diff --git "a/checkpoint-2129/trainer_state.json" "b/checkpoint-2129/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2129/trainer_state.json"
@@ -0,0 +1,12799 @@
+{
+  "best_metric": 0.5271788835525513,
+  "best_model_checkpoint": "./medium-article-titles-engagement/checkpoint-2129",
+  "epoch": 1.0,
+  "global_step": 2129,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6935,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7465,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 0.6845,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6868,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 0.708,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 0.7092,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.692,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7088,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-06,
+      "loss": 0.696,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.6625,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.6762,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6914,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6911,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.7021,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7048,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6843,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.6937,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6936,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6889,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 0.6974,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6827,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.698,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6906,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6866,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5e-05,
+      "loss": 0.6928,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.6814,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.6826,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.6831,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9e-05,
+      "loss": 0.6858,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 0.694,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1e-05,
+      "loss": 0.6806,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6793,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3e-05,
+      "loss": 0.6711,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.6556,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6697,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6e-05,
+      "loss": 0.6284,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7e-05,
+      "loss": 0.6936,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8e-05,
+      "loss": 0.6752,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.6749,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4e-05,
+      "loss": 0.6827,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1e-05,
+      "loss": 0.6976,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2e-05,
+      "loss": 0.67,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3e-05,
+      "loss": 0.735,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.7292,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6745,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6675,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7e-05,
+      "loss": 0.7007,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8e-05,
+      "loss": 0.7077,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9e-05,
+      "loss": 0.6693,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.6984,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.997594997594998e-05,
+      "loss": 0.6669,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.995189995189995e-05,
+      "loss": 0.6999,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.992784992784993e-05,
+      "loss": 0.7029,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.990379990379991e-05,
+      "loss": 0.6778,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.987974987974988e-05,
+      "loss": 0.704,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.985569985569986e-05,
+      "loss": 0.6622,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.983164983164983e-05,
+      "loss": 0.69,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.980759980759981e-05,
+      "loss": 0.6777,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.978354978354979e-05,
+      "loss": 0.6815,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9759499759499764e-05,
+      "loss": 0.659,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.973544973544973e-05,
+      "loss": 0.6293,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.971139971139971e-05,
+      "loss": 0.6304,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.968734968734969e-05,
+      "loss": 0.7037,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.966329966329967e-05,
+      "loss": 0.742,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.963924963924964e-05,
+      "loss": 0.5779,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.961519961519962e-05,
+      "loss": 0.7145,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9591149591149594e-05,
+      "loss": 0.6952,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.956709956709957e-05,
+      "loss": 0.6343,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9543049543049543e-05,
+      "loss": 0.7037,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.951899951899952e-05,
+      "loss": 0.6963,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.94949494949495e-05,
+      "loss": 0.6653,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9470899470899475e-05,
+      "loss": 0.6794,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.944684944684945e-05,
+      "loss": 0.6655,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9422799422799424e-05,
+      "loss": 0.703,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.93987493987494e-05,
+      "loss": 0.6642,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.937469937469938e-05,
+      "loss": 0.6709,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9350649350649355e-05,
+      "loss": 0.7228,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.932659932659932e-05,
+      "loss": 0.6628,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9302549302549305e-05,
+      "loss": 0.6729,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.927849927849928e-05,
+      "loss": 0.6084,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.925444925444926e-05,
+      "loss": 0.6587,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.923039923039923e-05,
+      "loss": 0.6244,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 0.6938,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9182299182299185e-05,
+      "loss": 0.7117,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.915824915824916e-05,
+      "loss": 0.6753,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9134199134199135e-05,
+      "loss": 0.6704,
+      "step": 86
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.911014911014911e-05,
+      "loss": 0.691,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.908609908609909e-05,
+      "loss": 0.6738,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9062049062049066e-05,
+      "loss": 0.6613,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.903799903799904e-05,
+      "loss": 0.6782,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9013949013949016e-05,
+      "loss": 0.6855,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.898989898989899e-05,
+      "loss": 0.73,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.896584896584897e-05,
+      "loss": 0.6897,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.894179894179895e-05,
+      "loss": 0.6697,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8917748917748915e-05,
+      "loss": 0.6764,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8893698893698896e-05,
+      "loss": 0.6575,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.886964886964887e-05,
+      "loss": 0.6608,
+      "step": 97
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.884559884559885e-05,
+      "loss": 0.6501,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.882154882154882e-05,
+      "loss": 0.6592,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8797498797498795e-05,
+      "loss": 0.7097,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.877344877344878e-05,
+      "loss": 0.645,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.874939874939875e-05,
+      "loss": 0.6545,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8725348725348726e-05,
+      "loss": 0.6626,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.87012987012987e-05,
+      "loss": 0.6545,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8677248677248676e-05,
+      "loss": 0.6749,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.865319865319866e-05,
+      "loss": 0.6668,
+      "step": 106
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.862914862914863e-05,
+      "loss": 0.6735,
+      "step": 107
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.860509860509861e-05,
+      "loss": 0.6934,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.858104858104858e-05,
+      "loss": 0.7042,
+      "step": 109
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8556998556998563e-05,
+      "loss": 0.6509,
+      "step": 110
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.853294853294854e-05,
+      "loss": 0.671,
+      "step": 111
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8508898508898506e-05,
+      "loss": 0.6566,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.6733,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.846079846079846e-05,
+      "loss": 0.6742,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.8436748436748444e-05,
+      "loss": 0.6458,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.7046,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.838864838864839e-05,
+      "loss": 0.6129,
+      "step": 117
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.836459836459837e-05,
+      "loss": 0.6944,
+      "step": 118
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.834054834054834e-05,
+      "loss": 0.6438,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.831649831649832e-05,
+      "loss": 0.596,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.829244829244829e-05,
+      "loss": 0.6511,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.826839826839827e-05,
+      "loss": 0.6645,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.824434824434825e-05,
+      "loss": 0.6965,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8220298220298224e-05,
+      "loss": 0.7112,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.81962481962482e-05,
+      "loss": 0.7012,
+      "step": 125
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8172198172198173e-05,
+      "loss": 0.7151,
+      "step": 126
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.6961,
+      "step": 127
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.812409812409813e-05,
+      "loss": 0.7191,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.81000481000481e-05,
+      "loss": 0.667,
+      "step": 129
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.807599807599808e-05,
+      "loss": 0.66,
+      "step": 130
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8051948051948054e-05,
+      "loss": 0.6604,
+      "step": 131
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8027898027898036e-05,
+      "loss": 0.678,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8003848003848004e-05,
+      "loss": 0.6344,
+      "step": 133
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.797979797979798e-05,
+      "loss": 0.6489,
+      "step": 134
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.795574795574796e-05,
+      "loss": 0.7045,
+      "step": 135
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7931697931697935e-05,
+      "loss": 0.6623,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.790764790764791e-05,
+      "loss": 0.6769,
+      "step": 137
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.7883597883597884e-05,
+      "loss": 0.676,
+      "step": 138
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.785954785954786e-05,
+      "loss": 0.655,
+      "step": 139
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.783549783549784e-05,
+      "loss": 0.6957,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7811447811447815e-05,
+      "loss": 0.679,
+      "step": 141
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.778739778739779e-05,
+      "loss": 0.6377,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7763347763347765e-05,
+      "loss": 0.6941,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.773929773929774e-05,
+      "loss": 0.6239,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.771524771524772e-05,
+      "loss": 0.6856,
+      "step": 145
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.769119769119769e-05,
+      "loss": 0.725,
+      "step": 146
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.766714766714767e-05,
+      "loss": 0.6533,
+      "step": 147
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7643097643097646e-05,
+      "loss": 0.624,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.6726,
+      "step": 149
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7594997594997595e-05,
+      "loss": 0.6692,
+      "step": 150
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.757094757094757e-05,
+      "loss": 0.6154,
+      "step": 151
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.754689754689755e-05,
+      "loss": 0.6247,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7522847522847526e-05,
+      "loss": 0.6856,
+      "step": 153
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.74987974987975e-05,
+      "loss": 0.6448,
+      "step": 154
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.7474747474747476e-05,
+      "loss": 0.6696,
+      "step": 155
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.745069745069745e-05,
+      "loss": 0.6549,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.742664742664743e-05,
+      "loss": 0.6853,
+      "step": 157
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.740259740259741e-05,
+      "loss": 0.6611,
+      "step": 158
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.737854737854738e-05,
+      "loss": 0.6509,
+      "step": 159
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.7354497354497356e-05,
+      "loss": 0.6566,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.733044733044733e-05,
+      "loss": 0.6764,
+      "step": 161
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.730639730639731e-05,
+      "loss": 0.726,
+      "step": 162
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.728234728234728e-05,
+      "loss": 0.6275,
+      "step": 163
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.725829725829726e-05,
+      "loss": 0.6662,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.723424723424724e-05,
+      "loss": 0.6747,
+      "step": 165
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.721019721019721e-05,
+      "loss": 0.7087,
+      "step": 166
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.718614718614719e-05,
+      "loss": 0.7048,
+      "step": 167
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.716209716209716e-05,
+      "loss": 0.6678,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.713804713804714e-05,
+      "loss": 0.6433,
+      "step": 169
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.711399711399712e-05,
+      "loss": 0.6199,
+      "step": 170
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.708994708994709e-05,
+      "loss": 0.6796,
+      "step": 171
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.706589706589707e-05,
+      "loss": 0.6756,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.704184704184704e-05,
+      "loss": 0.6196,
+      "step": 173
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.7017797017797024e-05,
+      "loss": 0.6146,
+      "step": 174
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6993746993747e-05,
+      "loss": 0.6573,
+      "step": 175
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.696969696969697e-05,
+      "loss": 0.6366,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.694564694564695e-05,
+      "loss": 0.5738,
+      "step": 177
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.692159692159692e-05,
+      "loss": 0.6311,
+      "step": 178
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.6897546897546904e-05,
+      "loss": 0.6142,
+      "step": 179
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.687349687349687e-05,
+      "loss": 0.6423,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6849446849446854e-05,
+      "loss": 0.7501,
+      "step": 181
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.7152,
+      "step": 182
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.68013468013468e-05,
+      "loss": 0.6557,
+      "step": 183
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.677729677729678e-05,
+      "loss": 0.6271,
+      "step": 184
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.675324675324675e-05,
+      "loss": 0.6788,
+      "step": 185
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6729196729196734e-05,
+      "loss": 0.6789,
+      "step": 186
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.670514670514671e-05,
+      "loss": 0.6399,
+      "step": 187
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6681096681096684e-05,
+      "loss": 0.6399,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.665704665704666e-05,
+      "loss": 0.661,
+      "step": 189
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6632996632996634e-05,
+      "loss": 0.6606,
+      "step": 190
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6608946608946615e-05,
+      "loss": 0.6252,
+      "step": 191
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.658489658489659e-05,
+      "loss": 0.7178,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.656084656084656e-05,
+      "loss": 0.6959,
+      "step": 193
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.653679653679654e-05,
+      "loss": 0.6675,
+      "step": 194
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6512746512746514e-05,
+      "loss": 0.6677,
+      "step": 195
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6488696488696496e-05,
+      "loss": 0.6591,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6464646464646464e-05,
+      "loss": 0.5972,
+      "step": 197
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6440596440596445e-05,
+      "loss": 0.6891,
+      "step": 198
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.641654641654642e-05,
+      "loss": 0.6233,
+      "step": 199
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6392496392496395e-05,
+      "loss": 0.6808,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.636844636844637e-05,
+      "loss": 0.701,
+      "step": 201
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.6344396344396344e-05,
+      "loss": 0.6508,
+      "step": 202
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.6320346320346326e-05,
+      "loss": 0.7082,
+      "step": 203
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.7773,
+      "step": 204
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.6272246272246276e-05,
+      "loss": 0.6586,
+      "step": 205
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.624819624819625e-05,
+      "loss": 0.6402,
+      "step": 206
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.6224146224146225e-05,
+      "loss": 0.723,
+      "step": 207
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.620009620009621e-05,
+      "loss": 0.6088,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.617604617604618e-05,
+      "loss": 0.7157,
+      "step": 209
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.615199615199615e-05,
+      "loss": 0.7086,
+      "step": 210
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.612794612794613e-05,
+      "loss": 0.6648,
+      "step": 211
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.6103896103896106e-05,
+      "loss": 0.6719,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.607984607984609e-05,
+      "loss": 0.6579,
+      "step": 213
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.6055796055796055e-05,
+      "loss": 0.6999,
+      "step": 214
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.6085,
+      "step": 215
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.600769600769601e-05,
+      "loss": 0.6819,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5983645983645986e-05,
+      "loss": 0.6423,
+      "step": 217
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.595959595959596e-05,
+      "loss": 0.686,
+      "step": 218
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5935545935545936e-05,
+      "loss": 0.635,
+      "step": 219
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.591149591149592e-05,
+      "loss": 0.6773,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.588744588744589e-05,
+      "loss": 0.6521,
+      "step": 221
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.586339586339587e-05,
+      "loss": 0.5995,
+      "step": 222
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.583934583934584e-05,
+      "loss": 0.6685,
+      "step": 223
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.5815295815295817e-05,
+      "loss": 0.6526,
+      "step": 224
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.57912457912458e-05,
+      "loss": 0.6598,
+      "step": 225
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.576719576719577e-05,
+      "loss": 0.7117,
+      "step": 226
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.574314574314574e-05,
+      "loss": 0.7277,
+      "step": 227
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.571909571909572e-05,
+      "loss": 0.6349,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.56950456950457e-05,
+      "loss": 0.6862,
+      "step": 229
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.567099567099568e-05,
+      "loss": 0.6279,
+      "step": 230
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.564694564694565e-05,
+      "loss": 0.7042,
+      "step": 231
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.562289562289562e-05,
+      "loss": 0.6482,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.55988455988456e-05,
+      "loss": 0.6692,
+      "step": 233
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.557479557479558e-05,
+      "loss": 0.602,
+      "step": 234
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.555074555074555e-05,
+      "loss": 0.6345,
+      "step": 235
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.552669552669553e-05,
+      "loss": 0.7093,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.55026455026455e-05,
+      "loss": 0.6626,
+      "step": 237
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.5478595478595484e-05,
+      "loss": 0.6935,
+      "step": 238
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.6689,
+      "step": 239
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.543049543049543e-05,
+      "loss": 0.6349,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.540644540644541e-05,
+      "loss": 0.6312,
+      "step": 241
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.538239538239538e-05,
+      "loss": 0.6249,
+      "step": 242
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.535834535834536e-05,
+      "loss": 0.6174,
+      "step": 243
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.533429533429533e-05,
+      "loss": 0.622,
+      "step": 244
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5310245310245314e-05,
+      "loss": 0.6176,
+      "step": 245
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.528619528619529e-05,
+      "loss": 0.649,
+      "step": 246
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5262145262145264e-05,
+      "loss": 0.6682,
+      "step": 247
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.6501,
+      "step": 248
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.521404521404521e-05,
+      "loss": 0.7281,
+      "step": 249
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5189995189995195e-05,
+      "loss": 0.6559,
+      "step": 250
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.516594516594517e-05,
+      "loss": 0.6048,
+      "step": 251
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5141895141895144e-05,
+      "loss": 0.618,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.511784511784512e-05,
+      "loss": 0.6391,
+      "step": 253
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5093795093795094e-05,
+      "loss": 0.6711,
+      "step": 254
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5069745069745075e-05,
+      "loss": 0.681,
+      "step": 255
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.504569504569504e-05,
+      "loss": 0.6134,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.5021645021645025e-05,
+      "loss": 0.6665,
+      "step": 257
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4997594997595e-05,
+      "loss": 0.7127,
+      "step": 258
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4973544973544974e-05,
+      "loss": 0.6303,
+      "step": 259
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.494949494949495e-05,
+      "loss": 0.6481,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4925444925444924e-05,
+      "loss": 0.6319,
+      "step": 261
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4901394901394906e-05,
+      "loss": 0.6908,
+      "step": 262
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.487734487734488e-05,
+      "loss": 0.6061,
+      "step": 263
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4853294853294855e-05,
+      "loss": 0.6348,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.482924482924483e-05,
+      "loss": 0.6309,
+      "step": 265
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.4805194805194805e-05,
+      "loss": 0.6806,
+      "step": 266
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4781144781144786e-05,
+      "loss": 0.6669,
+      "step": 267
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.475709475709476e-05,
+      "loss": 0.6119,
+      "step": 268
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4733044733044736e-05,
+      "loss": 0.6197,
+      "step": 269
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.470899470899471e-05,
+      "loss": 0.631,
+      "step": 270
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4684944684944685e-05,
+      "loss": 0.6865,
+      "step": 271
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.466089466089467e-05,
+      "loss": 0.6309,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4636844636844635e-05,
+      "loss": 0.6344,
+      "step": 273
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4612794612794616e-05,
+      "loss": 0.553,
+      "step": 274
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.458874458874459e-05,
+      "loss": 0.6579,
+      "step": 275
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4564694564694566e-05,
+      "loss": 0.6559,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.454064454064454e-05,
+      "loss": 0.6647,
+      "step": 277
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4516594516594515e-05,
+      "loss": 0.6889,
+      "step": 278
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.44925444925445e-05,
+      "loss": 0.6517,
+      "step": 279
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.446849446849447e-05,
+      "loss": 0.676,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5925,
+      "step": 281
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.442039442039442e-05,
+      "loss": 0.6424,
+      "step": 282
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4396344396344396e-05,
+      "loss": 0.6869,
+      "step": 283
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.437229437229438e-05,
+      "loss": 0.666,
+      "step": 284
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.434824434824435e-05,
+      "loss": 0.6466,
+      "step": 285
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.432419432419432e-05,
+      "loss": 0.647,
+      "step": 286
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.43001443001443e-05,
+      "loss": 0.6834,
+      "step": 287
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.427609427609428e-05,
+      "loss": 0.6097,
+      "step": 288
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.425204425204426e-05,
+      "loss": 0.6532,
+      "step": 289
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.4227994227994226e-05,
+      "loss": 0.5948,
+      "step": 290
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.420394420394421e-05,
+      "loss": 0.6226,
+      "step": 291
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.417989417989418e-05,
+      "loss": 0.6076,
+      "step": 292
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.415584415584416e-05,
+      "loss": 0.5432,
+      "step": 293
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.413179413179413e-05,
+      "loss": 0.6886,
+      "step": 294
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.410774410774411e-05,
+      "loss": 0.6683,
+      "step": 295
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.408369408369409e-05,
