diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13113 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 2182,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.633587786259542e-08,
+      "loss": 5.739,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5267175572519085e-07,
+      "loss": 5.8495,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2900763358778629e-07,
+      "loss": 5.654,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.053435114503817e-07,
+      "loss": 5.6536,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8167938931297716e-07,
+      "loss": 5.8335,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5801526717557257e-07,
+      "loss": 5.8259,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.34351145038168e-07,
+      "loss": 5.7934,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.106870229007634e-07,
+      "loss": 5.9135,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.870229007633589e-07,
+      "loss": 5.8733,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.633587786259543e-07,
+      "loss": 5.8098,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.396946564885497e-07,
+      "loss": 5.9146,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.160305343511451e-07,
+      "loss": 5.9152,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.923664122137404e-07,
+      "loss": 5.7155,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.068702290076336e-06,
+      "loss": 5.8148,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1450381679389313e-06,
+      "loss": 5.7195,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2213740458015268e-06,
+      "loss": 5.4608,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.297709923664122e-06,
+      "loss": 5.4997,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3740458015267178e-06,
+      "loss": 5.4569,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.450381679389313e-06,
+      "loss": 5.1754,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5267175572519086e-06,
+      "loss": 5.2463,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.603053435114504e-06,
+      "loss": 5.3135,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6793893129770995e-06,
+      "loss": 4.9404,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7557251908396948e-06,
+      "loss": 4.719,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8320610687022903e-06,
+      "loss": 4.6484,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.908396946564886e-06,
+      "loss": 4.5135,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.984732824427481e-06,
+      "loss": 4.5846,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0610687022900764e-06,
+      "loss": 4.3673,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.137404580152672e-06,
+      "loss": 4.3592,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.2137404580152674e-06,
+      "loss": 4.2058,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.2900763358778625e-06,
+      "loss": 4.146,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.3664122137404585e-06,
+      "loss": 3.9923,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.4427480916030536e-06,
+      "loss": 3.9277,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.5190839694656487e-06,
+      "loss": 3.8486,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.595419847328244e-06,
+      "loss": 3.7802,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.67175572519084e-06,
+      "loss": 3.7933,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.7480916030534356e-06,
+      "loss": 3.7421,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.824427480916031e-06,
+      "loss": 3.6219,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.900763358778626e-06,
+      "loss": 3.6658,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9770992366412218e-06,
+      "loss": 3.6386,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.0534351145038173e-06,
+      "loss": 3.5436,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.129770992366413e-06,
+      "loss": 3.4166,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.206106870229008e-06,
+      "loss": 3.4337,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.2824427480916034e-06,
+      "loss": 3.3053,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.358778625954199e-06,
+      "loss": 3.3307,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.4351145038167944e-06,
+      "loss": 3.3892,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.5114503816793895e-06,
+      "loss": 3.217,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.587786259541985e-06,
+      "loss": 3.1355,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.6641221374045806e-06,
+      "loss": 2.9934,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.740458015267176e-06,
+      "loss": 3.002,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.816793893129772e-06,
+      "loss": 2.9986,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.893129770992366e-06,
+      "loss": 2.8566,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 3.969465648854962e-06,
+      "loss": 2.7882,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.045801526717557e-06,
+      "loss": 2.8063,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.122137404580153e-06,
+      "loss": 2.8027,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.198473282442748e-06,
+      "loss": 2.7,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.274809160305344e-06,
+      "loss": 2.6548,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.351145038167939e-06,
+      "loss": 2.6475,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.427480916030535e-06,
+      "loss": 2.516,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.5038167938931296e-06,
+      "loss": 2.4829,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.580152671755725e-06,
+      "loss": 2.4672,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.656488549618321e-06,
+      "loss": 2.4217,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.732824427480917e-06,
+      "loss": 2.5097,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8091603053435125e-06,
+      "loss": 2.4857,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.885496183206107e-06,
+      "loss": 2.3667,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.961832061068703e-06,
+      "loss": 2.3572,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.038167938931297e-06,
+      "loss": 2.3964,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.114503816793893e-06,
+      "loss": 2.4392,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.190839694656488e-06,
+      "loss": 2.3106,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.267175572519084e-06,
+      "loss": 2.3751,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.34351145038168e-06,
+      "loss": 2.3084,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.419847328244276e-06,
+      "loss": 2.3557,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.496183206106871e-06,
+      "loss": 2.3583,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.572519083969467e-06,
+      "loss": 2.3424,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.648854961832062e-06,
+      "loss": 2.313,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.725190839694656e-06,
+      "loss": 2.2541,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.801526717557252e-06,
+      "loss": 2.2945,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.877862595419848e-06,
+      "loss": 2.2796,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9541984732824435e-06,
+      "loss": 2.2,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.030534351145039e-06,
+      "loss": 2.2954,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.1068702290076346e-06,
+      "loss": 2.209,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.18320610687023e-06,
+      "loss": 2.2771,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.259541984732826e-06,
+      "loss": 2.2492,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.335877862595419e-06,
+      "loss": 2.1282,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.412213740458016e-06,
+      "loss": 2.1563,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.488549618320611e-06,
+      "loss": 2.1825,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.564885496183207e-06,
+      "loss": 2.1714,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.641221374045802e-06,
+      "loss": 2.2773,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.717557251908398e-06,
+      "loss": 2.174,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.793893129770993e-06,
+      "loss": 2.213,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.870229007633589e-06,
+      "loss": 2.1664,
+      "step": 90
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.946564885496184e-06,
+      "loss": 2.1624,
+      "step": 91
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.022900763358779e-06,
+      "loss": 2.1852,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.0992366412213746e-06,
+      "loss": 2.2014,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.17557251908397e-06,
+      "loss": 2.1733,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.251908396946566e-06,
+      "loss": 2.1852,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.328244274809161e-06,
+      "loss": 2.1734,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.404580152671757e-06,
+      "loss": 2.1363,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.480916030534352e-06,
+      "loss": 2.1177,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.557251908396948e-06,
+      "loss": 2.1804,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.633587786259543e-06,
+      "loss": 2.1295,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.709923664122137e-06,
+      "loss": 2.1857,
+      "step": 101
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.786259541984733e-06,
+      "loss": 2.2074,
+      "step": 102
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.862595419847328e-06,
+      "loss": 2.1203,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 7.938931297709924e-06,
+      "loss": 2.0945,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.015267175572519e-06,
+      "loss": 2.1206,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.091603053435115e-06,
+      "loss": 2.1131,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.16793893129771e-06,
+      "loss": 2.1062,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.244274809160306e-06,
+      "loss": 2.1596,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.320610687022901e-06,
+      "loss": 2.105,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.396946564885497e-06,
