diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,12019 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.596169193934557,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 0.6487,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.6644,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.6141,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.6415,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.6281,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.6542,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.6392,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.6253,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.6373,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.6655,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.6521,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.6611,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.6569,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.6166,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.634,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.6268,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.633,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.5873,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.5754,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 0.6279,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.526315789473685e-06,
+      "loss": 0.6274,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.789473684210527e-06,
+      "loss": 0.6202,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.0526315789473685e-06,
+      "loss": 0.6231,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 0.6377,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.578947368421054e-06,
+      "loss": 0.6046,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.842105263157896e-06,
+      "loss": 0.5962,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.1052631578947375e-06,
+      "loss": 0.6332,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 0.618,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.631578947368423e-06,
+      "loss": 0.6363,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.894736842105265e-06,
+      "loss": 0.6094,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.157894736842106e-06,
+      "loss": 0.5766,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 0.5777,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 0.6267,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.947368421052632e-06,
+      "loss": 0.6147,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.210526315789474e-06,
+      "loss": 0.6288,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 0.6146,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.736842105263159e-06,
+      "loss": 0.5781,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.5902,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0263157894736844e-05,
+      "loss": 0.6087,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.672,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0789473684210528e-05,
+      "loss": 0.616,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.105263157894737e-05,
+      "loss": 0.5914,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.1315789473684212e-05,
+      "loss": 0.5995,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 0.619,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 0.5858,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2105263157894737e-05,
+      "loss": 0.663,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.236842105263158e-05,
+      "loss": 0.6458,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 0.5941,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2894736842105264e-05,
+      "loss": 0.6095,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3157894736842108e-05,
+      "loss": 0.6187,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3421052631578948e-05,
+      "loss": 0.6084,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3684210526315791e-05,
+      "loss": 0.6611,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3947368421052631e-05,
+      "loss": 0.5786,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4210526315789475e-05,
+      "loss": 0.6292,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4473684210526317e-05,
+      "loss": 0.6511,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 0.6452,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.6517,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5263157894736846e-05,
+      "loss": 0.6404,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5526315789473686e-05,
+      "loss": 0.6534,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.578947368421053e-05,
+      "loss": 0.6639,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.605263157894737e-05,
+      "loss": 0.5906,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6315789473684213e-05,
+      "loss": 0.6151,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6578947368421053e-05,
+      "loss": 0.6352,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 0.6549,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.7105263157894737e-05,
+      "loss": 0.6084,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 0.6229,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.763157894736842e-05,
+      "loss": 0.6251,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 0.6091,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8157894736842107e-05,
+      "loss": 0.5874,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.8421052631578947e-05,
+      "loss": 0.609,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.868421052631579e-05,
+      "loss": 0.5909,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 0.6391,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9210526315789474e-05,
+      "loss": 0.6419,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9473684210526318e-05,
+      "loss": 0.628,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 0.6238,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 0.686,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9991769547325103e-05,
+      "loss": 0.662,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.998353909465021e-05,
+      "loss": 0.6295,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.997530864197531e-05,
+      "loss": 0.65,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9967078189300412e-05,
+      "loss": 0.6087,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9958847736625514e-05,
+      "loss": 0.7357,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.995061728395062e-05,
+      "loss": 0.6881,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.994238683127572e-05,
+      "loss": 0.6753,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9934156378600823e-05,
+      "loss": 0.6538,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9925925925925928e-05,
+      "loss": 0.6558,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.991769547325103e-05,
+      "loss": 0.6554,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9909465020576132e-05,
+      "loss": 0.6577,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9901234567901237e-05,
+      "loss": 0.6316,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.989300411522634e-05,
+      "loss": 0.6142,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.988477366255144e-05,
+      "loss": 0.6853,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9876543209876546e-05,
+      "loss": 0.6294,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9868312757201648e-05,
+      "loss": 0.6505,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.986008230452675e-05,
+      "loss": 0.6358,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9851851851851855e-05,
+      "loss": 0.6874,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9843621399176956e-05,
+      "loss": 0.6253,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.983539094650206e-05,
+      "loss": 0.6699,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9827160493827163e-05,
+      "loss": 0.6514,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9818930041152265e-05,
+      "loss": 0.6626,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.981069958847737e-05,
+      "loss": 0.631,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9802469135802472e-05,
+      "loss": 0.6432,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9794238683127574e-05,
+      "loss": 0.6137,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9786008230452676e-05,
+      "loss": 0.625,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.977777777777778e-05,
+      "loss": 0.6344,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9769547325102883e-05,
+      "loss": 0.6638,
+      "step": 104
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9761316872427985e-05,
+      "loss": 0.6438,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.6541,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9744855967078192e-05,
+      "loss": 0.6406,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9736625514403294e-05,
+      "loss": 0.6225,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9728395061728395e-05,
+      "loss": 0.6389,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.97201646090535e-05,
+      "loss": 0.6679,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9711934156378602e-05,
+      "loss": 0.6647,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9703703703703704e-05,
+      "loss": 0.6491,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9695473251028806e-05,
+      "loss": 0.6349,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.968724279835391e-05,
+      "loss": 0.6354,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9679012345679013e-05,
+      "loss": 0.6326,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9670781893004115e-05,
+      "loss": 0.6437,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.966255144032922e-05,
+      "loss": 0.6462,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9654320987654322e-05,
+      "loss": 0.6322,
+      "step": 118
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9646090534979424e-05,
+      "loss": 0.6674,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.963786008230453e-05,
+      "loss": 0.6936,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 0.638,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9621399176954733e-05,
+      "loss": 0.6224,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9613168724279838e-05,
+      "loss": 0.6202,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.960493827160494e-05,
+      "loss": 0.6686,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.959670781893004e-05,
+      "loss": 0.6311,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9588477366255147e-05,
+      "loss": 0.604,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.958024691358025e-05,
+      "loss": 0.6507,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9572016460905354e-05,
+      "loss": 0.6805,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9563786008230455e-05,
+      "loss": 0.67,
+      "step": 129
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9555555555555557e-05,
+      "loss": 0.6694,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9547325102880663e-05,
+      "loss": 0.6603,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9539094650205764e-05,
+      "loss": 0.6699,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9530864197530866e-05,
+      "loss": 0.7279,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9522633744855968e-05,
+      "loss": 0.6208,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9514403292181073e-05,
+      "loss": 0.6499,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9506172839506175e-05,
+      "loss": 0.6685,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9497942386831277e-05,
+      "loss": 0.6557,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9489711934156382e-05,
+      "loss": 0.6737,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9481481481481484e-05,
+      "loss": 0.6111,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9473251028806586e-05,
+      "loss": 0.6348,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9465020576131687e-05,
+      "loss": 0.6718,
+      "step": 141
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9456790123456793e-05,
+      "loss": 0.6908,
+      "step": 142
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9448559670781894e-05,
+      "loss": 0.6281,
+      "step": 143
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9440329218106996e-05,
+      "loss": 0.6331,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.94320987654321e-05,
+      "loss": 0.5998,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9423868312757203e-05,
+      "loss": 0.6359,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9415637860082305e-05,
+      "loss": 0.6755,
+      "step": 147
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9407407407407407e-05,
+      "loss": 0.6374,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9399176954732512e-05,
+      "loss": 0.6638,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9390946502057614e-05,
+      "loss": 0.6071,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9382716049382716e-05,
+      "loss": 0.6809,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.937448559670782e-05,
+      "loss": 0.6175,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9366255144032923e-05,
+      "loss": 0.6628,
+      "step": 153
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9358024691358025e-05,
+      "loss": 0.6567,
+      "step": 154
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.934979423868313e-05,
+      "loss": 0.7288,
+      "step": 155
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.934156378600823e-05,
+      "loss": 0.6834,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 0.6118,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.932510288065844e-05,
+      "loss": 0.616,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.931687242798354e-05,
+      "loss": 0.6659,
+      "step": 159
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9308641975308646e-05,
+      "loss": 0.634,
+      "step": 160
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9300411522633747e-05,
+      "loss": 0.613,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.929218106995885e-05,
+      "loss": 0.6069,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9283950617283955e-05,
+      "loss": 0.7148,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9275720164609056e-05,
+      "loss": 0.6548,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9267489711934158e-05,
+      "loss": 0.6534,
+      "step": 165
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.6598,
+      "step": 166
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9251028806584365e-05,
+      "loss": 0.6485,
+      "step": 167
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9242798353909467e-05,
+      "loss": 0.6614,
+      "step": 168
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.923456790123457e-05,
+      "loss": 0.5956,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9226337448559674e-05,
+      "loss": 0.6292,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9218106995884776e-05,
+      "loss": 0.6557,
+      "step": 171
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9209876543209878e-05,
+      "loss": 0.6378,
+      "step": 172
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.920164609053498e-05,
+      "loss": 0.6515,
+      "step": 173
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9193415637860085e-05,
+      "loss": 0.6401,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9185185185185186e-05,
+      "loss": 0.6388,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9176954732510288e-05,
+      "loss": 0.6597,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9168724279835393e-05,
+      "loss": 0.6912,
+      "step": 177
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9160493827160495e-05,
+      "loss": 0.601,
+      "step": 178
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9152263374485597e-05,
+      "loss": 0.5757,
+      "step": 179
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.91440329218107e-05,
+      "loss": 0.6278,
+      "step": 180
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9135802469135804e-05,
+      "loss": 0.6985,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9127572016460906e-05,
+      "loss": 0.7056,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9119341563786008e-05,
+      "loss": 0.5969,
+      "step": 183
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9111111111111113e-05,
+      "loss": 0.624,
+      "step": 184
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9102880658436215e-05,
+      "loss": 0.7324,
+      "step": 185
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9094650205761317e-05,
+      "loss": 0.7273,
+      "step": 186
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9086419753086422e-05,
+      "loss": 0.6393,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9078189300411524e-05,
+      "loss": 0.669,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9069958847736625e-05,
+      "loss": 0.6635,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.906172839506173e-05,
+      "loss": 0.6529,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9053497942386832e-05,
+      "loss": 0.6548,
+      "step": 191
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9045267489711938e-05,
+      "loss": 0.6632,
+      "step": 192
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.903703703703704e-05,
+      "loss": 0.6497,
+      "step": 193
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.902880658436214e-05,
+      "loss": 0.6333,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9020576131687247e-05,
+      "loss": 0.6195,
+      "step": 195
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.901234567901235e-05,
+      "loss": 0.6771,
+      "step": 196
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.900411522633745e-05,
+      "loss": 0.6737,
+      "step": 197
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8995884773662555e-05,
+      "loss": 0.6388,
+      "step": 198
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8987654320987657e-05,
+      "loss": 0.6086,
+      "step": 199
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.897942386831276e-05,
+      "loss": 0.6954,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.897119341563786e-05,
+      "loss": 0.65,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8962962962962966e-05,
+      "loss": 0.6544,
+      "step": 202
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8954732510288068e-05,
+      "loss": 0.6824,
+      "step": 203
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.894650205761317e-05,
+      "loss": 0.6549,
+      "step": 204
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8938271604938275e-05,
+      "loss": 0.7288,
+      "step": 205
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.8930041152263377e-05,
+      "loss": 0.6738,
+      "step": 206
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.892181069958848e-05,
+      "loss": 0.7009,
+      "step": 207
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.891358024691358e-05,
+      "loss": 0.6515,
+      "step": 208
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8905349794238685e-05,
+      "loss": 0.6545,
+      "step": 209
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8897119341563787e-05,
+      "loss": 0.6139,
+      "step": 210
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 0.6256,
+      "step": 211
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8880658436213994e-05,
+      "loss": 0.6612,
+      "step": 212
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8872427983539096e-05,
+      "loss": 0.6823,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8864197530864198e-05,
+      "loss": 0.701,
+      "step": 214
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.88559670781893e-05,
+      "loss": 0.6589,
+      "step": 215
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8847736625514405e-05,
+      "loss": 0.6998,
+      "step": 216
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8839506172839507e-05,
+      "loss": 0.6361,
+      "step": 217
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.883127572016461e-05,
+      "loss": 0.6372,
+      "step": 218
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8823045267489714e-05,
+      "loss": 0.6627,
+      "step": 219
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8814814814814816e-05,
+      "loss": 0.6597,
+      "step": 220
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8806584362139917e-05,
+      "loss": 0.6513,
+      "step": 221
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8798353909465023e-05,
+      "loss": 0.6184,
+      "step": 222
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8790123456790124e-05,
+      "loss": 0.6736,
+      "step": 223
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.878189300411523e-05,
+      "loss": 0.6227,
+      "step": 224
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.877366255144033e-05,
+      "loss": 0.7008,
+      "step": 225
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.629,
+      "step": 226
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.875720164609054e-05,
+      "loss": 0.6751,
+      "step": 227
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.874897119341564e-05,
+      "loss": 0.6407,
+      "step": 228
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8740740740740742e-05,
+      "loss": 0.6166,
+      "step": 229
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8732510288065847e-05,
+      "loss": 0.6539,
+      "step": 230
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.872427983539095e-05,
+      "loss": 0.6517,
+      "step": 231
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.871604938271605e-05,
+      "loss": 0.6558,
+      "step": 232
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8707818930041153e-05,
+      "loss": 0.6422,
+      "step": 233
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8699588477366258e-05,
+      "loss": 0.6737,
+      "step": 234
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.869135802469136e-05,
+      "loss": 0.6326,
+      "step": 235
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.868312757201646e-05,
+      "loss": 0.7108,
+      "step": 236
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8674897119341567e-05,
+      "loss": 0.6744,
+      "step": 237
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 0.6607,
+      "step": 238
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.865843621399177e-05,
+      "loss": 0.6304,
+      "step": 239
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8650205761316872e-05,
+      "loss": 0.6559,
+      "step": 240
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8641975308641977e-05,
+      "loss": 0.6008,
+      "step": 241
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.863374485596708e-05,
+      "loss": 0.6412,
+      "step": 242
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.862551440329218e-05,
+      "loss": 0.6215,
+      "step": 243
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8617283950617286e-05,
+      "loss": 0.6417,
+      "step": 244
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8609053497942388e-05,
+      "loss": 0.6501,
+      "step": 245
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.860082304526749e-05,
+      "loss": 0.6536,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8592592592592592e-05,
+      "loss": 0.663,
+      "step": 247
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8584362139917697e-05,
+      "loss": 0.6357,
+      "step": 248
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.85761316872428e-05,
+      "loss": 0.602,
+      "step": 249
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.85679012345679e-05,
+      "loss": 0.6002,
+      "step": 250
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8559670781893006e-05,
+      "loss": 0.5942,
+      "step": 251
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8551440329218108e-05,
+      "loss": 0.6082,
+      "step": 252
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8543209876543213e-05,
+      "loss": 0.6408,
+      "step": 253
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8534979423868315e-05,
+      "loss": 0.6777,
+      "step": 254
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8526748971193416e-05,
+      "loss": 0.6613,
+      "step": 255
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.851851851851852e-05,
+      "loss": 0.6704,
+      "step": 256
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8510288065843623e-05,
+      "loss": 0.6037,
+      "step": 257
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8502057613168725e-05,
+      "loss": 0.6532,
+      "step": 258
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.849382716049383e-05,
+      "loss": 0.6123,
+      "step": 259
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8485596707818932e-05,
+      "loss": 0.6658,
+      "step": 260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8477366255144034e-05,
+      "loss": 0.6753,
+      "step": 261
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.846913580246914e-05,
+      "loss": 0.6448,
+      "step": 262
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.846090534979424e-05,
+      "loss": 0.5887,
+      "step": 263
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8452674897119343e-05,
+      "loss": 0.6264,
+      "step": 264
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8444444444444448e-05,
+      "loss": 0.6656,
+      "step": 265
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.843621399176955e-05,
+      "loss": 0.6447,
+      "step": 266
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8427983539094652e-05,
+      "loss": 0.5978,
+      "step": 267
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8419753086419754e-05,
+      "loss": 0.6323,
+      "step": 268
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.841152263374486e-05,
+      "loss": 0.6535,
+      "step": 269
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.840329218106996e-05,
+      "loss": 0.6056,
+      "step": 270
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8395061728395062e-05,
+      "loss": 0.6339,
+      "step": 271
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8386831275720164e-05,
+      "loss": 0.6336,
+      "step": 272
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.837860082304527e-05,
+      "loss": 0.645,
+      "step": 273
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.837037037037037e-05,
+      "loss": 0.6561,
+      "step": 274
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8362139917695473e-05,
+      "loss": 0.6263,
+      "step": 275
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.835390946502058e-05,
+      "loss": 0.6537,
+      "step": 276
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.834567901234568e-05,
+      "loss": 0.5949,
+      "step": 277
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8337448559670782e-05,
+      "loss": 0.6378,
+      "step": 278
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8329218106995884e-05,
+      "loss": 0.658,
+      "step": 279
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.832098765432099e-05,
+      "loss": 0.6538,
+      "step": 280
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.831275720164609e-05,
+      "loss": 0.6454,
+      "step": 281
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8304526748971193e-05,
+      "loss": 0.6609,
+      "step": 282
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8296296296296298e-05,
+      "loss": 0.669,
+      "step": 283
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.82880658436214e-05,