+      "loss": 0.6347,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.405964405964406e-05,
+      "loss": 0.6743,
+      "step": 297
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.403559403559404e-05,
+      "loss": 0.6558,
+      "step": 298
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.401154401154401e-05,
+      "loss": 0.6421,
+      "step": 299
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.398749398749399e-05,
+      "loss": 0.6614,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.396344396344397e-05,
+      "loss": 0.6536,
+      "step": 301
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3939393939393944e-05,
+      "loss": 0.5608,
+      "step": 302
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.391534391534391e-05,
+      "loss": 0.6557,
+      "step": 303
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.3891293891293894e-05,
+      "loss": 0.6794,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.386724386724387e-05,
+      "loss": 0.7131,
+      "step": 305
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.384319384319385e-05,
+      "loss": 0.6494,
+      "step": 306
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.381914381914382e-05,
+      "loss": 0.6849,
+      "step": 307
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.379509379509379e-05,
+      "loss": 0.679,
+      "step": 308
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3771043771043774e-05,
+      "loss": 0.6143,
+      "step": 309
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.374699374699375e-05,
+      "loss": 0.6622,
+      "step": 310
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3722943722943724e-05,
+      "loss": 0.65,
+      "step": 311
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.36988936988937e-05,
+      "loss": 0.6945,
+      "step": 312
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.367484367484368e-05,
+      "loss": 0.6396,
+      "step": 313
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.6403,
+      "step": 314
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.362674362674363e-05,
+      "loss": 0.6491,
+      "step": 315
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3602693602693604e-05,
+      "loss": 0.6274,
+      "step": 316
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.357864357864358e-05,
+      "loss": 0.6624,
+      "step": 317
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.355459355459356e-05,
+      "loss": 0.6903,
+      "step": 318
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3530543530543535e-05,
+      "loss": 0.6528,
+      "step": 319
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3506493506493503e-05,
+      "loss": 0.6226,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3482443482443485e-05,
+      "loss": 0.6591,
+      "step": 321
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.345839345839346e-05,
+      "loss": 0.6087,
+      "step": 322
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.343434343434344e-05,
+      "loss": 0.6646,
+      "step": 323
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.341029341029341e-05,
+      "loss": 0.624,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3386243386243384e-05,
+      "loss": 0.6495,
+      "step": 325
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3362193362193366e-05,
+      "loss": 0.6223,
+      "step": 326
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.333814333814334e-05,
+      "loss": 0.6165,
+      "step": 327
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.3314093314093315e-05,
+      "loss": 0.6515,
+      "step": 328
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.329004329004329e-05,
+      "loss": 0.5706,
+      "step": 329
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3265993265993265e-05,
+      "loss": 0.6959,
+      "step": 330
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3241943241943246e-05,
+      "loss": 0.6424,
+      "step": 331
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.321789321789322e-05,
+      "loss": 0.5986,
+      "step": 332
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3193843193843196e-05,
+      "loss": 0.6824,
+      "step": 333
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.316979316979317e-05,
+      "loss": 0.7028,
+      "step": 334
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.314574314574315e-05,
+      "loss": 0.7249,
+      "step": 335
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.312169312169313e-05,
+      "loss": 0.6908,
+      "step": 336
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3097643097643095e-05,
+      "loss": 0.5728,
+      "step": 337
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3073593073593077e-05,
+      "loss": 0.6195,
+      "step": 338
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.304954304954305e-05,
+      "loss": 0.68,
+      "step": 339
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.302549302549303e-05,
+      "loss": 0.6816,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.3001443001443e-05,
+      "loss": 0.6327,
+      "step": 341
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2977392977392976e-05,
+      "loss": 0.6417,
+      "step": 342
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.295334295334296e-05,
+      "loss": 0.6589,
+      "step": 343
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.292929292929293e-05,
+      "loss": 0.6895,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.290524290524291e-05,
+      "loss": 0.7106,
+      "step": 345
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.288119288119288e-05,
+      "loss": 0.662,
+      "step": 346
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.6387,
+      "step": 347
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.283309283309284e-05,
+      "loss": 0.606,
+      "step": 348
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.280904280904281e-05,
+      "loss": 0.6685,
+      "step": 349
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.278499278499279e-05,
+      "loss": 0.6482,
+      "step": 350
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.276094276094276e-05,
+      "loss": 0.6972,
+      "step": 351
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.273689273689274e-05,
+      "loss": 0.6413,
+      "step": 352
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.271284271284272e-05,
+      "loss": 0.6147,
+      "step": 353
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2688792688792686e-05,
+      "loss": 0.6859,
+      "step": 354
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.266474266474267e-05,
+      "loss": 0.666,
+      "step": 355
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.264069264069264e-05,
+      "loss": 0.6635,
+      "step": 356
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2616642616642624e-05,
+      "loss": 0.6955,
+      "step": 357
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.6853,
+      "step": 358
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.256854256854257e-05,
+      "loss": 0.6581,
+      "step": 359
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.254449254449255e-05,
+      "loss": 0.6113,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2520442520442523e-05,
+      "loss": 0.6279,
+      "step": 361
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.24963924963925e-05,
+      "loss": 0.6222,
+      "step": 362
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.247234247234247e-05,
+      "loss": 0.6719,
+      "step": 363
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.244829244829245e-05,
+      "loss": 0.6231,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.6435,
+      "step": 365
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2400192400192404e-05,
+      "loss": 0.7021,
+      "step": 366
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.237614237614238e-05,
+      "loss": 0.6278,
+      "step": 367
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.2352092352092354e-05,
+      "loss": 0.6128,
+      "step": 368
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.232804232804233e-05,
+      "loss": 0.6021,
+      "step": 369
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.230399230399231e-05,
+      "loss": 0.5992,
+      "step": 370
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.227994227994228e-05,
+      "loss": 0.7209,
+      "step": 371
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.225589225589226e-05,
+      "loss": 0.6713,
+      "step": 372
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.2231842231842234e-05,
+      "loss": 0.6105,
+      "step": 373
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.220779220779221e-05,
+      "loss": 0.6236,
+      "step": 374
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.2183742183742184e-05,
+      "loss": 0.6422,
+      "step": 375
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.215969215969216e-05,
+      "loss": 0.6745,
+      "step": 376
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.213564213564214e-05,
+      "loss": 0.675,
+      "step": 377
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.2111592111592115e-05,
+      "loss": 0.6201,
+      "step": 378
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.208754208754209e-05,
+      "loss": 0.6874,
+      "step": 379
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.6211,
+      "step": 380
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.203944203944204e-05,
+      "loss": 0.6249,
+      "step": 381
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.201539201539202e-05,
+      "loss": 0.6272,
+      "step": 382
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1991341991341996e-05,
+      "loss": 0.6479,
+      "step": 383
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.196729196729197e-05,
+      "loss": 0.6615,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1943241943241945e-05,
+      "loss": 0.5752,
+      "step": 385
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.191919191919192e-05,
+      "loss": 0.6305,
+      "step": 386
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.18951418951419e-05,
+      "loss": 0.5748,
+      "step": 387
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.187109187109187e-05,
+      "loss": 0.6582,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.184704184704185e-05,
+      "loss": 0.5888,
+      "step": 389
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1822991822991826e-05,
+      "loss": 0.5897,
+      "step": 390
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.17989417989418e-05,
+      "loss": 0.588,
+      "step": 391
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.1774891774891775e-05,
+      "loss": 0.6358,
+      "step": 392
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.175084175084175e-05,
+      "loss": 0.644,
+      "step": 393
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.172679172679173e-05,
+      "loss": 0.6771,
+      "step": 394
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.1702741702741707e-05,
+      "loss": 0.6212,
+      "step": 395
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.167869167869168e-05,
+      "loss": 0.608,
+      "step": 396
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.1654641654641656e-05,
+      "loss": 0.7034,
+      "step": 397
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.163059163059163e-05,
+      "loss": 0.6611,
+      "step": 398
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.160654160654161e-05,
+      "loss": 0.6317,
+      "step": 399
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.158249158249159e-05,
+      "loss": 0.6195,
+      "step": 400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.155844155844156e-05,
+      "loss": 0.6128,
+      "step": 401
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.153439153439154e-05,
+      "loss": 0.6482,
+      "step": 402
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.151034151034151e-05,
+      "loss": 0.5449,
+      "step": 403
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.148629148629149e-05,
+      "loss": 0.6385,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.146224146224146e-05,
+      "loss": 0.679,
+      "step": 405
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.143819143819144e-05,
+      "loss": 0.6386,
+      "step": 406
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.141414141414142e-05,
+      "loss": 0.6404,
+      "step": 407
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.139009139009139e-05,
+      "loss": 0.6538,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.136604136604137e-05,
+      "loss": 0.6033,
+      "step": 409
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.134199134199134e-05,
+      "loss": 0.5656,
+      "step": 410
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.131794131794132e-05,
+      "loss": 0.6424,
+      "step": 411
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.12938912938913e-05,
+      "loss": 0.5866,
+      "step": 412
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 0.5342,
+      "step": 413
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.124579124579125e-05,
+      "loss": 0.61,
+      "step": 414
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.122174122174122e-05,
+      "loss": 0.6428,
+      "step": 415
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.1197691197691204e-05,
+      "loss": 0.6719,
+      "step": 416
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.117364117364118e-05,
+      "loss": 0.7196,
+      "step": 417
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.114959114959115e-05,
+      "loss": 0.6948,
+      "step": 418
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.112554112554113e-05,
+      "loss": 0.6133,
+      "step": 419
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.11014911014911e-05,
+      "loss": 0.5917,
+      "step": 420
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.1077441077441085e-05,
+      "loss": 0.6082,
+      "step": 421
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.105339105339105e-05,
+      "loss": 0.6246,
+      "step": 422
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.1029341029341034e-05,
+      "loss": 0.6809,
+      "step": 423
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.100529100529101e-05,
+      "loss": 0.5781,
+      "step": 424
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0981240981240984e-05,
+      "loss": 0.673,
+      "step": 425
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.095719095719096e-05,
+      "loss": 0.5997,
+      "step": 426
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.093314093314093e-05,
+      "loss": 0.6125,
+      "step": 427
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.6633,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.088504088504089e-05,
+      "loss": 0.648,
+      "step": 429
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0860990860990864e-05,
+      "loss": 0.606,
+      "step": 430
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.083694083694084e-05,
+      "loss": 0.6816,
+      "step": 431
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0812890812890814e-05,
+      "loss": 0.7163,
+      "step": 432
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0788840788840795e-05,
+      "loss": 0.6251,
+      "step": 433
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.0764790764790763e-05,
+      "loss": 0.6879,
+      "step": 434
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.6717,
+      "step": 435
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.071669071669072e-05,
+      "loss": 0.6223,
+      "step": 436
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0692640692640695e-05,
+      "loss": 0.6942,
+      "step": 437
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.066859066859067e-05,
+      "loss": 0.6097,
+      "step": 438
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0644540644540644e-05,
+      "loss": 0.6749,
+      "step": 439
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.062049062049062e-05,
+      "loss": 0.6123,
+      "step": 440
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.05964405964406e-05,
+      "loss": 0.671,
+      "step": 441
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0572390572390575e-05,
+      "loss": 0.6367,
+      "step": 442
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.054834054834055e-05,
+      "loss": 0.6265,
+      "step": 443
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0524290524290525e-05,
+      "loss": 0.6193,
+      "step": 444
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.05002405002405e-05,
+      "loss": 0.6369,
+      "step": 445
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.6419,
+      "step": 446
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.045214045214045e-05,
+      "loss": 0.6197,
+      "step": 447
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.042809042809043e-05,
+      "loss": 0.6618,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 0.5955,
+      "step": 449
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.037999037999039e-05,
+      "loss": 0.6007,
+      "step": 450
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0355940355940355e-05,
+      "loss": 0.5996,
+      "step": 451
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.033189033189033e-05,
+      "loss": 0.5865,
+      "step": 452
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.030784030784031e-05,
+      "loss": 0.563,
+      "step": 453
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0283790283790286e-05,
+      "loss": 0.5815,
+      "step": 454
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.025974025974026e-05,
+      "loss": 0.577,
+      "step": 455
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.0235690235690236e-05,
+      "loss": 0.6537,
+      "step": 456
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.021164021164021e-05,
+      "loss": 0.5944,
+      "step": 457
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.018759018759019e-05,
+      "loss": 0.5609,
+      "step": 458
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.016354016354017e-05,
+      "loss": 0.6994,
+      "step": 459
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.013949013949014e-05,
+      "loss": 0.6261,
+      "step": 460
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.0115440115440116e-05,
+      "loss": 0.6641,
+      "step": 461
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.009139009139009e-05,
+      "loss": 0.6832,
+      "step": 462
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.006734006734007e-05,
+      "loss": 0.6188,
+      "step": 463
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.004329004329004e-05,
+      "loss": 0.6432,
+      "step": 464
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.001924001924002e-05,
+      "loss": 0.6733,
+      "step": 465
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.999518999519e-05,
+      "loss": 0.6412,
+      "step": 466
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.997113997113997e-05,
+      "loss": 0.6507,
+      "step": 467
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9947089947089946e-05,
+      "loss": 0.6327,
+      "step": 468
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.992303992303992e-05,
+      "loss": 0.6467,
+      "step": 469
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.98989898989899e-05,
+      "loss": 0.6107,
+      "step": 470
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.987493987493988e-05,
+      "loss": 0.6296,
+      "step": 471
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.985088985088985e-05,
+      "loss": 0.5792,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.982683982683983e-05,
+      "loss": 0.5954,
+      "step": 473
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.98027898027898e-05,
+      "loss": 0.6593,
+      "step": 474
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.9778739778739783e-05,
+      "loss": 0.6318,
+      "step": 475
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.975468975468976e-05,
+      "loss": 0.591,
+      "step": 476
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.973063973063973e-05,
+      "loss": 0.6104,
+      "step": 477
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.970658970658971e-05,
+      "loss": 0.667,
+      "step": 478
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.6284,
+      "step": 479
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9658489658489664e-05,
+      "loss": 0.6232,
+      "step": 480
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.963443963443963e-05,
+      "loss": 0.716,
+      "step": 481
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9610389610389614e-05,
+      "loss": 0.6479,
+      "step": 482
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.958633958633959e-05,
+      "loss": 0.612,
+      "step": 483
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.956228956228956e-05,
+      "loss": 0.6288,
+      "step": 484
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.953823953823954e-05,
+      "loss": 0.634,
+      "step": 485
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.951418951418951e-05,
+      "loss": 0.6738,
+      "step": 486
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9490139490139494e-05,
+      "loss": 0.6308,
+      "step": 487
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.946608946608947e-05,
+      "loss": 0.7108,
+      "step": 488
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9442039442039444e-05,
+      "loss": 0.6766,
+      "step": 489
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.941798941798942e-05,
+      "loss": 0.6178,
+      "step": 490
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.6399,
+      "step": 491
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9369889369889375e-05,
+      "loss": 0.6411,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.934583934583935e-05,
+      "loss": 0.6236,
+      "step": 493
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9321789321789324e-05,
+      "loss": 0.5669,
+      "step": 494
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.92977392977393e-05,
+      "loss": 0.6086,
+      "step": 495
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9273689273689274e-05,
+      "loss": 0.5822,
+      "step": 496
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9249639249639256e-05,
+      "loss": 0.6082,
+      "step": 497
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9225589225589224e-05,
+      "loss": 0.5991,
+      "step": 498
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.9201539201539205e-05,
+      "loss": 0.6422,
+      "step": 499
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 3.917748917748918e-05,
+      "loss": 0.6364,
+      "step": 500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.9153439153439155e-05,
+      "loss": 0.6001,
+      "step": 501
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.912938912938913e-05,
+      "loss": 0.5817,
+      "step": 502
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.9105339105339104e-05,
+      "loss": 0.504,
+      "step": 503
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.9081289081289086e-05,
+      "loss": 0.6364,
+      "step": 504
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.905723905723906e-05,
+      "loss": 0.7341,
+      "step": 505
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.9033189033189035e-05,
+      "loss": 0.5903,
+      "step": 506
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.900913900913901e-05,
+      "loss": 0.663,
+      "step": 507
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8985088985088985e-05,
+      "loss": 0.682,
+      "step": 508
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8961038961038966e-05,
+      "loss": 0.6491,
+      "step": 509
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.893698893698894e-05,
+      "loss": 0.5924,
+      "step": 510
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.891293891293891e-05,
+      "loss": 0.5685,
+      "step": 511
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.6252,
+      "step": 512
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8864838864838866e-05,
+      "loss": 0.6154,
+      "step": 513
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.884078884078885e-05,
+      "loss": 0.656,
+      "step": 514
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8816738816738815e-05,
+      "loss": 0.5544,
+      "step": 515
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.87926887926888e-05,
+      "loss": 0.6224,
+      "step": 516
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.876863876863877e-05,
+      "loss": 0.6418,
+      "step": 517
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8744588744588746e-05,
+      "loss": 0.6673,
+      "step": 518
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.872053872053872e-05,
+      "loss": 0.6268,
+      "step": 519
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.8696488696488696e-05,
+      "loss": 0.5976,
+      "step": 520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 3.867243867243868e-05,
+      "loss": 0.6346,
+      "step": 521
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.864838864838865e-05,
+      "loss": 0.635,
+      "step": 522
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.862433862433863e-05,
+      "loss": 0.683,
+      "step": 523
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.86002886002886e-05,
+      "loss": 0.6263,
+      "step": 524
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.8576238576238576e-05,
+      "loss": 0.643,
+      "step": 525
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.855218855218856e-05,
+      "loss": 0.6014,
+      "step": 526
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.852813852813853e-05,
+      "loss": 0.6479,
+      "step": 527
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.85040885040885e-05,
+      "loss": 0.7262,
+      "step": 528
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.848003848003848e-05,
+      "loss": 0.6006,
+      "step": 529
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.845598845598846e-05,
+      "loss": 0.6532,
+      "step": 530
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.843193843193844e-05,
+      "loss": 0.6606,
+      "step": 531
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.8407888407888407e-05,
+      "loss": 0.6805,
+      "step": 532
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.838383838383838e-05,
+      "loss": 0.623,
+      "step": 533
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.835978835978836e-05,
+      "loss": 0.6241,
+      "step": 534
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.833573833573834e-05,
+      "loss": 0.6244,
+      "step": 535
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.831168831168831e-05,
+      "loss": 0.597,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.828763828763829e-05,
+      "loss": 0.6401,
+      "step": 537
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.826358826358827e-05,
+      "loss": 0.6133,
+      "step": 538
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.8239538239538244e-05,
+      "loss": 0.6102,
+      "step": 539
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.821548821548822e-05,
+      "loss": 0.6892,
+      "step": 540
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.819143819143819e-05,
+      "loss": 0.6219,
+      "step": 541
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 3.816738816738817e-05,
+      "loss": 0.681,
+      "step": 542
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.814333814333815e-05,
+      "loss": 0.6581,
+      "step": 543
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.8119288119288124e-05,
+      "loss": 0.6593,
+      "step": 544
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.6434,
+      "step": 545
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.8071188071188074e-05,
+      "loss": 0.5519,
+      "step": 546
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.804713804713805e-05,
+      "loss": 0.7145,
+      "step": 547
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.802308802308803e-05,
+      "loss": 0.6602,
+      "step": 548
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7999037999038e-05,
+      "loss": 0.6344,
+      "step": 549
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.797498797498797e-05,
+      "loss": 0.6487,
+      "step": 550
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7950937950937954e-05,
+      "loss": 0.6102,
+      "step": 551
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.792688792688793e-05,
+      "loss": 0.6488,
+      "step": 552
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7902837902837904e-05,
+      "loss": 0.5618,
+      "step": 553
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.6205,
+      "step": 554
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7854737854737854e-05,
+      "loss": 0.5776,
+      "step": 555
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7830687830687835e-05,
+      "loss": 0.5677,
+      "step": 556
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.780663780663781e-05,
+      "loss": 0.6491,
+      "step": 557
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7782587782587785e-05,
+      "loss": 0.5867,
+      "step": 558
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.775853775853776e-05,
+      "loss": 0.6127,
+      "step": 559