+      "loss": 2.0863,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.473282442748092e-06,
+      "loss": 2.1503,
+      "step": 111
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.549618320610688e-06,
+      "loss": 2.145,
+      "step": 112
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.625954198473283e-06,
+      "loss": 2.087,
+      "step": 113
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.702290076335879e-06,
+      "loss": 1.9606,
+      "step": 114
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 8.778625954198474e-06,
+      "loss": 2.0476,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 8.85496183206107e-06,
+      "loss": 2.1205,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 8.931297709923665e-06,
+      "loss": 2.1009,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.007633587786259e-06,
+      "loss": 2.103,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.083969465648855e-06,
+      "loss": 2.098,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.16030534351145e-06,
+      "loss": 2.0877,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.236641221374046e-06,
+      "loss": 2.0249,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.312977099236641e-06,
+      "loss": 2.099,
+      "step": 122
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.389312977099237e-06,
+      "loss": 2.0018,
+      "step": 123
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.465648854961834e-06,
+      "loss": 2.0495,
+      "step": 124
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.54198473282443e-06,
+      "loss": 2.2098,
+      "step": 125
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.618320610687025e-06,
+      "loss": 2.0704,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.694656488549619e-06,
+      "loss": 2.0752,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.770992366412214e-06,
+      "loss": 2.0531,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.84732824427481e-06,
+      "loss": 2.0761,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 2.0345,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 2.0803,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0076335877862595e-05,
+      "loss": 2.0349,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0152671755725192e-05,
+      "loss": 2.0855,
+      "step": 133
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0229007633587786e-05,
+      "loss": 2.0987,
+      "step": 134
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0305343511450383e-05,
+      "loss": 2.0149,
+      "step": 135
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0381679389312977e-05,
+      "loss": 2.0191,
+      "step": 136
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0458015267175574e-05,
+      "loss": 2.0213,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0534351145038168e-05,
+      "loss": 2.0917,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0610687022900765e-05,
+      "loss": 2.0044,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.068702290076336e-05,
+      "loss": 1.9405,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0763358778625954e-05,
+      "loss": 2.0527,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0839694656488552e-05,
+      "loss": 1.9537,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0916030534351145e-05,
+      "loss": 2.0003,
+      "step": 143
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.0992366412213743e-05,
+      "loss": 1.9578,
+      "step": 144
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1068702290076336e-05,
+      "loss": 2.0776,
+      "step": 145
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1145038167938934e-05,
+      "loss": 2.0947,
+      "step": 146
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1221374045801527e-05,
+      "loss": 2.0241,
+      "step": 147
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1297709923664125e-05,
+      "loss": 1.9945,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1374045801526718e-05,
+      "loss": 2.0089,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1450381679389312e-05,
+      "loss": 2.0856,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.152671755725191e-05,
+      "loss": 2.0353,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1603053435114503e-05,
+      "loss": 2.0204,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.16793893129771e-05,
+      "loss": 2.0126,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1755725190839696e-05,
+      "loss": 2.0857,
+      "step": 154
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1832061068702292e-05,
+      "loss": 1.9218,
+      "step": 155
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1908396946564887e-05,
+      "loss": 2.0275,
+      "step": 156
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.1984732824427483e-05,
+      "loss": 1.9954,
+      "step": 157
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.2061068702290078e-05,
+      "loss": 2.045,
+      "step": 158
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2137404580152672e-05,
+      "loss": 1.9357,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2213740458015269e-05,
+      "loss": 1.9626,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2290076335877863e-05,
+      "loss": 1.9855,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.236641221374046e-05,
+      "loss": 1.8804,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2442748091603054e-05,
+      "loss": 1.8601,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2519083969465651e-05,
+      "loss": 1.9816,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2595419847328245e-05,
+      "loss": 1.9446,
+      "step": 165
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2671755725190839e-05,
+      "loss": 2.1042,
+      "step": 166
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2748091603053436e-05,
+      "loss": 1.9493,
+      "step": 167
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2824427480916032e-05,
+      "loss": 1.9244,
+      "step": 168
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.2900763358778627e-05,
+      "loss": 2.0119,
+      "step": 169
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.2977099236641223e-05,
+      "loss": 1.9248,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3053435114503818e-05,
+      "loss": 1.9271,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3129770992366414e-05,
+      "loss": 1.9563,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3206106870229009e-05,
+      "loss": 2.0182,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3282442748091605e-05,
+      "loss": 1.9954,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3358778625954198e-05,
+      "loss": 1.9741,
+      "step": 175
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3435114503816796e-05,
+      "loss": 2.0293,
+      "step": 176
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.351145038167939e-05,
+      "loss": 2.0193,
+      "step": 177
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3587786259541987e-05,
+      "loss": 2.0142,
+      "step": 178
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.366412213740458e-05,
+      "loss": 1.9701,
+      "step": 179
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.3740458015267178e-05,
+      "loss": 1.8442,
+      "step": 180
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.3816793893129772e-05,
+      "loss": 1.9596,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.3893129770992369e-05,
+      "loss": 1.9169,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.3969465648854963e-05,
+      "loss": 2.0052,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4045801526717558e-05,
+      "loss": 1.8829,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4122137404580154e-05,
+      "loss": 1.8761,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4198473282442749e-05,
+      "loss": 2.0034,
+      "step": 186
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4274809160305345e-05,
+      "loss": 2.01,
+      "step": 187
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.435114503816794e-05,
+      "loss": 1.9917,
+      "step": 188
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4427480916030536e-05,
+      "loss": 1.7963,
+      "step": 189
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.4503816793893131e-05,
+      "loss": 1.9143,
+      "step": 190
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4580152671755727e-05,
+      "loss": 1.9203,
+      "step": 191
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4656488549618322e-05,
+      "loss": 1.9458,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4732824427480916e-05,
+      "loss": 1.8744,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4809160305343513e-05,
+      "loss": 1.9719,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4885496183206107e-05,
+      "loss": 1.9789,
+      "step": 195
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.4961832061068704e-05,
+      "loss": 1.9464,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5038167938931298e-05,
+      "loss": 1.9092,
+      "step": 197
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5114503816793895e-05,
+      "loss": 1.9529,
+      "step": 198
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.519083969465649e-05,
+      "loss": 2.0258,
+      "step": 199
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5267175572519086e-05,
+      "loss": 1.8367,
+      "step": 200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5343511450381682e-05,
+      "loss": 1.9613,
+      "step": 201
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5419847328244274e-05,
+      "loss": 1.8596,
+      "step": 202
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5496183206106873e-05,
+      "loss": 1.9765,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5572519083969465e-05,
+      "loss": 1.9457,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5648854961832064e-05,
+      "loss": 1.8291,
+      "step": 205
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5725190839694656e-05,
+      "loss": 1.8676,
+      "step": 206
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5801526717557255e-05,
+      "loss": 1.865,
+      "step": 207
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5877862595419847e-05,
+      "loss": 1.9707,
+      "step": 208
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5954198473282446e-05,
+      "loss": 1.9273,
+      "step": 209
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6030534351145038e-05,
+      "loss": 2.0137,
+      "step": 210
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.6106870229007634e-05,
+      "loss": 2.0215,
+      "step": 211
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.618320610687023e-05,
