
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.974461782926299e-06,
+      "loss": 0.5626,
+      "step": 713
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.9715224900092895e-06,
+      "loss": 0.5879,
+      "step": 714
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.968580081351718e-06,
+      "loss": 0.5241,
+      "step": 715
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.96563456318372e-06,
+      "loss": 0.6111,
+      "step": 716
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.96268594174201e-06,
+      "loss": 0.5267,
+      "step": 717
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.959734223269879e-06,
+      "loss": 0.553,
+      "step": 718
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.956779414017173e-06,
+      "loss": 0.5159,
+      "step": 719
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.953821520240281e-06,
+      "loss": 0.4857,
+      "step": 720
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.950860548202125e-06,
+      "loss": 0.5817,
+      "step": 721
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.947896504172144e-06,
+      "loss": 0.5664,
+      "step": 722
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.944929394426283e-06,
+      "loss": 0.5236,
+      "step": 723
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.941959225246975e-06,
+      "loss": 0.5555,
+      "step": 724
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.938986002923133e-06,
+      "loss": 0.5173,
+      "step": 725
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.9360097337501355e-06,
+      "loss": 0.5483,
+      "step": 726
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.93303042402981e-06,
+      "loss": 0.6349,
+      "step": 727
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.930048080070424e-06,
+      "loss": 0.6136,
+      "step": 728
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.927062708186667e-06,
+      "loss": 0.517,
+      "step": 729
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.924074314699643e-06,
+      "loss": 0.5116,
+      "step": 730
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.921082905936852e-06,
+      "loss": 0.63,
+      "step": 731
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.918088488232178e-06,
+      "loss": 0.5495,
+      "step": 732
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.915091067925878e-06,
+      "loss": 0.5947,
+      "step": 733
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.912090651364563e-06,
+      "loss": 0.6379,
+      "step": 734
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.909087244901192e-06,
+      "loss": 0.5514,
+      "step": 735
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.906080854895051e-06,
+      "loss": 0.5258,
+      "step": 736
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.903071487711748e-06,
+      "loss": 0.4872,
+      "step": 737
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.90005914972319e-06,
+      "loss": 0.5287,
+      "step": 738
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.897043847307578e-06,
+      "loss": 0.5857,
+      "step": 739
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.894025586849386e-06,
+      "loss": 0.5029,
+      "step": 740
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.891004374739354e-06,
+      "loss": 0.5309,
+      "step": 741
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.887980217374472e-06,
+      "loss": 0.6516,
+      "step": 742
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.884953121157964e-06,
+      "loss": 0.554,
+      "step": 743
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.881923092499278e-06,
+      "loss": 0.5419,
+      "step": 744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87889013781407e-06,
+      "loss": 0.4948,
+      "step": 745
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.875854263524193e-06,
+      "loss": 0.5443,
+      "step": 746
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.872815476057682e-06,
+      "loss": 0.6264,
+      "step": 747
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.869773781848737e-06,
+      "loss": 0.579,
+      "step": 748
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.866729187337717e-06,
+      "loss": 0.572,
+      "step": 749
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8636816989711185e-06,
+      "loss": 0.5511,
+      "step": 750
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.860631323201568e-06,
+      "loss": 0.5141,
+      "step": 751
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.857578066487803e-06,
+      "loss": 0.5586,
+      "step": 752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.854521935294663e-06,
+      "loss": 0.5382,
+      "step": 753
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8514629360930725e-06,
+      "loss": 0.4989,
+      "step": 754
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8484010753600306e-06,
+      "loss": 0.6455,
+      "step": 755
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8453363595785935e-06,
+      "loss": 0.5253,
+      "step": 756
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.842268795237863e-06,
+      "loss": 0.5113,
+      "step": 757
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.839198388832972e-06,
+      "loss": 0.5168,
+      "step": 758
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8361251468650705e-06,
+      "loss": 0.5735,
+      "step": 759
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.833049075841316e-06,
+      "loss": 0.491,
+      "step": 760
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.8299701822748506e-06,
+      "loss": 0.5059,
+      "step": 761
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.826888472684798e-06,
+      "loss": 0.5321,
+      "step": 762
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.823803953596239e-06,
+      "loss": 0.5366,
+      "step": 763
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.820716631540209e-06,
+      "loss": 0.5039,
+      "step": 764
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.817626513053674e-06,
+      "loss": 0.5911,
+      "step": 765
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.8145336046795233e-06,
+      "loss": 0.5191,
+      "step": 766
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.8114379129665526e-06,
+      "loss": 0.4968,
+      "step": 767
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.8083394444694493e-06,
+      "loss": 0.5228,
+      "step": 768
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.8052382057487826e-06,
+      "loss": 0.4698,
+      "step": 769
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.8021342033709874e-06,
+      "loss": 0.5194,
+      "step": 770
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.7990274439083485e-06,
+      "loss": 0.568,
+      "step": 771
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.795917933938989e-06,
+      "loss": 0.5693,
+      "step": 772
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.792805680046856e-06,
+      "loss": 0.4882,
+      "step": 773
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.789690688821708e-06,
+      "loss": 0.5489,
+      "step": 774
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.7865729668590967e-06,
+      "loss": 0.5189,
+      "step": 775
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.7834525207603573e-06,
+      "loss": 0.4783,
+      "step": 776
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.780329357132592e-06,
+      "loss": 0.5052,
+      "step": 777
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.777203482588659e-06,
+      "loss": 0.4571,
+      "step": 778
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.7740749037471548e-06,
+      "loss": 0.4931,
+      "step": 779
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.7709436272324006e-06,
+      "loss": 0.4829,
+      "step": 780
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.7678096596744328e-06,
+      "loss": 0.5822,
+      "step": 781
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.7646730077089845e-06,
+      "loss": 0.4783,
+      "step": 782
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.76153367797747e-06,
+      "loss": 0.5901,
+      "step": 783
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.7583916771269784e-06,
+      "loss": 0.4686,
+      "step": 784
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.7552470118102493e-06,
+      "loss": 0.4926,
+      "step": 785
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.7520996886856674e-06,
+      "loss": 0.4903,
+      "step": 786
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.748949714417245e-06,
+      "loss": 0.5401,
+      "step": 787
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.745797095674604e-06,
+      "loss": 0.4422,
+      "step": 788
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.7426418391329704e-06,
+      "loss": 0.55,
+      "step": 789
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.739483951473153e-06,
+      "loss": 0.4528,
+      "step": 790
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.7363234393815313e-06,
+      "loss": 0.4996,
+      "step": 791
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.733160309550042e-06,
+      "loss": 0.4515,
+      "step": 792
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.7299945686761657e-06,
+      "loss": 0.474,
+      "step": 793
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.7268262234629094e-06,
+      "loss": 0.4305,
+      "step": 794
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.723655280618797e-06,
+      "loss": 0.4636,
+      "step": 795
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.7204817468578503e-06,
+      "loss": 0.4217,
+      "step": 796
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.7173056288995774e-06,
+      "loss": 0.3948,
+      "step": 797
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.7141269334689593e-06,
+      "loss": 0.438,
+      "step": 798
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.7109456672964332e-06,
+      "loss": 0.5027,
+      "step": 799
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.70776183711788e-06,
+      "loss": 0.4095,
+      "step": 800
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.704575449674609e-06,
+      "loss": 0.4639,
+      "step": 801
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.7013865117133446e-06,
+      "loss": 0.5047,
+      "step": 802
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.6981950299862124e-06,
+      "loss": 0.4354,
+      "step": 803
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.695001011250722e-06,
+      "loss": 0.5192,
+      "step": 804
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.691804462269756e-06,
+      "loss": 0.4742,
+      "step": 805
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6886053898115555e-06,
+      "loss": 0.467,
+      "step": 806
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6854038006497028e-06,
+      "loss": 0.4297,
+      "step": 807
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6821997015631094e-06,
+      "loss": 0.4801,
+      "step": 808
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6789930993360018e-06,