+      "loss": 0.6615,
+      "step": 284
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8279835390946505e-05,
+      "loss": 0.6236,
+      "step": 285
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8271604938271607e-05,
+      "loss": 0.672,
+      "step": 286
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.826337448559671e-05,
+      "loss": 0.6304,
+      "step": 287
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8255144032921814e-05,
+      "loss": 0.6671,
+      "step": 288
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8246913580246915e-05,
+      "loss": 0.6538,
+      "step": 289
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8238683127572017e-05,
+      "loss": 0.6119,
+      "step": 290
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8230452674897122e-05,
+      "loss": 0.6066,
+      "step": 291
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8222222222222224e-05,
+      "loss": 0.6064,
+      "step": 292
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8213991769547326e-05,
+      "loss": 0.6018,
+      "step": 293
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.820576131687243e-05,
+      "loss": 0.6307,
+      "step": 294
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8197530864197533e-05,
+      "loss": 0.6473,
+      "step": 295
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8189300411522635e-05,
+      "loss": 0.6149,
+      "step": 296
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.818106995884774e-05,
+      "loss": 0.6399,
+      "step": 297
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8172839506172842e-05,
+      "loss": 0.6297,
+      "step": 298
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8164609053497944e-05,
+      "loss": 0.7015,
+      "step": 299
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8156378600823046e-05,
+      "loss": 0.6196,
+      "step": 300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 0.6463,
+      "step": 301
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8139917695473253e-05,
+      "loss": 0.6232,
+      "step": 302
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8131687242798354e-05,
+      "loss": 0.6374,
+      "step": 303
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.812345679012346e-05,
+      "loss": 0.6684,
+      "step": 304
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.811522633744856e-05,
+      "loss": 0.6609,
+      "step": 305
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8106995884773663e-05,
+      "loss": 0.6217,
+      "step": 306
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8098765432098765e-05,
+      "loss": 0.67,
+      "step": 307
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.809053497942387e-05,
+      "loss": 0.63,
+      "step": 308
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8082304526748972e-05,
+      "loss": 0.5896,
+      "step": 309
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8074074074074074e-05,
+      "loss": 0.6012,
+      "step": 310
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.806584362139918e-05,
+      "loss": 0.632,
+      "step": 311
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.805761316872428e-05,
+      "loss": 0.6119,
+      "step": 312
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8049382716049383e-05,
+      "loss": 0.6451,
+      "step": 313
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8041152263374485e-05,
+      "loss": 0.5984,
+      "step": 314
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.803292181069959e-05,
+      "loss": 0.6617,
+      "step": 315
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 0.6145,
+      "step": 316
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8016460905349797e-05,
+      "loss": 0.6188,
+      "step": 317
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.80082304526749e-05,
+      "loss": 0.5804,
+      "step": 318
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6225,
+      "step": 319
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7991769547325106e-05,
+      "loss": 0.6267,
+      "step": 320
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7983539094650207e-05,
+      "loss": 0.6459,
+      "step": 321
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.797530864197531e-05,
+      "loss": 0.6723,
+      "step": 322
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7967078189300414e-05,
+      "loss": 0.6508,
+      "step": 323
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7958847736625516e-05,
+      "loss": 0.6537,
+      "step": 324
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7950617283950618e-05,
+      "loss": 0.6519,
+      "step": 325
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7942386831275723e-05,
+      "loss": 0.6578,
+      "step": 326
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7934156378600825e-05,
+      "loss": 0.6379,
+      "step": 327
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7925925925925927e-05,
+      "loss": 0.6127,
+      "step": 328
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7917695473251032e-05,
+      "loss": 0.6139,
+      "step": 329
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7909465020576134e-05,
+      "loss": 0.6128,
+      "step": 330
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7901234567901236e-05,
+      "loss": 0.5898,
+      "step": 331
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7893004115226338e-05,
+      "loss": 0.6081,
+      "step": 332
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7884773662551443e-05,
+      "loss": 0.6359,
+      "step": 333
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7876543209876545e-05,
+      "loss": 0.6268,
+      "step": 334
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7868312757201646e-05,
+      "loss": 0.6058,
+      "step": 335
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.786008230452675e-05,
+      "loss": 0.5287,
+      "step": 336
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7851851851851853e-05,
+      "loss": 0.6083,
+      "step": 337
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7843621399176955e-05,
+      "loss": 0.6445,
+      "step": 338
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7835390946502057e-05,
+      "loss": 0.6333,
+      "step": 339
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7827160493827162e-05,
+      "loss": 0.6612,
+      "step": 340
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7818930041152264e-05,
+      "loss": 0.6143,
+      "step": 341
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7810699588477366e-05,
+      "loss": 0.5507,
+      "step": 342
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.780246913580247e-05,
+      "loss": 0.6121,
+      "step": 343
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7794238683127573e-05,
+      "loss": 0.6195,
+      "step": 344
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7786008230452675e-05,
+      "loss": 0.6057,
+      "step": 345
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.6632,
+      "step": 346
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7769547325102882e-05,
+      "loss": 0.5683,
+      "step": 347
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7761316872427984e-05,
+      "loss": 0.5928,
+      "step": 348
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.775308641975309e-05,
+      "loss": 0.6613,
+      "step": 349
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.774485596707819e-05,
+      "loss": 0.6542,
+      "step": 350
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7736625514403292e-05,
+      "loss": 0.6588,
+      "step": 351
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7728395061728398e-05,
+      "loss": 0.5986,
+      "step": 352
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.77201646090535e-05,
+      "loss": 0.6309,
+      "step": 353
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.77119341563786e-05,
+      "loss": 0.6338,
+      "step": 354
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7703703703703706e-05,
+      "loss": 0.6144,
+      "step": 355
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.769547325102881e-05,
+      "loss": 0.6216,
+      "step": 356
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7687242798353914e-05,
+      "loss": 0.6183,
+      "step": 357
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7679012345679015e-05,
+      "loss": 0.6082,
+      "step": 358
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7670781893004117e-05,
+      "loss": 0.605,
+      "step": 359
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.766255144032922e-05,
+      "loss": 0.589,
+      "step": 360
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7654320987654324e-05,
+      "loss": 0.6644,
+      "step": 361
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7646090534979426e-05,
+      "loss": 0.6635,
+      "step": 362
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7637860082304528e-05,
+      "loss": 0.6392,
+      "step": 363
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7629629629629633e-05,
+      "loss": 0.5786,
+      "step": 364
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7621399176954735e-05,
+      "loss": 0.6126,
+      "step": 365
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7613168724279837e-05,
+      "loss": 0.5781,
+      "step": 366
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.760493827160494e-05,
+      "loss": 0.6108,
+      "step": 367
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7596707818930044e-05,
+      "loss": 0.5744,
+      "step": 368
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7588477366255145e-05,
+      "loss": 0.6323,
+      "step": 369
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7580246913580247e-05,
+      "loss": 0.6288,
+      "step": 370
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7572016460905352e-05,
+      "loss": 0.6054,
+      "step": 371
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7563786008230454e-05,
+      "loss": 0.6725,
+      "step": 372
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7555555555555556e-05,
+      "loss": 0.6441,
+      "step": 373
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7547325102880658e-05,
+      "loss": 0.5586,
+      "step": 374
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7539094650205763e-05,
+      "loss": 0.6179,
+      "step": 375
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7530864197530865e-05,
+      "loss": 0.6255,
+      "step": 376
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7522633744855967e-05,
+      "loss": 0.6096,
+      "step": 377
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7514403292181072e-05,
+      "loss": 0.6303,
+      "step": 378
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7506172839506174e-05,
+      "loss": 0.5581,
+      "step": 379
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7497942386831276e-05,
+      "loss": 0.6254,
+      "step": 380
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.748971193415638e-05,
+      "loss": 0.5763,
+      "step": 381
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7481481481481483e-05,
+      "loss": 0.6404,
+      "step": 382
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7473251028806584e-05,
+      "loss": 0.6309,
+      "step": 383
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.746502057613169e-05,
+      "loss": 0.5892,
+      "step": 384
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.745679012345679e-05,
+      "loss": 0.603,
+      "step": 385
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7448559670781893e-05,
+      "loss": 0.626,
+      "step": 386
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7440329218107e-05,
+      "loss": 0.6047,
+      "step": 387
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.74320987654321e-05,
+      "loss": 0.6067,
+      "step": 388
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7423868312757206e-05,
+      "loss": 0.6574,
+      "step": 389
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7415637860082307e-05,
+      "loss": 0.6354,
+      "step": 390
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 0.6345,
+      "step": 391
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.739917695473251e-05,
+      "loss": 0.6425,
+      "step": 392
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7390946502057616e-05,
+      "loss": 0.6034,
+      "step": 393
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7382716049382718e-05,
+      "loss": 0.5942,
+      "step": 394
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.737448559670782e-05,
+      "loss": 0.5661,
+      "step": 395
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7366255144032925e-05,
+      "loss": 0.6314,
+      "step": 396
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7358024691358027e-05,
+      "loss": 0.5899,
+      "step": 397
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.734979423868313e-05,
+      "loss": 0.6445,
+      "step": 398
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.734156378600823e-05,
+      "loss": 0.6052,
+      "step": 399
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 0.5993,
+      "step": 400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7325102880658437e-05,
+      "loss": 0.6319,
+      "step": 401
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.731687242798354e-05,
+      "loss": 0.6038,
+      "step": 402
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7308641975308644e-05,
+      "loss": 0.6356,
+      "step": 403
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7300411522633746e-05,
+      "loss": 0.6309,
+      "step": 404
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7292181069958848e-05,
+      "loss": 0.6172,
+      "step": 405
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.6102,
+      "step": 406
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7275720164609055e-05,
+      "loss": 0.606,
+      "step": 407
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7267489711934157e-05,
+      "loss": 0.5993,
+      "step": 408
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.725925925925926e-05,
+      "loss": 0.6291,
+      "step": 409
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7251028806584364e-05,
+      "loss": 0.6131,
+      "step": 410
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7242798353909466e-05,
+      "loss": 0.6307,
+      "step": 411
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7234567901234568e-05,
+      "loss": 0.6069,
+      "step": 412
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7226337448559673e-05,
+      "loss": 0.6129,
+      "step": 413
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7218106995884775e-05,
+      "loss": 0.5999,
+      "step": 414
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7209876543209876e-05,
+      "loss": 0.5973,
+      "step": 415
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.720164609053498e-05,
+      "loss": 0.6087,
+      "step": 416
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7193415637860083e-05,
+      "loss": 0.5843,
+      "step": 417
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7185185185185185e-05,
+      "loss": 0.662,
+      "step": 418
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.717695473251029e-05,
+      "loss": 0.5923,
+      "step": 419
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7168724279835392e-05,
+      "loss": 0.6279,
+      "step": 420
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7160493827160498e-05,
+      "loss": 0.6212,
+      "step": 421
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.71522633744856e-05,
+      "loss": 0.6264,
+      "step": 422
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.71440329218107e-05,
+      "loss": 0.6419,
+      "step": 423
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7135802469135806e-05,
+      "loss": 0.5774,
+      "step": 424
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7127572016460908e-05,
+      "loss": 0.614,
+      "step": 425
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.711934156378601e-05,
+      "loss": 0.6287,
+      "step": 426
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7111111111111112e-05,
+      "loss": 0.6377,
+      "step": 427
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7102880658436217e-05,
+      "loss": 0.6523,
+      "step": 428
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.709465020576132e-05,
+      "loss": 0.6033,
+      "step": 429
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.708641975308642e-05,
+      "loss": 0.6365,
+      "step": 430
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7078189300411526e-05,
+      "loss": 0.6646,
+      "step": 431
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.7069958847736628e-05,
+      "loss": 0.645,
+      "step": 432
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.706172839506173e-05,
+      "loss": 0.6217,
+      "step": 433
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.705349794238683e-05,
+      "loss": 0.6383,
+      "step": 434
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7045267489711936e-05,
+      "loss": 0.5929,
+      "step": 435
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.6395,
+      "step": 436
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.702880658436214e-05,
+      "loss": 0.579,
+      "step": 437
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7020576131687242e-05,
+      "loss": 0.6132,
+      "step": 438
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.7012345679012347e-05,
+      "loss": 0.6177,
+      "step": 439
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.700411522633745e-05,
+      "loss": 0.5766,
+      "step": 440
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.699588477366255e-05,
+      "loss": 0.6222,
+      "step": 441
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6987654320987656e-05,
+      "loss": 0.6346,
+      "step": 442
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6979423868312758e-05,
+      "loss": 0.5914,
+      "step": 443
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.697119341563786e-05,
+      "loss": 0.6204,
+      "step": 444
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6962962962962965e-05,
+      "loss": 0.6244,
+      "step": 445
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6954732510288067e-05,
+      "loss": 0.5821,
+      "step": 446
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.694650205761317e-05,
+      "loss": 0.5912,
+      "step": 447
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6938271604938274e-05,
+      "loss": 0.6165,
+      "step": 448
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6930041152263375e-05,
+      "loss": 0.5832,
+      "step": 449
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.692181069958848e-05,
+      "loss": 0.6354,
+      "step": 450
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6913580246913582e-05,
+      "loss": 0.6096,
+      "step": 451
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6905349794238684e-05,
+      "loss": 0.6405,
+      "step": 452
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.689711934156379e-05,
+      "loss": 0.5501,
+      "step": 453
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.688888888888889e-05,
+      "loss": 0.6346,
+      "step": 454
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6880658436213993e-05,
+      "loss": 0.6051,
+      "step": 455
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.68724279835391e-05,
+      "loss": 0.5807,
+      "step": 456
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.68641975308642e-05,
+      "loss": 0.5896,
+      "step": 457
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6855967078189302e-05,
+      "loss": 0.5758,
+      "step": 458
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6847736625514404e-05,
+      "loss": 0.5819,
+      "step": 459
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.683950617283951e-05,
+      "loss": 0.6367,
+      "step": 460
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.683127572016461e-05,
+      "loss": 0.638,
+      "step": 461
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6823045267489713e-05,
+      "loss": 0.634,
+      "step": 462
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6814814814814818e-05,
+      "loss": 0.6342,
+      "step": 463
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.680658436213992e-05,
+      "loss": 0.6491,
+      "step": 464
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.679835390946502e-05,
+      "loss": 0.5746,
+      "step": 465
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.5957,
+      "step": 466
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.678189300411523e-05,
+      "loss": 0.5942,
+      "step": 467
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.677366255144033e-05,
+      "loss": 0.5826,
+      "step": 468
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6765432098765432e-05,
+      "loss": 0.616,
+      "step": 469
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6757201646090537e-05,
+      "loss": 0.6195,
+      "step": 470
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.674897119341564e-05,
+      "loss": 0.6843,
+      "step": 471
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.674074074074074e-05,
+      "loss": 0.5883,
+      "step": 472
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6732510288065843e-05,
+      "loss": 0.6445,
+      "step": 473
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6724279835390948e-05,
+      "loss": 0.5817,
+      "step": 474
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.671604938271605e-05,
+      "loss": 0.576,
+      "step": 475
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.670781893004115e-05,
+      "loss": 0.5945,
+      "step": 476
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6699588477366257e-05,
+      "loss": 0.6095,
+      "step": 477
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.669135802469136e-05,
+      "loss": 0.6554,
+      "step": 478
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.668312757201646e-05,
+      "loss": 0.6044,
+      "step": 479
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6674897119341566e-05,
+      "loss": 0.615,
+      "step": 480
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.5805,
+      "step": 481
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.6658436213991773e-05,
+      "loss": 0.6138,
+      "step": 482
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6650205761316874e-05,
+      "loss": 0.5903,
+      "step": 483
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6641975308641976e-05,
+      "loss": 0.6034,
+      "step": 484
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.663374485596708e-05,
+      "loss": 0.6404,
+      "step": 485
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6625514403292183e-05,
+      "loss": 0.5957,
+      "step": 486
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6617283950617285e-05,
+      "loss": 0.6194,
+      "step": 487
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.660905349794239e-05,
+      "loss": 0.6032,
+      "step": 488
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6600823045267492e-05,
+      "loss": 0.5826,
+      "step": 489
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6592592592592594e-05,
+      "loss": 0.61,
+      "step": 490
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6584362139917696e-05,
+      "loss": 0.5975,
+      "step": 491
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.65761316872428e-05,
+      "loss": 0.5779,
+      "step": 492
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6567901234567903e-05,
+      "loss": 0.619,
+      "step": 493
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.6559670781893005e-05,
+      "loss": 0.5639,
+      "step": 494
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.655144032921811e-05,
+      "loss": 0.5855,
+      "step": 495
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.654320987654321e-05,
+      "loss": 0.6027,
+      "step": 496
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6534979423868313e-05,
+      "loss": 0.5665,
+      "step": 497
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6526748971193415e-05,
+      "loss": 0.598,
+      "step": 498
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.651851851851852e-05,
+      "loss": 0.626,
+      "step": 499
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6510288065843622e-05,
+      "loss": 0.5955,
+      "step": 500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6502057613168724e-05,
+      "loss": 0.6545,
+      "step": 501
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.649382716049383e-05,
+      "loss": 0.5632,
+      "step": 502
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.648559670781893e-05,
+      "loss": 0.6338,
+      "step": 503
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6477366255144033e-05,
+      "loss": 0.6542,
+      "step": 504
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6469135802469135e-05,
+      "loss": 0.648,
+      "step": 505
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.646090534979424e-05,
+      "loss": 0.5791,
+      "step": 506
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.6452674897119342e-05,
+      "loss": 0.569,
+      "step": 507
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6444444444444444e-05,
+      "loss": 0.6353,
+      "step": 508
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.643621399176955e-05,
+      "loss": 0.6266,
+      "step": 509
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.642798353909465e-05,
+      "loss": 0.5858,
+      "step": 510
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6419753086419752e-05,
+      "loss": 0.5775,
+      "step": 511
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6411522633744858e-05,
+      "loss": 0.6436,
+      "step": 512
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.640329218106996e-05,
+      "loss": 0.5894,
+      "step": 513
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6395061728395065e-05,
+      "loss": 0.6015,
+      "step": 514
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6386831275720166e-05,
+      "loss": 0.59,
+      "step": 515
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6378600823045268e-05,
+      "loss": 0.6318,
+      "step": 516
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6370370370370374e-05,
+      "loss": 0.5522,
+      "step": 517
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6362139917695475e-05,
+      "loss": 0.6141,
+      "step": 518
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.6353909465020577e-05,
+      "loss": 0.6704,
+      "step": 519
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6345679012345682e-05,
+      "loss": 0.5917,
+      "step": 520
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6337448559670784e-05,
+      "loss": 0.6677,
+      "step": 521
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6329218106995886e-05,
+      "loss": 0.5501,
+      "step": 522
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.632098765432099e-05,
+      "loss": 0.5735,
+      "step": 523
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6312757201646093e-05,
+      "loss": 0.6038,
+      "step": 524
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6304526748971195e-05,
+      "loss": 0.5849,
+      "step": 525
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.5568,
+      "step": 526
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6288065843621402e-05,
+      "loss": 0.6212,
+      "step": 527
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6279835390946504e-05,
+      "loss": 0.5629,
+      "step": 528
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6271604938271605e-05,