+      "loss": 1.8915,
+      "step": 212
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6259541984732825e-05,
+      "loss": 1.9244,
+      "step": 213
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.633587786259542e-05,
+      "loss": 1.9215,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6412213740458016e-05,
+      "loss": 1.9328,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.648854961832061e-05,
+      "loss": 1.9855,
+      "step": 216
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6564885496183207e-05,
+      "loss": 1.9385,
+      "step": 217
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6641221374045802e-05,
+      "loss": 1.8384,
+      "step": 218
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6717557251908398e-05,
+      "loss": 2.0026,
+      "step": 219
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6793893129770993e-05,
+      "loss": 1.9288,
+      "step": 220
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.687022900763359e-05,
+      "loss": 1.8725,
+      "step": 221
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.6946564885496184e-05,
+      "loss": 1.9284,
+      "step": 222
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.702290076335878e-05,
+      "loss": 1.9289,
+      "step": 223
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7099236641221375e-05,
+      "loss": 2.0223,
+      "step": 224
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.717557251908397e-05,
+      "loss": 1.8537,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7251908396946566e-05,
+      "loss": 1.9132,
+      "step": 226
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7328244274809162e-05,
+      "loss": 1.9822,
+      "step": 227
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7404580152671757e-05,
+      "loss": 1.9268,
+      "step": 228
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7480916030534353e-05,
+      "loss": 1.8708,
+      "step": 229
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.755725190839695e-05,
+      "loss": 1.9719,
+      "step": 230
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7633587786259544e-05,
+      "loss": 1.9634,
+      "step": 231
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.770992366412214e-05,
+      "loss": 2.0211,
+      "step": 232
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.7786259541984735e-05,
+      "loss": 1.8313,
+      "step": 233
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.786259541984733e-05,
+      "loss": 1.8407,
+      "step": 234
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.7938931297709926e-05,
+      "loss": 1.8726,
+      "step": 235
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8015267175572518e-05,
+      "loss": 1.9875,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8091603053435117e-05,
+      "loss": 1.9323,
+      "step": 237
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.816793893129771e-05,
+      "loss": 1.8038,
+      "step": 238
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8244274809160308e-05,
+      "loss": 1.9519,
+      "step": 239
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.83206106870229e-05,
+      "loss": 1.8213,
+      "step": 240
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.83969465648855e-05,
+      "loss": 1.9565,
+      "step": 241
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.847328244274809e-05,
+      "loss": 1.8746,
+      "step": 242
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.854961832061069e-05,
+      "loss": 1.8303,
+      "step": 243
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8625954198473282e-05,
+      "loss": 1.9107,
+      "step": 244
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8702290076335878e-05,
+      "loss": 1.9725,
+      "step": 245
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8778625954198473e-05,
+      "loss": 1.8625,
+      "step": 246
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.885496183206107e-05,
+      "loss": 1.8696,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8931297709923668e-05,
+      "loss": 1.835,
+      "step": 248
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.900763358778626e-05,
+      "loss": 1.9422,
+      "step": 249
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.908396946564886e-05,
+      "loss": 2.0395,
+      "step": 250
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.916030534351145e-05,
+      "loss": 1.902,
+      "step": 251
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.923664122137405e-05,
+      "loss": 2.0188,
+      "step": 252
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9312977099236642e-05,
+      "loss": 1.8627,
+      "step": 253
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9389312977099238e-05,
+      "loss": 1.9651,
+      "step": 254
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9465648854961833e-05,
+      "loss": 1.8167,
+      "step": 255
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.954198473282443e-05,
+      "loss": 1.8972,
+      "step": 256
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9618320610687024e-05,
+      "loss": 1.8112,
+      "step": 257
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.969465648854962e-05,
+      "loss": 1.804,
+      "step": 258
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9770992366412215e-05,
+      "loss": 1.9803,
+      "step": 259
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.984732824427481e-05,
+      "loss": 1.805,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9923664122137406e-05,
+      "loss": 1.897,
+      "step": 261
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.8232,
+      "step": 262
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9999999311485303e-05,
+      "loss": 1.7953,
+      "step": 263
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9999997245941302e-05,
+      "loss": 1.9135,
+      "step": 264
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.999999380336828e-05,
+      "loss": 1.7884,
+      "step": 265
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.999998898376671e-05,
+      "loss": 1.9319,
+      "step": 266
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9999982787137265e-05,
+      "loss": 2.0473,
+      "step": 267
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999975213480787e-05,
+      "loss": 1.8697,
+      "step": 268
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999966262798327e-05,
+      "loss": 1.9977,
+      "step": 269
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999955935091114e-05,
+      "loss": 1.8993,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.999994423036057e-05,
+      "loss": 1.9246,
+      "step": 271
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999931148608308e-05,
+      "loss": 1.877,
+      "step": 272
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999916689836126e-05,
+      "loss": 2.0014,
+      "step": 273
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999900854046022e-05,
+      "loss": 1.8775,
+      "step": 274
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.999988364124017e-05,
+      "loss": 1.8936,
+      "step": 275
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999865051420943e-05,
+      "loss": 1.8148,
+      "step": 276
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.99998450845909e-05,
+      "loss": 1.9981,
+      "step": 277
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9999823740752793e-05,
+      "loss": 1.8029,
+      "step": 278
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999801019909557e-05,
+      "loss": 1.841,
+      "step": 279
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999776922064324e-05,
+      "loss": 1.8639,
+      "step": 280
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999751447220413e-05,
+      "loss": 1.9861,
+      "step": 281
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999724595381325e-05,
+      "loss": 2.0126,
+      "step": 282
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999696366550767e-05,
+      "loss": 1.886,
+      "step": 283
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.999966676073262e-05,
+      "loss": 1.8643,
+      "step": 284
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999635777930962e-05,
+      "loss": 1.8576,
+      "step": 285
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999603418150065e-05,
+      "loss": 1.9738,
+      "step": 286
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999569681394376e-05,
+      "loss": 1.7583,
+      "step": 287
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.999953456766855e-05,
+      "loss": 1.9026,
+      "step": 288
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9999498076977407e-05,
+      "loss": 1.7891,
+      "step": 289
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.999946020932599e-05,
+      "loss": 1.8948,
+      "step": 290
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999420964719503e-05,
+      "loss": 1.9056,
+      "step": 291
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999380343163356e-05,
+      "loss": 1.8255,
+      "step": 292
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999338344663135e-05,
+      "loss": 1.9215,
+      "step": 293
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.999929496922463e-05,
+      "loss": 1.8972,
+      "step": 294
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.999925021685381e-05,
+      "loss": 1.9331,
+      "step": 295
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.999920408755684e-05,
+      "loss": 1.9299,
+      "step": 296
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.999915658134007e-05,
+      "loss": 1.8237,
+      "step": 297
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999107698210043e-05,
+      "loss": 1.7445,
+      "step": 298
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999057438173493e-05,
+      "loss": 1.8696,
+      "step": 299
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9999005801237338e-05,
+      "loss": 1.8326,
+      "step": 300
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998952787408687e-05,
+      "loss": 1.8971,
+      "step": 301
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.999889839669484e-05,
+      "loss": 1.8867,
+      "step": 302
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998842629103293e-05,
+      "loss": 1.7314,
+      "step": 303
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.999878548464172e-05,
+      "loss": 1.9405,
+      "step": 304
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.999872696331799e-05,
+      "loss": 1.9093,
+      "step": 305
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998667065140163e-05,
+      "loss": 1.9197,
+      "step": 306