+      "loss": 0.5009,
+      "step": 809
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6757840007579065e-06,
+      "loss": 0.4684,
+      "step": 810
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6725724126236357e-06,
+      "loss": 0.4725,
+      "step": 811
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.6693583417332717e-06,
+      "loss": 0.4635,
+      "step": 812
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.666141794892155e-06,
+      "loss": 0.4813,
+      "step": 813
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.662922778910869e-06,
+      "loss": 0.4749,
+      "step": 814
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.659701300605224e-06,
+      "loss": 0.4828,
+      "step": 815
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.6564773667962435e-06,
+      "loss": 0.4326,
+      "step": 816
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.653250984310151e-06,
+      "loss": 0.4867,
+      "step": 817
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.650022159978356e-06,
+      "loss": 0.4746,
+      "step": 818
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.646790900637436e-06,
+      "loss": 0.4464,
+      "step": 819
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6435572131291245e-06,
+      "loss": 0.4715,
+      "step": 820
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6403211043002986e-06,
+      "loss": 0.4402,
+      "step": 821
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6370825810029596e-06,
+      "loss": 0.4554,
+      "step": 822
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.633841650094223e-06,
+      "loss": 0.4136,
+      "step": 823
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6305983184363004e-06,
+      "loss": 0.4815,
+      "step": 824
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.627352592896487e-06,
+      "loss": 0.4664,
+      "step": 825
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6241044803471483e-06,
+      "loss": 0.4792,
+      "step": 826
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.6208539876657022e-06,
+      "loss": 0.4478,
+      "step": 827
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.6176011217346067e-06,
+      "loss": 0.4387,
+      "step": 828
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.6143458894413463e-06,
+      "loss": 0.471,
+      "step": 829
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.6110882976784122e-06,
+      "loss": 0.474,
+      "step": 830
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.6078283533432955e-06,
+      "loss": 0.4667,
+      "step": 831
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.604566063338467e-06,
+      "loss": 0.4932,
+      "step": 832
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.6013014345713627e-06,
+      "loss": 0.4579,
+      "step": 833
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.5980344739543733e-06,
+      "loss": 0.4456,
+      "step": 834
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.594765188404824e-06,
+      "loss": 0.4573,
+      "step": 835
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5914935848449656e-06,
+      "loss": 0.4501,
+      "step": 836
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.588219670201954e-06,
+      "loss": 0.4494,
+      "step": 837
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.584943451407841e-06,
+      "loss": 0.4308,
+      "step": 838
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.581664935399556e-06,
+      "loss": 0.4418,
+      "step": 839
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5783841291188923e-06,
+      "loss": 0.4301,
+      "step": 840
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.5751010395124926e-06,
+      "loss": 0.3994,
+      "step": 841
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.571815673531835e-06,
+      "loss": 0.4646,
+      "step": 842
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5685280381332168e-06,
+      "loss": 0.457,
+      "step": 843
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.565238140277739e-06,
+      "loss": 0.4585,
+      "step": 844
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5619459869312965e-06,
+      "loss": 0.4401,
+      "step": 845
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5586515850645582e-06,
+      "loss": 0.478,
+      "step": 846
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.555354941652952e-06,
+      "loss": 0.4814,
+      "step": 847
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.552056063676655e-06,
+      "loss": 0.4713,
+      "step": 848
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5487549581205727e-06,
+      "loss": 0.4332,
+      "step": 849
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5454516319743303e-06,
+      "loss": 0.4708,
+      "step": 850
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5421460922322537e-06,
+      "loss": 0.4572,
+      "step": 851
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5388383458933538e-06,
+      "loss": 0.45,
+      "step": 852
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.535528399961316e-06,
+      "loss": 0.4902,
+      "step": 853
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.532216261444483e-06,
+      "loss": 0.4909,
+      "step": 854
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5289019373558376e-06,
+      "loss": 0.3722,
+      "step": 855
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.525585434712993e-06,
+      "loss": 0.4256,
+      "step": 856
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.5222667605381732e-06,
+      "loss": 0.4836,
+      "step": 857
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5189459218582027e-06,
+      "loss": 0.4874,
+      "step": 858
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5156229257044857e-06,
+      "loss": 0.4543,
+      "step": 859
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.5122977791129962e-06,
+      "loss": 0.4805,
+      "step": 860
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.508970489124262e-06,
+      "loss": 0.4467,
+      "step": 861
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.505641062783349e-06,
+      "loss": 0.4223,
+      "step": 862
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.502309507139845e-06,
+      "loss": 0.467,
+      "step": 863
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.4989758292478484e-06,
+      "loss": 0.4843,
+      "step": 864
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4956400361659494e-06,
+      "loss": 0.4706,
+      "step": 865
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4923021349572183e-06,
+      "loss": 0.4363,
+      "step": 866
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.488962132689188e-06,
+      "loss": 0.454,
+      "step": 867
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.485620036433841e-06,
+      "loss": 0.4253,
+      "step": 868
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4822758532675926e-06,
+      "loss": 0.4748,
+      "step": 869
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4789295902712774e-06,
+      "loss": 0.4301,
+      "step": 870
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.4755812545301336e-06,
+      "loss": 0.4717,
+      "step": 871
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4722308531337884e-06,
+      "loss": 0.4264,
+      "step": 872
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4688783931762415e-06,
+      "loss": 0.4623,
+      "step": 873
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.465523881755853e-06,
+      "loss": 0.4282,
+      "step": 874
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4621673259753263e-06,
+      "loss": 0.4012,
+      "step": 875
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4588087329416924e-06,
+      "loss": 0.4439,
+      "step": 876
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4554481097662962e-06,
+      "loss": 0.4496,
+      "step": 877
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.452085463564783e-06,
+      "loss": 0.4873,
+      "step": 878
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.4487208014570794e-06,
+      "loss": 0.4755,
+      "step": 879
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.4453541305673815e-06,
+      "loss": 0.4256,
+      "step": 880
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.441985458024137e-06,
+      "loss": 0.4151,
+      "step": 881
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.4386147909600344e-06,
+      "loss": 0.5419,
+      "step": 882
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.4352421365119837e-06,
+      "loss": 0.4448,
+      "step": 883
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.431867501821103e-06,
+      "loss": 0.4386,
+      "step": 884
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.428490894032703e-06,
+      "loss": 0.4035,
+      "step": 885
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.425112320296274e-06,
+      "loss": 0.4531,
+      "step": 886
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.4217317877654675e-06,
+      "loss": 0.4316,
+      "step": 887
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.4183493035980825e-06,
+      "loss": 0.4891,
+      "step": 888
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.414964874956049e-06,
+      "loss": 0.4414,
+      "step": 889
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.411578509005417e-06,
+      "loss": 0.4654,
+      "step": 890
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.408190212916337e-06,
+      "loss": 0.4431,
+      "step": 891
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.404799993863045e-06,
+      "loss": 0.3852,
+      "step": 892
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.4014078590238507e-06,
+      "loss": 0.4753,
+      "step": 893
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.3980138155811202e-06,
+      "loss": 0.5182,
+      "step": 894
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3946178707212595e-06,
+      "loss": 0.5252,
+      "step": 895
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.391220031634701e-06,
+      "loss": 0.453,
+      "step": 896
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3878203055158874e-06,
+      "loss": 0.4376,
+      "step": 897
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.384418699563259e-06,
+      "loss": 0.4453,
+      "step": 898
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3810152209792324e-06,
+      "loss": 0.5098,
+      "step": 899
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3776098769701944e-06,
+      "loss": 0.4847,
+      "step": 900
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.3742026747464774e-06,
+      "loss": 0.499,
+      "step": 901
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.3707936215223496e-06,
+      "loss": 0.5069,
+      "step": 902
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.3673827245159995e-06,
+      "loss": 0.459,
+      "step": 903
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.363969990949518e-06,
+      "loss": 0.4301,
+      "step": 904
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.3605554280488862e-06,
+      "loss": 0.4132,
+      "step": 905