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.773448773448774e-05,
+      "loss": 0.6346,
+      "step": 560
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7710437710437716e-05,
+      "loss": 0.7072,
+      "step": 561
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7686387686387684e-05,
+      "loss": 0.5795,
+      "step": 562
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.7662337662337665e-05,
+      "loss": 0.6237,
+      "step": 563
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 3.763828763828764e-05,
+      "loss": 0.626,
+      "step": 564
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.761423761423762e-05,
+      "loss": 0.7194,
+      "step": 565
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.759018759018759e-05,
+      "loss": 0.4721,
+      "step": 566
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7566137566137564e-05,
+      "loss": 0.6833,
+      "step": 567
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7542087542087546e-05,
+      "loss": 0.6404,
+      "step": 568
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.751803751803752e-05,
+      "loss": 0.6086,
+      "step": 569
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7493987493987495e-05,
+      "loss": 0.5843,
+      "step": 570
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.746993746993747e-05,
+      "loss": 0.646,
+      "step": 571
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7445887445887445e-05,
+      "loss": 0.5741,
+      "step": 572
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7421837421837427e-05,
+      "loss": 0.646,
+      "step": 573
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.73977873977874e-05,
+      "loss": 0.6014,
+      "step": 574
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7373737373737376e-05,
+      "loss": 0.6566,
+      "step": 575
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.734968734968735e-05,
+      "loss": 0.622,
+      "step": 576
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7325637325637326e-05,
+      "loss": 0.5821,
+      "step": 577
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.6291,
+      "step": 578
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7277537277537275e-05,
+      "loss": 0.5052,
+      "step": 579
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.725348725348726e-05,
+      "loss": 0.6045,
+      "step": 580
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.722943722943723e-05,
+      "loss": 0.5614,
+      "step": 581
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.720538720538721e-05,
+      "loss": 0.6888,
+      "step": 582
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.718133718133718e-05,
+      "loss": 0.7168,
+      "step": 583
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.7157287157287156e-05,
+      "loss": 0.6599,
+      "step": 584
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 3.713323713323714e-05,
+      "loss": 0.5976,
+      "step": 585
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.710918710918711e-05,
+      "loss": 0.6097,
+      "step": 586
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.708513708513709e-05,
+      "loss": 0.5767,
+      "step": 587
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.706108706108706e-05,
+      "loss": 0.609,
+      "step": 588
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.5971,
+      "step": 589
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.701298701298702e-05,
+      "loss": 0.6781,
+      "step": 590
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.698893698893699e-05,
+      "loss": 0.6044,
+      "step": 591
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.696488696488697e-05,
+      "loss": 0.696,
+      "step": 592
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.694083694083694e-05,
+      "loss": 0.6149,
+      "step": 593
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.691678691678692e-05,
+      "loss": 0.6317,
+      "step": 594
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.68927368927369e-05,
+      "loss": 0.5926,
+      "step": 595
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.686868686868687e-05,
+      "loss": 0.6306,
+      "step": 596
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.684463684463685e-05,
+      "loss": 0.5687,
+      "step": 597
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.682058682058682e-05,
+      "loss": 0.5964,
+      "step": 598
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.67965367965368e-05,
+      "loss": 0.583,
+      "step": 599
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.677248677248677e-05,
+      "loss": 0.5716,
+      "step": 600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.674843674843675e-05,
+      "loss": 0.6092,
+      "step": 601
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.672438672438673e-05,
+      "loss": 0.5913,
+      "step": 602
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.6700336700336704e-05,
+      "loss": 0.6577,
+      "step": 603
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.667628667628668e-05,
+      "loss": 0.6016,
+      "step": 604
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.665223665223665e-05,
+      "loss": 0.6299,
+      "step": 605
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 3.662818662818663e-05,
+      "loss": 0.5966,
+      "step": 606
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.660413660413661e-05,
+      "loss": 0.7034,
+      "step": 607
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6580086580086584e-05,
+      "loss": 0.5454,
+      "step": 608
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.655603655603656e-05,
+      "loss": 0.6795,
+      "step": 609
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6531986531986534e-05,
+      "loss": 0.6884,
+      "step": 610
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.6546,
+      "step": 611
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.648388648388649e-05,
+      "loss": 0.6406,
+      "step": 612
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.645983645983646e-05,
+      "loss": 0.5857,
+      "step": 613
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.643578643578644e-05,
+      "loss": 0.5792,
+      "step": 614
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6411736411736415e-05,
+      "loss": 0.6893,
+      "step": 615
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.638768638768639e-05,
+      "loss": 0.6726,
+      "step": 616
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.6189,
+      "step": 617
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.633958633958634e-05,
+      "loss": 0.6772,
+      "step": 618
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.631553631553632e-05,
+      "loss": 0.6324,
+      "step": 619
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6291486291486295e-05,
+      "loss": 0.6231,
+      "step": 620
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.626743626743627e-05,
+      "loss": 0.619,
+      "step": 621
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6243386243386245e-05,
+      "loss": 0.5935,
+      "step": 622
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.621933621933622e-05,
+      "loss": 0.6436,
+      "step": 623
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.61952861952862e-05,
+      "loss": 0.6652,
+      "step": 624
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.617123617123617e-05,
+      "loss": 0.7586,
+      "step": 625
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6147186147186144e-05,
+      "loss": 0.5822,
+      "step": 626
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.6123136123136125e-05,
+      "loss": 0.6177,
+      "step": 627
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 3.60990860990861e-05,
+      "loss": 0.6376,
+      "step": 628
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.6075036075036075e-05,
+      "loss": 0.6175,
+      "step": 629
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.605098605098605e-05,
+      "loss": 0.614,
+      "step": 630
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.602693602693603e-05,
+      "loss": 0.6118,
+      "step": 631
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.6002886002886006e-05,
+      "loss": 0.6767,
+      "step": 632
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.597883597883598e-05,
+      "loss": 0.6448,
+      "step": 633
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5954785954785956e-05,
+      "loss": 0.6698,
+      "step": 634
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.593073593073593e-05,
+      "loss": 0.6125,
+      "step": 635
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.590668590668591e-05,
+      "loss": 0.6244,
+      "step": 636
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.588263588263589e-05,
+      "loss": 0.655,
+      "step": 637
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5858585858585855e-05,
+      "loss": 0.6197,
+      "step": 638
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5834535834535836e-05,
+      "loss": 0.6007,
+      "step": 639
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.581048581048581e-05,
+      "loss": 0.5863,
+      "step": 640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.578643578643579e-05,
+      "loss": 0.6277,
+      "step": 641
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.576238576238576e-05,
+      "loss": 0.6536,
+      "step": 642
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5738335738335735e-05,
+      "loss": 0.575,
+      "step": 643
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.6428,
+      "step": 644
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.569023569023569e-05,
+      "loss": 0.6733,
+      "step": 645
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5666185666185667e-05,
+      "loss": 0.6008,
+      "step": 646
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.564213564213564e-05,
+      "loss": 0.5621,
+      "step": 647
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.5618085618085616e-05,
+      "loss": 0.6806,
+      "step": 648
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 3.55940355940356e-05,
+      "loss": 0.6993,
+      "step": 649
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.556998556998557e-05,
+      "loss": 0.6837,
+      "step": 650
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.554593554593555e-05,
+      "loss": 0.7086,
+      "step": 651
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.552188552188552e-05,
+      "loss": 0.5537,
+      "step": 652
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.5497835497835503e-05,
+      "loss": 0.6135,
+      "step": 653
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.547378547378548e-05,
+      "loss": 0.624,
+      "step": 654
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.5449735449735446e-05,
+      "loss": 0.6544,
+      "step": 655
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.542568542568543e-05,
+      "loss": 0.6583,
+      "step": 656
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.54016354016354e-05,
+      "loss": 0.5435,
+      "step": 657
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.5377585377585384e-05,
+      "loss": 0.6673,
+      "step": 658
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 0.6043,
+      "step": 659
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.532948532948533e-05,
+      "loss": 0.5923,
+      "step": 660
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.530543530543531e-05,
+      "loss": 0.6374,
+      "step": 661
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.528138528138528e-05,
+      "loss": 0.6856,
+      "step": 662
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.525733525733526e-05,
+      "loss": 0.601,
+      "step": 663
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.523328523328523e-05,
+      "loss": 0.6796,
+      "step": 664
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.520923520923521e-05,
+      "loss": 0.6063,
+      "step": 665
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.5739,
+      "step": 666
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.5161135161135164e-05,
+      "loss": 0.6067,
+      "step": 667
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.513708513708514e-05,
+      "loss": 0.5615,
+      "step": 668
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.5113035113035113e-05,
+      "loss": 0.6302,
+      "step": 669
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 3.508898508898509e-05,
+      "loss": 0.5974,
+      "step": 670
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.506493506493507e-05,
+      "loss": 0.5828,
+      "step": 671
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.504088504088504e-05,
+      "loss": 0.5989,
+      "step": 672
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.501683501683502e-05,
+      "loss": 0.649,
+      "step": 673
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4992784992784994e-05,
+      "loss": 0.6579,
+      "step": 674
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4968734968734976e-05,
+      "loss": 0.5709,
+      "step": 675
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4944684944684944e-05,
+      "loss": 0.5912,
+      "step": 676
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.6462,
+      "step": 677
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.48965848965849e-05,
+      "loss": 0.6895,
+      "step": 678
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4872534872534875e-05,
+      "loss": 0.6061,
+      "step": 679
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.558,
+      "step": 680
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4824434824434824e-05,
+      "loss": 0.6109,
+      "step": 681
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.48003848003848e-05,
+      "loss": 0.5314,
+      "step": 682
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.477633477633478e-05,
+      "loss": 0.6396,
+      "step": 683
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4752284752284755e-05,
+      "loss": 0.6164,
+      "step": 684
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.472823472823473e-05,
+      "loss": 0.625,
+      "step": 685
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4704184704184705e-05,
+      "loss": 0.6393,
+      "step": 686
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.468013468013468e-05,
+      "loss": 0.7267,
+      "step": 687
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.465608465608466e-05,
+      "loss": 0.528,
+      "step": 688
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.463203463203463e-05,
+      "loss": 0.5681,
+      "step": 689
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.460798460798461e-05,
+      "loss": 0.5821,
+      "step": 690
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 3.4583934583934586e-05,
+      "loss": 0.5922,
+      "step": 691
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.455988455988456e-05,
+      "loss": 0.7121,
+      "step": 692
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.4535834535834535e-05,
+      "loss": 0.5754,
+      "step": 693
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.451178451178451e-05,
+      "loss": 0.6309,
+      "step": 694
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.448773448773449e-05,
+      "loss": 0.5787,
+      "step": 695
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.4463684463684466e-05,
+      "loss": 0.5855,
+      "step": 696
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.443963443963444e-05,
+      "loss": 0.6617,
+      "step": 697
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.4415584415584416e-05,
+      "loss": 0.5908,
+      "step": 698
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.439153439153439e-05,
+      "loss": 0.5981,
+      "step": 699
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.436748436748437e-05,
+      "loss": 0.5687,
+      "step": 700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.434343434343435e-05,
+      "loss": 0.6529,
+      "step": 701
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.431938431938432e-05,
+      "loss": 0.5419,
+      "step": 702
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.4295334295334296e-05,
+      "loss": 0.6504,
+      "step": 703
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.427128427128427e-05,
+      "loss": 0.6424,
+      "step": 704
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.424723424723425e-05,
+      "loss": 0.5736,
+      "step": 705
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.422318422318422e-05,
+      "loss": 0.6623,
+      "step": 706
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.41991341991342e-05,
+      "loss": 0.6377,
+      "step": 707
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.417508417508418e-05,
+      "loss": 0.6331,
+      "step": 708
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.415103415103415e-05,
+      "loss": 0.5459,
+      "step": 709
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.5513,
+      "step": 710
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.41029341029341e-05,
+      "loss": 0.6463,
+      "step": 711
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.407888407888408e-05,
+      "loss": 0.6458,
+      "step": 712
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.405483405483406e-05,
+      "loss": 0.6147,
+      "step": 713
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.403078403078403e-05,
+      "loss": 0.6572,
+      "step": 714
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.400673400673401e-05,
+      "loss": 0.5028,
+      "step": 715
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.398268398268398e-05,
+      "loss": 0.6424,
+      "step": 716
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3958633958633964e-05,
+      "loss": 0.5566,
+      "step": 717
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.393458393458394e-05,
+      "loss": 0.5505,
+      "step": 718
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.391053391053391e-05,
+      "loss": 0.5554,
+      "step": 719
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.388648388648389e-05,
+      "loss": 0.6083,
+      "step": 720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.386243386243386e-05,
+      "loss": 0.6233,
+      "step": 721
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3838383838383844e-05,
+      "loss": 0.7885,
+      "step": 722
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.381433381433381e-05,
+      "loss": 0.5911,
+      "step": 723
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3790283790283794e-05,
+      "loss": 0.6677,
+      "step": 724
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.376623376623377e-05,
+      "loss": 0.7168,
+      "step": 725
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3742183742183743e-05,
+      "loss": 0.7198,
+      "step": 726
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.371813371813372e-05,
+      "loss": 0.6669,
+      "step": 727
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.369408369408369e-05,
+      "loss": 0.6679,
+      "step": 728
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3670033670033675e-05,
+      "loss": 0.5628,
+      "step": 729
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.364598364598365e-05,
+      "loss": 0.5988,
+      "step": 730
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3621933621933624e-05,
+      "loss": 0.6698,
+      "step": 731
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.35978835978836e-05,
+      "loss": 0.5343,
+      "step": 732
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3573833573833574e-05,
+      "loss": 0.5546,
+      "step": 733
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.3549783549783555e-05,
+      "loss": 0.5699,
+      "step": 734
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.352573352573353e-05,
+      "loss": 0.6686,
+      "step": 735
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.35016835016835e-05,
+      "loss": 0.6025,
+      "step": 736
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.347763347763348e-05,
+      "loss": 0.6387,
+      "step": 737
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3453583453583454e-05,
+      "loss": 0.6997,
+      "step": 738
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3429533429533436e-05,
+      "loss": 0.5828,
+      "step": 739
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3405483405483404e-05,
+      "loss": 0.6296,
+      "step": 740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3381433381433385e-05,
+      "loss": 0.6382,
+      "step": 741
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.335738335738336e-05,
+      "loss": 0.5523,
+      "step": 742
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.6005,
+      "step": 743
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.330928330928331e-05,
+      "loss": 0.6235,
+      "step": 744
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3285233285233284e-05,
+      "loss": 0.6885,
+      "step": 745
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3261183261183266e-05,
+      "loss": 0.6364,
+      "step": 746
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.323713323713324e-05,
+      "loss": 0.5319,
+      "step": 747
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3213083213083216e-05,
+      "loss": 0.6407,
+      "step": 748
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.318903318903319e-05,
+      "loss": 0.6466,
+      "step": 749
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3164983164983165e-05,
+      "loss": 0.5511,
+      "step": 750
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.314093314093315e-05,
+      "loss": 0.6051,
+      "step": 751
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.311688311688312e-05,
+      "loss": 0.6094,
+      "step": 752
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.309283309283309e-05,
+      "loss": 0.6036,
+      "step": 753
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.306878306878307e-05,
+      "loss": 0.5995,
+      "step": 754
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.3044733044733046e-05,
+      "loss": 0.5927,
+      "step": 755
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.302068302068303e-05,
+      "loss": 0.5863,
+      "step": 756
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.2996632996632995e-05,
+      "loss": 0.681,
+      "step": 757
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.297258297258297e-05,
+      "loss": 0.6056,
+      "step": 758
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.294853294853295e-05,
+      "loss": 0.6063,
+      "step": 759
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.2924482924482926e-05,
+      "loss": 0.5796,
+      "step": 760
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.29004329004329e-05,
+      "loss": 0.6218,
+      "step": 761
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.2876382876382876e-05,
+      "loss": 0.579,
+      "step": 762
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.285233285233286e-05,
+      "loss": 0.625,
+      "step": 763
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.282828282828283e-05,
+      "loss": 0.6585,
+      "step": 764
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.280423280423281e-05,
+      "loss": 0.5336,
+      "step": 765
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.278018278018278e-05,
+      "loss": 0.6108,
+      "step": 766
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.275613275613276e-05,
+      "loss": 0.7248,
+      "step": 767
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.273208273208274e-05,
+      "loss": 0.5836,
+      "step": 768
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.270803270803271e-05,
+      "loss": 0.5846,
+      "step": 769
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.268398268398268e-05,
+      "loss": 0.5737,
+      "step": 770
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.265993265993266e-05,
+      "loss": 0.6745,
+      "step": 771
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.263588263588264e-05,
+      "loss": 0.6038,
+      "step": 772
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.261183261183262e-05,
+      "loss": 0.619,
+      "step": 773
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.258778258778259e-05,
+      "loss": 0.6024,
+      "step": 774
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.256373256373256e-05,
+      "loss": 0.5226,
+      "step": 775
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.6078,
+      "step": 776
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.251563251563252e-05,
+      "loss": 0.5884,
+      "step": 777
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.249158249158249e-05,
+      "loss": 0.6246,
+      "step": 778
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.246753246753247e-05,
+      "loss": 0.5796,
+      "step": 779
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.244348244348244e-05,
+      "loss": 0.5849,
+      "step": 780
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2419432419432424e-05,
+      "loss": 0.5309,
+      "step": 781
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.23953823953824e-05,
+      "loss": 0.6441,
+      "step": 782
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.237133237133237e-05,
+      "loss": 0.5986,
+      "step": 783
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.234728234728235e-05,
+      "loss": 0.6482,
+      "step": 784
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.232323232323233e-05,
+      "loss": 0.5993,
+      "step": 785
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2299182299182304e-05,
+      "loss": 0.5449,
+      "step": 786
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.227513227513227e-05,
+      "loss": 0.5898,
+      "step": 787
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2251082251082254e-05,
+      "loss": 0.5474,
+      "step": 788
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.222703222703223e-05,
+      "loss": 0.5381,
+      "step": 789
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.220298220298221e-05,
+      "loss": 0.7628,
+      "step": 790
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.217893217893218e-05,
+      "loss": 0.529,
+      "step": 791
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.215488215488215e-05,
+      "loss": 0.651,
+      "step": 792
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2130832130832135e-05,
+      "loss": 0.6471,
+      "step": 793
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.210678210678211e-05,
+      "loss": 0.6418,
+      "step": 794
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2082732082732084e-05,
+      "loss": 0.6321,
+      "step": 795
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.205868205868206e-05,
+      "loss": 0.6319,
+      "step": 796
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2034632034632034e-05,
+      "loss": 0.5769,
+      "step": 797
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.2010582010582015e-05,
+      "loss": 0.659,
+      "step": 798
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.198653198653199e-05,
+      "loss": 0.6225,
+      "step": 799
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1962481962481965e-05,
+      "loss": 0.761,
+      "step": 800
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.193843193843194e-05,
+      "loss": 0.5713,
+      "step": 801
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1914381914381914e-05,
+      "loss": 0.6076,
+      "step": 802
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1890331890331896e-05,
+      "loss": 0.6059,
+      "step": 803
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1866281866281864e-05,
+      "loss": 0.6253,
+      "step": 804
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1842231842231846e-05,
+      "loss": 0.5836,
+      "step": 805
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.5746,
+      "step": 806
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1794131794131795e-05,
+      "loss": 0.6029,
+      "step": 807
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.177008177008177e-05,
+      "loss": 0.6048,
+      "step": 808
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.6534,
+      "step": 809
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1721981721981726e-05,
+      "loss": 0.6249,
+      "step": 810
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.16979316979317e-05,
+      "loss": 0.6369,
+      "step": 811
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1673881673881676e-05,
+      "loss": 0.6809,
+      "step": 812
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.164983164983165e-05,
+      "loss": 0.6358,
+      "step": 813
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1625781625781625e-05,
+      "loss": 0.5915,
+      "step": 814
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.160173160173161e-05,
+      "loss": 0.5442,
+      "step": 815
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1577681577681575e-05,
+      "loss": 0.6105,
+      "step": 816
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1553631553631556e-05,
+      "loss": 0.619,
+      "step": 817
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.152958152958153e-05,
+      "loss": 0.6665,
+      "step": 818
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.1505531505531506e-05,
+      "loss": 0.6029,
+      "step": 819