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998605790116486e-05,
+      "loss": 1.8776,
+      "step": 307
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998543138255397e-05,
+      "loss": 1.7978,
+      "step": 308
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998479109565525e-05,
+      "loss": 1.9253,
+      "step": 309
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9998413704055687e-05,
+      "loss": 1.8054,
+      "step": 310
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9998346921734887e-05,
+      "loss": 1.8203,
+      "step": 311
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9998278762612322e-05,
+      "loss": 1.879,
+      "step": 312
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.999820922669738e-05,
+      "loss": 1.9619,
+      "step": 313
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9998138313999635e-05,
+      "loss": 1.983,
+      "step": 314
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9998066024528845e-05,
+      "loss": 1.9798,
+      "step": 315
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997992358294976e-05,
+      "loss": 1.9921,
+      "step": 316
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997917315308167e-05,
+      "loss": 1.909,
+      "step": 317
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997840895578753e-05,
+      "loss": 1.8917,
+      "step": 318
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997763099117253e-05,
+      "loss": 1.7703,
+      "step": 319
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997683925934385e-05,
+      "loss": 1.7923,
+      "step": 320
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9997603376041047e-05,
+      "loss": 1.9182,
+      "step": 321
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.999752144944833e-05,
+      "loss": 1.8829,
+      "step": 322
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9997438146167526e-05,
+      "loss": 1.8603,
+      "step": 323
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9997353466210093e-05,
+      "loss": 1.846,
+      "step": 324
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9997267409587697e-05,
+      "loss": 1.9414,
+      "step": 325
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9997179976312193e-05,
+      "loss": 1.8419,
+      "step": 326
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.999709116639561e-05,
+      "loss": 1.8432,
+      "step": 327
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.999700097985019e-05,
+      "loss": 1.7864,
+      "step": 328
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.999690941668834e-05,
+      "loss": 1.8797,
+      "step": 329
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.999681647692268e-05,
+      "loss": 2.0382,
+      "step": 330
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9996722160565996e-05,
+      "loss": 1.9409,
+      "step": 331
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9996626467631283e-05,
+      "loss": 1.8762,
+      "step": 332
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999652939813172e-05,
+      "loss": 1.8676,
+      "step": 333
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999643095208067e-05,
+      "loss": 1.8158,
+      "step": 334
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999633112949169e-05,
+      "loss": 1.8751,
+      "step": 335
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9996229930378525e-05,
+      "loss": 1.943,
+      "step": 336
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9996127354755112e-05,
+      "loss": 1.8143,
+      "step": 337
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9996023402635576e-05,
+      "loss": 1.7886,
+      "step": 338
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999591807403423e-05,
+      "loss": 1.8492,
+      "step": 339
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999581136896558e-05,
+      "loss": 1.848,
+      "step": 340
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9995703287444317e-05,
+      "loss": 1.9476,
+      "step": 341
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9995593829485328e-05,
+      "loss": 1.7402,
+      "step": 342
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.999548299510368e-05,
+      "loss": 1.9221,
+      "step": 343
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9995370784314643e-05,
+      "loss": 1.8502,
+      "step": 344
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999525719713366e-05,
+      "loss": 1.8553,
+      "step": 345
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999514223357638e-05,
+      "loss": 1.7584,
+      "step": 346
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9995025893658627e-05,
+      "loss": 1.8455,
+      "step": 347
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9994908177396427e-05,
+      "loss": 1.8864,
+      "step": 348
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999478908480599e-05,
+      "loss": 1.843,
+      "step": 349
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9994668615903707e-05,
+      "loss": 1.9445,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9994546770706177e-05,
+      "loss": 1.9506,
+      "step": 351
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9994423549230175e-05,
+      "loss": 1.843,
+      "step": 352
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.999429895149267e-05,
+      "loss": 1.7633,
+      "step": 353
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9994172977510813e-05,
+      "loss": 1.8948,
+      "step": 354
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9994045627301957e-05,
+      "loss": 1.893,
+      "step": 355
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.999391690088364e-05,
+      "loss": 1.8755,
+      "step": 356
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9993786798273584e-05,
+      "loss": 1.9384,
+      "step": 357
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9993655319489703e-05,
+      "loss": 1.9289,
+      "step": 358
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.999352246455011e-05,
+      "loss": 1.8559,
+      "step": 359
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.999338823347309e-05,
+      "loss": 1.8659,
+      "step": 360
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9993252626277133e-05,
+      "loss": 1.8458,
+      "step": 361
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9993115642980912e-05,
+      "loss": 1.892,
+      "step": 362
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.999297728360329e-05,
+      "loss": 1.9457,
+      "step": 363
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9992837548163315e-05,
+      "loss": 1.8572,
+      "step": 364
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9992696436680235e-05,
+      "loss": 1.884,
+      "step": 365
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9992553949173476e-05,
+      "loss": 1.85,
+      "step": 366
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9992410085662667e-05,
+      "loss": 1.7993,
+      "step": 367
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9992264846167608e-05,
+      "loss": 1.742,
+      "step": 368
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9992118230708306e-05,
+      "loss": 1.753,
+      "step": 369
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.999197023930495e-05,
+      "loss": 2.0057,
+      "step": 370
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991820871977917e-05,
+      "loss": 1.7954,
+      "step": 371
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991670128747773e-05,
+      "loss": 1.8778,
+      "step": 372
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.999151800963528e-05,
+      "loss": 1.8302,
+      "step": 373
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991364514661384e-05,
+      "loss": 1.8513,
+      "step": 374
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991209643847223e-05,
+      "loss": 1.8328,
+      "step": 375
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9991053397214122e-05,
+      "loss": 1.9222,
+      "step": 376
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9990895774783593e-05,
+      "loss": 1.8128,
+      "step": 377
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9990736776577346e-05,
+      "loss": 1.9326,
+      "step": 378
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9990576402617273e-05,
+      "loss": 2.0121,
+      "step": 379
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.999041465292546e-05,
+      "loss": 1.8682,
+      "step": 380
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9990251527524178e-05,
+      "loss": 1.8142,
+      "step": 381
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9990087026435893e-05,
+      "loss": 1.8229,
+      "step": 382
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9989921149683255e-05,
+      "loss": 2.0006,
+      "step": 383
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9989753897289104e-05,
+      "loss": 1.8617,
+      "step": 384
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9989585269276475e-05,
+      "loss": 1.8045,
+      "step": 385
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9989415265668587e-05,
+      "loss": 1.8581,
+      "step": 386
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9989243886488846e-05,
+      "loss": 1.8129,
+      "step": 387
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.998907113176086e-05,
+      "loss": 1.8394,
+      "step": 388
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.998889700150841e-05,
+      "loss": 1.8086,
+      "step": 389
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9988721495755477e-05,
+      "loss": 1.8848,
+      "step": 390
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9988544614526232e-05,
+      "loss": 1.9012,
+      "step": 391
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9988366357845027e-05,
+      "loss": 1.8159,
+      "step": 392
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.998818672573641e-05,
+      "loss": 1.699,
+      "step": 393
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9988005718225117e-05,
+      "loss": 1.9343,
+      "step": 394
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9987823335336073e-05,
+      "loss": 1.7926,
+      "step": 395
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9987639577094397e-05,
+      "loss": 1.9768,
+      "step": 396
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9987454443525386e-05,
+      "loss": 1.7706,
+      "step": 397
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.998726793465454e-05,
+      "loss": 1.7971,
+      "step": 398
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9987080050507534e-05,
+      "loss": 1.7945,
+      "step": 399
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.998689079111025e-05,