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.3571390430439566e-06,
+      "loss": 0.4532,
+      "step": 906
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.353720843168441e-06,
+      "loss": 0.4643,
+      "step": 907
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.3503008356598942e-06,
+      "loss": 0.4367,
+      "step": 908
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.3468790277596974e-06,
+      "loss": 0.3799,
+      "step": 909
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.343455426713045e-06,
+      "loss": 0.5535,
+      "step": 910
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.3400300397689276e-06,
+      "loss": 0.4929,
+      "step": 911
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.336602874180117e-06,
+      "loss": 0.4959,
+      "step": 912
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.33317393720315e-06,
+      "loss": 0.4741,
+      "step": 913
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.3297432360983174e-06,
+      "loss": 0.4413,
+      "step": 914
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.3263107781296424e-06,
+      "loss": 0.5093,
+      "step": 915
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.3228765705648675e-06,
+      "loss": 0.4652,
+      "step": 916
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.319440620675442e-06,
+      "loss": 0.3929,
+      "step": 917
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.3160029357365043e-06,
+      "loss": 0.4503,
+      "step": 918
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.3125635230268643e-06,
+      "loss": 0.4274,
+      "step": 919
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.309122389828992e-06,
+      "loss": 0.4499,
+      "step": 920
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.3056795434289997e-06,
+      "loss": 0.43,
+      "step": 921
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.3022349911166274e-06,
+      "loss": 0.4662,
+      "step": 922
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.298788740185227e-06,
+      "loss": 0.4481,
+      "step": 923
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.2953407979317474e-06,
+      "loss": 0.4092,
+      "step": 924
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.291891171656717e-06,
+      "loss": 0.4636,
+      "step": 925
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.2884398686642326e-06,
+      "loss": 0.5238,
+      "step": 926
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.2849868962619392e-06,
+      "loss": 0.494,
+      "step": 927
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.281532261761016e-06,
+      "loss": 0.4199,
+      "step": 928
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.2780759724761635e-06,
+      "loss": 0.4946,
+      "step": 929
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.274618035725585e-06,
+      "loss": 0.5422,
+      "step": 930
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.271158458830973e-06,
+      "loss": 0.4332,
+      "step": 931
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2676972491174895e-06,
+      "loss": 0.4638,
+      "step": 932
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2642344139137578e-06,
+      "loss": 0.3988,
+      "step": 933
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2607699605518408e-06,
+      "loss": 0.474,
+      "step": 934
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2573038963672276e-06,
+      "loss": 0.474,
+      "step": 935
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2538362286988193e-06,
+      "loss": 0.4844,
+      "step": 936
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2503669648889114e-06,
+      "loss": 0.5064,
+      "step": 937
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.2468961122831776e-06,
+      "loss": 0.4683,
+      "step": 938
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.2434236782306584e-06,
+      "loss": 0.4445,
+      "step": 939
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.2399496700837407e-06,
+      "loss": 0.4292,
+      "step": 940
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.236474095198144e-06,
+      "loss": 0.4297,
+      "step": 941
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.2329969609329083e-06,
+      "loss": 0.4752,
+      "step": 942
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.229518274650371e-06,
+      "loss": 0.4926,
+      "step": 943
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.2260380437161597e-06,
+      "loss": 0.4447,
+      "step": 944
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.222556275499169e-06,
+      "loss": 0.4409,
+      "step": 945
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.2190729773715518e-06,
+      "loss": 0.5051,
+      "step": 946
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.2155881567086976e-06,
+      "loss": 0.4297,
+      "step": 947
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.212101820889222e-06,
+      "loss": 0.5028,
+      "step": 948
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.208613977294945e-06,
+      "loss": 0.4916,
+      "step": 949
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.2051246333108856e-06,
+      "loss": 0.448,
+      "step": 950
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.201633796325233e-06,
+      "loss": 0.4684,
+      "step": 951
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.198141473729341e-06,
+      "loss": 0.4775,
+      "step": 952
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.1946476729177077e-06,
+      "loss": 0.4589,
+      "step": 953
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.1911524012879637e-06,
+      "loss": 0.4835,
+      "step": 954
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.187655666240849e-06,
+      "loss": 0.49,
+      "step": 955
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.184157475180208e-06,
+      "loss": 0.4572,
+      "step": 956
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.1806578355129627e-06,
+      "loss": 0.5014,
+      "step": 957
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.1771567546491054e-06,
+      "loss": 0.4769,
+      "step": 958
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.17365424000168e-06,
+      "loss": 0.4327,
+      "step": 959
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.1701502989867655e-06,
+      "loss": 0.4521,
+      "step": 960
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.166644939023459e-06,
+      "loss": 0.4867,
+      "step": 961
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.163138167533866e-06,
+      "loss": 0.453,
+      "step": 962
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.1596299919430783e-06,
+      "loss": 0.4223,
+      "step": 963
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.156120419679161e-06,
+      "loss": 0.4716,
+      "step": 964
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.152609458173136e-06,
+      "loss": 0.5006,
+      "step": 965
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.1490971148589676e-06,
+      "loss": 0.4581,
+      "step": 966
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.1455833971735463e-06,
+      "loss": 0.5333,
+      "step": 967
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.1420683125566703e-06,
+      "loss": 0.4713,
+      "step": 968
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.138551868451035e-06,
+      "loss": 0.4066,
+      "step": 969
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.1350340723022133e-06,
+      "loss": 0.4666,
+      "step": 970
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.1315149315586386e-06,
+      "loss": 0.487,
+      "step": 971
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.127994453671595e-06,
+      "loss": 0.4725,
+      "step": 972
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.124472646095196e-06,
+      "loss": 0.4595,
+      "step": 973
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.1209495162863697e-06,
+      "loss": 0.4726,
+      "step": 974
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.1174250717048453e-06,
+      "loss": 0.441,
+      "step": 975
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.1138993198131357e-06,
+      "loss": 0.4246,
+      "step": 976
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.11037226807652e-06,
+      "loss": 0.4562,
+      "step": 977
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.1068439239630327e-06,
+      "loss": 0.4201,
+      "step": 978
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.1033142949434418e-06,
+      "loss": 0.5002,
+      "step": 979
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0997833884912374e-06,
+      "loss": 0.4969,
+      "step": 980
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0962512120826136e-06,
+      "loss": 0.4834,
+      "step": 981
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.0927177731964554e-06,
+      "loss": 0.4432,
+      "step": 982
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.089183079314318e-06,
+      "loss": 0.4017,
+      "step": 983
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.0856471379204156e-06,
+      "loss": 0.4701,
+      "step": 984
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.082109956501604e-06,
+      "loss": 0.5251,
+      "step": 985
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.078571542547364e-06,
+      "loss": 0.5178,
+      "step": 986
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.0750319035497854e-06,
+      "loss": 0.5296,
+      "step": 987
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.071491047003554e-06,
+      "loss": 0.3877,
+      "step": 988
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.0679489804059313e-06,
+      "loss": 0.4216,
+      "step": 989
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.064405711256743e-06,
+      "loss": 0.4503,
+      "step": 990
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.060861247058359e-06,
+      "loss": 0.4858,
+      "step": 991
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.05731559531568e-06,
+      "loss": 0.5094,
+      "step": 992
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.053768763536123e-06,
+      "loss": 0.4501,
+      "step": 993
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.0502207592296015e-06,
+      "loss": 0.4836,
+      "step": 994
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.046671589908512e-06,
+      "loss": 0.425,
+      "step": 995
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.0431212630877195e-06,
+      "loss": 0.4508,
+      "step": 996
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.0395697862845357e-06,
+      "loss": 0.4303,
+      "step": 997
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0360171670187134e-06,
+      "loss": 0.3841,
+      "step": 998
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0324634128124185e-06,
+      "loss": 0.5045,
+      "step": 999
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0289085311902237e-06,
+      "loss": 0.4284,
+      "step": 1000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0253525296790866e-06,
+      "loss": 0.4899,
+      "step": 1001
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0217954158083384e-06,