+      "loss": 0.5877,
+      "step": 529
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.626337448559671e-05,
+      "loss": 0.6045,
+      "step": 530
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6255144032921812e-05,
+      "loss": 0.5969,
+      "step": 531
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.6246913580246914e-05,
+      "loss": 0.6483,
+      "step": 532
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6238683127572016e-05,
+      "loss": 0.6925,
+      "step": 533
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.623045267489712e-05,
+      "loss": 0.5737,
+      "step": 534
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6222222222222223e-05,
+      "loss": 0.5804,
+      "step": 535
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6213991769547325e-05,
+      "loss": 0.6547,
+      "step": 536
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.620576131687243e-05,
+      "loss": 0.5347,
+      "step": 537
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6197530864197532e-05,
+      "loss": 0.545,
+      "step": 538
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6189300411522634e-05,
+      "loss": 0.5384,
+      "step": 539
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6181069958847736e-05,
+      "loss": 0.6022,
+      "step": 540
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.617283950617284e-05,
+      "loss": 0.5841,
+      "step": 541
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6164609053497943e-05,
+      "loss": 0.6015,
+      "step": 542
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.6156378600823044e-05,
+      "loss": 0.6222,
+      "step": 543
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.614814814814815e-05,
+      "loss": 0.5661,
+      "step": 544
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.613991769547325e-05,
+      "loss": 0.5673,
+      "step": 545
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6131687242798357e-05,
+      "loss": 0.6028,
+      "step": 546
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.612345679012346e-05,
+      "loss": 0.5987,
+      "step": 547
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.611522633744856e-05,
+      "loss": 0.5724,
+      "step": 548
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6106995884773666e-05,
+      "loss": 0.6,
+      "step": 549
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6098765432098767e-05,
+      "loss": 0.5793,
+      "step": 550
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.609053497942387e-05,
+      "loss": 0.596,
+      "step": 551
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6082304526748974e-05,
+      "loss": 0.6053,
+      "step": 552
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6074074074074076e-05,
+      "loss": 0.6085,
+      "step": 553
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6065843621399178e-05,
+      "loss": 0.6197,
+      "step": 554
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6057613168724283e-05,
+      "loss": 0.5463,
+      "step": 555
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6049382716049385e-05,
+      "loss": 0.6217,
+      "step": 556
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.6041152263374487e-05,
+      "loss": 0.6048,
+      "step": 557
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.603292181069959e-05,
+      "loss": 0.5932,
+      "step": 558
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6024691358024694e-05,
+      "loss": 0.6031,
+      "step": 559
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6016460905349796e-05,
+      "loss": 0.5902,
+      "step": 560
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6008230452674897e-05,
+      "loss": 0.5749,
+      "step": 561
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6424,
+      "step": 562
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5991769547325104e-05,
+      "loss": 0.5602,
+      "step": 563
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5983539094650206e-05,
+      "loss": 0.5757,
+      "step": 564
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5975308641975308e-05,
+      "loss": 0.6082,
+      "step": 565
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5967078189300413e-05,
+      "loss": 0.648,
+      "step": 566
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5958847736625515e-05,
+      "loss": 0.6259,
+      "step": 567
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5950617283950617e-05,
+      "loss": 0.5971,
+      "step": 568
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5942386831275722e-05,
+      "loss": 0.5763,
+      "step": 569
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5934156378600824e-05,
+      "loss": 0.6119,
+      "step": 570
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 0.6069,
+      "step": 571
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5917695473251028e-05,
+      "loss": 0.5504,
+      "step": 572
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5909465020576133e-05,
+      "loss": 0.5809,
+      "step": 573
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5901234567901235e-05,
+      "loss": 0.5935,
+      "step": 574
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5893004115226336e-05,
+      "loss": 0.5749,
+      "step": 575
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.588477366255144e-05,
+      "loss": 0.5631,
+      "step": 576
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5876543209876543e-05,
+      "loss": 0.6188,
+      "step": 577
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.586831275720165e-05,
+      "loss": 0.5687,
+      "step": 578
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.586008230452675e-05,
+      "loss": 0.6583,
+      "step": 579
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5851851851851852e-05,
+      "loss": 0.6085,
+      "step": 580
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5843621399176957e-05,
+      "loss": 0.5445,
+      "step": 581
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.583539094650206e-05,
+      "loss": 0.5987,
+      "step": 582
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.582716049382716e-05,
+      "loss": 0.6199,
+      "step": 583
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5818930041152266e-05,
+      "loss": 0.6168,
+      "step": 584
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5810699588477368e-05,
+      "loss": 0.562,
+      "step": 585
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.580246913580247e-05,
+      "loss": 0.6126,
+      "step": 586
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5794238683127575e-05,
+      "loss": 0.6027,
+      "step": 587
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5786008230452677e-05,
+      "loss": 0.5782,
+      "step": 588
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.577777777777778e-05,
+      "loss": 0.6589,
+      "step": 589
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5769547325102884e-05,
+      "loss": 0.584,
+      "step": 590
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5761316872427986e-05,
+      "loss": 0.6236,
+      "step": 591
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5753086419753088e-05,
+      "loss": 0.578,
+      "step": 592
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.574485596707819e-05,
+      "loss": 0.6228,
+      "step": 593
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5736625514403295e-05,
+      "loss": 0.596,
+      "step": 594
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5728395061728396e-05,
+      "loss": 0.5976,
+      "step": 595
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5720164609053498e-05,
+      "loss": 0.5984,
+      "step": 596
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5711934156378603e-05,
+      "loss": 0.5889,
+      "step": 597
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5703703703703705e-05,
+      "loss": 0.6073,
+      "step": 598
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5695473251028807e-05,
+      "loss": 0.629,
+      "step": 599
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.568724279835391e-05,
+      "loss": 0.5446,
+      "step": 600
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5679012345679014e-05,
+      "loss": 0.6082,
+      "step": 601
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5670781893004116e-05,
+      "loss": 0.5761,
+      "step": 602
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5662551440329218e-05,
+      "loss": 0.5766,
+      "step": 603
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.565432098765432e-05,
+      "loss": 0.5862,
+      "step": 604
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5646090534979425e-05,
+      "loss": 0.5673,
+      "step": 605
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5637860082304527e-05,
+      "loss": 0.6084,
+      "step": 606
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5629629629629632e-05,
+      "loss": 0.523,
+      "step": 607
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5621399176954734e-05,
+      "loss": 0.5843,
+      "step": 608
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5613168724279835e-05,
+      "loss": 0.5654,
+      "step": 609
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.560493827160494e-05,
+      "loss": 0.5862,
+      "step": 610
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5596707818930042e-05,
+      "loss": 0.5862,
+      "step": 611
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5588477366255144e-05,
+      "loss": 0.5759,
+      "step": 612
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.558024691358025e-05,
+      "loss": 0.6136,
+      "step": 613
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.557201646090535e-05,
+      "loss": 0.5949,
+      "step": 614
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5563786008230453e-05,
+      "loss": 0.6093,
+      "step": 615
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.5808,
+      "step": 616
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.554732510288066e-05,
+      "loss": 0.5855,
+      "step": 617
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5539094650205762e-05,
+      "loss": 0.5835,
+      "step": 618
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5530864197530867e-05,
+      "loss": 0.5555,
+      "step": 619
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.552263374485597e-05,
+      "loss": 0.5674,
+      "step": 620
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.551440329218107e-05,
+      "loss": 0.6055,
+      "step": 621
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5506172839506176e-05,
+      "loss": 0.585,
+      "step": 622
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5497942386831278e-05,
+      "loss": 0.5657,
+      "step": 623
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.548971193415638e-05,
+      "loss": 0.6211,
+      "step": 624
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.548148148148148e-05,
+      "loss": 0.6239,
+      "step": 625
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5473251028806587e-05,
+      "loss": 0.5787,
+      "step": 626
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.546502057613169e-05,
+      "loss": 0.6414,
+      "step": 627
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.545679012345679e-05,
+      "loss": 0.6435,
+      "step": 628
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5448559670781895e-05,
+      "loss": 0.5709,
+      "step": 629
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5440329218106997e-05,
+      "loss": 0.5996,
+      "step": 630
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.54320987654321e-05,
+      "loss": 0.6423,
+      "step": 631
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.54238683127572e-05,
+      "loss": 0.5438,
+      "step": 632
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5415637860082306e-05,
+      "loss": 0.5748,
+      "step": 633
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5407407407407408e-05,
+      "loss": 0.6301,
+      "step": 634
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.539917695473251e-05,
+      "loss": 0.577,
+      "step": 635
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5390946502057615e-05,
+      "loss": 0.5845,
+      "step": 636
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5382716049382717e-05,
+      "loss": 0.6111,
+      "step": 637
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.537448559670782e-05,
+      "loss": 0.7054,
+      "step": 638
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5366255144032924e-05,
+      "loss": 0.5753,
+      "step": 639
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5358024691358026e-05,
+      "loss": 0.6248,
+      "step": 640
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5349794238683127e-05,
+      "loss": 0.5955,
+      "step": 641
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5341563786008233e-05,
+      "loss": 0.532,
+      "step": 642
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5333333333333334e-05,
+      "loss": 0.6073,
+      "step": 643
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5325102880658436e-05,
+      "loss": 0.6162,
+      "step": 644
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.531687242798354e-05,
+      "loss": 0.5404,
+      "step": 645
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5308641975308643e-05,
+      "loss": 0.6226,
+      "step": 646
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.530041152263375e-05,
+      "loss": 0.5838,
+      "step": 647
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.529218106995885e-05,
+      "loss": 0.5742,
+      "step": 648
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5283950617283952e-05,
+      "loss": 0.5691,
+      "step": 649
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5275720164609057e-05,
+      "loss": 0.6132,
+      "step": 650
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.526748971193416e-05,
+      "loss": 0.5553,
+      "step": 651
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.525925925925926e-05,
+      "loss": 0.6106,
+      "step": 652
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5251028806584363e-05,
+      "loss": 0.5598,
+      "step": 653
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5242798353909466e-05,
+      "loss": 0.5847,
+      "step": 654
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.523456790123457e-05,
+      "loss": 0.5495,
+      "step": 655
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5226337448559672e-05,
+      "loss": 0.5955,
+      "step": 656
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5218106995884773e-05,
+      "loss": 0.5789,
+      "step": 657
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5209876543209879e-05,
+      "loss": 0.5653,
+      "step": 658
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.520164609053498e-05,
+      "loss": 0.6177,
+      "step": 659
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5193415637860082e-05,
+      "loss": 0.5778,
+      "step": 660
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5185185185185187e-05,
+      "loss": 0.6116,
+      "step": 661
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.517695473251029e-05,
+      "loss": 0.5457,
+      "step": 662
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5168724279835391e-05,
+      "loss": 0.6262,
+      "step": 663
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5160493827160495e-05,
+      "loss": 0.5957,
+      "step": 664
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5152263374485598e-05,
+      "loss": 0.5743,
+      "step": 665
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.51440329218107e-05,
+      "loss": 0.5619,
+      "step": 666
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5135802469135803e-05,
+      "loss": 0.5688,
+      "step": 667
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5127572016460907e-05,
+      "loss": 0.5606,
+      "step": 668
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.511934156378601e-05,
+      "loss": 0.5687,
+      "step": 669
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5111111111111112e-05,
+      "loss": 0.5531,
+      "step": 670
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5102880658436214e-05,
+      "loss": 0.6193,
+      "step": 671
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.509465020576132e-05,
+      "loss": 0.5998,
+      "step": 672
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5086419753086421e-05,
+      "loss": 0.5382,
+      "step": 673
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5078189300411523e-05,
+      "loss": 0.5623,
+      "step": 674
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5069958847736628e-05,
+      "loss": 0.5439,
+      "step": 675
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.506172839506173e-05,
+      "loss": 0.5638,
+      "step": 676
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5053497942386832e-05,
+      "loss": 0.6169,
+      "step": 677
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5045267489711934e-05,
+      "loss": 0.5803,
+      "step": 678
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5037037037037039e-05,
+      "loss": 0.6376,
+      "step": 679
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.502880658436214e-05,
+      "loss": 0.5903,
+      "step": 680
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5020576131687244e-05,
+      "loss": 0.5923,
+      "step": 681
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5012345679012348e-05,
+      "loss": 0.6079,
+      "step": 682
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.500411522633745e-05,
+      "loss": 0.5773,
+      "step": 683
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4995884773662553e-05,
+      "loss": 0.5825,
+      "step": 684
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4987654320987655e-05,
+      "loss": 0.5935,
+      "step": 685
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4979423868312758e-05,
+      "loss": 0.641,
+      "step": 686
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4971193415637862e-05,
+      "loss": 0.615,
+      "step": 687
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4962962962962964e-05,
+      "loss": 0.7009,
+      "step": 688
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4954732510288069e-05,
+      "loss": 0.5626,
+      "step": 689
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.494650205761317e-05,
+      "loss": 0.6019,
+      "step": 690
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4938271604938272e-05,
+      "loss": 0.5909,
+      "step": 691
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4930041152263374e-05,
+      "loss": 0.5497,
+      "step": 692
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.492181069958848e-05,
+      "loss": 0.556,
+      "step": 693
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4913580246913581e-05,
+      "loss": 0.5475,
+      "step": 694
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.4905349794238683e-05,
+      "loss": 0.5869,
+      "step": 695
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4897119341563788e-05,
+      "loss": 0.5369,
+      "step": 696
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.488888888888889e-05,
+      "loss": 0.6049,
+      "step": 697
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4880658436213992e-05,
+      "loss": 0.5323,
+      "step": 698
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4872427983539095e-05,
+      "loss": 0.5991,
+      "step": 699
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4864197530864199e-05,
+      "loss": 0.558,
+      "step": 700
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4855967078189302e-05,
+      "loss": 0.5932,
+      "step": 701
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4847736625514404e-05,
+      "loss": 0.5857,
+      "step": 702
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4839506172839508e-05,
+      "loss": 0.5639,
+      "step": 703
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4831275720164611e-05,
+      "loss": 0.5657,
+      "step": 704
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4823045267489713e-05,
+      "loss": 0.5354,
+      "step": 705
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.6059,
+      "step": 706
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.480658436213992e-05,
+      "loss": 0.5289,
+      "step": 707
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4798353909465022e-05,
+      "loss": 0.5742,
+      "step": 708
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4790123456790124e-05,
+      "loss": 0.5648,
+      "step": 709
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4781893004115229e-05,
+      "loss": 0.6036,
+      "step": 710
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.477366255144033e-05,
+      "loss": 0.5734,
+      "step": 711
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4765432098765433e-05,
+      "loss": 0.5385,
+      "step": 712
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4757201646090536e-05,
+      "loss": 0.5672,
+      "step": 713
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.474897119341564e-05,
+      "loss": 0.6105,
+      "step": 714
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4740740740740741e-05,
+      "loss": 0.5796,
+      "step": 715
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4732510288065845e-05,
+      "loss": 0.5526,
+      "step": 716
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4724279835390947e-05,
+      "loss": 0.5916,
+      "step": 717
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.471604938271605e-05,
+      "loss": 0.632,
+      "step": 718
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4707818930041154e-05,
+      "loss": 0.6016,
+      "step": 719
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.4699588477366256e-05,
+      "loss": 0.5889,
+      "step": 720
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.469135802469136e-05,
+      "loss": 0.5636,
+      "step": 721
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4683127572016463e-05,
+      "loss": 0.5625,
+      "step": 722
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4674897119341564e-05,
+      "loss": 0.5556,
+      "step": 723
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 0.5558,
+      "step": 724
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4658436213991771e-05,
+      "loss": 0.5932,
+      "step": 725
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4650205761316873e-05,
+      "loss": 0.6073,
+      "step": 726
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4641975308641975e-05,
+      "loss": 0.5328,
+      "step": 727
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.463374485596708e-05,
+      "loss": 0.5402,
+      "step": 728
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4625514403292182e-05,
+      "loss": 0.5665,
+      "step": 729
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4617283950617284e-05,
+      "loss": 0.5847,
+      "step": 730
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4609053497942387e-05,
+      "loss": 0.5505,
+      "step": 731
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4600823045267491e-05,
+      "loss": 0.6071,
+      "step": 732
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4592592592592594e-05,
+      "loss": 0.5707,
+      "step": 733
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4584362139917696e-05,
+      "loss": 0.5394,
+      "step": 734
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.45761316872428e-05,
+      "loss": 0.5738,
+      "step": 735
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4567901234567903e-05,
+      "loss": 0.5812,
+      "step": 736
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4559670781893005e-05,
+      "loss": 0.5796,
+      "step": 737
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4551440329218107e-05,
+      "loss": 0.549,
+      "step": 738
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4543209876543212e-05,
+      "loss": 0.5823,
+      "step": 739
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4534979423868314e-05,
+      "loss": 0.5927,
+      "step": 740
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4526748971193416e-05,
+      "loss": 0.5396,
+      "step": 741
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4518518518518521e-05,
+      "loss": 0.5451,
+      "step": 742
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4510288065843623e-05,
+      "loss": 0.5531,
+      "step": 743
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4502057613168725e-05,
+      "loss": 0.6051,
+      "step": 744
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4493827160493828e-05,
+      "loss": 0.5756,
+      "step": 745
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4485596707818932e-05,
+      "loss": 0.5872,
+      "step": 746
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4477366255144033e-05,
+      "loss": 0.542,
+      "step": 747
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4469135802469137e-05,
+      "loss": 0.5696,
+      "step": 748
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.446090534979424e-05,
+      "loss": 0.5424,
+      "step": 749
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4452674897119342e-05,
+      "loss": 0.5662,
+      "step": 750
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 0.551,
+      "step": 751
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4436213991769548e-05,
+      "loss": 0.6475,
+      "step": 752
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4427983539094653e-05,
+      "loss": 0.6183,
+      "step": 753
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4419753086419755e-05,
+      "loss": 0.558,
+      "step": 754
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4411522633744856e-05,
+      "loss": 0.5884,
+      "step": 755
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4403292181069962e-05,
+      "loss": 0.5911,
+      "step": 756
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4395061728395063e-05,
+      "loss": 0.5816,
+      "step": 757
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4386831275720165e-05,
+      "loss": 0.5613,
+      "step": 758
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4378600823045267e-05,
+      "loss": 0.5654,
+      "step": 759
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4370370370370372e-05,
+      "loss": 0.58,
+      "step": 760
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4362139917695474e-05,
+      "loss": 0.5373,
+      "step": 761
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4353909465020578e-05,
+      "loss": 0.611,
+      "step": 762
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4345679012345681e-05,
+      "loss": 0.5067,
+      "step": 763
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4337448559670783e-05,
+      "loss": 0.5594,
+      "step": 764
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4329218106995886e-05,
+      "loss": 0.5922,
+      "step": 765
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4320987654320988e-05,
+      "loss": 0.5606,
+      "step": 766
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4312757201646092e-05,
+      "loss": 0.5485,
+      "step": 767
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4304526748971195e-05,
+      "loss": 0.4932,
+      "step": 768
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4296296296296297e-05,
+      "loss": 0.5634,
+      "step": 769
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4288065843621399e-05,
+      "loss": 0.5314,
+      "step": 770
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4279835390946504e-05,
+      "loss": 0.543,
+      "step": 771
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4271604938271606e-05,