+      "loss": 1.8076,
+      "step": 400
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.998670015648874e-05,
+      "loss": 1.7334,
+      "step": 401
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.998650814666926e-05,
+      "loss": 1.8735,
+      "step": 402
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.998631476167825e-05,
+      "loss": 1.7525,
+      "step": 403
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.998612000154234e-05,
+      "loss": 1.8791,
+      "step": 404
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9985923866288348e-05,
+      "loss": 1.8564,
+      "step": 405
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9985726355943284e-05,
+      "loss": 1.5983,
+      "step": 406
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9985527470534345e-05,
+      "loss": 1.8971,
+      "step": 407
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9985327210088915e-05,
+      "loss": 1.772,
+      "step": 408
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9985125574634573e-05,
+      "loss": 1.7375,
+      "step": 409
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9984922564199087e-05,
+      "loss": 1.8037,
+      "step": 410
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9984718178810408e-05,
+      "loss": 1.8834,
+      "step": 411
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9984512418496683e-05,
+      "loss": 1.8096,
+      "step": 412
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9984305283286244e-05,
+      "loss": 1.8545,
+      "step": 413
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.998409677320762e-05,
+      "loss": 1.8699,
+      "step": 414
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9983886888289515e-05,
+      "loss": 1.8503,
+      "step": 415
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9983675628560834e-05,
+      "loss": 1.8318,
+      "step": 416
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.998346299405067e-05,
+      "loss": 1.807,
+      "step": 417
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9983248984788304e-05,
+      "loss": 1.7335,
+      "step": 418
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.99830336008032e-05,
+      "loss": 1.7866,
+      "step": 419
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9982816842125023e-05,
+      "loss": 1.8669,
+      "step": 420
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.998259870878362e-05,
+      "loss": 1.8617,
+      "step": 421
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9982379200809024e-05,
+      "loss": 1.7006,
+      "step": 422
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.998215831823147e-05,
+      "loss": 1.8788,
+      "step": 423
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9981936061081366e-05,
+      "loss": 1.7763,
+      "step": 424
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9981712429389324e-05,
+      "loss": 1.7093,
+      "step": 425
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9981487423186134e-05,
+      "loss": 1.8861,
+      "step": 426
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9981261042502783e-05,
+      "loss": 1.8045,
+      "step": 427
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9981033287370443e-05,
+      "loss": 1.869,
+      "step": 428
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9980804157820474e-05,
+      "loss": 1.9033,
+      "step": 429
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9980573653884437e-05,
+      "loss": 1.8293,
+      "step": 430
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.998034177559406e-05,
+      "loss": 1.7934,
+      "step": 431
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9980108522981287e-05,
+      "loss": 1.8053,
+      "step": 432
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9979873896078226e-05,
+      "loss": 1.8873,
+      "step": 433
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9979637894917192e-05,
+      "loss": 1.7918,
+      "step": 434
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.997940051953068e-05,
+      "loss": 1.8949,
+      "step": 435
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9979161769951377e-05,
+      "loss": 1.8016,
+      "step": 436
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9978921646212166e-05,
+      "loss": 1.8338,
+      "step": 437
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9978680148346106e-05,
+      "loss": 1.8817,
+      "step": 438
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9978437276386454e-05,
+      "loss": 1.648,
+      "step": 439
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9978193030366655e-05,
+      "loss": 1.6715,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.997794741032034e-05,
+      "loss": 1.8645,
+      "step": 441
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9977700416281332e-05,
+      "loss": 1.7208,
+      "step": 442
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9977452048283648e-05,
+      "loss": 1.889,
+      "step": 443
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.997720230636148e-05,
+      "loss": 1.8453,
+      "step": 444
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9976951190549228e-05,
+      "loss": 1.8846,
+      "step": 445
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9976698700881464e-05,
+      "loss": 1.661,
+      "step": 446
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9976444837392958e-05,
+      "loss": 1.8927,
+      "step": 447
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9976189600118673e-05,
+      "loss": 1.6615,
+      "step": 448
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9975932989093746e-05,
+      "loss": 1.8102,
+      "step": 449
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9975675004353523e-05,
+      "loss": 1.7826,
+      "step": 450
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9975415645933525e-05,
+      "loss": 1.8427,
+      "step": 451
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9975154913869467e-05,
+      "loss": 1.8982,
+      "step": 452
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9974892808197248e-05,
+      "loss": 1.79,
+      "step": 453
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997462932895297e-05,
+      "loss": 1.8071,
+      "step": 454
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9974364476172908e-05,
+      "loss": 1.8722,
+      "step": 455
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997409824989353e-05,
+      "loss": 1.7876,
+      "step": 456
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997383065015151e-05,
+      "loss": 1.8915,
+      "step": 457
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997356167698368e-05,
+      "loss": 1.7092,
+      "step": 458
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997329133042709e-05,
+      "loss": 1.7807,
+      "step": 459
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9973019610518965e-05,
+      "loss": 1.8251,
+      "step": 460
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997274651729672e-05,
+      "loss": 1.6306,
+      "step": 461
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.997247205079796e-05,
+      "loss": 1.8234,
+      "step": 462
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9972196211060482e-05,
+      "loss": 1.8219,
+      "step": 463
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.997191899812227e-05,
+      "loss": 1.705,
+      "step": 464
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9971640412021494e-05,
+      "loss": 1.842,
+      "step": 465
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9971360452796523e-05,
+      "loss": 1.8278,
+      "step": 466
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.99710791204859e-05,
+      "loss": 1.704,
+      "step": 467
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9970796415128372e-05,
+      "loss": 1.7518,
+      "step": 468
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.997051233676286e-05,
+      "loss": 1.6605,
+      "step": 469
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.997022688542849e-05,
+      "loss": 1.8645,
+      "step": 470
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9969940061164567e-05,
+      "loss": 1.8299,
+      "step": 471
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.996965186401059e-05,
+      "loss": 1.7284,
+      "step": 472
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.996936229400624e-05,
+      "loss": 1.7175,
+      "step": 473
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9969071351191392e-05,
+      "loss": 1.6859,
+      "step": 474
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9968779035606118e-05,
+      "loss": 1.6966,
+      "step": 475
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9968485347290657e-05,
+      "loss": 1.7485,
+      "step": 476
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9968190286285463e-05,
+      "loss": 1.8697,
+      "step": 477
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.996789385263116e-05,
+      "loss": 1.835,
+      "step": 478
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.996759604636857e-05,
+      "loss": 1.6788,
+      "step": 479
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.99672968675387e-05,
+      "loss": 1.7785,
+      "step": 480
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9966996316182752e-05,
+      "loss": 1.7647,
+      "step": 481
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9966694392342107e-05,
+      "loss": 1.7192,
+      "step": 482
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9966391096058345e-05,
+      "loss": 1.7977,
+      "step": 483
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.996608642737323e-05,
+      "loss": 1.7716,
+      "step": 484
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.996578038632872e-05,
+      "loss": 1.815,
+      "step": 485
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9965472972966947e-05,
+      "loss": 1.7144,
+      "step": 486
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.996516418733025e-05,
+      "loss": 1.8677,
+      "step": 487
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9964854029461148e-05,
+      "loss": 1.7636,
+      "step": 488
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9964542499402352e-05,
+      "loss": 1.8092,
+      "step": 489
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9964229597196758e-05,
+      "loss": 1.7238,
+      "step": 490
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9963915322887457e-05,
+      "loss": 1.7433,
+      "step": 491
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9963599676517728e-05,
+      "loss": 1.8962,
+      "step": 492
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9963282658131024e-05,