+      "loss": 0.4855,
+      "step": 1002
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.018237197109663e-06,
+      "loss": 0.4771,
+      "step": 1003
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.0146778811170848e-06,
+      "loss": 0.4242,
+      "step": 1004
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.011117475366952e-06,
+      "loss": 0.4784,
+      "step": 1005
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.00755598739792e-06,
+      "loss": 0.4841,
+      "step": 1006
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.0039934247509343e-06,
+      "loss": 0.441,
+      "step": 1007
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.000429794969219e-06,
+      "loss": 0.4301,
+      "step": 1008
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9968651055982534e-06,
+      "loss": 0.4948,
+      "step": 1009
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9932993641857637e-06,
+      "loss": 0.4265,
+      "step": 1010
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9897325782817034e-06,
+      "loss": 0.5275,
+      "step": 1011
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9861647554382354e-06,
+      "loss": 0.4619,
+      "step": 1012
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9825959032097195e-06,
+      "loss": 0.4582,
+      "step": 1013
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9790260291526964e-06,
+      "loss": 0.4467,
+      "step": 1014
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9754551408258673e-06,
+      "loss": 0.4902,
+      "step": 1015
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9718832457900848e-06,
+      "loss": 0.4527,
+      "step": 1016
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9683103516083285e-06,
+      "loss": 0.4559,
+      "step": 1017
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.964736465845698e-06,
+      "loss": 0.4409,
+      "step": 1018
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.96116159606939e-06,
+      "loss": 0.4705,
+      "step": 1019
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.9575857498486852e-06,
+      "loss": 0.4667,
+      "step": 1020
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.9540089347549313e-06,
+      "loss": 0.4792,
+      "step": 1021
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.950431158361528e-06,
+      "loss": 0.4725,
+      "step": 1022
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.946852428243911e-06,
+      "loss": 0.4629,
+      "step": 1023
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.9432727519795345e-06,
+      "loss": 0.4286,
+      "step": 1024
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.939692137147855e-06,
+      "loss": 0.4437,
+      "step": 1025
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.9361105913303197e-06,
+      "loss": 0.4448,
+      "step": 1026
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.932528122110343e-06,
+      "loss": 0.4396,
+      "step": 1027
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9289447370732976e-06,
+      "loss": 0.4464,
+      "step": 1028
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9253604438064924e-06,
+      "loss": 0.46,
+      "step": 1029
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9217752498991636e-06,
+      "loss": 0.4675,
+      "step": 1030
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9181891629424497e-06,
+      "loss": 0.4385,
+      "step": 1031
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9146021905293826e-06,
+      "loss": 0.502,
+      "step": 1032
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9110143402548685e-06,
+      "loss": 0.4881,
+      "step": 1033
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.907425619715673e-06,
+      "loss": 0.5054,
+      "step": 1034
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.9038360365104033e-06,
+      "loss": 0.4798,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.9002455982394946e-06,
+      "loss": 0.4514,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.89665431250519e-06,
+      "loss": 0.4593,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8930621869115304e-06,
+      "loss": 0.4592,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8894692290643323e-06,
+      "loss": 0.4752,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8858754465711753e-06,
+      "loss": 0.4433,
+      "step": 1040
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8822808470413856e-06,
+      "loss": 0.4475,
+      "step": 1041
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8786854380860174e-06,
+      "loss": 0.4346,
+      "step": 1042
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8750892273178425e-06,
+      "loss": 0.4177,
+      "step": 1043
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.871492222351327e-06,
+      "loss": 0.4651,
+      "step": 1044
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8678944308026192e-06,
+      "loss": 0.4367,
+      "step": 1045
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.864295860289534e-06,
+      "loss": 0.4357,
+      "step": 1046
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8606965184315348e-06,
+      "loss": 0.4481,
+      "step": 1047
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.857096412849719e-06,
+      "loss": 0.4608,
+      "step": 1048
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.8534955511667994e-06,
+      "loss": 0.445,
+      "step": 1049
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8498939410070926e-06,
+      "loss": 0.4765,
+      "step": 1050
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8462915899964966e-06,
+      "loss": 0.4618,
+      "step": 1051
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8426885057624805e-06,
+      "loss": 0.4374,
+      "step": 1052
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.839084695934064e-06,
+      "loss": 0.4579,
+      "step": 1053
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8354801681418063e-06,
+      "loss": 0.4273,
+      "step": 1054
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8318749300177833e-06,
+      "loss": 0.4913,
+      "step": 1055
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.828268989195577e-06,
+      "loss": 0.4387,
+      "step": 1056
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.824662353310256e-06,
+      "loss": 0.4833,
+      "step": 1057
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8210550299983615e-06,
+      "loss": 0.5172,
+      "step": 1058
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8174470268978897e-06,
+      "loss": 0.4797,
+      "step": 1059
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8138383516482775e-06,
+      "loss": 0.5007,
+      "step": 1060
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8102290118903825e-06,
+      "loss": 0.5455,
+      "step": 1061
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8066190152664715e-06,
+      "loss": 0.5098,
+      "step": 1062
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.803008369420201e-06,
+      "loss": 0.4633,
+      "step": 1063
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.799397081996603e-06,
+      "loss": 0.462,
+      "step": 1064
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.795785160642066e-06,
+      "loss": 0.4373,
+      "step": 1065
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7921726130043243e-06,
+      "loss": 0.5169,
+      "step": 1066
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7885594467324344e-06,
+      "loss": 0.4685,
+      "step": 1067
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.784945669476765e-06,
+      "loss": 0.4916,
+      "step": 1068
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7813312888889775e-06,
+      "loss": 0.4162,
+      "step": 1069
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.777716312622012e-06,
+      "loss": 0.4744,
+      "step": 1070
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.774100748330068e-06,
+      "loss": 0.4954,
+      "step": 1071
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7704846036685925e-06,
+      "loss": 0.4926,
+      "step": 1072
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.766867886294257e-06,
+      "loss": 0.5414,
+      "step": 1073
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.763250603864952e-06,
+      "loss": 0.423,
+      "step": 1074
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.759632764039759e-06,
+      "loss": 0.4621,
+      "step": 1075
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.7560143744789423e-06,
+      "loss": 0.4318,
+      "step": 1076
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.7523954428439293e-06,
+      "loss": 0.4285,
+      "step": 1077
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.7487759767972955e-06,
+      "loss": 0.4846,
+      "step": 1078
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.745155984002748e-06,
+      "loss": 0.4695,
+      "step": 1079
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.741535472125109e-06,
+      "loss": 0.42,
+      "step": 1080
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.7379144488303e-06,
+      "loss": 0.5098,
+      "step": 1081
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.7342929217853248e-06,
+      "loss": 0.5338,
+      "step": 1082
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.730670898658255e-06,
+      "loss": 0.483,
+      "step": 1083
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.7270483871182117e-06,
+      "loss": 0.4917,
+      "step": 1084
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.723425394835349e-06,
+      "loss": 0.4755,
+      "step": 1085
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.719801929480842e-06,
+      "loss": 0.483,
+      "step": 1086
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.7161779987268643e-06,
+      "loss": 0.5473,
+      "step": 1087
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.712553610246578e-06,
+      "loss": 0.5265,
+      "step": 1088
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.7089287717141092e-06,
+      "loss": 0.5536,
+      "step": 1089
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.7053034908045444e-06,
+      "loss": 0.4776,
+      "step": 1090
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.7016777751939005e-06,
+      "loss": 0.4006,
+      "step": 1091
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.698051632559117e-06,
+      "loss": 0.458,
+      "step": 1092
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.694425070578038e-06,
+      "loss": 0.4272,
+      "step": 1093
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6907980969293955e-06,
+      "loss": 0.3955,
+      "step": 1094
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6871707192927927e-06,
+      "loss": 0.5046,
+      "step": 1095
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.683542945348687e-06,
+      "loss": 0.5155,
+      "step": 1096
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.679914782778378e-06,
+      "loss": 0.4663,
+      "step": 1097
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.676286239263985e-06,
+      "loss": 0.574,
+      "step": 1098