+      "loss": 0.5625,
+      "step": 772
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4263374485596708e-05,
+      "loss": 0.5022,
+      "step": 773
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4255144032921813e-05,
+      "loss": 0.5623,
+      "step": 774
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4246913580246915e-05,
+      "loss": 0.5766,
+      "step": 775
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4238683127572017e-05,
+      "loss": 0.608,
+      "step": 776
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.423045267489712e-05,
+      "loss": 0.5322,
+      "step": 777
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4222222222222224e-05,
+      "loss": 0.5835,
+      "step": 778
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4213991769547325e-05,
+      "loss": 0.5109,
+      "step": 779
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4205761316872429e-05,
+      "loss": 0.5938,
+      "step": 780
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4197530864197532e-05,
+      "loss": 0.546,
+      "step": 781
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4189300411522636e-05,
+      "loss": 0.5348,
+      "step": 782
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4181069958847738e-05,
+      "loss": 0.559,
+      "step": 783
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.417283950617284e-05,
+      "loss": 0.5268,
+      "step": 784
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4164609053497945e-05,
+      "loss": 0.6154,
+      "step": 785
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4156378600823047e-05,
+      "loss": 0.5661,
+      "step": 786
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4148148148148148e-05,
+      "loss": 0.5869,
+      "step": 787
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4139917695473254e-05,
+      "loss": 0.5425,
+      "step": 788
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4131687242798355e-05,
+      "loss": 0.6204,
+      "step": 789
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4123456790123457e-05,
+      "loss": 0.588,
+      "step": 790
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4115226337448559e-05,
+      "loss": 0.54,
+      "step": 791
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4106995884773664e-05,
+      "loss": 0.5868,
+      "step": 792
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4098765432098766e-05,
+      "loss": 0.5485,
+      "step": 793
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.409053497942387e-05,
+      "loss": 0.5508,
+      "step": 794
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4082304526748973e-05,
+      "loss": 0.5631,
+      "step": 795
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.5275,
+      "step": 796
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4065843621399178e-05,
+      "loss": 0.4996,
+      "step": 797
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.405761316872428e-05,
+      "loss": 0.5785,
+      "step": 798
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4049382716049384e-05,
+      "loss": 0.5726,
+      "step": 799
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4041152263374487e-05,
+      "loss": 0.5505,
+      "step": 800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4032921810699589e-05,
+      "loss": 0.5351,
+      "step": 801
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4024691358024694e-05,
+      "loss": 0.5527,
+      "step": 802
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4016460905349796e-05,
+      "loss": 0.6152,
+      "step": 803
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4008230452674898e-05,
+      "loss": 0.5502,
+      "step": 804
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4e-05,
+      "loss": 0.6209,
+      "step": 805
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.3991769547325105e-05,
+      "loss": 0.5697,
+      "step": 806
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.3983539094650207e-05,
+      "loss": 0.6031,
+      "step": 807
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.3975308641975309e-05,
+      "loss": 0.5595,
+      "step": 808
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3967078189300414e-05,
+      "loss": 0.5436,
+      "step": 809
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3958847736625516e-05,
+      "loss": 0.5315,
+      "step": 810
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3950617283950617e-05,
+      "loss": 0.5461,
+      "step": 811
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3942386831275721e-05,
+      "loss": 0.5997,
+      "step": 812
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3934156378600824e-05,
+      "loss": 0.5753,
+      "step": 813
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3925925925925928e-05,
+      "loss": 0.6,
+      "step": 814
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.391769547325103e-05,
+      "loss": 0.5839,
+      "step": 815
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3909465020576133e-05,
+      "loss": 0.5523,
+      "step": 816
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3901234567901237e-05,
+      "loss": 0.5073,
+      "step": 817
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3893004115226339e-05,
+      "loss": 0.5411,
+      "step": 818
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.388477366255144e-05,
+      "loss": 0.5323,
+      "step": 819
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.3876543209876546e-05,
+      "loss": 0.56,
+      "step": 820
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3868312757201647e-05,
+      "loss": 0.5745,
+      "step": 821
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.386008230452675e-05,
+      "loss": 0.564,
+      "step": 822
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3851851851851851e-05,
+      "loss": 0.5238,
+      "step": 823
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3843621399176956e-05,
+      "loss": 0.5504,
+      "step": 824
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3835390946502058e-05,
+      "loss": 0.5233,
+      "step": 825
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3827160493827162e-05,
+      "loss": 0.5974,
+      "step": 826
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3818930041152265e-05,
+      "loss": 0.5541,
+      "step": 827
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3810699588477367e-05,
+      "loss": 0.5876,
+      "step": 828
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.380246913580247e-05,
+      "loss": 0.5584,
+      "step": 829
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3794238683127572e-05,
+      "loss": 0.5608,
+      "step": 830
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3786008230452676e-05,
+      "loss": 0.574,
+      "step": 831
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.377777777777778e-05,
+      "loss": 0.5196,
+      "step": 832
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.3769547325102881e-05,
+      "loss": 0.6365,
+      "step": 833
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3761316872427986e-05,
+      "loss": 0.5549,
+      "step": 834
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3753086419753088e-05,
+      "loss": 0.5856,
+      "step": 835
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.374485596707819e-05,
+      "loss": 0.5147,
+      "step": 836
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3736625514403292e-05,
+      "loss": 0.6021,
+      "step": 837
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3728395061728397e-05,
+      "loss": 0.5549,
+      "step": 838
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3720164609053499e-05,
+      "loss": 0.5438,
+      "step": 839
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.37119341563786e-05,
+      "loss": 0.5106,
+      "step": 840
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3703703703703706e-05,
+      "loss": 0.5727,
+      "step": 841
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3695473251028808e-05,
+      "loss": 0.5395,
+      "step": 842
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.368724279835391e-05,
+      "loss": 0.5673,
+      "step": 843
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3679012345679013e-05,
+      "loss": 0.5574,
+      "step": 844
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.3670781893004116e-05,
+      "loss": 0.5215,
+      "step": 845
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.366255144032922e-05,
+      "loss": 0.5804,
+      "step": 846
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3654320987654322e-05,
+      "loss": 0.566,
+      "step": 847
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3646090534979425e-05,
+      "loss": 0.5701,
+      "step": 848
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3637860082304529e-05,
+      "loss": 0.5373,
+      "step": 849
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.362962962962963e-05,
+      "loss": 0.5228,
+      "step": 850
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3621399176954732e-05,
+      "loss": 0.5635,
+      "step": 851
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3613168724279838e-05,
+      "loss": 0.5801,
+      "step": 852
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.360493827160494e-05,
+      "loss": 0.5581,
+      "step": 853
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3596707818930041e-05,
+      "loss": 0.5835,
+      "step": 854
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3588477366255147e-05,
+      "loss": 0.5787,
+      "step": 855
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3580246913580248e-05,
+      "loss": 0.5476,
+      "step": 856
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.357201646090535e-05,
+      "loss": 0.5779,
+      "step": 857
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.3563786008230454e-05,
+      "loss": 0.5575,
+      "step": 858
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3555555555555557e-05,
+      "loss": 0.5214,
+      "step": 859
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3547325102880659e-05,
+      "loss": 0.563,
+      "step": 860
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3539094650205762e-05,
+      "loss": 0.6253,
+      "step": 861
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3530864197530866e-05,
+      "loss": 0.5631,
+      "step": 862
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3522633744855968e-05,
+      "loss": 0.6144,
+      "step": 863
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3514403292181071e-05,
+      "loss": 0.6165,
+      "step": 864
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3506172839506173e-05,
+      "loss": 0.5399,
+      "step": 865
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3497942386831278e-05,
+      "loss": 0.5727,
+      "step": 866
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.348971193415638e-05,
+      "loss": 0.5252,
+      "step": 867
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3481481481481482e-05,
+      "loss": 0.5758,
+      "step": 868
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3473251028806587e-05,
+      "loss": 0.6109,
+      "step": 869
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.3465020576131689e-05,
+      "loss": 0.5885,
+      "step": 870
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.345679012345679e-05,
+      "loss": 0.562,
+      "step": 871
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3448559670781893e-05,
+      "loss": 0.5528,
+      "step": 872
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3440329218106998e-05,
+      "loss": 0.5217,
+      "step": 873
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.34320987654321e-05,
+      "loss": 0.5424,
+      "step": 874
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3423868312757201e-05,
+      "loss": 0.538,
+      "step": 875
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3415637860082307e-05,
+      "loss": 0.5188,
+      "step": 876
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3407407407407408e-05,
+      "loss": 0.6183,
+      "step": 877
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3399176954732512e-05,
+      "loss": 0.5144,
+      "step": 878
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3390946502057614e-05,
+      "loss": 0.5709,
+      "step": 879
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3382716049382717e-05,
+      "loss": 0.5428,
+      "step": 880
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.337448559670782e-05,
+      "loss": 0.5923,
+      "step": 881
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3366255144032923e-05,
+      "loss": 0.5583,
+      "step": 882
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.3358024691358024e-05,
+      "loss": 0.5325,
+      "step": 883
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.334979423868313e-05,
+      "loss": 0.5393,
+      "step": 884
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3341563786008231e-05,
+      "loss": 0.565,
+      "step": 885
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5414,
+      "step": 886
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3325102880658438e-05,
+      "loss": 0.5507,
+      "step": 887
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.331687242798354e-05,
+      "loss": 0.5515,
+      "step": 888
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3308641975308642e-05,
+      "loss": 0.5435,
+      "step": 889
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3300411522633746e-05,
+      "loss": 0.4983,
+      "step": 890
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3292181069958849e-05,
+      "loss": 0.5572,
+      "step": 891
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3283950617283951e-05,
+      "loss": 0.5072,
+      "step": 892
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3275720164609054e-05,
+      "loss": 0.5471,
+      "step": 893
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.3267489711934158e-05,
+      "loss": 0.5981,
+      "step": 894
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.325925925925926e-05,
+      "loss": 0.6511,
+      "step": 895
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3251028806584363e-05,
+      "loss": 0.5418,
+      "step": 896
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3242798353909465e-05,
+      "loss": 0.5556,
+      "step": 897
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.323456790123457e-05,
+      "loss": 0.5495,
+      "step": 898
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3226337448559672e-05,
+      "loss": 0.5227,
+      "step": 899
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3218106995884774e-05,
+      "loss": 0.5318,
+      "step": 900
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.320987654320988e-05,
+      "loss": 0.5132,
+      "step": 901
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3201646090534981e-05,
+      "loss": 0.5707,
+      "step": 902
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3193415637860083e-05,
+      "loss": 0.5785,
+      "step": 903
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3185185185185185e-05,
+      "loss": 0.5222,
+      "step": 904
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.317695473251029e-05,
+      "loss": 0.5794,
+      "step": 905
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3168724279835392e-05,
+      "loss": 0.518,
+      "step": 906
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3160493827160493e-05,
+      "loss": 0.5476,
+      "step": 907
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.3152263374485599e-05,
+      "loss": 0.5577,
+      "step": 908
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.31440329218107e-05,
+      "loss": 0.5607,
+      "step": 909
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3135802469135804e-05,
+      "loss": 0.5307,
+      "step": 910
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3127572016460906e-05,
+      "loss": 0.5105,
+      "step": 911
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.311934156378601e-05,
+      "loss": 0.5454,
+      "step": 912
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3111111111111113e-05,
+      "loss": 0.5369,
+      "step": 913
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3102880658436215e-05,
+      "loss": 0.5066,
+      "step": 914
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3094650205761318e-05,
+      "loss": 0.541,
+      "step": 915
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3086419753086422e-05,
+      "loss": 0.5774,
+      "step": 916
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3078189300411523e-05,
+      "loss": 0.5904,
+      "step": 917
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3069958847736625e-05,
+      "loss": 0.5244,
+      "step": 918
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.306172839506173e-05,
+      "loss": 0.5106,
+      "step": 919
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.3053497942386832e-05,
+      "loss": 0.5299,
+      "step": 920
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3045267489711934e-05,
+      "loss": 0.5239,
+      "step": 921
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.303703703703704e-05,
+      "loss": 0.5468,
+      "step": 922
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3028806584362141e-05,
+      "loss": 0.5865,
+      "step": 923
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3020576131687243e-05,
+      "loss": 0.5563,
+      "step": 924
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.3012345679012346e-05,
+      "loss": 0.6072,
+      "step": 925
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.300411522633745e-05,
+      "loss": 0.5662,
+      "step": 926
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2995884773662552e-05,
+      "loss": 0.5635,
+      "step": 927
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2987654320987655e-05,
+      "loss": 0.5604,
+      "step": 928
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2979423868312759e-05,
+      "loss": 0.5058,
+      "step": 929
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2971193415637862e-05,
+      "loss": 0.5042,
+      "step": 930
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2962962962962964e-05,
+      "loss": 0.5613,
+      "step": 931
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2954732510288066e-05,
+      "loss": 0.5305,
+      "step": 932
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.2946502057613171e-05,
+      "loss": 0.5394,
+      "step": 933
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2938271604938273e-05,
+      "loss": 0.5814,
+      "step": 934
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2930041152263375e-05,
+      "loss": 0.5131,
+      "step": 935
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2921810699588477e-05,
+      "loss": 0.5241,
+      "step": 936
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2913580246913582e-05,
+      "loss": 0.5235,
+      "step": 937
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2905349794238684e-05,
+      "loss": 0.5118,
+      "step": 938
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2897119341563787e-05,
+      "loss": 0.5367,
+      "step": 939
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.288888888888889e-05,
+      "loss": 0.5856,
+      "step": 940
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2880658436213992e-05,
+      "loss": 0.5466,
+      "step": 941
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2872427983539096e-05,
+      "loss": 0.5816,
+      "step": 942
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2864197530864198e-05,
+      "loss": 0.583,
+      "step": 943
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2855967078189301e-05,
+      "loss": 0.5305,
+      "step": 944
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2847736625514405e-05,
+      "loss": 0.5382,
+      "step": 945
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.2839506172839507e-05,
+      "loss": 0.5664,
+      "step": 946
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.283127572016461e-05,
+      "loss": 0.5923,
+      "step": 947
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2823045267489714e-05,
+      "loss": 0.5363,
+      "step": 948
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2814814814814815e-05,
+      "loss": 0.6103,
+      "step": 949
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2806584362139917e-05,
+      "loss": 0.5257,
+      "step": 950
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2798353909465022e-05,
+      "loss": 0.5366,
+      "step": 951
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2790123456790124e-05,
+      "loss": 0.5086,
+      "step": 952
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2781893004115226e-05,
+      "loss": 0.5616,
+      "step": 953
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2773662551440331e-05,
+      "loss": 0.5561,
+      "step": 954
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2765432098765433e-05,
+      "loss": 0.5222,
+      "step": 955
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2757201646090535e-05,
+      "loss": 0.5518,
+      "step": 956
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2748971193415638e-05,
+      "loss": 0.535,
+      "step": 957
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.2740740740740742e-05,
+      "loss": 0.5305,
+      "step": 958
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2732510288065845e-05,
+      "loss": 0.5709,
+      "step": 959
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2724279835390947e-05,
+      "loss": 0.5262,
+      "step": 960
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.271604938271605e-05,
+      "loss": 0.5436,
+      "step": 961
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2707818930041154e-05,
+      "loss": 0.5176,
+      "step": 962
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2699588477366256e-05,
+      "loss": 0.4787,
+      "step": 963
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2691358024691358e-05,
+      "loss": 0.5263,
+      "step": 964
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2683127572016463e-05,
+      "loss": 0.5376,
+      "step": 965
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2674897119341565e-05,
+      "loss": 0.527,
+      "step": 966
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2666666666666667e-05,
+      "loss": 0.5277,
+      "step": 967
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2658436213991772e-05,
+      "loss": 0.5614,
+      "step": 968
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2650205761316874e-05,
+      "loss": 0.5321,
+      "step": 969
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2641975308641976e-05,
+      "loss": 0.5392,
+      "step": 970
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2633744855967079e-05,
+      "loss": 0.5672,
+      "step": 971
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2625514403292183e-05,
+      "loss": 0.5775,
+      "step": 972
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2617283950617284e-05,
+      "loss": 0.5114,
+      "step": 973
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2609053497942388e-05,
+      "loss": 0.5067,
+      "step": 974
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2600823045267491e-05,
+      "loss": 0.5355,
+      "step": 975
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2592592592592593e-05,
+      "loss": 0.5516,
+      "step": 976
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2584362139917697e-05,
+      "loss": 0.5554,
+      "step": 977
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2576131687242799e-05,
+      "loss": 0.5103,
+      "step": 978
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2567901234567904e-05,
+      "loss": 0.5419,
+      "step": 979
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2559670781893006e-05,
+      "loss": 0.5428,
+      "step": 980
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2551440329218107e-05,
+      "loss": 0.5481,
+      "step": 981
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2543209876543213e-05,
+      "loss": 0.584,
+      "step": 982
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2534979423868314e-05,
+      "loss": 0.5121,
+      "step": 983
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2526748971193416e-05,
+      "loss": 0.5165,
+      "step": 984
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2518518518518518e-05,
+      "loss": 0.5255,
+      "step": 985
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2510288065843623e-05,
+      "loss": 0.5154,
+      "step": 986
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2502057613168725e-05,
+      "loss": 0.5967,
+      "step": 987
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2493827160493827e-05,
+      "loss": 0.5582,
+      "step": 988
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.248559670781893e-05,
+      "loss": 0.5656,
+      "step": 989
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2477366255144034e-05,
+      "loss": 0.5534,
+      "step": 990
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2469135802469137e-05,
+      "loss": 0.5246,
+      "step": 991
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.246090534979424e-05,
+      "loss": 0.545,
+      "step": 992
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2452674897119343e-05,
+      "loss": 0.5125,
+      "step": 993
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2444444444444446e-05,
+      "loss": 0.6149,
+      "step": 994
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.2436213991769548e-05,
+      "loss": 0.537,
+      "step": 995
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.242798353909465e-05,
+      "loss": 0.5537,
+      "step": 996
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2419753086419755e-05,
+      "loss": 0.5295,
+      "step": 997
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2411522633744857e-05,
+      "loss": 0.5184,
+      "step": 998
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2403292181069959e-05,
+      "loss": 0.554,
+      "step": 999
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2395061728395064e-05,
+      "loss": 0.5095,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2386831275720166e-05,
+      "loss": 0.5396,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2378600823045268e-05,
+      "loss": 0.4929,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2370370370370371e-05,
+      "loss": 0.542,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2362139917695475e-05,
+      "loss": 0.5637,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2353909465020576e-05,
+      "loss": 0.588,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 0.537,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2337448559670783e-05,
+      "loss": 0.5304,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.2329218106995885e-05,
+      "loss": 0.553,
+      "step": 1008
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2320987654320989e-05,
+      "loss": 0.5283,
+      "step": 1009
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.231275720164609e-05,
+      "loss": 0.5774,
+      "step": 1010
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2304526748971196e-05,
+      "loss": 0.4815,
+      "step": 1011
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2296296296296298e-05,
+      "loss": 0.5064,
+      "step": 1012
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.22880658436214e-05,
+      "loss": 0.5605,
+      "step": 1013
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2279835390946505e-05,
+      "loss": 0.5459,
+      "step": 1014
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2271604938271606e-05,
+      "loss": 0.5687,
+      "step": 1015
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2263374485596708e-05,