+      "loss": 1.6997,
+      "step": 493
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9962964267771014e-05,
+      "loss": 1.8014,
+      "step": 494
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9962644505481532e-05,
+      "loss": 1.6852,
+      "step": 495
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9962323371306616e-05,
+      "loss": 1.7273,
+      "step": 496
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.996200086529048e-05,
+      "loss": 1.7405,
+      "step": 497
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9961676987477542e-05,
+      "loss": 1.7733,
+      "step": 498
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9961351737912394e-05,
+      "loss": 1.6887,
+      "step": 499
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.996102511663983e-05,
+      "loss": 1.6996,
+      "step": 500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9960697123704824e-05,
+      "loss": 1.8511,
+      "step": 501
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9960367759152543e-05,
+      "loss": 1.6562,
+      "step": 502
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9960037023028332e-05,
+      "loss": 1.7595,
+      "step": 503
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9959704915377748e-05,
+      "loss": 1.6627,
+      "step": 504
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9959371436246516e-05,
+      "loss": 1.8298,
+      "step": 505
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.995903658568056e-05,
+      "loss": 1.6343,
+      "step": 506
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.995870036372599e-05,
+      "loss": 1.7689,
+      "step": 507
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9958362770429098e-05,
+      "loss": 1.7343,
+      "step": 508
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.995802380583638e-05,
+      "loss": 1.8109,
+      "step": 509
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9957683469994507e-05,
+      "loss": 1.6798,
+      "step": 510
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9957341762950346e-05,
+      "loss": 1.5672,
+      "step": 511
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9956998684750952e-05,
+      "loss": 1.6998,
+      "step": 512
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9956654235443567e-05,
+      "loss": 1.65,
+      "step": 513
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9956308415075627e-05,
+      "loss": 1.6812,
+      "step": 514
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9955961223694744e-05,
+      "loss": 1.745,
+      "step": 515
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9955612661348735e-05,
+      "loss": 1.8008,
+      "step": 516
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9955262728085592e-05,
+      "loss": 1.6785,
+      "step": 517
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9954911423953503e-05,
+      "loss": 1.6658,
+      "step": 518
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9954558749000846e-05,
+      "loss": 1.7903,
+      "step": 519
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.995420470327619e-05,
+      "loss": 1.5605,
+      "step": 520
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9953849286828276e-05,
+      "loss": 1.7176,
+      "step": 521
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9953492499706057e-05,
+      "loss": 1.7637,
+      "step": 522
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9953134341958656e-05,
+      "loss": 1.7192,
+      "step": 523
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.99527748136354e-05,
+      "loss": 1.8086,
+      "step": 524
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9952413914785793e-05,
+      "loss": 1.7598,
+      "step": 525
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9952051645459524e-05,
+      "loss": 1.7109,
+      "step": 526
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.995168800570649e-05,
+      "loss": 1.6383,
+      "step": 527
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9951322995576766e-05,
+      "loss": 1.6923,
+      "step": 528
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9950956615120604e-05,
+      "loss": 1.719,
+      "step": 529
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9950588864388465e-05,
+      "loss": 1.6812,
+      "step": 530
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.995021974343099e-05,
+      "loss": 1.5432,
+      "step": 531
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9949849252298996e-05,
+      "loss": 1.6511,
+      "step": 532
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9949477391043514e-05,
+      "loss": 1.7873,
+      "step": 533
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9949104159715746e-05,
+      "loss": 1.5946,
+      "step": 534
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9948729558367084e-05,
+      "loss": 1.6482,
+      "step": 535
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9948353587049115e-05,
+      "loss": 1.6492,
+      "step": 536
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.994797624581361e-05,
+      "loss": 1.7376,
+      "step": 537
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9947597534712533e-05,
+      "loss": 1.6428,
+      "step": 538
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.994721745379803e-05,
+      "loss": 1.6131,
+      "step": 539
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9946836003122443e-05,
+      "loss": 1.6733,
+      "step": 540
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9946453182738296e-05,
+      "loss": 1.6912,
+      "step": 541
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.99460689926983e-05,
+      "loss": 1.627,
+      "step": 542
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9945683433055367e-05,
+      "loss": 1.6585,
+      "step": 543
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.994529650386259e-05,
+      "loss": 1.7827,
+      "step": 544
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.994490820517324e-05,
+      "loss": 1.6676,
+      "step": 545
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9944518537040797e-05,
+      "loss": 1.7347,
+      "step": 546
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.994412749951892e-05,
+      "loss": 1.5287,
+      "step": 547
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.994373509266145e-05,
+      "loss": 1.7976,
+      "step": 548
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.994334131652242e-05,
+      "loss": 1.6692,
+      "step": 549
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9942946171156062e-05,
+      "loss": 1.6323,
+      "step": 550
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.994254965661679e-05,
+      "loss": 1.7512,
+      "step": 551
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9942151772959197e-05,
+      "loss": 1.6259,
+      "step": 552
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9941752520238076e-05,
+      "loss": 1.7104,
+      "step": 553
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.994135189850841e-05,
+      "loss": 1.642,
+      "step": 554
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9940949907825357e-05,
+      "loss": 1.6445,
+      "step": 555
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.994054654824428e-05,
+      "loss": 1.6876,
+      "step": 556
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9940141819820717e-05,
+      "loss": 1.7104,
+      "step": 557
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9939735722610408e-05,
+      "loss": 1.6103,
+      "step": 558
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9939328256669266e-05,
+      "loss": 1.6323,
+      "step": 559
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9938919422053402e-05,
+      "loss": 1.5917,
+      "step": 560
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9938509218819114e-05,
+      "loss": 1.6899,
+      "step": 561
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9938097647022895e-05,
+      "loss": 1.7333,
+      "step": 562
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9937684706721408e-05,
+      "loss": 1.6612,
+      "step": 563
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9937270397971523e-05,
+      "loss": 1.7268,
+      "step": 564
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.993685472083029e-05,
+      "loss": 1.6013,
+      "step": 565
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9936437675354953e-05,
+      "loss": 1.6087,
+      "step": 566
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9936019261602933e-05,
+      "loss": 1.5958,
+      "step": 567
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.993559947963185e-05,
+      "loss": 1.6625,
+      "step": 568
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.993517832949951e-05,
+      "loss": 1.6453,
+      "step": 569
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.993475581126391e-05,
+      "loss": 1.63,
+      "step": 570
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.9934331924983222e-05,
+      "loss": 1.6132,
+      "step": 571
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.993390667071583e-05,
+      "loss": 1.5359,
+      "step": 572
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.993348004852028e-05,
+      "loss": 1.604,
+      "step": 573
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9933052058455325e-05,
+      "loss": 1.7028,
+      "step": 574
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9932622700579903e-05,
+      "loss": 1.633,
+      "step": 575
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9932191974953134e-05,
+      "loss": 1.551,
+      "step": 576
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9931759881634334e-05,
+      "loss": 1.6007,
+      "step": 577
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9931326420682997e-05,
+      "loss": 1.6543,
+      "step": 578
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9930891592158814e-05,
+      "loss": 1.556,
+      "step": 579
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9930455396121665e-05,
+      "loss": 1.6001,
+      "step": 580
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9930017832631616e-05,
+      "loss": 1.6563,
+      "step": 581
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9929578901748917e-05,
+      "loss": 1.5977,
+      "step": 582
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9929138603534013e-05,
+      "loss": 1.6498,
+      "step": 583
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9928696938047532e-05,
+      "loss": 1.6516,
+      "step": 584
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9928253905350297e-05,
+      "loss": 1.5765,
+      "step": 585
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9927809505503307e-05,