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6726573224884363e-06,
+      "loss": 0.4799,
+      "step": 1099
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6690280401354485e-06,
+      "loss": 0.4221,
+      "step": 1100
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6653983998895126e-06,
+      "loss": 0.4523,
+      "step": 1101
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6617684094358803e-06,
+      "loss": 0.4532,
+      "step": 1102
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6581380764605408e-06,
+      "loss": 0.4813,
+      "step": 1103
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.654507408650211e-06,
+      "loss": 0.5284,
+      "step": 1104
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.650876413692315e-06,
+      "loss": 0.4547,
+      "step": 1105
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6472450992749715e-06,
+      "loss": 0.4281,
+      "step": 1106
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.643613473086975e-06,
+      "loss": 0.4393,
+      "step": 1107
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.639981542817779e-06,
+      "loss": 0.4795,
+      "step": 1108
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6363493161574815e-06,
+      "loss": 0.5116,
+      "step": 1109
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6327168007968096e-06,
+      "loss": 0.5449,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6290840044270994e-06,
+      "loss": 0.4488,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6254509347402833e-06,
+      "loss": 0.4715,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.621817599428872e-06,
+      "loss": 0.4418,
+      "step": 1113
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6181840061859377e-06,
+      "loss": 0.4631,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6145501627051018e-06,
+      "loss": 0.4675,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.610916076680512e-06,
+      "loss": 0.4892,
+      "step": 1116
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.6072817558068302e-06,
+      "loss": 0.4002,
+      "step": 1117
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.6036472077792184e-06,
+      "loss": 0.5182,
+      "step": 1118
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.600012440293317e-06,
+      "loss": 0.4841,
+      "step": 1119
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5963774610452304e-06,
+      "loss": 0.4588,
+      "step": 1120
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.592742277731513e-06,
+      "loss": 0.4382,
+      "step": 1121
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.589106898049152e-06,
+      "loss": 0.4613,
+      "step": 1122
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5854713296955484e-06,
+      "loss": 0.4308,
+      "step": 1123
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.581835580368504e-06,
+      "loss": 0.4195,
+      "step": 1124
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5781996577662024e-06,
+      "loss": 0.5008,
+      "step": 1125
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5745635695871956e-06,
+      "loss": 0.483,
+      "step": 1126
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5709273235303857e-06,
+      "loss": 0.4916,
+      "step": 1127
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5672909272950086e-06,
+      "loss": 0.4556,
+      "step": 1128
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5636543885806185e-06,
+      "loss": 0.4009,
+      "step": 1129
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.560017715087072e-06,
+      "loss": 0.5185,
+      "step": 1130
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.55638091451451e-06,
+      "loss": 0.478,
+      "step": 1131
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5527439945633433e-06,
+      "loss": 0.411,
+      "step": 1132
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5491069629342336e-06,
+      "loss": 0.4087,
+      "step": 1133
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.545469827328082e-06,
+      "loss": 0.5582,
+      "step": 1134
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5418325954460083e-06,
+      "loss": 0.5244,
+      "step": 1135
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.538195274989335e-06,
+      "loss": 0.5002,
+      "step": 1136
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5345578736595738e-06,
+      "loss": 0.4716,
+      "step": 1137
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5309203991584076e-06,
+      "loss": 0.4425,
+      "step": 1138
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5272828591876734e-06,
+      "loss": 0.4819,
+      "step": 1139
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.523645261449347e-06,
+      "loss": 0.5313,
+      "step": 1140
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5200076136455263e-06,
+      "loss": 0.4399,
+      "step": 1141
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5163699234784165e-06,
+      "loss": 0.5089,
+      "step": 1142
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.512732198650311e-06,
+      "loss": 0.4911,
+      "step": 1143
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.509094446863577e-06,
+      "loss": 0.5013,
+      "step": 1144
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5054566758206395e-06,
+      "loss": 0.4678,
+      "step": 1145
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.501818893223963e-06,
+      "loss": 0.4627,
+      "step": 1146
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4981811067760375e-06,
+      "loss": 0.4783,
+      "step": 1147
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4945433241793614e-06,
+      "loss": 0.4276,
+      "step": 1148
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.490905553136424e-06,
+      "loss": 0.5088,
+      "step": 1149
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4872678013496902e-06,
+      "loss": 0.4814,
+      "step": 1150
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4836300765215843e-06,
+      "loss": 0.4905,
+      "step": 1151
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.479992386354474e-06,
+      "loss": 0.4468,
+      "step": 1152
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4763547385506536e-06,
+      "loss": 0.54,
+      "step": 1153
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4727171408123274e-06,
+      "loss": 0.5033,
+      "step": 1154
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4690796008415928e-06,
+      "loss": 0.4477,
+      "step": 1155
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4654421263404266e-06,
+      "loss": 0.4673,
+      "step": 1156
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4618047250106662e-06,
+      "loss": 0.5426,
+      "step": 1157
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.458167404553993e-06,
+      "loss": 0.437,
+      "step": 1158
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.454530172671918e-06,
+      "loss": 0.4141,
+      "step": 1159
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4508930370657668e-06,
+      "loss": 0.5401,
+      "step": 1160
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4472560054366576e-06,
+      "loss": 0.4495,
+      "step": 1161
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4436190854854904e-06,
+      "loss": 0.4844,
+      "step": 1162
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4399822849129288e-06,
+      "loss": 0.4214,
+      "step": 1163
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.436345611419382e-06,
+      "loss": 0.4774,
+      "step": 1164
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4327090727049926e-06,
+      "loss": 0.4115,
+      "step": 1165
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4290726764696156e-06,
+      "loss": 0.509,
+      "step": 1166
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.425436430412805e-06,
+      "loss": 0.4736,
+      "step": 1167
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.421800342233798e-06,
+      "loss": 0.4852,
+      "step": 1168
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4181644196314967e-06,
+      "loss": 0.4325,
+      "step": 1169
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.414528670304452e-06,
+      "loss": 0.5535,
+      "step": 1170
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.410893101950848e-06,
+      "loss": 0.3807,
+      "step": 1171
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4072577222684874e-06,
+      "loss": 0.5787,
+      "step": 1172
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.403622538954771e-06,
+      "loss": 0.5216,
+      "step": 1173
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3999875597066844e-06,
+      "loss": 0.4817,
+      "step": 1174
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3963527922207816e-06,
+      "loss": 0.4323,
+      "step": 1175
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3927182441931698e-06,
+      "loss": 0.4957,
+      "step": 1176
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.389083923319489e-06,
+      "loss": 0.4341,
+      "step": 1177
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.385449837294899e-06,
+      "loss": 0.4186,
+      "step": 1178
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3818159938140627e-06,
+      "loss": 0.4163,
+      "step": 1179
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.378182400571129e-06,
+      "loss": 0.5011,
+      "step": 1180
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3745490652597175e-06,
+      "loss": 0.5038,
+      "step": 1181
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3709159955729015e-06,
+      "loss": 0.4567,
+      "step": 1182
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.367283199203191e-06,
+      "loss": 0.4342,
+      "step": 1183
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.363650683842519e-06,
+      "loss": 0.4882,
+      "step": 1184
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3600184571822216e-06,
+      "loss": 0.5001,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3563865269130258e-06,
+      "loss": 0.5327,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3527549007250293e-06,
+      "loss": 0.4725,
+      "step": 1187
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.349123586307686e-06,
+      "loss": 0.4227,
+      "step": 1188
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3454925913497905e-06,
+      "loss": 0.4794,
+      "step": 1189
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.34186192353946e-06,
+      "loss": 0.447,
+      "step": 1190
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3382315905641196e-06,
+      "loss": 0.4644,
+      "step": 1191
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.334601600110487e-06,
+      "loss": 0.5373,
+      "step": 1192
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3309719598645524e-06,
+      "loss": 0.5481,
+      "step": 1193
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3273426775115645e-06,
+      "loss": 0.4904,
+      "step": 1194
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.323713760736016e-06,
+      "loss": 0.5049,
+      "step": 1195
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.320085217221623e-06,