+      "loss": 0.5535,
+      "step": 1016
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.225514403292181e-05,
+      "loss": 0.5241,
+      "step": 1017
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2246913580246915e-05,
+      "loss": 0.5543,
+      "step": 1018
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2238683127572017e-05,
+      "loss": 0.5427,
+      "step": 1019
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2230452674897119e-05,
+      "loss": 0.4904,
+      "step": 1020
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.2222222222222224e-05,
+      "loss": 0.5383,
+      "step": 1021
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2213991769547326e-05,
+      "loss": 0.5762,
+      "step": 1022
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.220576131687243e-05,
+      "loss": 0.5734,
+      "step": 1023
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2197530864197531e-05,
+      "loss": 0.5125,
+      "step": 1024
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2189300411522635e-05,
+      "loss": 0.5662,
+      "step": 1025
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2181069958847738e-05,
+      "loss": 0.5318,
+      "step": 1026
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.217283950617284e-05,
+      "loss": 0.5823,
+      "step": 1027
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2164609053497944e-05,
+      "loss": 0.5218,
+      "step": 1028
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2156378600823047e-05,
+      "loss": 0.5488,
+      "step": 1029
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2148148148148149e-05,
+      "loss": 0.5567,
+      "step": 1030
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.213991769547325e-05,
+      "loss": 0.5335,
+      "step": 1031
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2131687242798356e-05,
+      "loss": 0.5162,
+      "step": 1032
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.2123456790123458e-05,
+      "loss": 0.5105,
+      "step": 1033
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.211522633744856e-05,
+      "loss": 0.5231,
+      "step": 1034
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2106995884773665e-05,
+      "loss": 0.5246,
+      "step": 1035
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2098765432098767e-05,
+      "loss": 0.526,
+      "step": 1036
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2090534979423868e-05,
+      "loss": 0.5429,
+      "step": 1037
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2082304526748972e-05,
+      "loss": 0.4968,
+      "step": 1038
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2074074074074075e-05,
+      "loss": 0.5022,
+      "step": 1039
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2065843621399177e-05,
+      "loss": 0.4795,
+      "step": 1040
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.205761316872428e-05,
+      "loss": 0.5076,
+      "step": 1041
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2049382716049384e-05,
+      "loss": 0.5456,
+      "step": 1042
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2041152263374488e-05,
+      "loss": 0.5714,
+      "step": 1043
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.203292181069959e-05,
+      "loss": 0.5429,
+      "step": 1044
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2024691358024691e-05,
+      "loss": 0.5212,
+      "step": 1045
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.2016460905349797e-05,
+      "loss": 0.5386,
+      "step": 1046
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2008230452674898e-05,
+      "loss": 0.5516,
+      "step": 1047
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5593,
+      "step": 1048
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1991769547325102e-05,
+      "loss": 0.5362,
+      "step": 1049
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1983539094650207e-05,
+      "loss": 0.5232,
+      "step": 1050
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1975308641975309e-05,
+      "loss": 0.5048,
+      "step": 1051
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1967078189300411e-05,
+      "loss": 0.5344,
+      "step": 1052
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1958847736625516e-05,
+      "loss": 0.4771,
+      "step": 1053
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1950617283950618e-05,
+      "loss": 0.5103,
+      "step": 1054
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1942386831275721e-05,
+      "loss": 0.54,
+      "step": 1055
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1934156378600823e-05,
+      "loss": 0.5716,
+      "step": 1056
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.1925925925925927e-05,
+      "loss": 0.5293,
+      "step": 1057
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.191769547325103e-05,
+      "loss": 0.5157,
+      "step": 1058
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1909465020576132e-05,
+      "loss": 0.4781,
+      "step": 1059
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1901234567901236e-05,
+      "loss": 0.524,
+      "step": 1060
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.189300411522634e-05,
+      "loss": 0.4891,
+      "step": 1061
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1884773662551441e-05,
+      "loss": 0.4941,
+      "step": 1062
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1876543209876543e-05,
+      "loss": 0.5032,
+      "step": 1063
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1868312757201648e-05,
+      "loss": 0.5092,
+      "step": 1064
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.186008230452675e-05,
+      "loss": 0.5531,
+      "step": 1065
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.5453,
+      "step": 1066
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1843621399176957e-05,
+      "loss": 0.475,
+      "step": 1067
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1835390946502059e-05,
+      "loss": 0.533,
+      "step": 1068
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.182716049382716e-05,
+      "loss": 0.5161,
+      "step": 1069
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1818930041152264e-05,
+      "loss": 0.5217,
+      "step": 1070
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1810699588477367e-05,
+      "loss": 0.5246,
+      "step": 1071
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.180246913580247e-05,
+      "loss": 0.549,
+      "step": 1072
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1794238683127573e-05,
+      "loss": 0.5247,
+      "step": 1073
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1786008230452676e-05,
+      "loss": 0.4753,
+      "step": 1074
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.177777777777778e-05,
+      "loss": 0.5278,
+      "step": 1075
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1769547325102882e-05,
+      "loss": 0.5313,
+      "step": 1076
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1761316872427983e-05,
+      "loss": 0.5325,
+      "step": 1077
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1753086419753089e-05,
+      "loss": 0.5315,
+      "step": 1078
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.174485596707819e-05,
+      "loss": 0.498,
+      "step": 1079
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1736625514403292e-05,
+      "loss": 0.4958,
+      "step": 1080
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1728395061728398e-05,
+      "loss": 0.5529,
+      "step": 1081
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.17201646090535e-05,
+      "loss": 0.5157,
+      "step": 1082
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.1711934156378601e-05,
+      "loss": 0.5425,
+      "step": 1083
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1703703703703703e-05,
+      "loss": 0.5123,
+      "step": 1084
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1695473251028808e-05,
+      "loss": 0.5409,
+      "step": 1085
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.168724279835391e-05,
+      "loss": 0.5476,
+      "step": 1086
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1679012345679013e-05,
+      "loss": 0.5187,
+      "step": 1087
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1670781893004117e-05,
+      "loss": 0.4925,
+      "step": 1088
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1662551440329219e-05,
+      "loss": 0.5379,
+      "step": 1089
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1654320987654322e-05,
+      "loss": 0.5277,
+      "step": 1090
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1646090534979424e-05,
+      "loss": 0.5101,
+      "step": 1091
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1637860082304528e-05,
+      "loss": 0.5112,
+      "step": 1092
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1629629629629631e-05,
+      "loss": 0.5484,
+      "step": 1093
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1621399176954733e-05,
+      "loss": 0.5753,
+      "step": 1094
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1613168724279838e-05,
+      "loss": 0.5299,
+      "step": 1095
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.160493827160494e-05,
+      "loss": 0.494,
+      "step": 1096
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1596707818930042e-05,
+      "loss": 0.4865,
+      "step": 1097
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1588477366255144e-05,
+      "loss": 0.5303,
+      "step": 1098
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1580246913580249e-05,
+      "loss": 0.4744,
+      "step": 1099
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.157201646090535e-05,
+      "loss": 0.5507,
+      "step": 1100
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1563786008230452e-05,
+      "loss": 0.5021,
+      "step": 1101
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1555555555555556e-05,
+      "loss": 0.5513,
+      "step": 1102
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.154732510288066e-05,
+      "loss": 0.5133,
+      "step": 1103
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1539094650205761e-05,
+      "loss": 0.5431,
+      "step": 1104
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1530864197530865e-05,
+      "loss": 0.5259,
+      "step": 1105
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1522633744855968e-05,
+      "loss": 0.539,
+      "step": 1106
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1514403292181072e-05,
+      "loss": 0.493,
+      "step": 1107
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1506172839506174e-05,
+      "loss": 0.5295,
+      "step": 1108
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1497942386831275e-05,
+      "loss": 0.5026,
+      "step": 1109
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.148971193415638e-05,
+      "loss": 0.5472,
+      "step": 1110
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1481481481481482e-05,
+      "loss": 0.5196,
+      "step": 1111
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1473251028806584e-05,
+      "loss": 0.4716,
+      "step": 1112
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.146502057613169e-05,
+      "loss": 0.5255,
+      "step": 1113
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1456790123456791e-05,
+      "loss": 0.5915,
+      "step": 1114
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1448559670781893e-05,
+      "loss": 0.5526,
+      "step": 1115
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1440329218106997e-05,
+      "loss": 0.5044,
+      "step": 1116
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.14320987654321e-05,
+      "loss": 0.4965,
+      "step": 1117
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1423868312757202e-05,
+      "loss": 0.5098,
+      "step": 1118
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1415637860082305e-05,
+      "loss": 0.516,
+      "step": 1119
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1407407407407409e-05,
+      "loss": 0.5183,
+      "step": 1120
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.139917695473251e-05,
+      "loss": 0.4831,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1390946502057614e-05,
+      "loss": 0.5953,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1382716049382716e-05,
+      "loss": 0.5344,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.137448559670782e-05,
+      "loss": 0.5231,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1366255144032923e-05,
+      "loss": 0.5234,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1358024691358025e-05,
+      "loss": 0.5165,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.134979423868313e-05,
+      "loss": 0.5477,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1341563786008232e-05,
+      "loss": 0.5189,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1333333333333334e-05,
+      "loss": 0.5166,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1325102880658436e-05,
+      "loss": 0.5519,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.131687242798354e-05,
+      "loss": 0.5458,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1308641975308643e-05,
+      "loss": 0.5448,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1300411522633744e-05,
+      "loss": 0.5349,
+      "step": 1133
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.129218106995885e-05,
+      "loss": 0.4986,
+      "step": 1134
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1283950617283951e-05,
+      "loss": 0.557,
+      "step": 1135
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1275720164609055e-05,
+      "loss": 0.4802,
+      "step": 1136
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1267489711934157e-05,
+      "loss": 0.4882,
+      "step": 1137
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.125925925925926e-05,
+      "loss": 0.5489,
+      "step": 1138
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1251028806584364e-05,
+      "loss": 0.5104,
+      "step": 1139
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1242798353909466e-05,
+      "loss": 0.5004,
+      "step": 1140
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.123456790123457e-05,
+      "loss": 0.5383,
+      "step": 1141
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1226337448559673e-05,
+      "loss": 0.5246,
+      "step": 1142
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1218106995884774e-05,
+      "loss": 0.5334,
+      "step": 1143
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1209876543209876e-05,
+      "loss": 0.5431,
+      "step": 1144
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1201646090534982e-05,
+      "loss": 0.4977,
+      "step": 1145
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.1193415637860083e-05,
+      "loss": 0.5373,
+      "step": 1146
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1185185185185185e-05,
+      "loss": 0.5645,
+      "step": 1147
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.117695473251029e-05,
+      "loss": 0.5578,
+      "step": 1148
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1168724279835392e-05,
+      "loss": 0.5313,
+      "step": 1149
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1160493827160494e-05,
+      "loss": 0.5278,
+      "step": 1150
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1152263374485597e-05,
+      "loss": 0.5202,
+      "step": 1151
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1144032921810701e-05,
+      "loss": 0.5397,
+      "step": 1152
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1135802469135803e-05,
+      "loss": 0.5067,
+      "step": 1153
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1127572016460906e-05,
+      "loss": 0.5589,
+      "step": 1154
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1119341563786008e-05,
+      "loss": 0.5103,
+      "step": 1155
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.5159,
+      "step": 1156
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1102880658436215e-05,
+      "loss": 0.5444,
+      "step": 1157
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1094650205761317e-05,
+      "loss": 0.5595,
+      "step": 1158
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.1086419753086422e-05,
+      "loss": 0.515,
+      "step": 1159
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1078189300411524e-05,
+      "loss": 0.5088,
+      "step": 1160
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1069958847736626e-05,
+      "loss": 0.5397,
+      "step": 1161
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1061728395061728e-05,
+      "loss": 0.4955,
+      "step": 1162
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1053497942386833e-05,
+      "loss": 0.5262,
+      "step": 1163
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1045267489711935e-05,
+      "loss": 0.555,
+      "step": 1164
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1037037037037036e-05,
+      "loss": 0.5261,
+      "step": 1165
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1028806584362142e-05,
+      "loss": 0.5182,
+      "step": 1166
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1020576131687243e-05,
+      "loss": 0.5064,
+      "step": 1167
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1012345679012347e-05,
+      "loss": 0.6149,
+      "step": 1168
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1004115226337449e-05,
+      "loss": 0.5296,
+      "step": 1169
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.0995884773662552e-05,
+      "loss": 0.5306,
+      "step": 1170
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.0987654320987656e-05,
+      "loss": 0.5532,
+      "step": 1171
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0979423868312758e-05,
+      "loss": 0.5017,
+      "step": 1172
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0971193415637861e-05,
+      "loss": 0.5209,
+      "step": 1173
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0962962962962965e-05,
+      "loss": 0.4912,
+      "step": 1174
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0954732510288066e-05,
+      "loss": 0.5386,
+      "step": 1175
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0946502057613168e-05,
+      "loss": 0.4777,
+      "step": 1176
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0938271604938273e-05,
+      "loss": 0.5449,
+      "step": 1177
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0930041152263375e-05,
+      "loss": 0.5153,
+      "step": 1178
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0921810699588477e-05,
+      "loss": 0.5251,
+      "step": 1179
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0913580246913582e-05,
+      "loss": 0.4846,
+      "step": 1180
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0905349794238684e-05,
+      "loss": 0.4793,
+      "step": 1181
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0897119341563786e-05,
+      "loss": 0.4974,
+      "step": 1182
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.088888888888889e-05,
+      "loss": 0.5094,
+      "step": 1183
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0880658436213993e-05,
+      "loss": 0.5157,
+      "step": 1184
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0872427983539095e-05,
+      "loss": 0.493,
+      "step": 1185
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0864197530864198e-05,
+      "loss": 0.5304,
+      "step": 1186
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0855967078189302e-05,
+      "loss": 0.514,
+      "step": 1187
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0847736625514405e-05,
+      "loss": 0.5025,
+      "step": 1188
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0839506172839507e-05,
+      "loss": 0.5248,
+      "step": 1189
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0831275720164609e-05,
+      "loss": 0.5373,
+      "step": 1190
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0823045267489714e-05,
+      "loss": 0.5149,
+      "step": 1191
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0814814814814816e-05,
+      "loss": 0.5614,
+      "step": 1192
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0806584362139918e-05,
+      "loss": 0.5571,
+      "step": 1193
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0798353909465023e-05,
+      "loss": 0.512,
+      "step": 1194
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0790123456790125e-05,
+      "loss": 0.5479,
+      "step": 1195
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0781893004115227e-05,
+      "loss": 0.5233,
+      "step": 1196
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0773662551440328e-05,
+      "loss": 0.5727,
+      "step": 1197
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0765432098765434e-05,
+      "loss": 0.5786,
+      "step": 1198
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0757201646090535e-05,
+      "loss": 0.519,
+      "step": 1199
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0748971193415639e-05,
+      "loss": 0.4767,
+      "step": 1200
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0740740740740742e-05,
+      "loss": 0.5268,
+      "step": 1201
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0732510288065844e-05,
+      "loss": 0.5637,
+      "step": 1202
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0724279835390948e-05,
+      "loss": 0.4883,
+      "step": 1203
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.071604938271605e-05,
+      "loss": 0.5147,
+      "step": 1204
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0707818930041153e-05,
+      "loss": 0.5135,
+      "step": 1205
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0699588477366257e-05,
+      "loss": 0.5313,
+      "step": 1206
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0691358024691358e-05,
+      "loss": 0.4877,
+      "step": 1207
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0683127572016464e-05,
+      "loss": 0.513,
+      "step": 1208
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0674897119341565e-05,
+      "loss": 0.4865,
+      "step": 1209
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.5286,
+      "step": 1210
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0658436213991769e-05,
+      "loss": 0.5058,
+      "step": 1211
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0650205761316874e-05,
+      "loss": 0.4998,
+      "step": 1212
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0641975308641976e-05,
+      "loss": 0.5163,
+      "step": 1213
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0633744855967078e-05,
+      "loss": 0.5149,
+      "step": 1214
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0625514403292181e-05,
+      "loss": 0.5005,
+      "step": 1215
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0617283950617285e-05,
+      "loss": 0.4753,
+      "step": 1216
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0609053497942387e-05,
+      "loss": 0.5009,
+      "step": 1217
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.060082304526749e-05,
+      "loss": 0.5569,
+      "step": 1218
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0592592592592594e-05,
+      "loss": 0.4954,
+      "step": 1219
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0584362139917697e-05,
+      "loss": 0.4952,
+      "step": 1220
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0576131687242799e-05,
+      "loss": 0.4762,
+      "step": 1221
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0567901234567901e-05,
+      "loss": 0.5076,
+      "step": 1222
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0559670781893006e-05,
+      "loss": 0.486,
+      "step": 1223
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0551440329218108e-05,
+      "loss": 0.502,
+      "step": 1224
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.054320987654321e-05,
+      "loss": 0.5802,
+      "step": 1225
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0534979423868315e-05,
+      "loss": 0.5708,
+      "step": 1226
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0526748971193417e-05,
+      "loss": 0.5051,
+      "step": 1227
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0518518518518519e-05,
+      "loss": 0.506,
+      "step": 1228
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.051028806584362e-05,
+      "loss": 0.5356,
+      "step": 1229
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0502057613168726e-05,
+      "loss": 0.5325,
+      "step": 1230
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0493827160493827e-05,
+      "loss": 0.5199,
+      "step": 1231
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0485596707818931e-05,
+      "loss": 0.5346,
+      "step": 1232
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0477366255144034e-05,
+      "loss": 0.5627,
+      "step": 1233
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0469135802469136e-05,
+      "loss": 0.5338,
+      "step": 1234
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.046090534979424e-05,
+      "loss": 0.5085,
+      "step": 1235
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0452674897119342e-05,
+      "loss": 0.5182,
+      "step": 1236
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0444444444444445e-05,
+      "loss": 0.5225,
+      "step": 1237
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0436213991769549e-05,
+      "loss": 0.5226,
+      "step": 1238
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.042798353909465e-05,
+      "loss": 0.4991,
+      "step": 1239
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0419753086419756e-05,
+      "loss": 0.5049,
+      "step": 1240
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0411522633744857e-05,
+      "loss": 0.5078,
+      "step": 1241
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.040329218106996e-05,
+      "loss": 0.6137,
+      "step": 1242
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0395061728395061e-05,
+      "loss": 0.5137,
+      "step": 1243
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0386831275720166e-05,
+      "loss": 0.5034,
+      "step": 1244
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0378600823045268e-05,
+      "loss": 0.5338,
+      "step": 1245
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.5104,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0362139917695475e-05,
+      "loss": 0.5049,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0353909465020577e-05,
+      "loss": 0.4552,
+      "step": 1248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0345679012345679e-05,
+      "loss": 0.5168,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0337448559670782e-05,
+      "loss": 0.4915,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0329218106995886e-05,
+      "loss": 0.504,
+      "step": 1251
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.032098765432099e-05,
+      "loss": 0.4946,
+      "step": 1252
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0312757201646091e-05,
+      "loss": 0.5057,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0304526748971195e-05,
+      "loss": 0.5022,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0296296296296298e-05,
+      "loss": 0.5264,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.02880658436214e-05,
+      "loss": 0.4802,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0279835390946502e-05,
+      "loss": 0.4995,
+      "step": 1257
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0271604938271607e-05,
+      "loss": 0.4696,
+      "step": 1258
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0263374485596709e-05,
+      "loss": 0.4868,
+      "step": 1259
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.025514403292181e-05,
+      "loss": 0.474,