+      "loss": 1.7407,
+      "step": 586
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9927363738567767e-05,
+      "loss": 1.687,
+      "step": 587
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9926916604605055e-05,
+      "loss": 1.6431,
+      "step": 588
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.992646810367674e-05,
+      "loss": 1.6885,
+      "step": 589
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9926018235844586e-05,
+      "loss": 1.5775,
+      "step": 590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9925567001170543e-05,
+      "loss": 1.6362,
+      "step": 591
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.992511439971674e-05,
+      "loss": 1.5617,
+      "step": 592
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.9924660431545512e-05,
+      "loss": 1.591,
+      "step": 593
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.992420509671936e-05,
+      "loss": 1.642,
+      "step": 594
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.992374839530099e-05,
+      "loss": 1.4616,
+      "step": 595
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9923290327353296e-05,
+      "loss": 1.6068,
+      "step": 596
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9922830892939352e-05,
+      "loss": 1.5341,
+      "step": 597
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.992237009212242e-05,
+      "loss": 1.6299,
+      "step": 598
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9921907924965954e-05,
+      "loss": 1.6563,
+      "step": 599
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9921444391533603e-05,
+      "loss": 1.57,
+      "step": 600
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.992097949188919e-05,
+      "loss": 1.5214,
+      "step": 601
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9920513226096735e-05,
+      "loss": 1.5948,
+      "step": 602
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9920045594220442e-05,
+      "loss": 1.6581,
+      "step": 603
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.991957659632471e-05,
+      "loss": 1.5873,
+      "step": 604
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.991910623247412e-05,
+      "loss": 1.5893,
+      "step": 605
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9918634502733437e-05,
+      "loss": 1.6928,
+      "step": 606
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9918161407167628e-05,
+      "loss": 1.5026,
+      "step": 607
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9917686945841833e-05,
+      "loss": 1.5989,
+      "step": 608
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9917211118821392e-05,
+      "loss": 1.5625,
+      "step": 609
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9916733926171824e-05,
+      "loss": 1.5139,
+      "step": 610
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.991625536795884e-05,
+      "loss": 1.5568,
+      "step": 611
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.991577544424834e-05,
+      "loss": 1.5304,
+      "step": 612
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.991529415510641e-05,
+      "loss": 1.5601,
+      "step": 613
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9914811500599326e-05,
+      "loss": 1.5091,
+      "step": 614
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.991432748079355e-05,
+      "loss": 1.6967,
+      "step": 615
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9913842095755737e-05,
+      "loss": 1.6833,
+      "step": 616
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9913355345552717e-05,
+      "loss": 1.6055,
+      "step": 617
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9912867230251528e-05,
+      "loss": 1.5077,
+      "step": 618
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9912377749919373e-05,
+      "loss": 1.5261,
+      "step": 619
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9911886904623666e-05,
+      "loss": 1.5246,
+      "step": 620
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.991139469443199e-05,
+      "loss": 1.5397,
+      "step": 621
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9910901119412132e-05,
+      "loss": 1.5477,
+      "step": 622
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.991040617963205e-05,
+      "loss": 1.6274,
+      "step": 623
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.99099098751599e-05,
+      "loss": 1.5744,
+      "step": 624
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.990941220606403e-05,
+      "loss": 1.4178,
+      "step": 625
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.9908913172412968e-05,
+      "loss": 1.5711,
+      "step": 626
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.990841277427543e-05,
+      "loss": 1.6248,
+      "step": 627
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9907911011720323e-05,
+      "loss": 1.6031,
+      "step": 628
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.990740788481674e-05,
+      "loss": 1.4831,
+      "step": 629
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.990690339363397e-05,
+      "loss": 1.4069,
+      "step": 630
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9906397538241474e-05,
+      "loss": 1.5534,
+      "step": 631
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9905890318708918e-05,
+      "loss": 1.6032,
+      "step": 632
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9905381735106145e-05,
+      "loss": 1.5415,
+      "step": 633
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.990487178750318e-05,
+      "loss": 1.5904,
+      "step": 634
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9904360475970255e-05,
+      "loss": 1.5134,
+      "step": 635
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9903847800577777e-05,
+      "loss": 1.5733,
+      "step": 636
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.990333376139634e-05,
+      "loss": 1.6063,
+      "step": 637
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9902818358496728e-05,
+      "loss": 1.5073,
+      "step": 638
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.990230159194992e-05,
+      "loss": 1.529,
+      "step": 639
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9901783461827066e-05,
+      "loss": 1.651,
+      "step": 640
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9901263968199524e-05,
+      "loss": 1.5224,
+      "step": 641
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9900743111138822e-05,
+      "loss": 1.4537,
+      "step": 642
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.990022089071669e-05,
+      "loss": 1.4578,
+      "step": 643
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9899697307005036e-05,
+      "loss": 1.5631,
+      "step": 644
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.989917236007596e-05,
+      "loss": 1.5858,
+      "step": 645
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.989864605000175e-05,
+      "loss": 1.6115,
+      "step": 646
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9898118376854873e-05,
+      "loss": 1.5673,
+      "step": 647
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9897589340708002e-05,
+      "loss": 1.6368,
+      "step": 648
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.989705894163398e-05,
+      "loss": 1.5352,
+      "step": 649
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.989652717970584e-05,
+      "loss": 1.4412,
+      "step": 650
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.989599405499682e-05,
+      "loss": 1.5853,
+      "step": 651
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9895459567580326e-05,
+      "loss": 1.6206,
+      "step": 652
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9894923717529956e-05,
+      "loss": 1.5007,
+      "step": 653
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9894386504919507e-05,
+      "loss": 1.4655,
+      "step": 654
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.989384792982294e-05,
+      "loss": 1.4174,
+      "step": 655
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9893307992314433e-05,
+      "loss": 1.4529,
+      "step": 656
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9892766692468327e-05,
+      "loss": 1.509,
+      "step": 657
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9892224030359167e-05,
+      "loss": 1.5375,
+      "step": 658
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9891680006061673e-05,
+      "loss": 1.5534,
+      "step": 659
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9891134619650764e-05,
+      "loss": 1.5251,
+      "step": 660
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.989058787120154e-05,
+      "loss": 1.513,
+      "step": 661
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9890039760789294e-05,
+      "loss": 1.4764,
+      "step": 662
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.988949028848949e-05,
+      "loss": 1.5605,
+      "step": 663
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9888939454377806e-05,
+      "loss": 1.5591,
+      "step": 664
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9888387258530084e-05,
+      "loss": 1.5744,
+      "step": 665
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9887833701022366e-05,
+      "loss": 1.5326,
+      "step": 666
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.988727878193088e-05,
+      "loss": 1.4804,
+      "step": 667
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.988672250133204e-05,
+      "loss": 1.4859,
+      "step": 668
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9886164859302447e-05,
+      "loss": 1.499,
+      "step": 669
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9885605855918887e-05,
+      "loss": 1.4716,
+      "step": 670
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.988504549125834e-05,
+      "loss": 1.5594,
+      "step": 671
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9884483765397968e-05,
+      "loss": 1.5236,
+      "step": 672
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9883920678415124e-05,
+      "loss": 1.4642,
+      "step": 673
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9883356230387344e-05,
+      "loss": 1.4754,
+      "step": 674
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9882790421392354e-05,
+      "loss": 1.5043,
+      "step": 675
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9882223251508074e-05,
+      "loss": 1.5576,
+      "step": 676
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9881654720812594e-05,
+      "loss": 1.4564,
+      "step": 677
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9881084829384214e-05,
+      "loss": 1.4957,
+      "step": 678
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9880513577301403e-05,