+      "loss": 0.4484,
+      "step": 1196
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3164570546513137e-06,
+      "loss": 0.5044,
+      "step": 1197
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.312829280707209e-06,
+      "loss": 0.5162,
+      "step": 1198
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.309201903070605e-06,
+      "loss": 0.4684,
+      "step": 1199
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.305574929421962e-06,
+      "loss": 0.4827,
+      "step": 1200
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3019483674408833e-06,
+      "loss": 0.4944,
+      "step": 1201
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2983222248061003e-06,
+      "loss": 0.4311,
+      "step": 1202
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.2946965091954564e-06,
+      "loss": 0.4637,
+      "step": 1203
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.291071228285891e-06,
+      "loss": 0.4731,
+      "step": 1204
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.287446389753424e-06,
+      "loss": 0.4691,
+      "step": 1205
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2838220012731365e-06,
+      "loss": 0.4159,
+      "step": 1206
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.280198070519158e-06,
+      "loss": 0.4777,
+      "step": 1207
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.276574605164651e-06,
+      "loss": 0.5147,
+      "step": 1208
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.272951612881789e-06,
+      "loss": 0.4338,
+      "step": 1209
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.269329101341745e-06,
+      "loss": 0.5076,
+      "step": 1210
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2657070782146756e-06,
+      "loss": 0.4277,
+      "step": 1211
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.262085551169701e-06,
+      "loss": 0.4229,
+      "step": 1212
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2584645278748917e-06,
+      "loss": 0.4802,
+      "step": 1213
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2548440159972526e-06,
+      "loss": 0.5048,
+      "step": 1214
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.251224023202705e-06,
+      "loss": 0.4287,
+      "step": 1215
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.247604557156071e-06,
+      "loss": 0.4499,
+      "step": 1216
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2439856255210585e-06,
+      "loss": 0.4827,
+      "step": 1217
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.240367235960242e-06,
+      "loss": 0.4686,
+      "step": 1218
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2367493961350485e-06,
+      "loss": 0.4762,
+      "step": 1219
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2331321137057434e-06,
+      "loss": 0.5297,
+      "step": 1220
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.229515396331409e-06,
+      "loss": 0.4682,
+      "step": 1221
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.225899251669933e-06,
+      "loss": 0.461,
+      "step": 1222
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.222283687377989e-06,
+      "loss": 0.447,
+      "step": 1223
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.218668711111023e-06,
+      "loss": 0.5136,
+      "step": 1224
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2150543305232354e-06,
+      "loss": 0.5263,
+      "step": 1225
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2114405532675664e-06,
+      "loss": 0.4519,
+      "step": 1226
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2078273869956765e-06,
+      "loss": 0.4641,
+      "step": 1227
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2042148393579343e-06,
+      "loss": 0.4422,
+      "step": 1228
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2006029180033982e-06,
+      "loss": 0.4808,
+      "step": 1229
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1969916305798e-06,
+      "loss": 0.4258,
+      "step": 1230
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.193380984733529e-06,
+      "loss": 0.5415,
+      "step": 1231
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.189770988109618e-06,
+      "loss": 0.5136,
+      "step": 1232
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1861616483517234e-06,
+      "loss": 0.4523,
+      "step": 1233
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1825529731021107e-06,
+      "loss": 0.5178,
+      "step": 1234
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1789449700016397e-06,
+      "loss": 0.5336,
+      "step": 1235
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1753376466897454e-06,
+      "loss": 0.4963,
+      "step": 1236
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.171731010804424e-06,
+      "loss": 0.3947,
+      "step": 1237
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.168125069982218e-06,
+      "loss": 0.4813,
+      "step": 1238
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.164519831858194e-06,
+      "loss": 0.465,
+      "step": 1239
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.160915304065936e-06,
+      "loss": 0.4103,
+      "step": 1240
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.1573114942375204e-06,
+      "loss": 0.4576,
+      "step": 1241
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.153708410003504e-06,
+      "loss": 0.4572,
+      "step": 1242
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.1501060589929087e-06,
+      "loss": 0.4302,
+      "step": 1243
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.146504448833201e-06,
+      "loss": 0.4622,
+      "step": 1244
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.142903587150282e-06,
+      "loss": 0.4313,
+      "step": 1245
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.1393034815684656e-06,
+      "loss": 0.4521,
+      "step": 1246
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.1357041397104663e-06,
+      "loss": 0.5088,
+      "step": 1247
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.132105569197381e-06,
+      "loss": 0.4564,
+      "step": 1248
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.1285077776486737e-06,
+      "loss": 0.424,
+      "step": 1249
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.124910772682158e-06,
+      "loss": 0.4589,
+      "step": 1250
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.121314561913983e-06,
+      "loss": 0.5132,
+      "step": 1251
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.1177191529586156e-06,
+      "loss": 0.508,
+      "step": 1252
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.114124553428826e-06,
+      "loss": 0.4152,
+      "step": 1253
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.110530770935669e-06,
+      "loss": 0.501,
+      "step": 1254
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.1069378130884704e-06,
+      "loss": 0.4752,
+      "step": 1255
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.103345687494811e-06,
+      "loss": 0.4588,
+      "step": 1256
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.0997544017605063e-06,
+      "loss": 0.4574,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.096163963489597e-06,
+      "loss": 0.5141,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.092574380284328e-06,
+      "loss": 0.4532,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0889856597451323e-06,
+      "loss": 0.4393,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0853978094706187e-06,
+      "loss": 0.4687,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.081810837057552e-06,
+      "loss": 0.4859,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.078224750100837e-06,
+      "loss": 0.4164,
+      "step": 1263
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.074639556193507e-06,
+      "loss": 0.536,
+      "step": 1264
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.0710552629267033e-06,
+      "loss": 0.4338,
+      "step": 1265
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.067471877889658e-06,
+      "loss": 0.44,
+      "step": 1266
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.0638894086696816e-06,
+      "loss": 0.5146,
+      "step": 1267
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.0603078628521456e-06,
+      "loss": 0.4594,
+      "step": 1268
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.056727248020467e-06,
+      "loss": 0.4478,
+      "step": 1269
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.05314757175609e-06,
+      "loss": 0.4862,
+      "step": 1270
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.049568841638472e-06,
+      "loss": 0.465,
+      "step": 1271
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.045991065245069e-06,
+      "loss": 0.5096,
+      "step": 1272
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.0424142501513156e-06,
+      "loss": 0.463,
+      "step": 1273
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.0388384039306105e-06,
+      "loss": 0.4457,
+      "step": 1274
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.0352635341543026e-06,
+      "loss": 0.398,
+      "step": 1275
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.0316896483916724e-06,
+      "loss": 0.4651,
+      "step": 1276
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.028116754209917e-06,
+      "loss": 0.5716,
+      "step": 1277
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.024544859174134e-06,
+      "loss": 0.4113,
+      "step": 1278
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.020973970847304e-06,
+      "loss": 0.4814,
+      "step": 1279
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.017404096790281e-06,
+      "loss": 0.4918,
+      "step": 1280
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.0138352445617654e-06,
+      "loss": 0.4379,
+      "step": 1281
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.0102674217182974e-06,
+      "loss": 0.4345,
+      "step": 1282
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.0067006358142367e-06,
+      "loss": 0.4182,
+      "step": 1283
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.0031348944017474e-06,
+      "loss": 0.4849,
+      "step": 1284
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9995702050307817e-06,
+      "loss": 0.4282,
+      "step": 1285
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.996006575249066e-06,
+      "loss": 0.4582,
+      "step": 1286
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.9924440126020805e-06,
+      "loss": 0.4859,
+      "step": 1287
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9888825246330486e-06,
+      "loss": 0.5555,
+      "step": 1288
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.985322118882916e-06,
+      "loss": 0.5336,
+      "step": 1289
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.981762802890338e-06,
+      "loss": 0.455,
+      "step": 1290
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9782045841916625e-06,
+      "loss": 0.4537,
+      "step": 1291
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.974647470320914e-06,
+      "loss": 0.4446,
+      "step": 1292
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9710914688097775e-06,