+      "step": 1260
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0246913580246916e-05,
+      "loss": 0.4845,
+      "step": 1261
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0238683127572018e-05,
+      "loss": 0.5053,
+      "step": 1262
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.023045267489712e-05,
+      "loss": 0.4865,
+      "step": 1263
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0222222222222223e-05,
+      "loss": 0.4582,
+      "step": 1264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0213991769547326e-05,
+      "loss": 0.4529,
+      "step": 1265
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0205761316872428e-05,
+      "loss": 0.4882,
+      "step": 1266
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0197530864197532e-05,
+      "loss": 0.4321,
+      "step": 1267
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0189300411522634e-05,
+      "loss": 0.4572,
+      "step": 1268
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0181069958847737e-05,
+      "loss": 0.4207,
+      "step": 1269
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.017283950617284e-05,
+      "loss": 0.4671,
+      "step": 1270
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0164609053497942e-05,
+      "loss": 0.4195,
+      "step": 1271
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0156378600823048e-05,
+      "loss": 0.4259,
+      "step": 1272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.014814814814815e-05,
+      "loss": 0.4434,
+      "step": 1273
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0139917695473251e-05,
+      "loss": 0.4646,
+      "step": 1274
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0131687242798353e-05,
+      "loss": 0.4251,
+      "step": 1275
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0123456790123458e-05,
+      "loss": 0.4445,
+      "step": 1276
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.011522633744856e-05,
+      "loss": 0.4697,
+      "step": 1277
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0106995884773662e-05,
+      "loss": 0.3973,
+      "step": 1278
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0098765432098767e-05,
+      "loss": 0.4281,
+      "step": 1279
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0090534979423869e-05,
+      "loss": 0.4355,
+      "step": 1280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.008230452674897e-05,
+      "loss": 0.4301,
+      "step": 1281
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0074074074074074e-05,
+      "loss": 0.417,
+      "step": 1282
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0065843621399178e-05,
+      "loss": 0.4208,
+      "step": 1283
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0057613168724281e-05,
+      "loss": 0.4019,
+      "step": 1284
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0049382716049383e-05,
+      "loss": 0.3739,
+      "step": 1285
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0041152263374487e-05,
+      "loss": 0.3848,
+      "step": 1286
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.003292181069959e-05,
+      "loss": 0.3868,
+      "step": 1287
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0024691358024692e-05,
+      "loss": 0.4228,
+      "step": 1288
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0016460905349794e-05,
+      "loss": 0.3993,
+      "step": 1289
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.0008230452674899e-05,
+      "loss": 0.3704,
+      "step": 1290
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.3627,
+      "step": 1291
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.991769547325104e-06,
+      "loss": 0.3864,
+      "step": 1292
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.983539094650206e-06,
+      "loss": 0.4302,
+      "step": 1293
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.97530864197531e-06,
+      "loss": 0.4156,
+      "step": 1294
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.967078189300411e-06,
+      "loss": 0.3256,
+      "step": 1295
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.958847736625515e-06,
+      "loss": 0.3605,
+      "step": 1296
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.950617283950618e-06,
+      "loss": 0.3694,
+      "step": 1297
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.94238683127572e-06,
+      "loss": 0.3475,
+      "step": 1298
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.934156378600824e-06,
+      "loss": 0.3735,
+      "step": 1299
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.925925925925927e-06,
+      "loss": 0.397,
+      "step": 1300
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.91769547325103e-06,
+      "loss": 0.3492,
+      "step": 1301
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.909465020576133e-06,
+      "loss": 0.3531,
+      "step": 1302
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.901234567901236e-06,
+      "loss": 0.3487,
+      "step": 1303
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.893004115226338e-06,
+      "loss": 0.3534,
+      "step": 1304
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.884773662551441e-06,
+      "loss": 0.3829,
+      "step": 1305
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.3128,
+      "step": 1306
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.868312757201647e-06,
+      "loss": 0.3337,
+      "step": 1307
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.86008230452675e-06,
+      "loss": 0.3631,
+      "step": 1308
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.851851851851852e-06,
+      "loss": 0.3663,
+      "step": 1309
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.843621399176956e-06,
+      "loss": 0.3512,
+      "step": 1310
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.835390946502057e-06,
+      "loss": 0.3322,
+      "step": 1311
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.827160493827161e-06,
+      "loss": 0.3696,
+      "step": 1312
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.818930041152264e-06,
+      "loss": 0.3548,
+      "step": 1313
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.810699588477366e-06,
+      "loss": 0.299,
+      "step": 1314
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.80246913580247e-06,
+      "loss": 0.3024,
+      "step": 1315
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.794238683127573e-06,
+      "loss": 0.3428,
+      "step": 1316
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.786008230452677e-06,
+      "loss": 0.3382,
+      "step": 1317
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.3013,
+      "step": 1318
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.769547325102882e-06,
+      "loss": 0.2947,
+      "step": 1319
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.761316872427984e-06,
+      "loss": 0.3174,
+      "step": 1320
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.753086419753087e-06,
+      "loss": 0.2942,
+      "step": 1321
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.744855967078191e-06,
+      "loss": 0.2798,
+      "step": 1322
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.736625514403293e-06,
+      "loss": 0.3103,
+      "step": 1323
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.728395061728396e-06,
+      "loss": 0.2699,
+      "step": 1324
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.720164609053498e-06,
+      "loss": 0.3081,
+      "step": 1325
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.711934156378602e-06,
+      "loss": 0.2962,
+      "step": 1326
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.703703703703703e-06,
+      "loss": 0.3059,
+      "step": 1327
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.695473251028807e-06,
+      "loss": 0.3101,
+      "step": 1328
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.68724279835391e-06,
+      "loss": 0.3246,
+      "step": 1329
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.679012345679012e-06,
+      "loss": 0.3194,
+      "step": 1330
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.670781893004116e-06,
+      "loss": 0.2728,
+      "step": 1331
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.66255144032922e-06,
+      "loss": 0.2989,
+      "step": 1332
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.654320987654323e-06,
+      "loss": 0.2716,
+      "step": 1333
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.646090534979425e-06,
+      "loss": 0.3631,
+      "step": 1334
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.637860082304528e-06,
+      "loss": 0.3426,
+      "step": 1335
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.3272,
+      "step": 1336
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.621399176954733e-06,
+      "loss": 0.2949,
+      "step": 1337
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.613168724279837e-06,
+      "loss": 0.3826,
+      "step": 1338
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.604938271604939e-06,
+      "loss": 0.3356,
+      "step": 1339
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.596707818930042e-06,
+      "loss": 0.3237,
+      "step": 1340
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.588477366255144e-06,
+      "loss": 0.3324,
+      "step": 1341
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.580246913580248e-06,
+      "loss": 0.2777,
+      "step": 1342
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.57201646090535e-06,
+      "loss": 0.3447,
+      "step": 1343
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.563786008230453e-06,
+      "loss": 0.3217,
+      "step": 1344
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 0.2932,
+      "step": 1345
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.547325102880658e-06,
+      "loss": 0.2861,
+      "step": 1346
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.539094650205762e-06,
+      "loss": 0.326,
+      "step": 1347
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.530864197530865e-06,
+      "loss": 0.2734,
+      "step": 1348
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.522633744855969e-06,
+      "loss": 0.3054,
+      "step": 1349
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.51440329218107e-06,
+      "loss": 0.2998,
+      "step": 1350
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.506172839506174e-06,
+      "loss": 0.3123,
+      "step": 1351
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.497942386831278e-06,
+      "loss": 0.2947,
+      "step": 1352
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.48971193415638e-06,
+      "loss": 0.3058,
+      "step": 1353
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.481481481481483e-06,
+      "loss": 0.2916,
+      "step": 1354
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.473251028806585e-06,
+      "loss": 0.2898,
+      "step": 1355
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.465020576131688e-06,
+      "loss": 0.2776,
+      "step": 1356
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.45679012345679e-06,
+      "loss": 0.3094,
+      "step": 1357
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.448559670781894e-06,
+      "loss": 0.2882,
+      "step": 1358
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.440329218106997e-06,
+      "loss": 0.2911,
+      "step": 1359
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.432098765432099e-06,
+      "loss": 0.2657,
+      "step": 1360
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.423868312757202e-06,
+      "loss": 0.2731,
+      "step": 1361
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.415637860082304e-06,
+      "loss": 0.2729,
+      "step": 1362
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.407407407407408e-06,
+      "loss": 0.3034,
+      "step": 1363
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.399176954732511e-06,
+      "loss": 0.2899,
+      "step": 1364
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.390946502057615e-06,
+      "loss": 0.2782,
+      "step": 1365
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.2776,
+      "step": 1366
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.37448559670782e-06,
+      "loss": 0.2737,
+      "step": 1367
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.366255144032924e-06,
+      "loss": 0.2793,
+      "step": 1368
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.358024691358025e-06,
+      "loss": 0.2729,
+      "step": 1369
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.349794238683129e-06,
+      "loss": 0.3065,
+      "step": 1370
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.34156378600823e-06,
+      "loss": 0.2739,
+      "step": 1371
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.2853,
+      "step": 1372
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.325102880658436e-06,
+      "loss": 0.3306,
+      "step": 1373
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.31687242798354e-06,
+      "loss": 0.286,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.308641975308643e-06,
+      "loss": 0.2735,
+      "step": 1375
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.300411522633745e-06,
+      "loss": 0.2606,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.292181069958848e-06,
+      "loss": 0.2836,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.28395061728395e-06,
+      "loss": 0.2723,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.275720164609054e-06,
+      "loss": 0.2472,
+      "step": 1379
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.267489711934157e-06,
+      "loss": 0.3006,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.3048,
+      "step": 1381
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.251028806584363e-06,
+      "loss": 0.2849,
+      "step": 1382
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.242798353909466e-06,
+      "loss": 0.2946,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.23456790123457e-06,
+      "loss": 0.2828,
+      "step": 1384
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.226337448559671e-06,
+      "loss": 0.2904,
+      "step": 1385
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.218106995884775e-06,
+      "loss": 0.324,
+      "step": 1386
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.209876543209877e-06,
+      "loss": 0.2671,
+      "step": 1387
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.20164609053498e-06,
+      "loss": 0.273,
+      "step": 1388
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.193415637860082e-06,
+      "loss": 0.2793,
+      "step": 1389
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.185185185185186e-06,
+      "loss": 0.2938,
+      "step": 1390
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.17695473251029e-06,
+      "loss": 0.3201,
+      "step": 1391
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.168724279835391e-06,
+      "loss": 0.2743,
+      "step": 1392
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.160493827160494e-06,
+      "loss": 0.2627,
+      "step": 1393
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.152263374485596e-06,
+      "loss": 0.3088,
+      "step": 1394
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.1440329218107e-06,
+      "loss": 0.3134,
+      "step": 1395
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 0.2767,
+      "step": 1396
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.127572016460907e-06,
+      "loss": 0.2863,
+      "step": 1397
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.119341563786009e-06,
+      "loss": 0.2322,
+      "step": 1398
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 0.2918,
+      "step": 1399
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.102880658436216e-06,
+      "loss": 0.2813,
+      "step": 1400
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.094650205761317e-06,
+      "loss": 0.2957,
+      "step": 1401
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.086419753086421e-06,
+      "loss": 0.3009,
+      "step": 1402
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.078189300411523e-06,
+      "loss": 0.2558,
+      "step": 1403
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.069958847736626e-06,
+      "loss": 0.2943,
+      "step": 1404
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.06172839506173e-06,
+      "loss": 0.3802,
+      "step": 1405
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.053497942386832e-06,
+      "loss": 0.2981,
+      "step": 1406
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.045267489711935e-06,
+      "loss": 0.2869,
+      "step": 1407
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.037037037037037e-06,
+      "loss": 0.3341,
+      "step": 1408
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.02880658436214e-06,
+      "loss": 0.2696,
+      "step": 1409
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.020576131687242e-06,
+      "loss": 0.2485,
+      "step": 1410
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.012345679012346e-06,
+      "loss": 0.2582,
+      "step": 1411
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.00411522633745e-06,
+      "loss": 0.259,
+      "step": 1412
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.995884773662553e-06,
+      "loss": 0.2742,
+      "step": 1413
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.987654320987655e-06,
+      "loss": 0.2547,
+      "step": 1414
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.979423868312758e-06,
+      "loss": 0.248,
+      "step": 1415
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.971193415637862e-06,
+      "loss": 0.3047,
+      "step": 1416
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.962962962962963e-06,
+      "loss": 0.2786,
+      "step": 1417
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.954732510288067e-06,
+      "loss": 0.2941,
+      "step": 1418
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.946502057613169e-06,
+      "loss": 0.2915,
+      "step": 1419
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.938271604938272e-06,
+      "loss": 0.2721,
+      "step": 1420
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.930041152263376e-06,
+      "loss": 0.2718,
+      "step": 1421
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 8.921810699588478e-06,
+      "loss": 0.2706,
+      "step": 1422
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.913580246913581e-06,
+      "loss": 0.2611,
+      "step": 1423
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.905349794238683e-06,
+      "loss": 0.2819,
+      "step": 1424
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.897119341563786e-06,
+      "loss": 0.2635,
+      "step": 1425
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.2758,
+      "step": 1426
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.880658436213992e-06,
+      "loss": 0.2855,
+      "step": 1427
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.872427983539095e-06,
+      "loss": 0.2823,
+      "step": 1428
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.864197530864199e-06,
+      "loss": 0.2946,
+      "step": 1429
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.8559670781893e-06,
+      "loss": 0.3063,
+      "step": 1430
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.847736625514404e-06,
+      "loss": 0.2328,
+      "step": 1431
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.839506172839508e-06,
+      "loss": 0.2344,
+      "step": 1432
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.83127572016461e-06,
+      "loss": 0.2858,
+      "step": 1433
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 8.823045267489713e-06,
+      "loss": 0.3126,
+      "step": 1434
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.814814814814817e-06,
+      "loss": 0.2863,
+      "step": 1435
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.806584362139918e-06,
+      "loss": 0.2392,
+      "step": 1436
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.798353909465022e-06,
+      "loss": 0.2569,
+      "step": 1437
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.790123456790124e-06,
+      "loss": 0.3039,
+      "step": 1438
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.781893004115227e-06,
+      "loss": 0.3257,
+      "step": 1439
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.773662551440329e-06,
+      "loss": 0.2793,
+      "step": 1440
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.765432098765432e-06,
+      "loss": 0.262,
+      "step": 1441
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.757201646090536e-06,
+      "loss": 0.2747,
+      "step": 1442
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.748971193415638e-06,
+      "loss": 0.2523,
+      "step": 1443
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.740740740740741e-06,
+      "loss": 0.2514,
+      "step": 1444
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.732510288065845e-06,
+      "loss": 0.2906,
+      "step": 1445
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.724279835390947e-06,
+      "loss": 0.2631,
+      "step": 1446
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.71604938271605e-06,
+      "loss": 0.2677,
+      "step": 1447
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.707818930041154e-06,
+      "loss": 0.2718,
+      "step": 1448
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.699588477366255e-06,
+      "loss": 0.2903,
+      "step": 1449
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.691358024691359e-06,
+      "loss": 0.3198,
+      "step": 1450
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.683127572016463e-06,
+      "loss": 0.2662,
+      "step": 1451
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.674897119341564e-06,
+      "loss": 0.282,
+      "step": 1452
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.2757,
+      "step": 1453
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.65843621399177e-06,
+      "loss": 0.2874,
+      "step": 1454
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.650205761316873e-06,
+      "loss": 0.2843,
+      "step": 1455
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.2909,
+      "step": 1456
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.633744855967078e-06,
+      "loss": 0.2664,
+      "step": 1457
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.625514403292182e-06,
+      "loss": 0.3829,
+      "step": 1458
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.617283950617284e-06,
+      "loss": 0.3079,
+      "step": 1459
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.609053497942387e-06,
+      "loss": 0.3027,
+      "step": 1460
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.60082304526749e-06,
+      "loss": 0.2687,
+      "step": 1461
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.592592592592593e-06,
+      "loss": 0.2745,
+      "step": 1462
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.584362139917696e-06,
+      "loss": 0.251,
+      "step": 1463
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.5761316872428e-06,
+      "loss": 0.251,
+      "step": 1464
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.567901234567903e-06,
+      "loss": 0.2884,
+      "step": 1465
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.559670781893005e-06,
+      "loss": 0.2972,
+      "step": 1466
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.551440329218108e-06,
+      "loss": 0.316,
+      "step": 1467
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.54320987654321e-06,
+      "loss": 0.2755,
+      "step": 1468
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.534979423868314e-06,
+      "loss": 0.3401,
+      "step": 1469
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.526748971193416e-06,
+      "loss": 0.2628,
+      "step": 1470
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.2902,
+      "step": 1471
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.510288065843621e-06,
+      "loss": 0.2714,
+      "step": 1472
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.502057613168724e-06,
+      "loss": 0.2779,
+      "step": 1473
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.493827160493828e-06,
+      "loss": 0.2788,
+      "step": 1474
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.48559670781893e-06,
+      "loss": 0.26,
+      "step": 1475
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.477366255144033e-06,
+      "loss": 0.2766,
+      "step": 1476
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.469135802469137e-06,
+      "loss": 0.23,
+      "step": 1477
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.46090534979424e-06,
+      "loss": 0.3091,
+      "step": 1478
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.452674897119342e-06,
+      "loss": 0.2483,
+      "step": 1479
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 0.2824,
+      "step": 1480
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.43621399176955e-06,
+      "loss": 0.2732,
+      "step": 1481
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.427983539094651e-06,
+      "loss": 0.2535,
+      "step": 1482
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.419753086419754e-06,
+      "loss": 0.2917,
+      "step": 1483
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.411522633744856e-06,
+      "loss": 0.2779,
+      "step": 1484
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.40329218106996e-06,
+      "loss": 0.2714,
+      "step": 1485
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.2835,
+      "step": 1486
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.386831275720165e-06,
+      "loss": 0.2868,
+      "step": 1487
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.378600823045269e-06,
+      "loss": 0.277,
+      "step": 1488
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.37037037037037e-06,
+      "loss": 0.3242,
+      "step": 1489
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.362139917695474e-06,
+      "loss": 0.2791,
+      "step": 1490
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.353909465020576e-06,
+      "loss": 0.2954,
+      "step": 1491
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.34567901234568e-06,
+      "loss": 0.3437,
+      "step": 1492
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.337448559670783e-06,
+      "loss": 0.3047,
+      "step": 1493
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.329218106995886e-06,
+      "loss": 0.2606,
+      "step": 1494
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.320987654320988e-06,
+      "loss": 0.2764,
+      "step": 1495
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.312757201646092e-06,
+      "loss": 0.3562,
+      "step": 1496
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.304526748971195e-06,
+      "loss": 0.2678,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.296296296296297e-06,
+      "loss": 0.2884,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.2880658436214e-06,
+      "loss": 0.2849,
+      "step": 1499
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.279835390946502e-06,
+      "loss": 0.2832,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.271604938271606e-06,
+      "loss": 0.2761,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.263374485596708e-06,
+      "loss": 0.2366,
+      "step": 1502
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.255144032921811e-06,
+      "loss": 0.2619,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.246913580246915e-06,
+      "loss": 0.2568,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.238683127572016e-06,
+      "loss": 0.2505,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.23045267489712e-06,
+      "loss": 0.272,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 0.3069,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.213991769547325e-06,