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.5395,
+      "step": 820
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.1457431457431456e-05,
+      "loss": 0.6057,
+      "step": 821
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.143338143338144e-05,
+      "loss": 0.5499,
+      "step": 822
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.140933140933141e-05,
+      "loss": 0.586,
+      "step": 823
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.1385281385281387e-05,
+      "loss": 0.5984,
+      "step": 824
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.136123136123136e-05,
+      "loss": 0.649,
+      "step": 825
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.1337181337181336e-05,
+      "loss": 0.5849,
+      "step": 826
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 0.6476,
+      "step": 827
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.128908128908129e-05,
+      "loss": 0.6093,
+      "step": 828
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.126503126503126e-05,
+      "loss": 0.5944,
+      "step": 829
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.124098124098124e-05,
+      "loss": 0.6555,
+      "step": 830
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.121693121693122e-05,
+      "loss": 0.575,
+      "step": 831
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.11928811928812e-05,
+      "loss": 0.6448,
+      "step": 832
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.1168831168831166e-05,
+      "loss": 0.5506,
+      "step": 833
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.114478114478115e-05,
+      "loss": 0.5528,
+      "step": 834
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.112073112073112e-05,
+      "loss": 0.6237,
+      "step": 835
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.10966810966811e-05,
+      "loss": 0.6577,
+      "step": 836
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.107263107263107e-05,
+      "loss": 0.6109,
+      "step": 837
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.104858104858105e-05,
+      "loss": 0.6219,
+      "step": 838
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.102453102453103e-05,
+      "loss": 0.6372,
+      "step": 839
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.1000481000481e-05,
+      "loss": 0.6254,
+      "step": 840
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.097643097643098e-05,
+      "loss": 0.5241,
+      "step": 841
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.7298,
+      "step": 842
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.092833092833093e-05,
+      "loss": 0.6936,
+      "step": 843
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.090428090428091e-05,
+      "loss": 0.6414,
+      "step": 844
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0880230880230884e-05,
+      "loss": 0.6379,
+      "step": 845
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.085618085618085e-05,
+      "loss": 0.6467,
+      "step": 846
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0832130832130834e-05,
+      "loss": 0.6608,
+      "step": 847
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.080808080808081e-05,
+      "loss": 0.5921,
+      "step": 848
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.078403078403079e-05,
+      "loss": 0.5618,
+      "step": 849
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.075998075998076e-05,
+      "loss": 0.6527,
+      "step": 850
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.073593073593073e-05,
+      "loss": 0.6437,
+      "step": 851
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0711880711880714e-05,
+      "loss": 0.6518,
+      "step": 852
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.068783068783069e-05,
+      "loss": 0.6535,
+      "step": 853
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0663780663780664e-05,
+      "loss": 0.5865,
+      "step": 854
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.063973063973064e-05,
+      "loss": 0.5705,
+      "step": 855
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.061568061568062e-05,
+      "loss": 0.5575,
+      "step": 856
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0591630591630595e-05,
+      "loss": 0.636,
+      "step": 857
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.056758056758057e-05,
+      "loss": 0.6461,
+      "step": 858
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0543530543530544e-05,
+      "loss": 0.5965,
+      "step": 859
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.051948051948052e-05,
+      "loss": 0.6352,
+      "step": 860
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.04954304954305e-05,
+      "loss": 0.6339,
+      "step": 861
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.0471380471380472e-05,
+      "loss": 0.6234,
+      "step": 862
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0447330447330447e-05,
+      "loss": 0.5826,
+      "step": 863
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0423280423280425e-05,
+      "loss": 0.5596,
+      "step": 864
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.03992303992304e-05,
+      "loss": 0.5818,
+      "step": 865
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0375180375180378e-05,
+      "loss": 0.5903,
+      "step": 866
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0351130351130353e-05,
+      "loss": 0.6095,
+      "step": 867
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0327080327080328e-05,
+      "loss": 0.623,
+      "step": 868
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.71,
+      "step": 869
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.027898027898028e-05,
+      "loss": 0.6727,
+      "step": 870
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.025493025493026e-05,
+      "loss": 0.6683,
+      "step": 871
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0230880230880233e-05,
+      "loss": 0.5799,
+      "step": 872
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0206830206830205e-05,
+      "loss": 0.6862,
+      "step": 873
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0182780182780186e-05,
+      "loss": 0.6355,
+      "step": 874
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 0.6856,
+      "step": 875
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.013468013468014e-05,
+      "loss": 0.5529,
+      "step": 876
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.011063011063011e-05,
+      "loss": 0.6358,
+      "step": 877
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0086580086580092e-05,
+      "loss": 0.6674,
+      "step": 878
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0062530062530064e-05,
+      "loss": 0.6802,
+      "step": 879
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.003848003848004e-05,
+      "loss": 0.5824,
+      "step": 880
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.0014430014430017e-05,
+      "loss": 0.6291,
+      "step": 881
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.999037999037999e-05,
+      "loss": 0.7381,
+      "step": 882
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.996632996632997e-05,
+      "loss": 0.6458,
+      "step": 883
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9942279942279944e-05,
+      "loss": 0.5949,
+      "step": 884
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.991822991822992e-05,
+      "loss": 0.567,
+      "step": 885
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9894179894179897e-05,
+      "loss": 0.5597,
+      "step": 886
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9870129870129872e-05,
+      "loss": 0.6076,
+      "step": 887
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.984607984607985e-05,
+      "loss": 0.551,
+      "step": 888
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9822029822029825e-05,
+      "loss": 0.6641,
+      "step": 889
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9797979797979796e-05,
+      "loss": 0.5848,
+      "step": 890
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9773929773929778e-05,
+      "loss": 0.6555,
+      "step": 891
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.974987974987975e-05,
+      "loss": 0.629,
+      "step": 892
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.972582972582973e-05,
+      "loss": 0.6514,
+      "step": 893
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9701779701779702e-05,
+      "loss": 0.5599,
+      "step": 894
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9677729677729677e-05,
+      "loss": 0.5905,
+      "step": 895
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9653679653679655e-05,
+      "loss": 0.696,
+      "step": 896
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.6151,
+      "step": 897
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9605579605579608e-05,
+      "loss": 0.5887,
+      "step": 898
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9581529581529583e-05,
+      "loss": 0.5859,
+      "step": 899
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.955747955747956e-05,
+      "loss": 0.6392,
+      "step": 900
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9533429533429536e-05,
+      "loss": 0.571,
+      "step": 901
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.950937950937951e-05,
+      "loss": 0.6049,
+      "step": 902
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.948532948532949e-05,
+      "loss": 0.5122,
+      "step": 903
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.946127946127946e-05,
+      "loss": 0.6905,
+      "step": 904
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.943722943722944e-05,
+      "loss": 0.6921,
+      "step": 905
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9413179413179413e-05,
+      "loss": 0.5658,
+      "step": 906
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9389129389129388e-05,
+      "loss": 0.5533,
+      "step": 907
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.757,
+      "step": 908
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.934102934102934e-05,
+      "loss": 0.6191,
+      "step": 909
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.931697931697932e-05,
+      "loss": 0.6521,
+      "step": 910
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9292929292929294e-05,
+      "loss": 0.6367,
+      "step": 911
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.926887926887927e-05,
+      "loss": 0.613,
+      "step": 912
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9244829244829247e-05,
+      "loss": 0.5809,
+      "step": 913
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.922077922077922e-05,
+      "loss": 0.597,
+      "step": 914
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.91967291967292e-05,
+      "loss": 0.5557,
+      "step": 915
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9172679172679174e-05,
+      "loss": 0.5492,
+      "step": 916
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9148629148629146e-05,
+      "loss": 0.5803,
+      "step": 917
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9124579124579127e-05,
+      "loss": 0.6613,
+      "step": 918
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.91005291005291e-05,
+      "loss": 0.5651,
+      "step": 919
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.907647907647908e-05,
+      "loss": 0.6794,
+      "step": 920
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.905242905242905e-05,
+      "loss": 0.5786,
+      "step": 921
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9028379028379033e-05,
+      "loss": 0.5992,
+      "step": 922
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9004329004329005e-05,
+      "loss": 0.5899,
+      "step": 923
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.898027898027898e-05,
+      "loss": 0.5949,
+      "step": 924
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8956228956228958e-05,
+      "loss": 0.6468,
+      "step": 925
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8932178932178932e-05,
+      "loss": 0.5553,
+      "step": 926
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.890812890812891e-05,
+      "loss": 0.5763,
+      "step": 927
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8884078884078885e-05,
+      "loss": 0.5554,
+      "step": 928
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.886002886002886e-05,
+      "loss": 0.5378,
+      "step": 929
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8835978835978838e-05,
+      "loss": 0.5617,
+      "step": 930
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8811928811928813e-05,
+      "loss": 0.65,
+      "step": 931
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.878787878787879e-05,
+      "loss": 0.6805,
+      "step": 932
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8763828763828766e-05,
+      "loss": 0.5234,
+      "step": 933
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8739778739778737e-05,
+      "loss": 0.6347,
+      "step": 934
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.871572871572872e-05,
+      "loss": 0.5879,
+      "step": 935
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.869167869167869e-05,
+      "loss": 0.5816,
+      "step": 936
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8667628667628672e-05,
+      "loss": 0.6027,
+      "step": 937
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8643578643578643e-05,
+      "loss": 0.4949,
+      "step": 938
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8619528619528618e-05,
+      "loss": 0.606,
+      "step": 939
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8595478595478596e-05,
+      "loss": 0.5943,
+      "step": 940
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.6511,
+      "step": 941
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.854737854737855e-05,
+      "loss": 0.5393,
+      "step": 942
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8523328523328524e-05,
+      "loss": 0.558,
+      "step": 943
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8499278499278502e-05,
+      "loss": 0.555,
+      "step": 944
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8475228475228477e-05,
+      "loss": 0.6163,
+      "step": 945
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.845117845117845e-05,
+      "loss": 0.6244,
+      "step": 946
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.842712842712843e-05,
+      "loss": 0.5859,
+      "step": 947
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8403078403078404e-05,
+      "loss": 0.6096,
+      "step": 948
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8379028379028383e-05,
+      "loss": 0.5758,
+      "step": 949
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8354978354978357e-05,
+      "loss": 0.5548,
+      "step": 950
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.833092833092833e-05,
+      "loss": 0.5739,
+      "step": 951
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.830687830687831e-05,
+      "loss": 0.6564,
+      "step": 952
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8282828282828282e-05,
+      "loss": 0.6407,
+      "step": 953
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8258778258778263e-05,
+      "loss": 0.6097,
+      "step": 954
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8234728234728235e-05,
+      "loss": 0.646,
+      "step": 955
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.821067821067821e-05,
+      "loss": 0.6538,
+      "step": 956
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8186628186628188e-05,
+      "loss": 0.5563,
+      "step": 957
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8162578162578162e-05,
+      "loss": 0.6612,
+      "step": 958
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.813852813852814e-05,
+      "loss": 0.5626,
+      "step": 959
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8114478114478115e-05,
+      "loss": 0.5763,
+      "step": 960
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.809042809042809e-05,
+      "loss": 0.5835,
+      "step": 961
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8066378066378068e-05,
+      "loss": 0.6207,
+      "step": 962
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8042328042328043e-05,
+      "loss": 0.5648,
+      "step": 963
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.801827801827802e-05,
+      "loss": 0.5015,
+      "step": 964
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.7994227994227996e-05,
+      "loss": 0.5105,
+      "step": 965
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.7970177970177974e-05,
+      "loss": 0.5712,
+      "step": 966
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.794612794612795e-05,
+      "loss": 0.6482,
+      "step": 967
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.792207792207792e-05,
+      "loss": 0.5847,
+      "step": 968
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7898027898027902e-05,
+      "loss": 0.5143,
+      "step": 969
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7873977873977873e-05,
+      "loss": 0.5007,
+      "step": 970
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7849927849927855e-05,
+      "loss": 0.5174,
+      "step": 971
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7825877825877826e-05,
+      "loss": 0.5996,
+      "step": 972
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.78018278018278e-05,
+      "loss": 0.6033,
+      "step": 973
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.6048,
+      "step": 974
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7753727753727754e-05,
+      "loss": 0.6424,
+      "step": 975
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7729677729677732e-05,
+      "loss": 0.498,
+      "step": 976
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7705627705627707e-05,
+      "loss": 0.6618,
+      "step": 977
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.768157768157768e-05,
+      "loss": 0.621,
+      "step": 978
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.765752765752766e-05,
+      "loss": 0.6672,
+      "step": 979
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7633477633477635e-05,
+      "loss": 0.5236,
+      "step": 980
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7609427609427613e-05,
+      "loss": 0.7001,
+      "step": 981
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7585377585377587e-05,
+      "loss": 0.618,
+      "step": 982
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.756132756132756e-05,
+      "loss": 0.6535,
+      "step": 983
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.753727753727754e-05,
+      "loss": 0.6013,
+      "step": 984
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7513227513227512e-05,
+      "loss": 0.5862,
+      "step": 985
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7489177489177493e-05,
+      "loss": 0.5488,
+      "step": 986
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7465127465127465e-05,
+      "loss": 0.5016,
+      "step": 987
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7441077441077446e-05,
+      "loss": 0.5794,
+      "step": 988
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.7417027417027418e-05,
+      "loss": 0.6258,
+      "step": 989
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7392977392977392e-05,
+      "loss": 0.5818,
+      "step": 990
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.736892736892737e-05,
+      "loss": 0.5905,
+      "step": 991
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7344877344877345e-05,
+      "loss": 0.5233,
+      "step": 992
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7320827320827324e-05,
+      "loss": 0.5707,
+      "step": 993
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.72967772967773e-05,
+      "loss": 0.7378,
+      "step": 994
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.5061,
+      "step": 995
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.724867724867725e-05,
+      "loss": 0.6302,
+      "step": 996
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7224627224627226e-05,
+      "loss": 0.68,
+      "step": 997
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7200577200577204e-05,
+      "loss": 0.6005,
+      "step": 998
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.717652717652718e-05,
+      "loss": 0.5596,
+      "step": 999
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.715247715247715e-05,
+      "loss": 0.5923,
+      "step": 1000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7128427128427132e-05,
+      "loss": 0.5402,
+      "step": 1001
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7104377104377103e-05,
+      "loss": 0.5907,
+      "step": 1002
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7080327080327085e-05,
+      "loss": 0.5319,
+      "step": 1003
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.7056277056277056e-05,
+      "loss": 0.5688,
+      "step": 1004
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.703222703222703e-05,
+      "loss": 0.6467,
+      "step": 1005
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.700817700817701e-05,
+      "loss": 0.6701,
+      "step": 1006
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.6542,
+      "step": 1007
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.6960076960076962e-05,
+      "loss": 0.6534,
+      "step": 1008
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.6936026936026937e-05,
+      "loss": 0.571,
+      "step": 1009
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.691197691197691e-05,
+      "loss": 0.6061,
+      "step": 1010
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.688792688792689e-05,
+      "loss": 0.6137,
+      "step": 1011
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6863876863876865e-05,
+      "loss": 0.6259,
+      "step": 1012
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6839826839826843e-05,
+      "loss": 0.5517,
+      "step": 1013
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6815776815776818e-05,
+      "loss": 0.5911,
+      "step": 1014
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6791726791726796e-05,
+      "loss": 0.5821,
+      "step": 1015
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.676767676767677e-05,
+      "loss": 0.5731,
+      "step": 1016
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6743626743626742e-05,
+      "loss": 0.5351,
+      "step": 1017
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6719576719576723e-05,
+      "loss": 0.5502,
+      "step": 1018
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6695526695526695e-05,
+      "loss": 0.5816,
+      "step": 1019
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6671476671476676e-05,
+      "loss": 0.6028,
+      "step": 1020
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6647426647426648e-05,
+      "loss": 0.6368,
+      "step": 1021
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6623376623376623e-05,
+      "loss": 0.5602,
+      "step": 1022
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.65993265993266e-05,
+      "loss": 0.6341,
+      "step": 1023
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6575276575276575e-05,
+      "loss": 0.5365,
+      "step": 1024
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6551226551226554e-05,
+      "loss": 0.4991,
+      "step": 1025
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.652717652717653e-05,
+      "loss": 0.619,
+      "step": 1026
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6503126503126503e-05,
+      "loss": 0.6652,
+      "step": 1027
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.647907647907648e-05,
+      "loss": 0.6085,
+      "step": 1028
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6455026455026456e-05,
+      "loss": 0.69,
+      "step": 1029
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.6430976430976434e-05,
+      "loss": 0.6032,
+      "step": 1030
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.640692640692641e-05,
+      "loss": 0.6387,
+      "step": 1031
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.638287638287638e-05,
+      "loss": 0.5783,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6358826358826362e-05,
+      "loss": 0.5989,
+      "step": 1033
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6334776334776333e-05,
+      "loss": 0.6094,
+      "step": 1034
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6310726310726315e-05,
+      "loss": 0.5763,
+      "step": 1035
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6286676286676286e-05,
+      "loss": 0.602,
+      "step": 1036
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6262626262626268e-05,
+      "loss": 0.5402,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.623857623857624e-05,
+      "loss": 0.5396,
+      "step": 1038
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6214526214526214e-05,
+      "loss": 0.5809,
+      "step": 1039
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.6115,
+      "step": 1040
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6166426166426167e-05,
+      "loss": 0.6055,
+      "step": 1041
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6142376142376145e-05,
+      "loss": 0.6132,
+      "step": 1042
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.611832611832612e-05,
+      "loss": 0.6103,
+      "step": 1043
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6094276094276095e-05,
+      "loss": 0.617,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6070226070226073e-05,
+      "loss": 0.6092,
+      "step": 1045
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6046176046176048e-05,
+      "loss": 0.5182,
+      "step": 1046
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.6022126022126026e-05,
+      "loss": 0.4857,
+      "step": 1047
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5998075998076e-05,
+      "loss": 0.5451,
+      "step": 1048
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5974025974025972e-05,
+      "loss": 0.5239,
+      "step": 1049
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5949975949975954e-05,
+      "loss": 0.5901,
+      "step": 1050
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.6925,
+      "step": 1051
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5901875901875906e-05,
+      "loss": 0.6426,
+      "step": 1052
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.5877825877825878e-05,
+      "loss": 0.6611,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5853775853775853e-05,
+      "loss": 0.6474,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.582972582972583e-05,
+      "loss": 0.6248,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5805675805675806e-05,
+      "loss": 0.6444,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5781625781625784e-05,
+      "loss": 0.6301,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.575757575757576e-05,
+      "loss": 0.5666,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5733525733525737e-05,
+      "loss": 0.5229,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.570947570947571e-05,
+      "loss": 0.6014,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5685425685425686e-05,
+      "loss": 0.5385,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5661375661375664e-05,
+      "loss": 0.5481,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.563732563732564e-05,
+      "loss": 0.6046,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5613275613275617e-05,
+      "loss": 0.6216,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5589225589225592e-05,
+      "loss": 0.6261,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5565175565175563e-05,
+      "loss": 0.6177,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5541125541125545e-05,
+      "loss": 0.6145,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5517075517075516e-05,
+      "loss": 0.5767,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5493025493025498e-05,
+      "loss": 0.576,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.546897546897547e-05,
+      "loss": 0.5037,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5444925444925444e-05,
+      "loss": 0.5683,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5420875420875422e-05,
+      "loss": 0.58,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.6229,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.5372775372775375e-05,
+      "loss": 0.6487,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.534872534872535e-05,
+      "loss": 0.5228,
+      "step": 1075
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5324675324675325e-05,
+      "loss": 0.5797,
+      "step": 1076
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5300625300625303e-05,
+      "loss": 0.6058,
+      "step": 1077
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5276575276575278e-05,
+      "loss": 0.5933,
+      "step": 1078
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 0.6271,
+      "step": 1079
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.522847522847523e-05,
+      "loss": 0.5521,
+      "step": 1080
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.520442520442521e-05,
+      "loss": 0.6151,
+      "step": 1081
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5180375180375184e-05,
+      "loss": 0.6316,
+      "step": 1082
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5156325156325155e-05,
+      "loss": 0.5709,
+      "step": 1083