+      "loss": 0.4922,
+      "step": 1293
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.9675365871875827e-06,
+      "loss": 0.4738,
+      "step": 1294
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.963982832981287e-06,
+      "loss": 0.4916,
+      "step": 1295
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.960430213715464e-06,
+      "loss": 0.4835,
+      "step": 1296
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9568787369122814e-06,
+      "loss": 0.4826,
+      "step": 1297
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9533284100914884e-06,
+      "loss": 0.4993,
+      "step": 1298
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9497792407703994e-06,
+      "loss": 0.5285,
+      "step": 1299
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.946231236463878e-06,
+      "loss": 0.5115,
+      "step": 1300
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9426844046843207e-06,
+      "loss": 0.45,
+      "step": 1301
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.9391387529416425e-06,
+      "loss": 0.4645,
+      "step": 1302
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.935594288743257e-06,
+      "loss": 0.4742,
+      "step": 1303
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.9320510195940687e-06,
+      "loss": 0.485,
+      "step": 1304
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.9285089529964467e-06,
+      "loss": 0.4599,
+      "step": 1305
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.924968096450215e-06,
+      "loss": 0.4331,
+      "step": 1306
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.921428457452637e-06,
+      "loss": 0.4596,
+      "step": 1307
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.917890043498397e-06,
+      "loss": 0.5332,
+      "step": 1308
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.9143528620795852e-06,
+      "loss": 0.497,
+      "step": 1309
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.9108169206856833e-06,
+      "loss": 0.5316,
+      "step": 1310
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.907282226803545e-06,
+      "loss": 0.5171,
+      "step": 1311
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.9037487879173866e-06,
+      "loss": 0.4177,
+      "step": 1312
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.9002166115087634e-06,
+      "loss": 0.5167,
+      "step": 1313
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.896685705056559e-06,
+      "loss": 0.4589,
+      "step": 1314
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8931560760369682e-06,
+      "loss": 0.4782,
+      "step": 1315
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8896277319234806e-06,
+      "loss": 0.4445,
+      "step": 1316
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8861006801868656e-06,
+      "loss": 0.4836,
+      "step": 1317
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8825749282951553e-06,
+      "loss": 0.5183,
+      "step": 1318
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8790504837136305e-06,
+      "loss": 0.4726,
+      "step": 1319
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8755273539048047e-06,
+      "loss": 0.5342,
+      "step": 1320
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8720055463284053e-06,
+      "loss": 0.4408,
+      "step": 1321
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8684850684413622e-06,
+      "loss": 0.4668,
+      "step": 1322
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8649659276977878e-06,
+      "loss": 0.4443,
+      "step": 1323
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8614481315489655e-06,
+      "loss": 0.4375,
+      "step": 1324
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8579316874433306e-06,
+      "loss": 0.5268,
+      "step": 1325
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8544166028264548e-06,
+      "loss": 0.4627,
+      "step": 1326
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.850902885141033e-06,
+      "loss": 0.4351,
+      "step": 1327
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8473905418268649e-06,
+      "loss": 0.499,
+      "step": 1328
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8438795803208394e-06,
+      "loss": 0.4845,
+      "step": 1329
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.840370008056922e-06,
+      "loss": 0.4637,
+      "step": 1330
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8368618324661345e-06,
+      "loss": 0.5191,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8333550609765417e-06,
+      "loss": 0.4301,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.829849701013236e-06,
+      "loss": 0.4472,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.826345759998321e-06,
+      "loss": 0.4687,
+      "step": 1334
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8228432453508942e-06,
+      "loss": 0.4919,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.819342164487038e-06,
+      "loss": 0.5069,
+      "step": 1336
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8158425248197931e-06,
+      "loss": 0.4394,
+      "step": 1337
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8123443337591511e-06,
+      "loss": 0.4474,
+      "step": 1338
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8088475987120373e-06,
+      "loss": 0.4324,
+      "step": 1339
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.805352327082293e-06,
+      "loss": 0.4461,
+      "step": 1340
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.80185852627066e-06,
+      "loss": 0.4395,
+      "step": 1341
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7983662036747682e-06,
+      "loss": 0.5062,
+      "step": 1342
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7948753666891144e-06,
+      "loss": 0.462,
+      "step": 1343
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7913860227050544e-06,
+      "loss": 0.5352,
+      "step": 1344
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7878981791107791e-06,
+      "loss": 0.4158,
+      "step": 1345
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.784411843291303e-06,
+      "loss": 0.4624,
+      "step": 1346
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7809270226284493e-06,
+      "loss": 0.4923,
+      "step": 1347
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7774437245008319e-06,
+      "loss": 0.469,
+      "step": 1348
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7739619562838417e-06,
+      "loss": 0.5253,
+      "step": 1349
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.77048172534963e-06,
+      "loss": 0.4501,
+      "step": 1350
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7670030390670923e-06,
+      "loss": 0.4399,
+      "step": 1351
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7635259048018561e-06,
+      "loss": 0.497,
+      "step": 1352
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7600503299162603e-06,
+      "loss": 0.4803,
+      "step": 1353
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7565763217693427e-06,
+      "loss": 0.441,
+      "step": 1354
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7531038877168232e-06,
+      "loss": 0.4071,
+      "step": 1355
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7496330351110892e-06,
+      "loss": 0.492,
+      "step": 1356
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.746163771301181e-06,
+      "loss": 0.4948,
+      "step": 1357
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.742696103632773e-06,
+      "loss": 0.4436,
+      "step": 1358
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7392300394481599e-06,
+      "loss": 0.4811,
+      "step": 1359
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.735765586086243e-06,
+      "loss": 0.523,
+      "step": 1360
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7323027508825113e-06,
+      "loss": 0.502,
+      "step": 1361
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7288415411690279e-06,
+      "loss": 0.4596,
+      "step": 1362
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7253819642744152e-06,
+      "loss": 0.503,
+      "step": 1363
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.721924027523837e-06,
+      "loss": 0.4281,
+      "step": 1364
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.718467738238985e-06,
+      "loss": 0.4821,
+      "step": 1365
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7150131037380625e-06,
+      "loss": 0.5685,
+      "step": 1366
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7115601313357676e-06,
+      "loss": 0.4378,
+      "step": 1367
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.708108828343283e-06,
+      "loss": 0.471,
+      "step": 1368
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7046592020682535e-06,
+      "loss": 0.5385,
+      "step": 1369
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7012112598147735e-06,
+      "loss": 0.5294,
+      "step": 1370
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6977650088833734e-06,
+      "loss": 0.4645,
+      "step": 1371
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6943204565710013e-06,
+      "loss": 0.4618,
+      "step": 1372
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6908776101710093e-06,
+      "loss": 0.4226,
+      "step": 1373
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.6874364769731372e-06,
+      "loss": 0.504,
+      "step": 1374
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.683997064263496e-06,
+      "loss": 0.491,
+      "step": 1375
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.680559379324558e-06,
+      "loss": 0.5104,
+      "step": 1376
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6771234294351329e-06,
+      "loss": 0.4912,
+      "step": 1377
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6736892218703587e-06,
+      "loss": 0.535,
+      "step": 1378
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6702567639016832e-06,
+      "loss": 0.5251,
+      "step": 1379
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6668260627968502e-06,
+      "loss": 0.45,
+      "step": 1380
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6633971258198844e-06,
+      "loss": 0.4906,
+      "step": 1381
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.6599699602310734e-06,
+      "loss": 0.3786,
+      "step": 1382
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.656544573286955e-06,
+      "loss": 0.5122,
+      "step": 1383
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.653120972240303e-06,
+      "loss": 0.4583,
+      "step": 1384
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6496991643401064e-06,
+      "loss": 0.4812,
+      "step": 1385
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6462791568315598e-06,
+      "loss": 0.4433,
+      "step": 1386
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6428609569560444e-06,
+      "loss": 0.4194,
+      "step": 1387
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6394445719511148e-06,
+      "loss": 0.4787,
+      "step": 1388
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6360300090504827e-06,
+      "loss": 0.5076,
+      "step": 1389