+      "loss": 0.2972,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.205761316872429e-06,
+      "loss": 0.3149,
+      "step": 1509
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.197530864197532e-06,
+      "loss": 0.271,
+      "step": 1510
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.189300411522634e-06,
+      "loss": 0.2989,
+      "step": 1511
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.181069958847738e-06,
+      "loss": 0.2547,
+      "step": 1512
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.172839506172841e-06,
+      "loss": 0.2726,
+      "step": 1513
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.164609053497943e-06,
+      "loss": 0.3303,
+      "step": 1514
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.156378600823046e-06,
+      "loss": 0.3004,
+      "step": 1515
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.259,
+      "step": 1516
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.139917695473252e-06,
+      "loss": 0.2723,
+      "step": 1517
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.131687242798355e-06,
+      "loss": 0.2705,
+      "step": 1518
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.123456790123457e-06,
+      "loss": 0.3152,
+      "step": 1519
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.11522633744856e-06,
+      "loss": 0.2565,
+      "step": 1520
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.106995884773662e-06,
+      "loss": 0.3491,
+      "step": 1521
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.098765432098766e-06,
+      "loss": 0.274,
+      "step": 1522
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.090534979423868e-06,
+      "loss": 0.2731,
+      "step": 1523
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.082304526748971e-06,
+      "loss": 0.2679,
+      "step": 1524
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.074074074074075e-06,
+      "loss": 0.2645,
+      "step": 1525
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.065843621399178e-06,
+      "loss": 0.4237,
+      "step": 1526
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.05761316872428e-06,
+      "loss": 0.2785,
+      "step": 1527
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.049382716049384e-06,
+      "loss": 0.2532,
+      "step": 1528
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.041152263374487e-06,
+      "loss": 0.2598,
+      "step": 1529
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.032921810699589e-06,
+      "loss": 0.2415,
+      "step": 1530
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.024691358024692e-06,
+      "loss": 0.2627,
+      "step": 1531
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.016460905349794e-06,
+      "loss": 0.2991,
+      "step": 1532
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.008230452674898e-06,
+      "loss": 0.2986,
+      "step": 1533
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.267,
+      "step": 1534
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.991769547325103e-06,
+      "loss": 0.2931,
+      "step": 1535
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.983539094650207e-06,
+      "loss": 0.2944,
+      "step": 1536
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.975308641975308e-06,
+      "loss": 0.451,
+      "step": 1537
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.967078189300412e-06,
+      "loss": 0.2745,
+      "step": 1538
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.958847736625514e-06,
+      "loss": 0.301,
+      "step": 1539
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.950617283950617e-06,
+      "loss": 0.2736,
+      "step": 1540
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.94238683127572e-06,
+      "loss": 0.3158,
+      "step": 1541
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.934156378600824e-06,
+      "loss": 0.3097,
+      "step": 1542
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.925925925925926e-06,
+      "loss": 0.2732,
+      "step": 1543
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.91769547325103e-06,
+      "loss": 0.2548,
+      "step": 1544
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.909465020576133e-06,
+      "loss": 0.2691,
+      "step": 1545
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 0.2645,
+      "step": 1546
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.893004115226338e-06,
+      "loss": 0.3265,
+      "step": 1547
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.884773662551442e-06,
+      "loss": 0.3058,
+      "step": 1548
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.876543209876544e-06,
+      "loss": 0.2806,
+      "step": 1549
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.868312757201647e-06,
+      "loss": 0.2949,
+      "step": 1550
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.860082304526749e-06,
+      "loss": 0.3447,
+      "step": 1551
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.851851851851853e-06,
+      "loss": 0.4751,
+      "step": 1552
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.843621399176954e-06,
+      "loss": 0.3162,
+      "step": 1553
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.835390946502058e-06,
+      "loss": 0.2954,
+      "step": 1554
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.82716049382716e-06,
+      "loss": 0.2666,
+      "step": 1555
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.818930041152263e-06,
+      "loss": 0.409,
+      "step": 1556
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.810699588477367e-06,
+      "loss": 0.3183,
+      "step": 1557
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.80246913580247e-06,
+      "loss": 0.2919,
+      "step": 1558
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 7.794238683127572e-06,
+      "loss": 0.2778,
+      "step": 1559
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.786008230452676e-06,
+      "loss": 0.2945,
+      "step": 1560
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.2922,
+      "step": 1561
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.769547325102881e-06,
+      "loss": 0.2354,
+      "step": 1562
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.761316872427984e-06,
+      "loss": 0.2466,
+      "step": 1563
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.753086419753088e-06,
+      "loss": 0.3021,
+      "step": 1564
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.74485596707819e-06,
+      "loss": 0.3685,
+      "step": 1565
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.736625514403293e-06,
+      "loss": 0.2941,
+      "step": 1566
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.728395061728395e-06,
+      "loss": 0.2417,
+      "step": 1567
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.720164609053499e-06,
+      "loss": 0.2903,
+      "step": 1568
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.7119341563786e-06,
+      "loss": 0.2588,
+      "step": 1569
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.703703703703704e-06,
+      "loss": 0.2874,
+      "step": 1570
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.695473251028807e-06,
+      "loss": 0.2368,
+      "step": 1571
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.68724279835391e-06,
+      "loss": 0.2746,
+      "step": 1572
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.679012345679013e-06,
+      "loss": 0.275,
+      "step": 1573
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.670781893004116e-06,
+      "loss": 0.3024,
+      "step": 1574
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.662551440329218e-06,
+      "loss": 0.4758,
+      "step": 1575
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 0.2927,
+      "step": 1576
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.646090534979425e-06,
+      "loss": 0.2678,
+      "step": 1577
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.637860082304529e-06,
+      "loss": 0.2603,
+      "step": 1578
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.62962962962963e-06,
+      "loss": 0.3039,
+      "step": 1579
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.621399176954733e-06,
+      "loss": 0.2833,
+      "step": 1580
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.613168724279836e-06,
+      "loss": 0.2665,
+      "step": 1581
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.604938271604939e-06,
+      "loss": 0.2614,
+      "step": 1582
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.596707818930041e-06,
+      "loss": 0.2517,
+      "step": 1583
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.588477366255145e-06,
+      "loss": 0.2541,
+      "step": 1584
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.580246913580247e-06,
+      "loss": 0.3003,
+      "step": 1585
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.57201646090535e-06,
+      "loss": 0.2794,
+      "step": 1586
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.5637860082304535e-06,
+      "loss": 0.2739,
+      "step": 1587
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 0.28,
+      "step": 1588
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.54732510288066e-06,
+      "loss": 0.2259,
+      "step": 1589
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.5390946502057615e-06,
+      "loss": 0.2573,
+      "step": 1590
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.530864197530865e-06,
+      "loss": 0.344,
+      "step": 1591
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.522633744855967e-06,
+      "loss": 0.2811,
+      "step": 1592
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.51440329218107e-06,
+      "loss": 0.2761,
+      "step": 1593
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.506172839506174e-06,
+      "loss": 0.2735,
+      "step": 1594
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.4979423868312765e-06,
+      "loss": 0.229,
+      "step": 1595
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.489711934156379e-06,
+      "loss": 0.2578,
+      "step": 1596
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.481481481481482e-06,
+      "loss": 0.2816,
+      "step": 1597
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.473251028806585e-06,
+      "loss": 0.2523,
+      "step": 1598
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.465020576131687e-06,
+      "loss": 0.2721,
+      "step": 1599
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.456790123456791e-06,
+      "loss": 0.2323,
+      "step": 1600
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.448559670781894e-06,
+      "loss": 0.2423,
+      "step": 1601
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.440329218106996e-06,
+      "loss": 0.3098,
+      "step": 1602
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.4320987654320995e-06,
+      "loss": 0.36,
+      "step": 1603
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.423868312757202e-06,
+      "loss": 0.2811,
+      "step": 1604
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.415637860082306e-06,
+      "loss": 0.2509,
+      "step": 1605
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.2693,
+      "step": 1606
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.399176954732511e-06,
+      "loss": 0.2706,
+      "step": 1607
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.3909465020576145e-06,
+      "loss": 0.259,
+      "step": 1608
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.382716049382716e-06,
+      "loss": 0.2704,
+      "step": 1609
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.37448559670782e-06,
+      "loss": 0.2847,
+      "step": 1610
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.3662551440329225e-06,
+      "loss": 0.2656,
+      "step": 1611
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.358024691358025e-06,
+      "loss": 0.2641,
+      "step": 1612
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.349794238683128e-06,
+      "loss": 0.3288,
+      "step": 1613
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.341563786008231e-06,
+      "loss": 0.3277,
+      "step": 1614
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 0.3049,
+      "step": 1615
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.325102880658437e-06,
+      "loss": 0.3834,
+      "step": 1616
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.31687242798354e-06,
+      "loss": 0.2356,
+      "step": 1617
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.308641975308642e-06,
+      "loss": 0.274,
+      "step": 1618
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.3004115226337455e-06,
+      "loss": 0.286,
+      "step": 1619
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.292181069958848e-06,
+      "loss": 0.3289,
+      "step": 1620
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.283950617283952e-06,
+      "loss": 0.2502,
+      "step": 1621
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.2757201646090535e-06,
+      "loss": 0.3189,
+      "step": 1622
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.267489711934157e-06,
+      "loss": 0.2945,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.2592592592592605e-06,
+      "loss": 0.3027,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.251028806584362e-06,
+      "loss": 0.3287,
+      "step": 1625
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.242798353909466e-06,
+      "loss": 0.4754,
+      "step": 1626
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.2345679012345685e-06,
+      "loss": 0.2683,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.226337448559671e-06,
+      "loss": 0.265,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.218106995884774e-06,
+      "loss": 0.2786,
+      "step": 1629
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.209876543209877e-06,
+      "loss": 0.3648,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.201646090534981e-06,
+      "loss": 0.2753,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.193415637860083e-06,
+      "loss": 0.2185,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.185185185185186e-06,
+      "loss": 0.2811,
+      "step": 1633
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.176954732510289e-06,
+      "loss": 0.216,
+      "step": 1634
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.1687242798353915e-06,
+      "loss": 0.2594,
+      "step": 1635
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 0.2797,
+      "step": 1636
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.152263374485598e-06,
+      "loss": 0.2505,
+      "step": 1637
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.1440329218106995e-06,
+      "loss": 0.2536,
+      "step": 1638
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.135802469135803e-06,
+      "loss": 0.2855,
+      "step": 1639
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.1275720164609065e-06,
+      "loss": 0.266,
+      "step": 1640
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.119341563786008e-06,
+      "loss": 0.2501,
+      "step": 1641
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 0.3099,
+      "step": 1642
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.1028806584362145e-06,
+      "loss": 0.2851,
+      "step": 1643
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.094650205761318e-06,
+      "loss": 0.2726,
+      "step": 1644
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.08641975308642e-06,
+      "loss": 0.2673,
+      "step": 1645
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.078189300411523e-06,
+      "loss": 0.2446,
+      "step": 1646
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.069958847736627e-06,
+      "loss": 0.2295,
+      "step": 1647
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.061728395061729e-06,
+      "loss": 0.3588,
+      "step": 1648
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.053497942386832e-06,
+      "loss": 0.2556,
+      "step": 1649
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.045267489711935e-06,
+      "loss": 0.2297,
+      "step": 1650
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.2432,
+      "step": 1651
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.02880658436214e-06,
+      "loss": 0.2508,
+      "step": 1652
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.020576131687244e-06,
+      "loss": 0.2527,
+      "step": 1653
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.012345679012347e-06,
+      "loss": 0.2569,
+      "step": 1654
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 7.004115226337449e-06,
+      "loss": 0.2654,
+      "step": 1655
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.9958847736625525e-06,
+      "loss": 0.2859,
+      "step": 1656
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.987654320987654e-06,
+      "loss": 0.2399,
+      "step": 1657
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.979423868312758e-06,
+      "loss": 0.2492,
+      "step": 1658
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.9711934156378605e-06,
+      "loss": 0.245,
+      "step": 1659
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.962962962962964e-06,
+      "loss": 0.2688,
+      "step": 1660
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.954732510288067e-06,
+      "loss": 0.2423,
+      "step": 1661
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.946502057613169e-06,
+      "loss": 0.2782,
+      "step": 1662
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.938271604938273e-06,
+      "loss": 0.2659,
+      "step": 1663
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.930041152263375e-06,
+      "loss": 0.3015,
+      "step": 1664
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.921810699588478e-06,
+      "loss": 0.2669,
+      "step": 1665
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 0.2551,
+      "step": 1666
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.9053497942386835e-06,
+      "loss": 0.2698,
+      "step": 1667
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.897119341563786e-06,
+      "loss": 0.2514,
+      "step": 1668
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.2637,
+      "step": 1669
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.880658436213993e-06,
+      "loss": 0.241,
+      "step": 1670
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.872427983539095e-06,
+      "loss": 0.2954,
+      "step": 1671
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.8641975308641985e-06,
+      "loss": 0.2362,
+      "step": 1672
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.8559670781893e-06,
+      "loss": 0.3149,
+      "step": 1673
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.847736625514404e-06,
+      "loss": 0.2798,
+      "step": 1674
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.8395061728395065e-06,
+      "loss": 0.2677,
+      "step": 1675
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.83127572016461e-06,
+      "loss": 0.3053,
+      "step": 1676
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.823045267489713e-06,
+      "loss": 0.2534,
+      "step": 1677
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.814814814814815e-06,
+      "loss": 0.262,
+      "step": 1678
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.806584362139919e-06,
+      "loss": 0.2628,
+      "step": 1679
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.798353909465021e-06,
+      "loss": 0.2769,
+      "step": 1680
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.790123456790124e-06,
+      "loss": 0.2958,
+      "step": 1681
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.781893004115227e-06,
+      "loss": 0.2534,
+      "step": 1682
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.7736625514403295e-06,
+      "loss": 0.3071,
+      "step": 1683
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.765432098765433e-06,
+      "loss": 0.2658,
+      "step": 1684
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.757201646090536e-06,
+      "loss": 0.2977,
+      "step": 1685
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.748971193415639e-06,
+      "loss": 0.2461,
+      "step": 1686
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.740740740740741e-06,
+      "loss": 0.2745,
+      "step": 1687
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.7325102880658445e-06,
+      "loss": 0.2304,
+      "step": 1688
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.724279835390946e-06,
+      "loss": 0.2657,
+      "step": 1689
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.71604938271605e-06,
+      "loss": 0.2386,
+      "step": 1690
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.707818930041153e-06,
+      "loss": 0.2398,
+      "step": 1691
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.699588477366256e-06,
+      "loss": 0.2585,
+      "step": 1692
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.691358024691359e-06,
+      "loss": 0.2314,
+      "step": 1693
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.683127572016461e-06,
+      "loss": 0.277,
+      "step": 1694
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.674897119341565e-06,
+      "loss": 0.2823,
+      "step": 1695
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2378,
+      "step": 1696
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 6.65843621399177e-06,
+      "loss": 0.2814,
+      "step": 1697
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.650205761316873e-06,
+      "loss": 0.2841,
+      "step": 1698
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.6419753086419755e-06,
+      "loss": 0.2474,
+      "step": 1699
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.633744855967079e-06,
+      "loss": 0.2631,
+      "step": 1700
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.625514403292182e-06,
+      "loss": 0.2627,
+      "step": 1701
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.617283950617285e-06,
+      "loss": 0.2239,
+      "step": 1702
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.609053497942387e-06,
+      "loss": 0.2745,
+      "step": 1703
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.6008230452674905e-06,
+      "loss": 0.2687,
+      "step": 1704
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.592592592592592e-06,
+      "loss": 0.2776,
+      "step": 1705
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.584362139917696e-06,
+      "loss": 0.2357,
+      "step": 1706
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.576131687242799e-06,
+      "loss": 0.366,
+      "step": 1707
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.567901234567902e-06,
+      "loss": 0.2441,
+      "step": 1708
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.559670781893005e-06,
+      "loss": 0.2377,
+      "step": 1709
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 6.551440329218107e-06,
+      "loss": 0.2497,
+      "step": 1710
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.543209876543211e-06,
+      "loss": 0.2438,
+      "step": 1711
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.534979423868313e-06,
+      "loss": 0.2606,
+      "step": 1712
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.526748971193416e-06,
+      "loss": 0.291,
+      "step": 1713
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.51851851851852e-06,
+      "loss": 0.2583,
+      "step": 1714
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.5102880658436215e-06,
+      "loss": 0.283,
+      "step": 1715
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.502057613168725e-06,
+      "loss": 0.2787,
+      "step": 1716
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.493827160493828e-06,
+      "loss": 0.2844,
+      "step": 1717
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.485596707818931e-06,
+      "loss": 0.2465,
+      "step": 1718
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.477366255144033e-06,
+      "loss": 0.2394,
+      "step": 1719
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.4691358024691365e-06,
+      "loss": 0.2609,
+      "step": 1720
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.460905349794238e-06,
+      "loss": 0.2292,
+      "step": 1721
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.452674897119342e-06,
+      "loss": 0.2665,
+      "step": 1722
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 0.2704,
+      "step": 1723
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.436213991769548e-06,
+      "loss": 0.3177,
+      "step": 1724
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.427983539094651e-06,
+      "loss": 0.2499,
+      "step": 1725
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 0.2739,
+      "step": 1726
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.411522633744857e-06,
+      "loss": 0.2745,
+      "step": 1727
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.403292181069959e-06,
+      "loss": 0.2273,
+      "step": 1728
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.395061728395062e-06,
+      "loss": 0.243,
+      "step": 1729
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.386831275720166e-06,
+      "loss": 0.2467,
+      "step": 1730
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.3786008230452675e-06,
+      "loss": 0.304,
+      "step": 1731
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.370370370370371e-06,
+      "loss": 0.2474,
+      "step": 1732
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.362139917695474e-06,
+      "loss": 0.2651,
+      "step": 1733
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.353909465020577e-06,
+      "loss": 0.2351,
+      "step": 1734
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 6.345679012345679e-06,
+      "loss": 0.2662,
+      "step": 1735
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.3374485596707825e-06,
+      "loss": 0.2575,
+      "step": 1736
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.329218106995886e-06,
+      "loss": 0.2544,
+      "step": 1737
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.320987654320988e-06,
+      "loss": 0.2951,
+      "step": 1738
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.312757201646091e-06,
+      "loss": 0.2719,
+      "step": 1739
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.304526748971194e-06,
+      "loss": 0.2594,
+      "step": 1740
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.2644,
+      "step": 1741
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.288065843621399e-06,
+      "loss": 0.2658,
+      "step": 1742
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.279835390946503e-06,
+      "loss": 0.2566,
+      "step": 1743
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.271604938271606e-06,
+      "loss": 0.2543,
+      "step": 1744
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.263374485596708e-06,
+      "loss": 0.2477,
+      "step": 1745
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.255144032921812e-06,
+      "loss": 0.2705,
+      "step": 1746
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 6.2469135802469135e-06,
+      "loss": 0.243,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.238683127572017e-06,
+      "loss": 0.2465,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.23045267489712e-06,
+      "loss": 0.2333,
+      "step": 1749
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 0.2198,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.213991769547325e-06,
+      "loss": 0.2459,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.2057613168724285e-06,
+      "loss": 0.2639,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.197530864197532e-06,
+      "loss": 0.2528,
+      "step": 1753
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.189300411522634e-06,
+      "loss": 0.2856,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.181069958847737e-06,