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5132275132275137e-05,
+      "loss": 0.7207,
+      "step": 1084
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5108225108225108e-05,
+      "loss": 0.596,
+      "step": 1085
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5084175084175086e-05,
+      "loss": 0.5144,
+      "step": 1086
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.506012506012506e-05,
+      "loss": 0.5285,
+      "step": 1087
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5036075036075036e-05,
+      "loss": 0.5903,
+      "step": 1088
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.5012025012025014e-05,
+      "loss": 0.5036,
+      "step": 1089
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.498797498797499e-05,
+      "loss": 0.5586,
+      "step": 1090
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.4963924963924963e-05,
+      "loss": 0.5737,
+      "step": 1091
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.493987493987494e-05,
+      "loss": 0.5714,
+      "step": 1092
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.4915824915824916e-05,
+      "loss": 0.5695,
+      "step": 1093
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.4891774891774894e-05,
+      "loss": 0.6015,
+      "step": 1094
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.4867724867724866e-05,
+      "loss": 0.6421,
+      "step": 1095
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.4843674843674844e-05,
+      "loss": 0.5816,
+      "step": 1096
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.481962481962482e-05,
+      "loss": 0.5713,
+      "step": 1097
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4795574795574797e-05,
+      "loss": 0.5713,
+      "step": 1098
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4771524771524772e-05,
+      "loss": 0.6099,
+      "step": 1099
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.474747474747475e-05,
+      "loss": 0.6334,
+      "step": 1100
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4723424723424725e-05,
+      "loss": 0.5137,
+      "step": 1101
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.46993746993747e-05,
+      "loss": 0.5815,
+      "step": 1102
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4675324675324678e-05,
+      "loss": 0.616,
+      "step": 1103
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4651274651274652e-05,
+      "loss": 0.6214,
+      "step": 1104
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.462722462722463e-05,
+      "loss": 0.5263,
+      "step": 1105
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.7654,
+      "step": 1106
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.457912457912458e-05,
+      "loss": 0.5505,
+      "step": 1107
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4555074555074555e-05,
+      "loss": 0.5988,
+      "step": 1108
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4531024531024533e-05,
+      "loss": 0.5751,
+      "step": 1109
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4506974506974508e-05,
+      "loss": 0.5669,
+      "step": 1110
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4482924482924486e-05,
+      "loss": 0.6344,
+      "step": 1111
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4458874458874457e-05,
+      "loss": 0.5263,
+      "step": 1112
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4434824434824436e-05,
+      "loss": 0.6725,
+      "step": 1113
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.441077441077441e-05,
+      "loss": 0.6111,
+      "step": 1114
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.438672438672439e-05,
+      "loss": 0.5955,
+      "step": 1115
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4362674362674363e-05,
+      "loss": 0.547,
+      "step": 1116
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.4338624338624338e-05,
+      "loss": 0.5795,
+      "step": 1117
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4314574314574316e-05,
+      "loss": 0.5935,
+      "step": 1118
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.429052429052429e-05,
+      "loss": 0.5986,
+      "step": 1119
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.426647426647427e-05,
+      "loss": 0.5601,
+      "step": 1120
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.5468,
+      "step": 1121
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4218374218374222e-05,
+      "loss": 0.5621,
+      "step": 1122
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4194324194324193e-05,
+      "loss": 0.5679,
+      "step": 1123
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.417027417027417e-05,
+      "loss": 0.6124,
+      "step": 1124
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4146224146224146e-05,
+      "loss": 0.61,
+      "step": 1125
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4122174122174125e-05,
+      "loss": 0.634,
+      "step": 1126
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.40981240981241e-05,
+      "loss": 0.5696,
+      "step": 1127
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.5852,
+      "step": 1128
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.405002405002405e-05,
+      "loss": 0.5742,
+      "step": 1129
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4025974025974027e-05,
+      "loss": 0.613,
+      "step": 1130
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.4001924001924002e-05,
+      "loss": 0.641,
+      "step": 1131
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.397787397787398e-05,
+      "loss": 0.5683,
+      "step": 1132
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.3953823953823955e-05,
+      "loss": 0.5118,
+      "step": 1133
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.392977392977393e-05,
+      "loss": 0.5148,
+      "step": 1134
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.3905723905723908e-05,
+      "loss": 0.5954,
+      "step": 1135
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.3881673881673882e-05,
+      "loss": 0.5533,
+      "step": 1136
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.385762385762386e-05,
+      "loss": 0.5908,
+      "step": 1137
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.3833573833573835e-05,
+      "loss": 0.5897,
+      "step": 1138
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.6477,
+      "step": 1139
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3785473785473785e-05,
+      "loss": 0.6786,
+      "step": 1140
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3761423761423763e-05,
+      "loss": 0.5087,
+      "step": 1141
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3737373737373738e-05,
+      "loss": 0.5997,
+      "step": 1142
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3713323713323716e-05,
+      "loss": 0.5937,
+      "step": 1143
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.368927368927369e-05,
+      "loss": 0.6698,
+      "step": 1144
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3665223665223666e-05,
+      "loss": 0.5612,
+      "step": 1145
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.364117364117364e-05,
+      "loss": 0.5724,
+      "step": 1146
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.361712361712362e-05,
+      "loss": 0.6231,
+      "step": 1147
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3593073593073593e-05,
+      "loss": 0.6317,
+      "step": 1148
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.356902356902357e-05,
+      "loss": 0.6,
+      "step": 1149
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3544973544973546e-05,
+      "loss": 0.5668,
+      "step": 1150
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.352092352092352e-05,
+      "loss": 0.6445,
+      "step": 1151
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.34968734968735e-05,
+      "loss": 0.58,
+      "step": 1152
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3472823472823474e-05,
+      "loss": 0.5646,
+      "step": 1153
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3448773448773452e-05,
+      "loss": 0.5785,
+      "step": 1154
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3424723424723427e-05,
+      "loss": 0.6086,
+      "step": 1155
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.34006734006734e-05,
+      "loss": 0.6353,
+      "step": 1156
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3376623376623376e-05,
+      "loss": 0.577,
+      "step": 1157
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3352573352573355e-05,
+      "loss": 0.6481,
+      "step": 1158
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.332852332852333e-05,
+      "loss": 0.5723,
+      "step": 1159
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.3304473304473308e-05,
+      "loss": 0.5079,
+      "step": 1160
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.328042328042328e-05,
+      "loss": 0.5516,
+      "step": 1161
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3256373256373257e-05,
+      "loss": 0.556,
+      "step": 1162
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3232323232323232e-05,
+      "loss": 0.5305,
+      "step": 1163
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.320827320827321e-05,
+      "loss": 0.6933,
+      "step": 1164
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3184223184223185e-05,
+      "loss": 0.627,
+      "step": 1165
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3160173160173163e-05,
+      "loss": 0.666,
+      "step": 1166
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3136123136123138e-05,
+      "loss": 0.5633,
+      "step": 1167
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3112073112073113e-05,
+      "loss": 0.5343,
+      "step": 1168
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.308802308802309e-05,
+      "loss": 0.581,
+      "step": 1169
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3063973063973065e-05,
+      "loss": 0.5577,
+      "step": 1170
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3039923039923044e-05,
+      "loss": 0.6325,
+      "step": 1171
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.5988,
+      "step": 1172
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.2991822991822993e-05,
+      "loss": 0.6179,
+      "step": 1173
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.2967772967772968e-05,
+      "loss": 0.537,
+      "step": 1174
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.2943722943722946e-05,
+      "loss": 0.5799,
+      "step": 1175
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.291967291967292e-05,
+      "loss": 0.5869,
+      "step": 1176
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.28956228956229e-05,
+      "loss": 0.6989,
+      "step": 1177
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.287157287157287e-05,
+      "loss": 0.644,
+      "step": 1178
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.284752284752285e-05,
+      "loss": 0.6221,
+      "step": 1179
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.2823472823472823e-05,
+      "loss": 0.4945,
+      "step": 1180
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.27994227994228e-05,
+      "loss": 0.6039,
+      "step": 1181
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2775372775372776e-05,
+      "loss": 0.4756,
+      "step": 1182
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.275132275132275e-05,
+      "loss": 0.6358,
+      "step": 1183
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.6142,
+      "step": 1184
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2703222703222704e-05,
+      "loss": 0.6017,
+      "step": 1185
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.267917267917268e-05,
+      "loss": 0.5109,
+      "step": 1186
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2655122655122657e-05,
+      "loss": 0.5398,
+      "step": 1187
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2631072631072632e-05,
+      "loss": 0.5652,
+      "step": 1188
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2607022607022607e-05,
+      "loss": 0.5754,
+      "step": 1189
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2582972582972585e-05,
+      "loss": 0.5761,
+      "step": 1190
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.255892255892256e-05,
+      "loss": 0.608,
+      "step": 1191
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2534872534872538e-05,
+      "loss": 0.6665,
+      "step": 1192
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2510822510822512e-05,
+      "loss": 0.6785,
+      "step": 1193
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2486772486772487e-05,
+      "loss": 0.641,
+      "step": 1194
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2462722462722462e-05,
+      "loss": 0.607,
+      "step": 1195
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.243867243867244e-05,
+      "loss": 0.5877,
+      "step": 1196
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2414622414622415e-05,
+      "loss": 0.5946,
+      "step": 1197
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2390572390572393e-05,
+      "loss": 0.506,
+      "step": 1198
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2366522366522368e-05,
+      "loss": 0.5196,
+      "step": 1199
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2342472342472343e-05,
+      "loss": 0.5969,
+      "step": 1200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2318422318422317e-05,
+      "loss": 0.6032,
+      "step": 1201
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.2294372294372296e-05,
+      "loss": 0.5103,
+      "step": 1202
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.227032227032227e-05,
+      "loss": 0.5921,
+      "step": 1203
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.224627224627225e-05,
+      "loss": 0.5516,
+      "step": 1204
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.7495,
+      "step": 1205
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2198172198172198e-05,
+      "loss": 0.6484,
+      "step": 1206
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2174122174122176e-05,
+      "loss": 0.5402,
+      "step": 1207
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.215007215007215e-05,
+      "loss": 0.6497,
+      "step": 1208
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.212602212602213e-05,
+      "loss": 0.5618,
+      "step": 1209
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2101972101972104e-05,
+      "loss": 0.5185,
+      "step": 1210
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.207792207792208e-05,
+      "loss": 0.6739,
+      "step": 1211
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2053872053872053e-05,
+      "loss": 0.6285,
+      "step": 1212
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.202982202982203e-05,
+      "loss": 0.5977,
+      "step": 1213
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.2005772005772006e-05,
+      "loss": 0.5822,
+      "step": 1214
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1981721981721985e-05,
+      "loss": 0.5335,
+      "step": 1215
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1957671957671956e-05,
+      "loss": 0.5863,
+      "step": 1216
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1933621933621934e-05,
+      "loss": 0.5942,
+      "step": 1217
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.190957190957191e-05,
+      "loss": 0.6119,
+      "step": 1218
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1885521885521887e-05,
+      "loss": 0.5068,
+      "step": 1219
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1861471861471862e-05,
+      "loss": 0.5997,
+      "step": 1220
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.183742183742184e-05,
+      "loss": 0.658,
+      "step": 1221
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1813371813371815e-05,
+      "loss": 0.5331,
+      "step": 1222
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.178932178932179e-05,
+      "loss": 0.5902,
+      "step": 1223
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.1765271765271768e-05,
+      "loss": 0.572,
+      "step": 1224
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1741221741221743e-05,
+      "loss": 0.5603,
+      "step": 1225
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.171717171717172e-05,
+      "loss": 0.6551,
+      "step": 1226
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1693121693121692e-05,
+      "loss": 0.6163,
+      "step": 1227
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.166907166907167e-05,
+      "loss": 0.5319,
+      "step": 1228
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1645021645021645e-05,
+      "loss": 0.5836,
+      "step": 1229
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1620971620971623e-05,
+      "loss": 0.5312,
+      "step": 1230
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1596921596921598e-05,
+      "loss": 0.5912,
+      "step": 1231
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1572871572871576e-05,
+      "loss": 0.5586,
+      "step": 1232
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1548821548821547e-05,
+      "loss": 0.5716,
+      "step": 1233
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1524771524771526e-05,
+      "loss": 0.5687,
+      "step": 1234
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.15007215007215e-05,
+      "loss": 0.6008,
+      "step": 1235
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.147667147667148e-05,
+      "loss": 0.638,
+      "step": 1236
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1452621452621453e-05,
+      "loss": 0.5737,
+      "step": 1237
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.5301,
+      "step": 1238
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1404521404521406e-05,
+      "loss": 0.6131,
+      "step": 1239
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.138047138047138e-05,
+      "loss": 0.5245,
+      "step": 1240
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.135642135642136e-05,
+      "loss": 0.6166,
+      "step": 1241
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1332371332371334e-05,
+      "loss": 0.54,
+      "step": 1242
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1308321308321312e-05,
+      "loss": 0.4971,
+      "step": 1243
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1284271284271284e-05,
+      "loss": 0.5388,
+      "step": 1244
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.1260221260221262e-05,
+      "loss": 0.5811,
+      "step": 1245
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1236171236171237e-05,
+      "loss": 0.5244,
+      "step": 1246
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.5003,
+      "step": 1247
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.118807118807119e-05,
+      "loss": 0.5098,
+      "step": 1248
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1164021164021164e-05,
+      "loss": 0.5447,
+      "step": 1249
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.113997113997114e-05,
+      "loss": 0.5913,
+      "step": 1250
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1115921115921117e-05,
+      "loss": 0.5574,
+      "step": 1251
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1091871091871092e-05,
+      "loss": 0.5339,
+      "step": 1252
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.106782106782107e-05,
+      "loss": 0.5948,
+      "step": 1253
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.1043771043771045e-05,
+      "loss": 0.5929,
+      "step": 1254
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.101972101972102e-05,
+      "loss": 0.5072,
+      "step": 1255
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0995670995670998e-05,
+      "loss": 0.5054,
+      "step": 1256
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0971620971620973e-05,
+      "loss": 0.6383,
+      "step": 1257
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.094757094757095e-05,
+      "loss": 0.534,
+      "step": 1258
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0923520923520926e-05,
+      "loss": 0.5384,
+      "step": 1259
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.08994708994709e-05,
+      "loss": 0.5929,
+      "step": 1260
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0875420875420875e-05,
+      "loss": 0.4985,
+      "step": 1261
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0851370851370853e-05,
+      "loss": 0.563,
+      "step": 1262
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0827320827320828e-05,
+      "loss": 0.5459,
+      "step": 1263
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0803270803270806e-05,
+      "loss": 0.559,
+      "step": 1264
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.077922077922078e-05,
+      "loss": 0.6835,
+      "step": 1265
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.0755170755170756e-05,
+      "loss": 0.6066,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.073112073112073e-05,
+      "loss": 0.538,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.070707070707071e-05,
+      "loss": 0.694,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0683020683020683e-05,
+      "loss": 0.5813,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.065897065897066e-05,
+      "loss": 0.5509,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.786,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.061087061087061e-05,
+      "loss": 0.5837,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.058682058682059e-05,
+      "loss": 0.6485,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0562770562770564e-05,
+      "loss": 0.7161,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0538720538720542e-05,
+      "loss": 0.6768,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0514670514670517e-05,
+      "loss": 0.6081,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0490620490620492e-05,
+      "loss": 0.5353,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0466570466570467e-05,
+      "loss": 0.5178,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0442520442520445e-05,
+      "loss": 0.5862,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.041847041847042e-05,
+      "loss": 0.5986,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0394420394420398e-05,
+      "loss": 0.5803,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.5845,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0346320346320347e-05,
+      "loss": 0.5889,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0322270322270322e-05,
+      "loss": 0.5274,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.02982202982203e-05,
+      "loss": 0.5834,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0274170274170275e-05,
+      "loss": 0.5228,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.025012025012025e-05,
+      "loss": 0.5619,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.0226070226070225e-05,
+      "loss": 0.5735,
+      "step": 1288
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 0.5475,
+      "step": 1289
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0177970177970177e-05,
+      "loss": 0.5635,
+      "step": 1290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0153920153920156e-05,
+      "loss": 0.6249,
+      "step": 1291
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.012987012987013e-05,
+      "loss": 0.61,
+      "step": 1292
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0105820105820105e-05,
+      "loss": 0.6051,
+      "step": 1293
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0081770081770083e-05,
+      "loss": 0.6042,
+      "step": 1294
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0057720057720058e-05,
+      "loss": 0.5726,
+      "step": 1295
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.0033670033670036e-05,
+      "loss": 0.4402,
+      "step": 1296
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.000962000962001e-05,
+      "loss": 0.6102,
+      "step": 1297
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9985569985569986e-05,
+      "loss": 0.6354,
+      "step": 1298
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.996151996151996e-05,
+      "loss": 0.5174,
+      "step": 1299
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.993746993746994e-05,
+      "loss": 0.6049,
+      "step": 1300
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9913419913419914e-05,
+      "loss": 0.5977,
+      "step": 1301
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9889369889369892e-05,
+      "loss": 0.5778,
+      "step": 1302
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9865319865319866e-05,
+      "loss": 0.5875,
+      "step": 1303
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.6054,
+      "step": 1304
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9817219817219816e-05,
+      "loss": 0.559,
+      "step": 1305
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9793169793169794e-05,
+      "loss": 0.5135,
+      "step": 1306
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.976911976911977e-05,
+      "loss": 0.5817,
+      "step": 1307
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9745069745069747e-05,
+      "loss": 0.6049,
+      "step": 1308
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9721019721019722e-05,
+      "loss": 0.4793,
+      "step": 1309
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 0.6013,
+      "step": 1310
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9672919672919675e-05,
+      "loss": 0.5564,
+      "step": 1311
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.964886964886965e-05,
+      "loss": 0.6089,
+      "step": 1312
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9624819624819628e-05,
+      "loss": 0.5878,
+      "step": 1313
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9600769600769603e-05,
+      "loss": 0.5901,
+      "step": 1314
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9576719576719577e-05,
+      "loss": 0.4948,
+      "step": 1315
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9552669552669552e-05,
+      "loss": 0.6545,
+      "step": 1316
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.952861952861953e-05,
+      "loss": 0.5229,
+      "step": 1317
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9504569504569505e-05,
+      "loss": 0.6251,
+      "step": 1318
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9480519480519483e-05,
+      "loss": 0.65,
+      "step": 1319
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9456469456469455e-05,
+      "loss": 0.6832,
+      "step": 1320
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9432419432419433e-05,
+      "loss": 0.6322,
+      "step": 1321
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9408369408369408e-05,
+      "loss": 0.6063,
+      "step": 1322
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9384319384319386e-05,
+      "loss": 0.5351,
+      "step": 1323
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.936026936026936e-05,
+      "loss": 0.6032,
+      "step": 1324
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.933621933621934e-05,
+      "loss": 0.4981,
+      "step": 1325
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9312169312169313e-05,
+      "loss": 0.5529,
+      "step": 1326
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9288119288119288e-05,
+      "loss": 0.615,
+      "step": 1327
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9264069264069266e-05,
+      "loss": 0.6231,
+      "step": 1328
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.924001924001924e-05,
+      "loss": 0.5888,
+      "step": 1329
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.921596921596922e-05,
+      "loss": 0.5604,
+      "step": 1330
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.919191919191919e-05,
+      "loss": 0.5791,
+      "step": 1331
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.916786916786917e-05,
+      "loss": 0.518,
+      "step": 1332
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9143819143819144e-05,
+      "loss": 0.6004,
+      "step": 1333
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9119769119769122e-05,
+      "loss": 0.6559,
+      "step": 1334
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9095719095719097e-05,
+      "loss": 0.615,
+      "step": 1335
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9071669071669075e-05,
+      "loss": 0.5483,
+      "step": 1336
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.6347,
+      "step": 1337
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9023569023569024e-05,
+      "loss": 0.545,
+      "step": 1338
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8999518999519e-05,
+      "loss": 0.6577,
+      "step": 1339
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8975468975468977e-05,
+      "loss": 0.627,
+      "step": 1340
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8951418951418952e-05,
+      "loss": 0.5216,
+      "step": 1341
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8927368927368927e-05,
+      "loss": 0.5443,
+      "step": 1342
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8903318903318905e-05,
+      "loss": 0.6069,
+      "step": 1343
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.887926887926888e-05,
+      "loss": 0.6572,
+      "step": 1344
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8855218855218858e-05,
+      "loss": 0.6141,
+      "step": 1345
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8831168831168833e-05,