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6326172754840011e-06,
+      "loss": 0.4698,
+      "step": 1390
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.6292063784776508e-06,
+      "loss": 0.4747,
+      "step": 1391
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.625797325253523e-06,
+      "loss": 0.4298,
+      "step": 1392
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6223901230298062e-06,
+      "loss": 0.5069,
+      "step": 1393
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.618984779020768e-06,
+      "loss": 0.6336,
+      "step": 1394
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.615581300436742e-06,
+      "loss": 0.5049,
+      "step": 1395
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.612179694484113e-06,
+      "loss": 0.5074,
+      "step": 1396
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6087799683653e-06,
+      "loss": 0.4639,
+      "step": 1397
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.6053821292787414e-06,
+      "loss": 0.5134,
+      "step": 1398
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.60198618441888e-06,
+      "loss": 0.5235,
+      "step": 1399
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.598592140976149e-06,
+      "loss": 0.5051,
+      "step": 1400
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5952000061369555e-06,
+      "loss": 0.4739,
+      "step": 1401
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5918097870836642e-06,
+      "loss": 0.4798,
+      "step": 1402
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.588421490994584e-06,
+      "loss": 0.4877,
+      "step": 1403
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5850351250439518e-06,
+      "loss": 0.4887,
+      "step": 1404
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.581650696401919e-06,
+      "loss": 0.5062,
+      "step": 1405
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.5782682122345333e-06,
+      "loss": 0.5253,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5748876797037258e-06,
+      "loss": 0.4418,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.571509105967297e-06,
+      "loss": 0.4748,
+      "step": 1408
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5681324981788978e-06,
+      "loss": 0.5507,
+      "step": 1409
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5647578634880174e-06,
+      "loss": 0.5254,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5613852090399666e-06,
+      "loss": 0.493,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.5580145419758642e-06,
+      "loss": 0.4661,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.55464586943262e-06,
+      "loss": 0.5035,
+      "step": 1413
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5512791985429214e-06,
+      "loss": 0.4757,
+      "step": 1414
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5479145364352166e-06,
+      "loss": 0.5146,
+      "step": 1415
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5445518902337036e-06,
+      "loss": 0.4561,
+      "step": 1416
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5411912670583084e-06,
+      "loss": 0.4807,
+      "step": 1417
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5378326740246746e-06,
+      "loss": 0.4434,
+      "step": 1418
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5344761182441476e-06,
+      "loss": 0.4957,
+      "step": 1419
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.5311216068237595e-06,
+      "loss": 0.4948,
+      "step": 1420
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5277691468662128e-06,
+      "loss": 0.4725,
+      "step": 1421
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5244187454698673e-06,
+      "loss": 0.5028,
+      "step": 1422
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5210704097287226e-06,
+      "loss": 0.4492,
+      "step": 1423
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5177241467324078e-06,
+      "loss": 0.4737,
+      "step": 1424
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5143799635661593e-06,
+      "loss": 0.5261,
+      "step": 1425
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5110378673108124e-06,
+      "loss": 0.4802,
+      "step": 1426
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5076978650427825e-06,
+      "loss": 0.5315,
+      "step": 1427
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.5043599638340512e-06,
+      "loss": 0.4941,
+      "step": 1428
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.5010241707521529e-06,
+      "loss": 0.5148,
+      "step": 1429
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.497690492860156e-06,
+      "loss": 0.5048,
+      "step": 1430
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4943589372166516e-06,
+      "loss": 0.5391,
+      "step": 1431
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4910295108757384e-06,
+      "loss": 0.4637,
+      "step": 1432
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.487702220887004e-06,
+      "loss": 0.556,
+      "step": 1433
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.484377074295515e-06,
+      "loss": 0.414,
+      "step": 1434
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4810540781417981e-06,
+      "loss": 0.5288,
+      "step": 1435
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4777332394618272e-06,
+      "loss": 0.5309,
+      "step": 1436
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4744145652870078e-06,
+      "loss": 0.4785,
+      "step": 1437
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4710980626441634e-06,
+      "loss": 0.5112,
+      "step": 1438
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4677837385555177e-06,
+      "loss": 0.4808,
+      "step": 1439
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.464471600038684e-06,
+      "loss": 0.5043,
+      "step": 1440
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4611616541066464e-06,
+      "loss": 0.4916,
+      "step": 1441
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4578539077677469e-06,
+      "loss": 0.4415,
+      "step": 1442
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.4545483680256705e-06,
+      "loss": 0.4154,
+      "step": 1443
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.451245041879428e-06,
+      "loss": 0.5162,
+      "step": 1444
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.447943936323346e-06,
+      "loss": 0.4572,
+      "step": 1445
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4446450583470492e-06,
+      "loss": 0.4697,
+      "step": 1446
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4413484149354428e-06,
+      "loss": 0.4425,
+      "step": 1447
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4380540130687033e-06,
+      "loss": 0.5209,
+      "step": 1448
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4347618597222619e-06,
+      "loss": 0.466,
+      "step": 1449
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.4314719618667843e-06,
+      "loss": 0.499,
+      "step": 1450
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4281843264681655e-06,
+      "loss": 0.5118,
+      "step": 1451
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4248989604875083e-06,
+      "loss": 0.531,
+      "step": 1452
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4216158708811085e-06,
+      "loss": 0.5211,
+      "step": 1453
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4183350646004451e-06,
+      "loss": 0.4538,
+      "step": 1454
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4150565485921586e-06,
+      "loss": 0.4864,
+      "step": 1455
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4117803297980464e-06,
+      "loss": 0.5141,
+      "step": 1456
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4085064151550348e-06,
+      "loss": 0.4534,
+      "step": 1457
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.4052348115951765e-06,
+      "loss": 0.5364,
+      "step": 1458
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.401965526045627e-06,
+      "loss": 0.4573,
+      "step": 1459
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3986985654286379e-06,
+      "loss": 0.4805,
+      "step": 1460
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3954339366615335e-06,
+      "loss": 0.4415,
+      "step": 1461
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3921716466567053e-06,
+      "loss": 0.4189,
+      "step": 1462
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3889117023215875e-06,
+      "loss": 0.5209,
+      "step": 1463
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.3856541105586545e-06,
+      "loss": 0.4862,
+      "step": 1464
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.382398878265393e-06,
+      "loss": 0.4511,
+      "step": 1465
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3791460123342986e-06,
+      "loss": 0.4852,
+      "step": 1466
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3758955196528523e-06,
+      "loss": 0.4454,
+      "step": 1467
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3726474071035145e-06,
+      "loss": 0.4932,
+      "step": 1468
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3694016815637013e-06,
+      "loss": 0.5615,
+      "step": 1469
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.366158349905779e-06,
+      "loss": 0.5356,
+      "step": 1470
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3629174189970408e-06,
+      "loss": 0.4442,
+      "step": 1471
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.3596788956997022e-06,
+      "loss": 0.4413,
+      "step": 1472
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3564427868708757e-06,
+      "loss": 0.5499,
+      "step": 1473
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3532090993625654e-06,
+      "loss": 0.481,
+      "step": 1474
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3499778400216447e-06,
+      "loss": 0.5212,
+      "step": 1475
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3467490156898499e-06,
+      "loss": 0.5635,
+      "step": 1476
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3435226332037576e-06,
+      "loss": 0.476,
+      "step": 1477
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.340298699394777e-06,
+      "loss": 0.4587,
+      "step": 1478
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.337077221089131e-06,
+      "loss": 0.4198,
+      "step": 1479
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.3338582051078455e-06,
+      "loss": 0.4588,
+      "step": 1480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.3306416582667287e-06,
+      "loss": 0.5145,
+      "step": 1481
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.3274275873763655e-06,
+      "loss": 0.4311,
+      "step": 1482
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.3242159992420942e-06,
+      "loss": 0.4604,
+      "step": 1483
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.3210069006639986e-06,
+      "loss": 0.5721,
+      "step": 1484
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2226,
+  "num_train_epochs": 3,
+  "save_steps": 10000.0,
+  "total_flos": 2.1128979567712666e+18,
+  "trial_name": null,
+  "trial_params": null
+}