+      "loss": 0.2288,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.17283950617284e-06,
+      "loss": 0.2705,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.164609053497943e-06,
+      "loss": 0.2938,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.156378600823045e-06,
+      "loss": 0.2778,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.148148148148149e-06,
+      "loss": 0.2585,
+      "step": 1759
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 6.139917695473252e-06,
+      "loss": 0.2379,
+      "step": 1760
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.131687242798354e-06,
+      "loss": 0.2672,
+      "step": 1761
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.123456790123458e-06,
+      "loss": 0.2544,
+      "step": 1762
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.1152263374485595e-06,
+      "loss": 0.2563,
+      "step": 1763
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.106995884773663e-06,
+      "loss": 0.2344,
+      "step": 1764
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.098765432098766e-06,
+      "loss": 0.311,
+      "step": 1765
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.090534979423869e-06,
+      "loss": 0.2623,
+      "step": 1766
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.082304526748972e-06,
+      "loss": 0.2345,
+      "step": 1767
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.0740740740740745e-06,
+      "loss": 0.2498,
+      "step": 1768
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.065843621399178e-06,
+      "loss": 0.2848,
+      "step": 1769
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.05761316872428e-06,
+      "loss": 0.2318,
+      "step": 1770
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.049382716049383e-06,
+      "loss": 0.258,
+      "step": 1771
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 6.041152263374486e-06,
+      "loss": 0.3022,
+      "step": 1772
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.032921810699589e-06,
+      "loss": 0.2558,
+      "step": 1773
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.024691358024692e-06,
+      "loss": 0.3015,
+      "step": 1774
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.016460905349795e-06,
+      "loss": 0.2086,
+      "step": 1775
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6.008230452674898e-06,
+      "loss": 0.2304,
+      "step": 1776
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 6e-06,
+      "loss": 0.2591,
+      "step": 1777
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.991769547325104e-06,
+      "loss": 0.2306,
+      "step": 1778
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.9835390946502055e-06,
+      "loss": 0.2252,
+      "step": 1779
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.975308641975309e-06,
+      "loss": 0.2824,
+      "step": 1780
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.967078189300412e-06,
+      "loss": 0.2137,
+      "step": 1781
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.958847736625515e-06,
+      "loss": 0.2446,
+      "step": 1782
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.950617283950618e-06,
+      "loss": 0.2717,
+      "step": 1783
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.9423868312757205e-06,
+      "loss": 0.2344,
+      "step": 1784
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.934156378600824e-06,
+      "loss": 0.2899,
+      "step": 1785
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.302,
+      "step": 1786
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.917695473251029e-06,
+      "loss": 0.2488,
+      "step": 1787
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.909465020576132e-06,
+      "loss": 0.2415,
+      "step": 1788
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.901234567901235e-06,
+      "loss": 0.2803,
+      "step": 1789
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.893004115226338e-06,
+      "loss": 0.1996,
+      "step": 1790
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.884773662551441e-06,
+      "loss": 0.2424,
+      "step": 1791
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.876543209876544e-06,
+      "loss": 0.222,
+      "step": 1792
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.868312757201646e-06,
+      "loss": 0.3162,
+      "step": 1793
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.86008230452675e-06,
+      "loss": 0.2855,
+      "step": 1794
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.8518518518518515e-06,
+      "loss": 0.275,
+      "step": 1795
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.843621399176955e-06,
+      "loss": 0.2865,
+      "step": 1796
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.8353909465020585e-06,
+      "loss": 0.2369,
+      "step": 1797
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.827160493827161e-06,
+      "loss": 0.2382,
+      "step": 1798
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.818930041152264e-06,
+      "loss": 0.27,
+      "step": 1799
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.8106995884773665e-06,
+      "loss": 0.2703,
+      "step": 1800
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.80246913580247e-06,
+      "loss": 0.25,
+      "step": 1801
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.794238683127572e-06,
+      "loss": 0.2547,
+      "step": 1802
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.786008230452675e-06,
+      "loss": 0.2467,
+      "step": 1803
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 0.2602,
+      "step": 1804
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.769547325102881e-06,
+      "loss": 0.2528,
+      "step": 1805
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.761316872427984e-06,
+      "loss": 0.2527,
+      "step": 1806
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.753086419753087e-06,
+      "loss": 0.2861,
+      "step": 1807
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.74485596707819e-06,
+      "loss": 0.2252,
+      "step": 1808
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.736625514403292e-06,
+      "loss": 0.2899,
+      "step": 1809
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.728395061728396e-06,
+      "loss": 0.2609,
+      "step": 1810
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.720164609053498e-06,
+      "loss": 0.2716,
+      "step": 1811
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.711934156378601e-06,
+      "loss": 0.2581,
+      "step": 1812
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.7037037037037045e-06,
+      "loss": 0.2699,
+      "step": 1813
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.695473251028807e-06,
+      "loss": 0.2363,
+      "step": 1814
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.68724279835391e-06,
+      "loss": 0.2789,
+      "step": 1815
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.2448,
+      "step": 1816
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.670781893004116e-06,
+      "loss": 0.2535,
+      "step": 1817
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.662551440329218e-06,
+      "loss": 0.2712,
+      "step": 1818
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.654320987654321e-06,
+      "loss": 0.2785,
+      "step": 1819
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.646090534979425e-06,
+      "loss": 0.267,
+      "step": 1820
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.6378600823045275e-06,
+      "loss": 0.272,
+      "step": 1821
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.62962962962963e-06,
+      "loss": 0.2355,
+      "step": 1822
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.621399176954733e-06,
+      "loss": 0.2676,
+      "step": 1823
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.613168724279836e-06,
+      "loss": 0.2593,
+      "step": 1824
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.604938271604938e-06,
+      "loss": 0.2324,
+      "step": 1825
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.596707818930042e-06,
+      "loss": 0.2456,
+      "step": 1826
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.588477366255145e-06,
+      "loss": 0.2471,
+      "step": 1827
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.580246913580247e-06,
+      "loss": 0.2658,
+      "step": 1828
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5720164609053505e-06,
+      "loss": 0.2509,
+      "step": 1829
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.563786008230453e-06,
+      "loss": 0.2893,
+      "step": 1830
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.2642,
+      "step": 1831
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5473251028806585e-06,
+      "loss": 0.2983,
+      "step": 1832
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.539094650205762e-06,
+      "loss": 0.2747,
+      "step": 1833
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.530864197530864e-06,
+      "loss": 0.2233,
+      "step": 1834
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.522633744855967e-06,
+      "loss": 0.2528,
+      "step": 1835
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.514403292181071e-06,
+      "loss": 0.2838,
+      "step": 1836
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5061728395061735e-06,
+      "loss": 0.2569,
+      "step": 1837
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.497942386831276e-06,
+      "loss": 0.2461,
+      "step": 1838
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.489711934156379e-06,
+      "loss": 0.2687,
+      "step": 1839
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.481481481481482e-06,
+      "loss": 0.2651,
+      "step": 1840
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.473251028806584e-06,
+      "loss": 0.2372,
+      "step": 1841
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.465020576131688e-06,
+      "loss": 0.2937,
+      "step": 1842
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.456790123456791e-06,
+      "loss": 0.2497,
+      "step": 1843
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.448559670781893e-06,
+      "loss": 0.27,
+      "step": 1844
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.4403292181069965e-06,
+      "loss": 0.2497,
+      "step": 1845
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.2759,
+      "step": 1846
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.423868312757203e-06,
+      "loss": 0.2654,
+      "step": 1847
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.4156378600823045e-06,
+      "loss": 0.2582,
+      "step": 1848
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.407407407407408e-06,
+      "loss": 0.2627,
+      "step": 1849
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.3991769547325115e-06,
+      "loss": 0.2594,
+      "step": 1850
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.390946502057613e-06,
+      "loss": 0.2633,
+      "step": 1851
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.382716049382717e-06,
+      "loss": 0.2781,
+      "step": 1852
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.3744855967078195e-06,
+      "loss": 0.2156,
+      "step": 1853
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.366255144032922e-06,
+      "loss": 0.2623,
+      "step": 1854
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.358024691358025e-06,
+      "loss": 0.2532,
+      "step": 1855
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.349794238683128e-06,
+      "loss": 0.2402,
+      "step": 1856
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.341563786008232e-06,
+      "loss": 0.2577,
+      "step": 1857
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.2353,
+      "step": 1858
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.325102880658437e-06,
+      "loss": 0.2861,
+      "step": 1859
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.316872427983539e-06,
+      "loss": 0.2117,
+      "step": 1860
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.3086419753086425e-06,
+      "loss": 0.2399,
+      "step": 1861
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.300411522633745e-06,
+      "loss": 0.251,
+      "step": 1862
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.292181069958849e-06,
+      "loss": 0.2707,
+      "step": 1863
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.2839506172839505e-06,
+      "loss": 0.2547,
+      "step": 1864
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.275720164609054e-06,
+      "loss": 0.2431,
+      "step": 1865
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.2674897119341575e-06,
+      "loss": 0.26,
+      "step": 1866
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.259259259259259e-06,
+      "loss": 0.27,
+      "step": 1867
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.251028806584363e-06,
+      "loss": 0.2737,
+      "step": 1868
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.2427983539094655e-06,
+      "loss": 0.2615,
+      "step": 1869
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.234567901234568e-06,
+      "loss": 0.2572,
+      "step": 1870
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.226337448559671e-06,
+      "loss": 0.2597,
+      "step": 1871
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.218106995884774e-06,
+      "loss": 0.2537,
+      "step": 1872
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.209876543209878e-06,
+      "loss": 0.252,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.20164609053498e-06,
+      "loss": 0.2793,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.193415637860083e-06,
+      "loss": 0.2711,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.2445,
+      "step": 1876
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1769547325102885e-06,
+      "loss": 0.266,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.168724279835391e-06,
+      "loss": 0.2833,
+      "step": 1878
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.160493827160495e-06,
+      "loss": 0.2508,
+      "step": 1879
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.152263374485597e-06,
+      "loss": 0.2998,
+      "step": 1880
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1440329218107e-06,
+      "loss": 0.2842,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1358024691358035e-06,
+      "loss": 0.2595,
+      "step": 1882
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.127572016460905e-06,
+      "loss": 0.2716,
+      "step": 1883
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.119341563786009e-06,
+      "loss": 0.2879,
+      "step": 1884
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 0.2694,
+      "step": 1885
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.102880658436214e-06,
+      "loss": 0.2453,
+      "step": 1886
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.094650205761317e-06,
+      "loss": 0.2823,
+      "step": 1887
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.08641975308642e-06,
+      "loss": 0.2519,
+      "step": 1888
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.078189300411524e-06,
+      "loss": 0.2511,
+      "step": 1889
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.069958847736626e-06,
+      "loss": 0.2627,
+      "step": 1890
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.061728395061729e-06,
+      "loss": 0.3092,
+      "step": 1891
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.053497942386831e-06,
+      "loss": 0.2493,
+      "step": 1892
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.0452674897119345e-06,
+      "loss": 0.271,
+      "step": 1893
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.037037037037037e-06,
+      "loss": 0.2664,
+      "step": 1894
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.028806584362141e-06,
+      "loss": 0.2246,
+      "step": 1895
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.020576131687243e-06,
+      "loss": 0.2533,
+      "step": 1896
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.012345679012346e-06,
+      "loss": 0.2793,
+      "step": 1897
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.0041152263374495e-06,
+      "loss": 0.2333,
+      "step": 1898
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.995884773662552e-06,
+      "loss": 0.2839,
+      "step": 1899
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.987654320987655e-06,
+      "loss": 0.2498,
+      "step": 1900
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9794238683127575e-06,
+      "loss": 0.2473,
+      "step": 1901
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.97119341563786e-06,
+      "loss": 0.2323,
+      "step": 1902
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.962962962962964e-06,
+      "loss": 0.3006,
+      "step": 1903
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.954732510288066e-06,
+      "loss": 0.2506,
+      "step": 1904
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.946502057613169e-06,
+      "loss": 0.2769,
+      "step": 1905
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.2464,
+      "step": 1906
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.930041152263375e-06,
+      "loss": 0.2727,
+      "step": 1907
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.921810699588478e-06,
+      "loss": 0.2357,
+      "step": 1908
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.9135802469135805e-06,
+      "loss": 0.2728,
+      "step": 1909
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.905349794238683e-06,
+      "loss": 0.2477,
+      "step": 1910
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.897119341563787e-06,
+      "loss": 0.2493,
+      "step": 1911
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 0.284,
+      "step": 1912
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.880658436213992e-06,
+      "loss": 0.2573,
+      "step": 1913
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.8724279835390955e-06,
+      "loss": 0.2838,
+      "step": 1914
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.864197530864198e-06,
+      "loss": 0.2588,
+      "step": 1915
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.855967078189301e-06,
+      "loss": 0.2998,
+      "step": 1916
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.8477366255144035e-06,
+      "loss": 0.2665,
+      "step": 1917
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.839506172839506e-06,
+      "loss": 0.2754,
+      "step": 1918
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.83127572016461e-06,
+      "loss": 0.2484,
+      "step": 1919
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.823045267489712e-06,
+      "loss": 0.2821,
+      "step": 1920
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.2555,
+      "step": 1921
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.8065843621399185e-06,
+      "loss": 0.2807,
+      "step": 1922
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.798353909465021e-06,
+      "loss": 0.2413,
+      "step": 1923
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.790123456790124e-06,
+      "loss": 0.2827,
+      "step": 1924
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.7818930041152265e-06,
+      "loss": 0.258,
+      "step": 1925
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.773662551440329e-06,
+      "loss": 0.2261,
+      "step": 1926
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.765432098765433e-06,
+      "loss": 0.243,
+      "step": 1927
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.757201646090535e-06,
+      "loss": 0.2352,
+      "step": 1928
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.748971193415639e-06,
+      "loss": 0.24,
+      "step": 1929
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.7407407407407415e-06,
+      "loss": 0.2931,
+      "step": 1930
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.732510288065844e-06,
+      "loss": 0.2676,
+      "step": 1931
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.724279835390947e-06,
+      "loss": 0.3159,
+      "step": 1932
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.7160493827160495e-06,
+      "loss": 0.2811,
+      "step": 1933
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.707818930041152e-06,
+      "loss": 0.2532,
+      "step": 1934
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.699588477366256e-06,
+      "loss": 0.2756,
+      "step": 1935
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.27,
+      "step": 1936
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.683127572016462e-06,
+      "loss": 0.2452,
+      "step": 1937
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.6748971193415645e-06,
+      "loss": 0.2569,
+      "step": 1938
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 0.2983,
+      "step": 1939
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.65843621399177e-06,
+      "loss": 0.2712,
+      "step": 1940
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.6502057613168725e-06,
+      "loss": 0.382,
+      "step": 1941
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.641975308641975e-06,
+      "loss": 0.2679,
+      "step": 1942
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.633744855967079e-06,
+      "loss": 0.2697,
+      "step": 1943
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.625514403292181e-06,
+      "loss": 0.2801,
+      "step": 1944
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.617283950617285e-06,
+      "loss": 0.2229,
+      "step": 1945
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.6090534979423875e-06,
+      "loss": 0.2478,
+      "step": 1946
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.60082304526749e-06,
+      "loss": 0.2377,
+      "step": 1947
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.592592592592593e-06,
+      "loss": 0.3088,
+      "step": 1948
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.5843621399176955e-06,
+      "loss": 0.2409,
+      "step": 1949
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.576131687242798e-06,
+      "loss": 0.2787,
+      "step": 1950
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.567901234567902e-06,
+      "loss": 0.2176,
+      "step": 1951
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.559670781893004e-06,
+      "loss": 0.3024,
+      "step": 1952
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.551440329218108e-06,
+      "loss": 0.2576,
+      "step": 1953
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.5432098765432105e-06,
+      "loss": 0.259,
+      "step": 1954
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.534979423868313e-06,
+      "loss": 0.288,
+      "step": 1955
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.526748971193416e-06,
+      "loss": 0.2489,
+      "step": 1956
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.5185185185185185e-06,
+      "loss": 0.2566,
+      "step": 1957
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.510288065843621e-06,
+      "loss": 0.2359,
+      "step": 1958
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.502057613168725e-06,
+      "loss": 0.2746,
+      "step": 1959
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.493827160493827e-06,
+      "loss": 0.2377,
+      "step": 1960
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.485596707818931e-06,
+      "loss": 0.2431,
+      "step": 1961
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.4773662551440335e-06,
+      "loss": 0.2478,
+      "step": 1962
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.469135802469136e-06,
+      "loss": 0.288,
+      "step": 1963
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.460905349794239e-06,
+      "loss": 0.2531,
+      "step": 1964
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.4526748971193415e-06,
+      "loss": 0.24,
+      "step": 1965
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.2583,
+      "step": 1966
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.436213991769548e-06,
+      "loss": 0.3077,
+      "step": 1967
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.42798353909465e-06,
+      "loss": 0.2674,
+      "step": 1968
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.419753086419754e-06,
+      "loss": 0.2602,
+      "step": 1969
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.4115226337448565e-06,
+      "loss": 0.2894,
+      "step": 1970
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.403292181069959e-06,
+      "loss": 0.3005,
+      "step": 1971
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.395061728395062e-06,
+      "loss": 0.2807,
+      "step": 1972
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.3868312757201645e-06,
+      "loss": 0.285,
+      "step": 1973
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.378600823045268e-06,
+      "loss": 0.2473,
+      "step": 1974
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.370370370370371e-06,
+      "loss": 0.2729,
+      "step": 1975
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.362139917695473e-06,
+      "loss": 0.2445,
+      "step": 1976
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.353909465020577e-06,
+      "loss": 0.2545,
+      "step": 1977
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.3456790123456795e-06,
+      "loss": 0.2791,
+      "step": 1978
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.337448559670782e-06,
+      "loss": 0.2772,
+      "step": 1979
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.329218106995885e-06,
+      "loss": 0.2498,
+      "step": 1980
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.3209876543209875e-06,
+      "loss": 0.2375,
+      "step": 1981
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.312757201646091e-06,
+      "loss": 0.269,
+      "step": 1982
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.304526748971194e-06,
+      "loss": 0.2423,
+      "step": 1983
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.296296296296296e-06,
+      "loss": 0.277,
+      "step": 1984
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.2880658436214e-06,
+      "loss": 0.2989,
+      "step": 1985
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.2798353909465025e-06,
+      "loss": 0.2781,
+      "step": 1986
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.271604938271605e-06,
+      "loss": 0.2403,
+      "step": 1987
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.263374485596708e-06,
+      "loss": 0.2797,
+      "step": 1988
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2551440329218105e-06,
+      "loss": 0.3414,
+      "step": 1989
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.246913580246914e-06,
+      "loss": 0.273,
+      "step": 1990
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.238683127572017e-06,
+      "loss": 0.2491,
+      "step": 1991
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.23045267489712e-06,
+      "loss": 0.2778,
+      "step": 1992
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.2598,
+      "step": 1993
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.2139917695473255e-06,
+      "loss": 0.2426,
+      "step": 1994
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.205761316872428e-06,
+      "loss": 0.2425,
+      "step": 1995
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.2548,
+      "step": 1996
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.189300411522634e-06,
+      "loss": 0.2949,
+      "step": 1997
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.181069958847737e-06,
+      "loss": 0.2693,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.17283950617284e-06,
+      "loss": 0.2733,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.164609053497943e-06,
+      "loss": 0.2524,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2506,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "total_flos": 0.0,
+  "trial_name": null,
+  "trial_params": null
+}