+      "loss": 0.5435,
+      "step": 1346
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.880711880711881e-05,
+      "loss": 0.5757,
+      "step": 1347
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8783068783068782e-05,
+      "loss": 0.656,
+      "step": 1348
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.875901875901876e-05,
+      "loss": 0.6017,
+      "step": 1349
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8734968734968735e-05,
+      "loss": 0.6813,
+      "step": 1350
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8710918710918713e-05,
+      "loss": 0.6338,
+      "step": 1351
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8686868686868688e-05,
+      "loss": 0.6165,
+      "step": 1352
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8662818662818663e-05,
+      "loss": 0.591,
+      "step": 1353
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8638768638768638e-05,
+      "loss": 0.5521,
+      "step": 1354
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8614718614718616e-05,
+      "loss": 0.5629,
+      "step": 1355
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.859066859066859e-05,
+      "loss": 0.6117,
+      "step": 1356
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.856661856661857e-05,
+      "loss": 0.5357,
+      "step": 1357
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8542568542568544e-05,
+      "loss": 0.6292,
+      "step": 1358
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.58,
+      "step": 1359
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8494468494468496e-05,
+      "loss": 0.5643,
+      "step": 1360
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.847041847041847e-05,
+      "loss": 0.5723,
+      "step": 1361
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.844636844636845e-05,
+      "loss": 0.577,
+      "step": 1362
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8422318422318424e-05,
+      "loss": 0.5581,
+      "step": 1363
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.83982683982684e-05,
+      "loss": 0.6122,
+      "step": 1364
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8374218374218374e-05,
+      "loss": 0.5178,
+      "step": 1365
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8350168350168352e-05,
+      "loss": 0.5759,
+      "step": 1366
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8326118326118327e-05,
+      "loss": 0.6772,
+      "step": 1367
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8302068302068305e-05,
+      "loss": 0.5625,
+      "step": 1368
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.827801827801828e-05,
+      "loss": 0.611,
+      "step": 1369
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.552,
+      "step": 1370
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.822991822991823e-05,
+      "loss": 0.5713,
+      "step": 1371
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8205868205868207e-05,
+      "loss": 0.5791,
+      "step": 1372
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.5451,
+      "step": 1373
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.815776815776816e-05,
+      "loss": 0.5704,
+      "step": 1374
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8133718133718135e-05,
+      "loss": 0.656,
+      "step": 1375
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.810966810966811e-05,
+      "loss": 0.5926,
+      "step": 1376
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8085618085618085e-05,
+      "loss": 0.545,
+      "step": 1377
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8061568061568063e-05,
+      "loss": 0.5471,
+      "step": 1378
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8037518037518038e-05,
+      "loss": 0.5853,
+      "step": 1379
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8013468013468016e-05,
+      "loss": 0.6211,
+      "step": 1380
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.798941798941799e-05,
+      "loss": 0.5551,
+      "step": 1381
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7965367965367965e-05,
+      "loss": 0.6679,
+      "step": 1382
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7941317941317943e-05,
+      "loss": 0.6205,
+      "step": 1383
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7917267917267918e-05,
+      "loss": 0.4969,
+      "step": 1384
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7893217893217896e-05,
+      "loss": 0.5381,
+      "step": 1385
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7869167869167868e-05,
+      "loss": 0.5606,
+      "step": 1386
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7845117845117846e-05,
+      "loss": 0.7007,
+      "step": 1387
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.782106782106782e-05,
+      "loss": 0.6537,
+      "step": 1388
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.77970177970178e-05,
+      "loss": 0.6,
+      "step": 1389
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7772967772967774e-05,
+      "loss": 0.6578,
+      "step": 1390
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7748917748917752e-05,
+      "loss": 0.5849,
+      "step": 1391
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7724867724867723e-05,
+      "loss": 0.5273,
+      "step": 1392
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.77008177008177e-05,
+      "loss": 0.5632,
+      "step": 1393
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.7676767676767676e-05,
+      "loss": 0.588,
+      "step": 1394
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7652717652717654e-05,
+      "loss": 0.5082,
+      "step": 1395
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.762866762866763e-05,
+      "loss": 0.5865,
+      "step": 1396
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7604617604617604e-05,
+      "loss": 0.5774,
+      "step": 1397
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7580567580567582e-05,
+      "loss": 0.6227,
+      "step": 1398
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7556517556517557e-05,
+      "loss": 0.6261,
+      "step": 1399
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7532467532467535e-05,
+      "loss": 0.6227,
+      "step": 1400
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.750841750841751e-05,
+      "loss": 0.5311,
+      "step": 1401
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7484367484367488e-05,
+      "loss": 0.4874,
+      "step": 1402
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.6275,
+      "step": 1403
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7436267436267437e-05,
+      "loss": 0.5285,
+      "step": 1404
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7412217412217412e-05,
+      "loss": 0.6103,
+      "step": 1405
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.738816738816739e-05,
+      "loss": 0.5389,
+      "step": 1406
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7364117364117365e-05,
+      "loss": 0.6095,
+      "step": 1407
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.734006734006734e-05,
+      "loss": 0.5406,
+      "step": 1408
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7316017316017315e-05,
+      "loss": 0.4554,
+      "step": 1409
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7291967291967293e-05,
+      "loss": 0.5461,
+      "step": 1410
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7267917267917268e-05,
+      "loss": 0.5513,
+      "step": 1411
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7243867243867246e-05,
+      "loss": 0.6076,
+      "step": 1412
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.721981721981722e-05,
+      "loss": 0.5239,
+      "step": 1413
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7195767195767195e-05,
+      "loss": 0.6529,
+      "step": 1414
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.7171717171717173e-05,
+      "loss": 0.6179,
+      "step": 1415
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7147667147667148e-05,
+      "loss": 0.5524,
+      "step": 1416
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7123617123617126e-05,
+      "loss": 0.5523,
+      "step": 1417
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.70995670995671e-05,
+      "loss": 0.586,
+      "step": 1418
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7075517075517076e-05,
+      "loss": 0.5851,
+      "step": 1419
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.705146705146705e-05,
+      "loss": 0.6076,
+      "step": 1420
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.702741702741703e-05,
+      "loss": 0.472,
+      "step": 1421
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.7003367003367004e-05,
+      "loss": 0.609,
+      "step": 1422
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6979316979316982e-05,
+      "loss": 0.4656,
+      "step": 1423
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6955266955266957e-05,
+      "loss": 0.5685,
+      "step": 1424
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.693121693121693e-05,
+      "loss": 0.6327,
+      "step": 1425
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6907166907166906e-05,
+      "loss": 0.5692,
+      "step": 1426
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6883116883116884e-05,
+      "loss": 0.5362,
+      "step": 1427
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.685906685906686e-05,
+      "loss": 0.499,
+      "step": 1428
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6835016835016837e-05,
+      "loss": 0.5999,
+      "step": 1429
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6810966810966812e-05,
+      "loss": 0.6133,
+      "step": 1430
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6786916786916787e-05,
+      "loss": 0.5957,
+      "step": 1431
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6762866762866765e-05,
+      "loss": 0.4752,
+      "step": 1432
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.673881673881674e-05,
+      "loss": 0.5791,
+      "step": 1433
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6714766714766718e-05,
+      "loss": 0.5621,
+      "step": 1434
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6690716690716693e-05,
+      "loss": 0.5953,
+      "step": 1435
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.5868,
+      "step": 1436
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.6642616642616642e-05,
+      "loss": 0.5896,
+      "step": 1437
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.661856661856662e-05,
+      "loss": 0.6117,
+      "step": 1438
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6594516594516595e-05,
+      "loss": 0.5812,
+      "step": 1439
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6570466570466573e-05,
+      "loss": 0.5626,
+      "step": 1440
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6546416546416545e-05,
+      "loss": 0.6178,
+      "step": 1441
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6522366522366523e-05,
+      "loss": 0.5907,
+      "step": 1442
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6498316498316498e-05,
+      "loss": 0.5158,
+      "step": 1443
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6474266474266476e-05,
+      "loss": 0.5296,
+      "step": 1444
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.645021645021645e-05,
+      "loss": 0.6603,
+      "step": 1445
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.642616642616643e-05,
+      "loss": 0.613,
+      "step": 1446
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6402116402116404e-05,
+      "loss": 0.5198,
+      "step": 1447
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.637806637806638e-05,
+      "loss": 0.6874,
+      "step": 1448
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6354016354016356e-05,
+      "loss": 0.6339,
+      "step": 1449
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.632996632996633e-05,
+      "loss": 0.6178,
+      "step": 1450
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.630591630591631e-05,
+      "loss": 0.5547,
+      "step": 1451
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.628186628186628e-05,
+      "loss": 0.6198,
+      "step": 1452
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.625781625781626e-05,
+      "loss": 0.5205,
+      "step": 1453
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6233766233766234e-05,
+      "loss": 0.7739,
+      "step": 1454
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6209716209716212e-05,
+      "loss": 0.5333,
+      "step": 1455
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6185666185666187e-05,
+      "loss": 0.5881,
+      "step": 1456
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6161616161616165e-05,
+      "loss": 0.514,
+      "step": 1457
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.6137566137566136e-05,
+      "loss": 0.4917,
+      "step": 1458
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6113516113516114e-05,
+      "loss": 0.5392,
+      "step": 1459
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.608946608946609e-05,
+      "loss": 0.5813,
+      "step": 1460
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6065416065416067e-05,
+      "loss": 0.527,
+      "step": 1461
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6041366041366042e-05,
+      "loss": 0.5332,
+      "step": 1462
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.6017316017316017e-05,
+      "loss": 0.6546,
+      "step": 1463
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5993265993265995e-05,
+      "loss": 0.4607,
+      "step": 1464
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.596921596921597e-05,
+      "loss": 0.5209,
+      "step": 1465
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5945165945165948e-05,
+      "loss": 0.539,
+      "step": 1466
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5921115921115923e-05,
+      "loss": 0.5871,
+      "step": 1467
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5897065897065898e-05,
+      "loss": 0.5765,
+      "step": 1468
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.6952,
+      "step": 1469
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.584896584896585e-05,
+      "loss": 0.6188,
+      "step": 1470
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5824915824915825e-05,
+      "loss": 0.6805,
+      "step": 1471
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5800865800865803e-05,
+      "loss": 0.5076,
+      "step": 1472
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5776815776815778e-05,
+      "loss": 0.5314,
+      "step": 1473
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5752765752765753e-05,
+      "loss": 0.5023,
+      "step": 1474
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5728715728715728e-05,
+      "loss": 0.5523,
+      "step": 1475
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.5704665704665706e-05,
+      "loss": 0.5379,
+      "step": 1476
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.568061568061568e-05,
+      "loss": 0.5919,
+      "step": 1477
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.565656565656566e-05,
+      "loss": 0.529,
+      "step": 1478
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.563251563251563e-05,
+      "loss": 0.5872,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.560846560846561e-05,
+      "loss": 0.5799,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5584415584415583e-05,
+      "loss": 0.5685,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.556036556036556e-05,
+      "loss": 0.7085,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5536315536315536e-05,
+      "loss": 0.5002,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5512265512265514e-05,
+      "loss": 0.5871,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.548821548821549e-05,
+      "loss": 0.5623,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5464165464165464e-05,
+      "loss": 0.5371,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5440115440115442e-05,
+      "loss": 0.5061,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5416065416065417e-05,
+      "loss": 0.5819,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5392015392015395e-05,
+      "loss": 0.6551,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5367965367965366e-05,
+      "loss": 0.574,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5343915343915344e-05,
+      "loss": 0.6173,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.531986531986532e-05,
+      "loss": 0.489,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5295815295815297e-05,
+      "loss": 0.5328,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5271765271765272e-05,
+      "loss": 0.6448,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.524771524771525e-05,
+      "loss": 0.7005,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5223665223665223e-05,
+      "loss": 0.5298,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.51996151996152e-05,
+      "loss": 0.5759,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5175565175565176e-05,
+      "loss": 0.6547,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.4891,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.512746512746513e-05,
+      "loss": 0.505,
+      "step": 1500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5103415103415102e-05,
+      "loss": 0.5906,
+      "step": 1501
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.5736,
+      "step": 1502
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5055315055315055e-05,
+      "loss": 0.5896,
+      "step": 1503
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5031265031265032e-05,
+      "loss": 0.5726,
+      "step": 1504
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.5007215007215008e-05,
+      "loss": 0.6761,
+      "step": 1505
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4983164983164985e-05,
+      "loss": 0.5338,
+      "step": 1506
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.495911495911496e-05,
+      "loss": 0.552,
+      "step": 1507
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4935064935064936e-05,
+      "loss": 0.5533,
+      "step": 1508
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4911014911014912e-05,
+      "loss": 0.5431,
+      "step": 1509
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4886964886964889e-05,
+      "loss": 0.605,
+      "step": 1510
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4862914862914865e-05,
+      "loss": 0.4795,
+      "step": 1511
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4838864838864838e-05,
+      "loss": 0.5596,
+      "step": 1512
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.4755,
+      "step": 1513
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4790764790764791e-05,
+      "loss": 0.5466,
+      "step": 1514
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4766714766714768e-05,
+      "loss": 0.5698,
+      "step": 1515
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4742664742664744e-05,
+      "loss": 0.5602,
+      "step": 1516
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.471861471861472e-05,
+      "loss": 0.6138,
+      "step": 1517
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4694564694564694e-05,
+      "loss": 0.5708,
+      "step": 1518
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.467051467051467e-05,
+      "loss": 0.6296,
+      "step": 1519
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4646464646464647e-05,
+      "loss": 0.5705,
+      "step": 1520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4622414622414623e-05,
+      "loss": 0.5524,
+      "step": 1521
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.45983645983646e-05,
+      "loss": 0.5203,
+      "step": 1522
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4574314574314573e-05,
+      "loss": 0.5503,
+      "step": 1523
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.455026455026455e-05,
+      "loss": 0.5604,
+      "step": 1524
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4526214526214526e-05,
+      "loss": 0.5223,
+      "step": 1525
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4502164502164502e-05,
+      "loss": 0.6359,
+      "step": 1526
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4478114478114479e-05,
+      "loss": 0.5481,
+      "step": 1527
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4454064454064455e-05,
+      "loss": 0.476,
+      "step": 1528
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.443001443001443e-05,
+      "loss": 0.4672,
+      "step": 1529
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4405964405964406e-05,
+      "loss": 0.5436,
+      "step": 1530
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4381914381914383e-05,
+      "loss": 0.5814,
+      "step": 1531
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.435786435786436e-05,
+      "loss": 0.5058,
+      "step": 1532
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4333814333814336e-05,
+      "loss": 0.4944,
+      "step": 1533
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4309764309764309e-05,
+      "loss": 0.5984,
+      "step": 1534
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.4705,
+      "step": 1535
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4261664261664262e-05,
+      "loss": 0.4316,
+      "step": 1536
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4237614237614238e-05,
+      "loss": 0.6038,
+      "step": 1537
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4213564213564215e-05,
+      "loss": 0.5716,
+      "step": 1538
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4189514189514191e-05,
+      "loss": 0.5512,
+      "step": 1539
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4165464165464164e-05,
+      "loss": 0.6395,
+      "step": 1540
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4141414141414141e-05,
+      "loss": 0.548,
+      "step": 1541
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4117364117364117e-05,
+      "loss": 0.645,
+      "step": 1542
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4093314093314094e-05,
+      "loss": 0.5535,
+      "step": 1543
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.406926406926407e-05,
+      "loss": 0.644,
+      "step": 1544
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4045214045214045e-05,
+      "loss": 0.5338,
+      "step": 1545
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4021164021164022e-05,
+      "loss": 0.6834,
+      "step": 1546
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3997113997113998e-05,
+      "loss": 0.5792,
+      "step": 1547
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3973063973063974e-05,
+      "loss": 0.5548,
+      "step": 1548
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3949013949013951e-05,
+      "loss": 0.52,
+      "step": 1549
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3924963924963927e-05,
+      "loss": 0.6705,
+      "step": 1550
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.39009139009139e-05,
+      "loss": 0.5587,
+      "step": 1551
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3876863876863877e-05,
+      "loss": 0.4603,
+      "step": 1552
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3852813852813853e-05,
+      "loss": 0.6382,
+      "step": 1553
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.382876382876383e-05,
+      "loss": 0.5633,
+      "step": 1554
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3804713804713806e-05,
+      "loss": 0.6838,
+      "step": 1555
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.378066378066378e-05,
+      "loss": 0.7206,
+      "step": 1556
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3756613756613756e-05,
+      "loss": 0.6036,
+      "step": 1557
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3732563732563732e-05,
+      "loss": 0.4934,
+      "step": 1558
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3708513708513709e-05,
+      "loss": 0.4921,
+      "step": 1559
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3684463684463685e-05,
+      "loss": 0.6349,
+      "step": 1560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3660413660413662e-05,
+      "loss": 0.6319,
+      "step": 1561
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.5772,
+      "step": 1562
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3612313612313613e-05,
+      "loss": 0.5981,
+      "step": 1563
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.358826358826359e-05,
+      "loss": 0.5655,
+      "step": 1564
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3564213564213566e-05,
+      "loss": 0.6731,
+      "step": 1565
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3540163540163542e-05,
+      "loss": 0.5841,
+      "step": 1566
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3516113516113516e-05,
+      "loss": 0.5725,
+      "step": 1567
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.5188,
+      "step": 1568
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3468013468013468e-05,
+      "loss": 0.4954,
+      "step": 1569
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3443963443963445e-05,
+      "loss": 0.5932,
+      "step": 1570
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3419913419913421e-05,
+      "loss": 0.4811,
+      "step": 1571
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3395863395863398e-05,
+      "loss": 0.5511,
+      "step": 1572
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3371813371813371e-05,
+      "loss": 0.5234,
+      "step": 1573
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3347763347763347e-05,
+      "loss": 0.5244,
+      "step": 1574
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3323713323713324e-05,
+      "loss": 0.5831,
+      "step": 1575
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.32996632996633e-05,
+      "loss": 0.5435,
+      "step": 1576
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3275613275613277e-05,
+      "loss": 0.5754,
+      "step": 1577
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3251563251563252e-05,
+      "loss": 0.544,
+      "step": 1578
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3227513227513228e-05,
+      "loss": 0.5661,
+      "step": 1579
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3203463203463205e-05,
+      "loss": 0.5347,
+      "step": 1580
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3179413179413181e-05,
+      "loss": 0.5351,
+      "step": 1581
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3155363155363157e-05,
+      "loss": 0.5602,
+      "step": 1582
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3131313131313134e-05,
+      "loss": 0.6315,
+      "step": 1583
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3107263107263107e-05,
+      "loss": 0.5109,
+      "step": 1584
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3083213083213083e-05,
+      "loss": 0.5808,
+      "step": 1585
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.305916305916306e-05,
+      "loss": 0.6144,
+      "step": 1586
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3035113035113036e-05,
+      "loss": 0.5395,
+      "step": 1587
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.3011063011063013e-05,
+      "loss": 0.5093,
+      "step": 1588
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2987012987012986e-05,
+      "loss": 0.7379,
+      "step": 1589
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.5193,
+      "step": 1590
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2938912938912939e-05,
+      "loss": 0.5503,
+      "step": 1591
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2914862914862915e-05,
+      "loss": 0.6004,
+      "step": 1592
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2890812890812892e-05,
+      "loss": 0.5611,
+      "step": 1593
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2866762866762868e-05,
+      "loss": 0.632,
+      "step": 1594
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2842712842712843e-05,
+      "loss": 0.5992,
+      "step": 1595
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.281866281866282e-05,
+      "loss": 0.4747,
+      "step": 1596
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2794612794612796e-05,
+      "loss": 0.5865,
+      "step": 1597
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2770562770562773e-05,
+      "loss": 0.5508,
+      "step": 1598
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2746512746512749e-05,
+      "loss": 0.622,
+      "step": 1599
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2722462722462722e-05,
+      "loss": 0.5058,
+      "step": 1600
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.5496,
+      "step": 1601
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2674362674362675e-05,
+      "loss": 0.5323,
+      "step": 1602
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2650312650312651e-05,
+      "loss": 0.4477,
+      "step": 1603
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2626262626262628e-05,
+      "loss": 0.6631,
+      "step": 1604
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2602212602212604e-05,
+      "loss": 0.5275,
+      "step": 1605
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2578162578162577e-05,
+      "loss": 0.6476,
+      "step": 1606
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2554112554112554e-05,
+      "loss": 0.4736,
+      "step": 1607
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.253006253006253e-05,
+      "loss": 0.6321,
+      "step": 1608
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2506012506012507e-05,
+      "loss": 0.5449,
+      "step": 1609
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2481962481962482e-05,
+      "loss": 0.5878,
+      "step": 1610
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2457912457912458e-05,
+      "loss": 0.5955,
+      "step": 1611
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2433862433862433e-05,
+      "loss": 0.5935,
+      "step": 1612
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.240981240981241e-05,
+      "loss": 0.5341,
+      "step": 1613
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2385762385762386e-05,
+      "loss": 0.4634,
+      "step": 1614
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2361712361712362e-05,
+      "loss": 0.4771,
+      "step": 1615
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2337662337662339e-05,
+      "loss": 0.5328,
+      "step": 1616
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2313612313612315e-05,
+      "loss": 0.5125,
+      "step": 1617
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.228956228956229e-05,
+      "loss": 0.6462,
+      "step": 1618
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2265512265512267e-05,
+      "loss": 0.5712,
+      "step": 1619
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2241462241462243e-05,
+      "loss": 0.5838,
+      "step": 1620
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2217412217412218e-05,
+      "loss": 0.5308,
+      "step": 1621
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2193362193362194e-05,
+      "loss": 0.5879,
+      "step": 1622
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2169312169312169e-05,
+      "loss": 0.591,
+      "step": 1623
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2145262145262145e-05,
+      "loss": 0.6062,
+      "step": 1624
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.5263,
+      "step": 1625
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2097162097162097e-05,
+      "loss": 0.6074,
+      "step": 1626
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2073112073112073e-05,
+      "loss": 0.4953,
+      "step": 1627
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.204906204906205e-05,
+      "loss": 0.5474,
+      "step": 1628
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2025012025012024e-05,
+      "loss": 0.5876,
+      "step": 1629
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2000962000962001e-05,
+      "loss": 0.6011,
+      "step": 1630
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1976911976911977e-05,
+      "loss": 0.5539,
+      "step": 1631
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1952861952861954e-05,
+      "loss": 0.5341,
+      "step": 1632
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.192881192881193e-05,
+      "loss": 0.7118,
+      "step": 1633
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.4899,
+      "step": 1634
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1880711880711882e-05,
+      "loss": 0.546,
+      "step": 1635
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1856661856661858e-05,
+      "loss": 0.4721,
+      "step": 1636
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1832611832611833e-05,
+      "loss": 0.6183,
+      "step": 1637
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.180856180856181e-05,
+      "loss": 0.4948,
+      "step": 1638
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1784511784511786e-05,
+      "loss": 0.4466,
+      "step": 1639
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.176046176046176e-05,
+      "loss": 0.5864,
+      "step": 1640
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1736411736411737e-05,
+      "loss": 0.604,
+      "step": 1641
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1712361712361713e-05,
+      "loss": 0.6098,
+      "step": 1642
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1688311688311688e-05,
+      "loss": 0.5865,
+      "step": 1643
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1664261664261665e-05,
+      "loss": 0.5409,
+      "step": 1644
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.164021164021164e-05,
+      "loss": 0.5426,
+      "step": 1645
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1616161616161616e-05,
+      "loss": 0.4545,
+      "step": 1646
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1592111592111592e-05,
+      "loss": 0.6357,
+      "step": 1647
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1568061568061569e-05,
+      "loss": 0.5427,
+      "step": 1648
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.1544011544011545e-05,
+      "loss": 0.4858,
+      "step": 1649
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1519961519961522e-05,
+      "loss": 0.4807,
+      "step": 1650
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1495911495911497e-05,
+      "loss": 0.5331,
+      "step": 1651
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1471861471861473e-05,
+      "loss": 0.4778,
+      "step": 1652
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.144781144781145e-05,
+      "loss": 0.5949,
+      "step": 1653
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1423761423761424e-05,
+      "loss": 0.5184,
+      "step": 1654
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.13997113997114e-05,
+      "loss": 0.5217,
+      "step": 1655
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1375661375661376e-05,
+      "loss": 0.5584,
+      "step": 1656
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1351611351611352e-05,
+      "loss": 0.5637,
+      "step": 1657
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1327561327561329e-05,
+      "loss": 0.6354,
+      "step": 1658
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1303511303511303e-05,
+      "loss": 0.6338,
+      "step": 1659
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.127946127946128e-05,
+      "loss": 0.5096,
+      "step": 1660
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1255411255411256e-05,
+      "loss": 0.6057,
+      "step": 1661
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1231361231361231e-05,
+      "loss": 0.5228,
+      "step": 1662
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1207311207311207e-05,
+      "loss": 0.6037,
+      "step": 1663
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1183261183261184e-05,
+      "loss": 0.5388,
+      "step": 1664
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1159211159211159e-05,
+      "loss": 0.4856,
+      "step": 1665
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1135161135161135e-05,
+      "loss": 0.5084,
+      "step": 1666
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.4903,
+      "step": 1667
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1087061087061088e-05,
+      "loss": 0.6455,
+      "step": 1668
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1063011063011065e-05,
+      "loss": 0.5867,
+      "step": 1669
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.103896103896104e-05,
+      "loss": 0.6626,
+      "step": 1670
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1014911014911016e-05,
+      "loss": 0.6223,
+      "step": 1671
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0990860990860992e-05,
+      "loss": 0.4428,
+      "step": 1672
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0966810966810967e-05,
+      "loss": 0.5545,
+      "step": 1673
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0942760942760944e-05,
+      "loss": 0.5805,
+      "step": 1674
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.091871091871092e-05,
+      "loss": 0.5472,
+      "step": 1675
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0894660894660895e-05,
+      "loss": 0.6305,
+      "step": 1676
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0870610870610871e-05,
+      "loss": 0.5659,
+      "step": 1677
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0846560846560846e-05,
+      "loss": 0.5619,
+      "step": 1678
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0822510822510823e-05,
+      "loss": 0.5599,
+      "step": 1679
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0798460798460799e-05,
+      "loss": 0.606,
+      "step": 1680
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0774410774410774e-05,
+      "loss": 0.6268,
+      "step": 1681
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.075036075036075e-05,
+      "loss": 0.4547,
+      "step": 1682
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0726310726310727e-05,
+      "loss": 0.6137,
+      "step": 1683
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0702260702260703e-05,
+      "loss": 0.6192,
+      "step": 1684
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.067821067821068e-05,
+      "loss": 0.5139,
+      "step": 1685
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0654160654160656e-05,
+      "loss": 0.5799,
+      "step": 1686
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0630110630110631e-05,
+      "loss": 0.5718,
+      "step": 1687
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0606060606060607e-05,
+      "loss": 0.6441,
+      "step": 1688
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0582010582010582e-05,
+      "loss": 0.5673,
+      "step": 1689
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0557960557960559e-05,
+      "loss": 0.5037,
+      "step": 1690
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0533910533910535e-05,
+      "loss": 0.5549,
+      "step": 1691
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.050986050986051e-05,
+      "loss": 0.5677,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0485810485810486e-05,
+      "loss": 0.4753,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0461760461760463e-05,
+      "loss": 0.5608,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0437710437710438e-05,
+      "loss": 0.7183,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0413660413660414e-05,
+      "loss": 0.6544,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.038961038961039e-05,
+      "loss": 0.5765,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0365560365560365e-05,
+      "loss": 0.559,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0341510341510342e-05,
+      "loss": 0.4791,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.6109,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0293410293410295e-05,
+      "loss": 0.5058,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0269360269360271e-05,
+      "loss": 0.5351,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0245310245310246e-05,
+      "loss": 0.4618,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0221260221260222e-05,
+      "loss": 0.5658,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0197210197210199e-05,
+      "loss": 0.5627,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0173160173160174e-05,
+      "loss": 0.5971,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.014911014911015e-05,
+      "loss": 0.5349,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0125060125060125e-05,
+      "loss": 0.6091,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 0.6132,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0076960076960078e-05,
+      "loss": 0.5681,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0052910052910053e-05,
+      "loss": 0.5634,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0028860028860029e-05,
+      "loss": 0.5856,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0004810004810006e-05,
+      "loss": 0.5842,
+      "step": 1713
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.98075998075998e-06,
+      "loss": 0.6588,
+      "step": 1714
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.956709956709957e-06,
+      "loss": 0.5167,
+      "step": 1715
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.932659932659933e-06,
+      "loss": 0.6123,
+      "step": 1716
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.908609908609908e-06,
+      "loss": 0.6064,
+      "step": 1717
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.884559884559884e-06,
+      "loss": 0.5394,
+      "step": 1718
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.860509860509861e-06,
+      "loss": 0.4706,
+      "step": 1719
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.836459836459837e-06,
+      "loss": 0.5751,
+      "step": 1720
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.812409812409814e-06,
+      "loss": 0.6832,
+      "step": 1721
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.788359788359789e-06,
+      "loss": 0.7178,
+      "step": 1722
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.764309764309765e-06,
+      "loss": 0.554,
+      "step": 1723
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.740259740259742e-06,
+      "loss": 0.7026,
+      "step": 1724
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.716209716209716e-06,
+      "loss": 0.5458,
+      "step": 1725
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.692159692159693e-06,
+      "loss": 0.5557,
+      "step": 1726
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.66810966810967e-06,
+      "loss": 0.5833,
+      "step": 1727
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.644059644059644e-06,
+      "loss": 0.4683,
+      "step": 1728
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.62000962000962e-06,
+      "loss": 0.5413,
+      "step": 1729
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.595959595959595e-06,
+      "loss": 0.69,
+      "step": 1730
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.571909571909572e-06,
+      "loss": 0.6008,
+      "step": 1731
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.547859547859548e-06,
+      "loss": 0.558,
+      "step": 1732
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.4891,
+      "step": 1733
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.4997594997595e-06,
+      "loss": 0.4966,
+      "step": 1734
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.475709475709476e-06,
+      "loss": 0.5538,
+      "step": 1735
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.451659451659452e-06,
+      "loss": 0.4838,
+      "step": 1736
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.427609427609429e-06,
+      "loss": 0.622,
+      "step": 1737
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.403559403559405e-06,
+      "loss": 0.5843,
+      "step": 1738
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.37950937950938e-06,
+      "loss": 0.562,
+      "step": 1739
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.355459355459357e-06,
+      "loss": 0.5712,
+      "step": 1740
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.331409331409331e-06,
+      "loss": 0.6185,
+      "step": 1741
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.307359307359308e-06,
+      "loss": 0.5126,
+      "step": 1742
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.283309283309284e-06,
+      "loss": 0.6268,
+      "step": 1743
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.5099,
+      "step": 1744
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.235209235209236e-06,
+      "loss": 0.5972,
+      "step": 1745
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.211159211159212e-06,
+      "loss": 0.4467,
+      "step": 1746
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.187109187109187e-06,
+      "loss": 0.6146,
+      "step": 1747
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.163059163059163e-06,
+      "loss": 0.6405,
+      "step": 1748
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.13900913900914e-06,
+      "loss": 0.5499,
+      "step": 1749
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.114959114959115e-06,
+      "loss": 0.5107,
+      "step": 1750
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.6749,
+      "step": 1751
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.066859066859068e-06,
+      "loss": 0.643,
+      "step": 1752
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.042809042809042e-06,
+      "loss": 0.4986,
+      "step": 1753
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.018759018759019e-06,
+      "loss": 0.5686,
+      "step": 1754
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.994708994708995e-06,
+      "loss": 0.5242,
+      "step": 1755
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.970658970658972e-06,
+      "loss": 0.576,
+      "step": 1756
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.946608946608948e-06,
+      "loss": 0.5131,
+      "step": 1757
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.922558922558923e-06,
+      "loss": 0.613,
+      "step": 1758
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.8985088985089e-06,
+      "loss": 0.5125,
+      "step": 1759
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.874458874458876e-06,
+      "loss": 0.563,
+      "step": 1760
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.85040885040885e-06,
+      "loss": 0.6282,
+      "step": 1761
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.826358826358827e-06,
+      "loss": 0.5569,
+      "step": 1762
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.802308802308802e-06,
+      "loss": 0.5853,
+      "step": 1763
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.778258778258778e-06,
+      "loss": 0.5174,
+      "step": 1764
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.754208754208755e-06,
+      "loss": 0.5563,
+      "step": 1765
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.5508,
+      "step": 1766
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.706108706108706e-06,
+      "loss": 0.6126,
+      "step": 1767
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.682058682058683e-06,
+      "loss": 0.5815,
+      "step": 1768
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.658008658008657e-06,
+      "loss": 0.6425,
+      "step": 1769
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.633958633958634e-06,
+      "loss": 0.4888,
+      "step": 1770
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.60990860990861e-06,
+      "loss": 0.5107,
+      "step": 1771
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 0.5234,
+      "step": 1772
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.561808561808563e-06,
+      "loss": 0.5916,
+      "step": 1773
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.537758537758538e-06,
+      "loss": 0.5641,
+      "step": 1774
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.513708513708514e-06,
+      "loss": 0.6548,
+      "step": 1775
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.489658489658491e-06,
+      "loss": 0.5354,
+      "step": 1776
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.465608465608466e-06,
+      "loss": 0.5724,
+      "step": 1777
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.441558441558442e-06,
+      "loss": 0.511,
+      "step": 1778
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.417508417508419e-06,
+      "loss": 0.6357,
+      "step": 1779
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.393458393458393e-06,
+      "loss": 0.4993,
+      "step": 1780
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.36940836940837e-06,
+      "loss": 0.6106,
+      "step": 1781
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.345358345358346e-06,
+      "loss": 0.5953,
+      "step": 1782
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.321308321308321e-06,
+      "loss": 0.5657,
+      "step": 1783
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.297258297258298e-06,
+      "loss": 0.5351,
+      "step": 1784
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.273208273208272e-06,
+      "loss": 0.5797,
+      "step": 1785
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.249158249158249e-06,
+      "loss": 0.5435,
+      "step": 1786
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.225108225108225e-06,
+      "loss": 0.5813,
+      "step": 1787
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.201058201058202e-06,
+      "loss": 0.6156,
+      "step": 1788
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.177008177008178e-06,
+      "loss": 0.6364,
+      "step": 1789
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.152958152958155e-06,
+      "loss": 0.5727,
+      "step": 1790
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.12890812890813e-06,
+      "loss": 0.5769,
+      "step": 1791
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.104858104858106e-06,
+      "loss": 0.5171,
+      "step": 1792
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.080808080808082e-06,
+      "loss": 0.53,
+      "step": 1793
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.056758056758057e-06,
+      "loss": 0.5309,
+      "step": 1794
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.032708032708034e-06,
+      "loss": 0.7615,
+      "step": 1795
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.008658008658008e-06,
+      "loss": 0.5525,
+      "step": 1796
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 7.984607984607985e-06,
+      "loss": 0.5287,
+      "step": 1797
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 7.960557960557961e-06,
+      "loss": 0.4706,
+      "step": 1798
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.5128,
+      "step": 1799
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.912457912457913e-06,
+      "loss": 0.5996,
+      "step": 1800
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.888407888407889e-06,
+      "loss": 0.4725,
+      "step": 1801
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.864357864357864e-06,
+      "loss": 0.519,
+      "step": 1802
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.84030784030784e-06,
+      "loss": 0.6211,
+      "step": 1803
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.816257816257815e-06,
+      "loss": 0.541,
+      "step": 1804
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.792207792207792e-06,
+      "loss": 0.5543,
+      "step": 1805
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.768157768157768e-06,
+      "loss": 0.5273,
+      "step": 1806
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.744107744107745e-06,
+      "loss": 0.5808,
+      "step": 1807
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.720057720057721e-06,
+      "loss": 0.5666,
+      "step": 1808
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.696007696007697e-06,
+      "loss": 0.5681,
+      "step": 1809
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.671957671957672e-06,
+      "loss": 0.6064,
+      "step": 1810
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.647907647907649e-06,
+      "loss": 0.4872,
+      "step": 1811
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.623857623857625e-06,
+      "loss": 0.609,
+      "step": 1812
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.5998075998076e-06,
+      "loss": 0.6087,
+      "step": 1813
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.6033,
+      "step": 1814
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.551707551707551e-06,
+      "loss": 0.5125,
+      "step": 1815
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.527657527657528e-06,
+      "loss": 0.5655,
+      "step": 1816
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.503607503607504e-06,
+      "loss": 0.4423,
+      "step": 1817
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.47955747955748e-06,
+      "loss": 0.6386,
+      "step": 1818
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.455507455507456e-06,
+      "loss": 0.5831,
+      "step": 1819
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 7.431457431457433e-06,
+      "loss": 0.5973,
+      "step": 1820
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.4867,
+      "step": 1821
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.383357383357384e-06,
+      "loss": 0.4015,
+      "step": 1822
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.35930735930736e-06,
+      "loss": 0.5765,
+      "step": 1823
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.335257335257335e-06,
+      "loss": 0.5645,
+      "step": 1824
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.311207311207312e-06,
+      "loss": 0.6565,
+      "step": 1825
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.2871572871572864e-06,
+      "loss": 0.5194,
+      "step": 1826
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.263107263107263e-06,
+      "loss": 0.4814,
+      "step": 1827
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.239057239057239e-06,
+      "loss": 0.5124,
+      "step": 1828
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.215007215007215e-06,
+      "loss": 0.5731,
+      "step": 1829
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.1909571909571915e-06,
+      "loss": 0.545,
+      "step": 1830
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.166907166907168e-06,
+      "loss": 0.6512,
+      "step": 1831
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.4826,
+      "step": 1832
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.118807118807119e-06,
+      "loss": 0.4529,
+      "step": 1833
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.094757094757096e-06,
+      "loss": 0.5116,
+      "step": 1834
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.0707070707070704e-06,
+      "loss": 0.6472,
+      "step": 1835
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.046657046657047e-06,
+      "loss": 0.512,
+      "step": 1836
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 7.0226070226070225e-06,
+      "loss": 0.4962,
+      "step": 1837
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.998556998556999e-06,
+      "loss": 0.5435,
+      "step": 1838
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.9745069745069755e-06,
+      "loss": 0.6568,
+      "step": 1839
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.95045695045695e-06,
+      "loss": 0.4923,
+      "step": 1840
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 6.926406926406927e-06,
+      "loss": 0.5545,
+      "step": 1841
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.902356902356903e-06,
+      "loss": 0.5135,
+      "step": 1842
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.878306878306878e-06,
+      "loss": 0.6168,
+      "step": 1843
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.854256854256854e-06,
+      "loss": 0.5509,
+      "step": 1844
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.830206830206831e-06,
+      "loss": 0.636,
+      "step": 1845
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.8061568061568065e-06,
+      "loss": 0.5497,
+      "step": 1846
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.782106782106783e-06,
+      "loss": 0.5691,
+      "step": 1847
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.758056758056758e-06,
+      "loss": 0.5028,
+      "step": 1848
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.734006734006734e-06,
+      "loss": 0.476,
+      "step": 1849
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.709956709956711e-06,
+      "loss": 0.4925,
+      "step": 1850
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.6859066859066855e-06,
+      "loss": 0.5209,
+      "step": 1851
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.661856661856662e-06,
+      "loss": 0.5412,
+      "step": 1852
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.637806637806638e-06,
+      "loss": 0.6505,
+      "step": 1853
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.613756613756614e-06,
+      "loss": 0.4326,
+      "step": 1854
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.5897065897065905e-06,
+      "loss": 0.4538,
+      "step": 1855
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.565656565656567e-06,
+      "loss": 0.5426,
+      "step": 1856
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.541606541606542e-06,
+      "loss": 0.4955,
+      "step": 1857
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.517556517556518e-06,
+      "loss": 0.518,
+      "step": 1858
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.493506493506493e-06,
+      "loss": 0.5521,
+      "step": 1859
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.4694564694564695e-06,
+      "loss": 0.5641,
+      "step": 1860
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.445406445406446e-06,
+      "loss": 0.4224,
+      "step": 1861
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 6.4213564213564216e-06,
+      "loss": 0.5755,
+      "step": 1862
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.397306397306398e-06,
+      "loss": 0.566,
+      "step": 1863
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.3732563732563745e-06,
+      "loss": 0.5528,
+      "step": 1864
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.5636,
+      "step": 1865
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.325156325156326e-06,
+      "loss": 0.5933,
+      "step": 1866
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.301106301106302e-06,
+      "loss": 0.6666,
+      "step": 1867
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.277056277056277e-06,
+      "loss": 0.5533,
+      "step": 1868
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.2530062530062535e-06,
+      "loss": 0.5326,
+      "step": 1869
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.228956228956229e-06,
+      "loss": 0.495,
+      "step": 1870
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.204906204906205e-06,
+      "loss": 0.4661,
+      "step": 1871
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.180856180856181e-06,
+      "loss": 0.5008,
+      "step": 1872
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.156806156806158e-06,
+      "loss": 0.5437,
+      "step": 1873
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.132756132756133e-06,
+      "loss": 0.581,
+      "step": 1874
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.108706108706109e-06,
+      "loss": 0.5869,
+      "step": 1875
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.0846560846560845e-06,
+      "loss": 0.6146,
+      "step": 1876
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.6531,
+      "step": 1877
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.036556036556037e-06,
+      "loss": 0.5169,
+      "step": 1878
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.012506012506012e-06,
+      "loss": 0.5723,
+      "step": 1879
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.988455988455989e-06,
+      "loss": 0.5669,
+      "step": 1880
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.964405964405965e-06,
+      "loss": 0.5969,
+      "step": 1881
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.940355940355941e-06,
+      "loss": 0.5656,
+      "step": 1882
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.916305916305916e-06,
+      "loss": 0.5303,
+      "step": 1883
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.892255892255893e-06,
+      "loss": 0.5289,
+      "step": 1884
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.8682058682058685e-06,
+      "loss": 0.5032,
+      "step": 1885
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.844155844155844e-06,
+      "loss": 0.5539,
+      "step": 1886
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.82010582010582e-06,
+      "loss": 0.6089,
+      "step": 1887
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.796055796055796e-06,
+      "loss": 0.6146,
+      "step": 1888
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.772005772005773e-06,
+      "loss": 0.4867,
+      "step": 1889
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.747955747955748e-06,
+      "loss": 0.5426,
+      "step": 1890
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.723905723905725e-06,
+      "loss": 0.4856,
+      "step": 1891
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.6998556998557e-06,
+      "loss": 0.4991,
+      "step": 1892
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.675805675805676e-06,
+      "loss": 0.6276,
+      "step": 1893
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.651755651755652e-06,
+      "loss": 0.5877,
+      "step": 1894
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.627705627705628e-06,
+      "loss": 0.4771,
+      "step": 1895
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.603655603655604e-06,
+      "loss": 0.6002,
+      "step": 1896
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.579605579605579e-06,
+      "loss": 0.5492,
+      "step": 1897
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.536,
+      "step": 1898
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.531505531505532e-06,
+      "loss": 0.56,
+      "step": 1899
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.507455507455508e-06,
+      "loss": 0.4295,
+      "step": 1900
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.4834054834054835e-06,
+      "loss": 0.4668,
+      "step": 1901
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.45935545935546e-06,
+      "loss": 0.5517,
+      "step": 1902
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.435305435305436e-06,
+      "loss": 0.5497,
+      "step": 1903
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.411255411255411e-06,
+      "loss": 0.5071,
+      "step": 1904
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.387205387205387e-06,
+      "loss": 0.6097,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.363155363155363e-06,
+      "loss": 0.5362,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.33910533910534e-06,
+      "loss": 0.5276,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.3150553150553154e-06,
+      "loss": 0.5189,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.291005291005291e-06,
+      "loss": 0.5624,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2669552669552675e-06,
+      "loss": 0.5291,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.242905242905243e-06,
+      "loss": 0.5132,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.218855218855219e-06,
+      "loss": 0.6354,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.194805194805195e-06,
+      "loss": 0.58,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.170755170755171e-06,
+      "loss": 0.5756,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.146705146705147e-06,
+      "loss": 0.5261,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.122655122655123e-06,
+      "loss": 0.5557,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.0986050986050994e-06,
+      "loss": 0.4822,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.074555074555075e-06,
+      "loss": 0.6373,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 0.5548,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.026455026455026e-06,
+      "loss": 0.5279,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.002405002405003e-06,
+      "loss": 0.4521,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.978354978354978e-06,
+      "loss": 0.7385,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.954304954304954e-06,
+      "loss": 0.5023,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.9302549302549305e-06,
+      "loss": 0.5486,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.906204906204907e-06,
+      "loss": 0.5703,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.8821548821548826e-06,
+      "loss": 0.4461,
+      "step": 1926
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.858104858104858e-06,
+      "loss": 0.5666,
+      "step": 1927
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.834054834054835e-06,
+      "loss": 0.5303,
+      "step": 1928
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.81000481000481e-06,
+      "loss": 0.5771,
+      "step": 1929
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.785954785954786e-06,
+      "loss": 0.592,
+      "step": 1930
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.5879,
+      "step": 1931
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.737854737854738e-06,
+      "loss": 0.5322,
+      "step": 1932
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.7138047138047145e-06,
+      "loss": 0.6585,
+      "step": 1933
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.68975468975469e-06,
+      "loss": 0.4755,
+      "step": 1934
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.665704665704666e-06,
+      "loss": 0.6204,
+      "step": 1935
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.641654641654642e-06,
+      "loss": 0.5138,
+      "step": 1936
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.617604617604618e-06,
+      "loss": 0.5636,
+      "step": 1937
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.5935545935545934e-06,
+      "loss": 0.5719,
+      "step": 1938
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.56950456950457e-06,
+      "loss": 0.6195,
+      "step": 1939
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.4663,
+      "step": 1940
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.521404521404521e-06,
+      "loss": 0.5741,
+      "step": 1941
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.497354497354498e-06,
+      "loss": 0.4642,
+      "step": 1942
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.473304473304474e-06,
+      "loss": 0.5606,
+      "step": 1943
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.44925444925445e-06,
+      "loss": 0.4868,
+      "step": 1944
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.425204425204425e-06,
+      "loss": 0.5555,
+      "step": 1945
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.401154401154401e-06,
+      "loss": 0.4905,
+      "step": 1946
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.377104377104377e-06,
+      "loss": 0.5488,
+      "step": 1947
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.353054353054353e-06,
+      "loss": 0.5482,
+      "step": 1948
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.329004329004329e-06,
+      "loss": 0.4635,
+      "step": 1949
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.304954304954305e-06,
+      "loss": 0.4476,
+      "step": 1950
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.280904280904282e-06,
+      "loss": 0.5674,
+      "step": 1951
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.256854256854257e-06,
+      "loss": 0.642,
+      "step": 1952
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.232804232804233e-06,
+      "loss": 0.571,
+      "step": 1953
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.208754208754209e-06,
+      "loss": 0.5452,
+      "step": 1954
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.184704184704185e-06,
+      "loss": 0.527,
+      "step": 1955
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.1606541606541606e-06,
+      "loss": 0.4758,
+      "step": 1956
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.136604136604136e-06,
+      "loss": 0.6521,
+      "step": 1957
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.112554112554113e-06,
+      "loss": 0.4782,
+      "step": 1958
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.088504088504089e-06,
+      "loss": 0.7381,
+      "step": 1959
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.064454064454065e-06,
+      "loss": 0.4931,
+      "step": 1960
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.040404040404041e-06,
+      "loss": 0.4991,
+      "step": 1961
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.016354016354017e-06,
+      "loss": 0.658,
+      "step": 1962
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.9923039923039925e-06,
+      "loss": 0.4418,
+      "step": 1963
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.6207,
+      "step": 1964
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.9442039442039446e-06,
+      "loss": 0.592,
+      "step": 1965
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.92015392015392e-06,
+      "loss": 0.523,
+      "step": 1966
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.896103896103896e-06,
+      "loss": 0.5366,
+      "step": 1967
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.872053872053872e-06,
+      "loss": 0.6349,
+      "step": 1968
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.848003848003849e-06,
+      "loss": 0.5642,
+      "step": 1969
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.823953823953824e-06,
+      "loss": 0.5613,
+      "step": 1970
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7999037999038e-06,
+      "loss": 0.5487,
+      "step": 1971
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7758537758537756e-06,
+      "loss": 0.4929,
+      "step": 1972
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.751803751803752e-06,
+      "loss": 0.6536,
+      "step": 1973
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.727753727753728e-06,
+      "loss": 0.6371,
+      "step": 1974
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.5328,
+      "step": 1975
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.67965367965368e-06,
+      "loss": 0.6117,
+      "step": 1976
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.655603655603656e-06,
+      "loss": 0.577,
+      "step": 1977
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.6315536315536315e-06,
+      "loss": 0.5762,
+      "step": 1978
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.6075036075036075e-06,
+      "loss": 0.5083,
+      "step": 1979
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.583453583453584e-06,
+      "loss": 0.5506,
+      "step": 1980
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5594035594035596e-06,
+      "loss": 0.622,
+      "step": 1981
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5353535353535352e-06,
+      "loss": 0.565,
+      "step": 1982
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5113035113035113e-06,
+      "loss": 0.663,
+      "step": 1983
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4872534872534877e-06,
+      "loss": 0.5141,
+      "step": 1984
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4632034632034634e-06,
+      "loss": 0.5757,
+      "step": 1985
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.439153439153439e-06,
+      "loss": 0.4751,
+      "step": 1986
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4151034151034154e-06,
+      "loss": 0.5925,
+      "step": 1987
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3910533910533915e-06,
+      "loss": 0.504,
+      "step": 1988
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.367003367003367e-06,
+      "loss": 0.4189,
+      "step": 1989
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3429533429533427e-06,
+      "loss": 0.5398,
+      "step": 1990
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.318903318903319e-06,
+      "loss": 0.4812,
+      "step": 1991
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2948532948532953e-06,
+      "loss": 0.5106,
+      "step": 1992
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.270803270803271e-06,
+      "loss": 0.5397,
+      "step": 1993
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2467532467532465e-06,
+      "loss": 0.5217,
+      "step": 1994
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.222703222703223e-06,
+      "loss": 0.5342,
+      "step": 1995
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.198653198653199e-06,
+      "loss": 0.5578,
+      "step": 1996
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.5525,
+      "step": 1997
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.150553150553151e-06,
+      "loss": 0.6375,
+      "step": 1998
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1265031265031267e-06,
+      "loss": 0.5295,
+      "step": 1999
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1024531024531023e-06,
+      "loss": 0.5733,
+      "step": 2000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.078403078403079e-06,
+      "loss": 0.6174,
+      "step": 2001
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.0543530543530544e-06,
+      "loss": 0.5292,
+      "step": 2002
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.5161,
+      "step": 2003
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.006253006253006e-06,
+      "loss": 0.5646,
+      "step": 2004
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9822029822029826e-06,
+      "loss": 0.5115,
+      "step": 2005
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.958152958152958e-06,
+      "loss": 0.5226,
+      "step": 2006
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9341029341029342e-06,
+      "loss": 0.5357,
+      "step": 2007
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.91005291005291e-06,
+      "loss": 0.5078,
+      "step": 2008
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.8860028860028863e-06,
+      "loss": 0.5435,
+      "step": 2009
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.8619528619528624e-06,
+      "loss": 0.5302,
+      "step": 2010
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.837902837902838e-06,
+      "loss": 0.4819,
+      "step": 2011
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.813852813852814e-06,
+      "loss": 0.5196,
+      "step": 2012
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7898027898027897e-06,
+      "loss": 0.4209,
+      "step": 2013
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.765752765752766e-06,
+      "loss": 0.5686,
+      "step": 2014
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7417027417027418e-06,
+      "loss": 0.5357,
+      "step": 2015
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.717652717652718e-06,
+      "loss": 0.5609,
+      "step": 2016
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6936026936026934e-06,
+      "loss": 0.541,
+      "step": 2017
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.66955266955267e-06,
+      "loss": 0.5246,
+      "step": 2018
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6455026455026455e-06,
+      "loss": 0.5684,
+      "step": 2019
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6214526214526216e-06,
+      "loss": 0.5495,
+      "step": 2020
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5974025974025976e-06,
+      "loss": 0.5549,
+      "step": 2021
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5733525733525737e-06,
+      "loss": 0.5249,
+      "step": 2022
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5493025493025497e-06,
+      "loss": 0.4281,
+      "step": 2023
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5252525252525253e-06,
+      "loss": 0.494,
+      "step": 2024
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5012025012025014e-06,
+      "loss": 0.5505,
+      "step": 2025
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.477152477152477e-06,
+      "loss": 0.5031,
+      "step": 2026
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.4531024531024535e-06,
+      "loss": 0.6041,
+      "step": 2027
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.429052429052429e-06,
+      "loss": 0.5288,
+      "step": 2028
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.405002405002405e-06,
+      "loss": 0.6107,
+      "step": 2029
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.4908,
+      "step": 2030
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3569023569023572e-06,
+      "loss": 0.6089,
+      "step": 2031
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.332852332852333e-06,
+      "loss": 0.6813,
+      "step": 2032
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.308802308802309e-06,
+      "loss": 0.5298,
+      "step": 2033
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.284752284752285e-06,
+      "loss": 0.5167,
+      "step": 2034
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.2607022607022606e-06,
+      "loss": 0.5876,
+      "step": 2035
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.236652236652237e-06,
+      "loss": 0.5211,
+      "step": 2036
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.2126022126022127e-06,
+      "loss": 0.5655,
+      "step": 2037
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1885521885521887e-06,
+      "loss": 0.6772,
+      "step": 2038
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1645021645021643e-06,
+      "loss": 0.5497,
+      "step": 2039
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.140452140452141e-06,
+      "loss": 0.522,
+      "step": 2040
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.1164021164021164e-06,
+      "loss": 0.4973,
+      "step": 2041
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0923520923520925e-06,
+      "loss": 0.5367,
+      "step": 2042
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.068302068302068e-06,
+      "loss": 0.5901,
+      "step": 2043
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0442520442520446e-06,
+      "loss": 0.5298,
+      "step": 2044
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0202020202020206e-06,
+      "loss": 0.5091,
+      "step": 2045
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9961519961519962e-06,
+      "loss": 0.595,
+      "step": 2046
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9721019721019723e-06,
+      "loss": 0.5809,
+      "step": 2047
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.948051948051948e-06,
+      "loss": 0.5442,
+      "step": 2048
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9240019240019244e-06,
+      "loss": 0.5194,
+      "step": 2049
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.8999518999519e-06,
+      "loss": 0.5724,
+      "step": 2050
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.875901875901876e-06,
+      "loss": 0.5791,
+      "step": 2051
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.5045,
+      "step": 2052
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.827801827801828e-06,
+      "loss": 0.4662,
+      "step": 2053
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.8037518037518038e-06,
+      "loss": 0.606,
+      "step": 2054
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7797017797017798e-06,
+      "loss": 0.5654,
+      "step": 2055
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7556517556517556e-06,
+      "loss": 0.5191,
+      "step": 2056
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7316017316017317e-06,
+      "loss": 0.6071,
+      "step": 2057
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7075517075517077e-06,
+      "loss": 0.6138,
+      "step": 2058
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6835016835016836e-06,
+      "loss": 0.5339,
+      "step": 2059
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6594516594516596e-06,
+      "loss": 0.5351,
+      "step": 2060
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6354016354016354e-06,
+      "loss": 0.5588,
+      "step": 2061
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.6113516113516115e-06,
+      "loss": 0.6608,
+      "step": 2062
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.5637,
+      "step": 2063
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5632515632515634e-06,
+      "loss": 0.5279,
+      "step": 2064
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5392015392015394e-06,
+      "loss": 0.5457,
+      "step": 2065
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.5733,
+      "step": 2066
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4911014911014913e-06,
+      "loss": 0.583,
+      "step": 2067
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4670514670514671e-06,
+      "loss": 0.4447,
+      "step": 2068
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4430014430014432e-06,
+      "loss": 0.5769,
+      "step": 2069
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.418951418951419e-06,
+      "loss": 0.4559,
+      "step": 2070
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3949013949013948e-06,
+      "loss": 0.5212,
+      "step": 2071
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3708513708513709e-06,
+      "loss": 0.5696,
+      "step": 2072
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3468013468013467e-06,
+      "loss": 0.5156,
+      "step": 2073
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.3227513227513228e-06,
+      "loss": 0.6569,
+      "step": 2074
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2987012987012988e-06,
+      "loss": 0.5507,
+      "step": 2075
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2746512746512749e-06,
+      "loss": 0.6478,
+      "step": 2076
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2506012506012507e-06,
+      "loss": 0.5749,
+      "step": 2077
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2265512265512267e-06,
+      "loss": 0.5515,
+      "step": 2078
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2025012025012026e-06,
+      "loss": 0.5575,
+      "step": 2079
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1784511784511786e-06,
+      "loss": 0.5535,
+      "step": 2080
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1544011544011545e-06,
+      "loss": 0.4983,
+      "step": 2081
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1303511303511303e-06,
+      "loss": 0.6102,
+      "step": 2082
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.1063011063011063e-06,
+      "loss": 0.6102,
+      "step": 2083
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0822510822510822e-06,
+      "loss": 0.6008,
+      "step": 2084
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0582010582010582e-06,
+      "loss": 0.4785,
+      "step": 2085
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.034151034151034e-06,
+      "loss": 0.6137,
+      "step": 2086
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0101010101010103e-06,
+      "loss": 0.5189,
+      "step": 2087
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.860509860509861e-07,
+      "loss": 0.4601,
+      "step": 2088
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.620009620009622e-07,
+      "loss": 0.5339,
+      "step": 2089
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.37950937950938e-07,
+      "loss": 0.5913,
+      "step": 2090
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.13900913900914e-07,
+      "loss": 0.4763,
+      "step": 2091
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.898508898508899e-07,
+      "loss": 0.5596,
+      "step": 2092
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.658008658008658e-07,
+      "loss": 0.5218,
+      "step": 2093
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.417508417508418e-07,
+      "loss": 0.535,
+      "step": 2094
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.177008177008177e-07,
+      "loss": 0.5078,
+      "step": 2095
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.6476,
+      "step": 2096
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.696007696007697e-07,
+      "loss": 0.5788,
+      "step": 2097
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.455507455507456e-07,
+      "loss": 0.5566,
+      "step": 2098
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.215007215007216e-07,
+      "loss": 0.5525,
+      "step": 2099
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.974506974506974e-07,
+      "loss": 0.6142,
+      "step": 2100
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.734006734006734e-07,
+      "loss": 0.4986,
+      "step": 2101
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.493506493506494e-07,
+      "loss": 0.5031,
+      "step": 2102
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.253006253006253e-07,
+      "loss": 0.5728,
+      "step": 2103
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.012506012506013e-07,
+      "loss": 0.6026,
+      "step": 2104
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.772005772005772e-07,
+      "loss": 0.5608,
+      "step": 2105
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.531505531505532e-07,
+      "loss": 0.5934,
+      "step": 2106
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.291005291005291e-07,
+      "loss": 0.5774,
+      "step": 2107
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.050505050505052e-07,
+      "loss": 0.5442,
+      "step": 2108
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.810004810004811e-07,
+      "loss": 0.525,
+      "step": 2109
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.56950456950457e-07,
+      "loss": 0.4985,
+      "step": 2110
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.329004329004329e-07,
+      "loss": 0.5417,
+      "step": 2111
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.0885040885040886e-07,
+      "loss": 0.488,
+      "step": 2112
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8480038480038485e-07,
+      "loss": 0.4861,
+      "step": 2113
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.607503607503608e-07,
+      "loss": 0.5976,
+      "step": 2114
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.367003367003367e-07,
+      "loss": 0.5361,
+      "step": 2115
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.1265031265031267e-07,
+      "loss": 0.5624,
+      "step": 2116
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.886002886002886e-07,
+      "loss": 0.5513,
+      "step": 2117
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.6455026455026455e-07,
+      "loss": 0.4521,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4050024050024055e-07,
+      "loss": 0.5759,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1645021645021646e-07,
+      "loss": 0.5405,
+      "step": 2120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9240019240019243e-07,
+      "loss": 0.6285,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6835016835016834e-07,
+      "loss": 0.6091,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.443001443001443e-07,
+      "loss": 0.5633,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2025012025012027e-07,
+      "loss": 0.609,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.620009620009621e-08,
+      "loss": 0.5454,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.215007215007215e-08,
+      "loss": 0.5319,
+      "step": 2126
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8100048100048107e-08,
+      "loss": 0.5182,
+      "step": 2127
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4050024050024053e-08,
+      "loss": 0.4324,
+      "step": 2128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.4479,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.7365577188500279,
+      "eval_loss": 0.5271788835525513,
+      "eval_runtime": 23.4488,
+      "eval_samples_per_second": 1452.226,
+      "eval_steps_per_second": 22.73,
+      "step": 2129
+    }
+  ],
+  "max_steps": 2129,
+  "num_train_epochs": 1,
+  "total_flos": 1369268670611844.0,
+  "trial_name": null,
+  "trial_params": null
+}