+      "loss": 1.4646,
+      "step": 679
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.9879940964642827e-05,
+      "loss": 1.5312,
+      "step": 680
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.987936699148733e-05,
+      "loss": 1.5723,
+      "step": 681
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.987879165791396e-05,
+      "loss": 1.4411,
+      "step": 682
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9878214964001933e-05,
+      "loss": 1.5163,
+      "step": 683
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9877636909830665e-05,
+      "loss": 1.506,
+      "step": 684
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9877057495479758e-05,
+      "loss": 1.3842,
+      "step": 685
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9876476721028996e-05,
+      "loss": 1.5068,
+      "step": 686
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9875894586558357e-05,
+      "loss": 1.4416,
+      "step": 687
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9875311092148e-05,
+      "loss": 1.5619,
+      "step": 688
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.987472623787827e-05,
+      "loss": 1.5357,
+      "step": 689
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9874140023829706e-05,
+      "loss": 1.4855,
+      "step": 690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9873552450083036e-05,
+      "loss": 1.5012,
+      "step": 691
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9872963516719164e-05,
+      "loss": 1.4785,
+      "step": 692
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9872373223819196e-05,
+      "loss": 1.4763,
+      "step": 693
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.987178157146441e-05,
+      "loss": 1.5018,
+      "step": 694
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.987118855973628e-05,
+      "loss": 1.5105,
+      "step": 695
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9870594188716463e-05,
+      "loss": 1.4728,
+      "step": 696
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.986999845848681e-05,
+      "loss": 1.4932,
+      "step": 697
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9869401369129355e-05,
+      "loss": 1.5351,
+      "step": 698
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.986880292072632e-05,
+      "loss": 1.4757,
+      "step": 699
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9868203113360102e-05,
+      "loss": 1.45,
+      "step": 700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.986760194711331e-05,
+      "loss": 1.5697,
+      "step": 701
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.986699942206872e-05,
+      "loss": 1.5292,
+      "step": 702
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9866395538309302e-05,
+      "loss": 1.4349,
+      "step": 703
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9865790295918212e-05,
+      "loss": 1.4685,
+      "step": 704
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9865183694978798e-05,
+      "loss": 1.5684,
+      "step": 705
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9864575735574583e-05,
+      "loss": 1.4827,
+      "step": 706
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.986396641778929e-05,
+      "loss": 1.4268,
+      "step": 707
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9863355741706823e-05,
+      "loss": 1.4084,
+      "step": 708
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9862743707411275e-05,
+      "loss": 1.4641,
+      "step": 709
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9862130314986924e-05,
+      "loss": 1.4583,
+      "step": 710
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9861515564518234e-05,
+      "loss": 1.4326,
+      "step": 711
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9860899456089863e-05,
+      "loss": 1.4253,
+      "step": 712
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9860281989786644e-05,
+      "loss": 1.5222,
+      "step": 713
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.985966316569361e-05,
+      "loss": 1.3944,
+      "step": 714
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9859042983895967e-05,
+      "loss": 1.5203,
+      "step": 715
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9858421444479126e-05,
+      "loss": 1.367,
+      "step": 716
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.985779854752867e-05,
+      "loss": 1.4695,
+      "step": 717
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9857174293130374e-05,
+      "loss": 1.5416,
+      "step": 718
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.98565486813702e-05,
+      "loss": 1.469,
+      "step": 719
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9855921712334293e-05,
+      "loss": 1.4466,
+      "step": 720
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9855293386108995e-05,
+      "loss": 1.5391,
+      "step": 721
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.985466370278082e-05,
+      "loss": 1.4579,
+      "step": 722
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9854032662436486e-05,
+      "loss": 1.4852,
+      "step": 723
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9853400265162884e-05,
+      "loss": 1.4893,
+      "step": 724
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9852766511047098e-05,
+      "loss": 1.4622,
+      "step": 725
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.98521314001764e-05,
+      "loss": 1.4853,
+      "step": 726
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9851494932638246e-05,
+      "loss": 1.5525,
+      "step": 727
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.985085710852027e-05,
+      "loss": 1.3856,
+      "step": 728
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9850217927910318e-05,
+      "loss": 1.4709,
+      "step": 729
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.98495773908964e-05,
+      "loss": 1.4148,
+      "step": 730
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.984893549756671e-05,
+      "loss": 1.3798,
+      "step": 731
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9848292248009657e-05,
+      "loss": 1.4639,
+      "step": 732
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9847647642313806e-05,
+      "loss": 1.5146,
+      "step": 733
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9847001680567925e-05,
+      "loss": 1.4585,
+      "step": 734
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9846354362860964e-05,
+      "loss": 1.381,
+      "step": 735
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.984570568928206e-05,
+      "loss": 1.5215,
+      "step": 736
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9845055659920538e-05,
+      "loss": 1.444,
+      "step": 737
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.984440427486591e-05,
+      "loss": 1.4516,
+      "step": 738
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9843751534207868e-05,
+      "loss": 1.4637,
+      "step": 739
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9843097438036307e-05,
+      "loss": 1.3517,
+      "step": 740
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9842441986441286e-05,
+      "loss": 1.451,
+      "step": 741
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9841785179513074e-05,
+      "loss": 1.3485,
+      "step": 742
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9841127017342108e-05,
+      "loss": 1.4986,
+      "step": 743
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.984046750001902e-05,
+      "loss": 1.4334,
+      "step": 744
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9839806627634628e-05,
+      "loss": 1.3787,
+      "step": 745
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9839144400279936e-05,
+      "loss": 1.4509,
+      "step": 746
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9838480818046137e-05,
+      "loss": 1.4206,
+      "step": 747
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.98378158810246e-05,
+      "loss": 1.5761,
+      "step": 748
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9837149589306897e-05,
+      "loss": 1.4572,
+      "step": 749
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.983648194298478e-05,
+      "loss": 1.484,
+      "step": 750
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.983581294215018e-05,
+      "loss": 1.4801,
+      "step": 751
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9835142586895218e-05,
+      "loss": 1.4168,
+      "step": 752
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9834470877312213e-05,
+      "loss": 1.4463,
+      "step": 753
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9833797813493656e-05,
+      "loss": 1.471,
+      "step": 754
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9833123395532227e-05,
+      "loss": 1.344,
+      "step": 755
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9832447623520802e-05,
+      "loss": 1.3702,
+      "step": 756
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9831770497552432e-05,
+      "loss": 1.4666,
+      "step": 757
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9831092017720365e-05,
+      "loss": 1.4521,
+      "step": 758
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.983041218411802e-05,
+      "loss": 1.3586,
+      "step": 759
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 1.5519,
+      "step": 760
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9829048455977166e-05,
+      "loss": 1.4506,
+      "step": 761
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.982836456162644e-05,
+      "loss": 1.4545,
+      "step": 762
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.982767931388102e-05,
+      "loss": 1.5028,
+      "step": 763
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9826992712835268e-05,
+      "loss": 1.4372,
+      "step": 764
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9826304758583732e-05,
+      "loss": 1.4239,
+      "step": 765
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9825615451221142e-05,
+      "loss": 1.4746,
+      "step": 766
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9824924790842418e-05,
+      "loss": 1.4814,
+      "step": 767
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9824232777542666e-05,
+      "loss": 1.352,
+      "step": 768
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9823539411417178e-05,
+      "loss": 1.4911,
+      "step": 769
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.982284469256144e-05,
+      "loss": 1.4084,
+      "step": 770
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9822148621071102e-05,
+      "loss": 1.4143,
+      "step": 771
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9821451197042028e-05,
+      "loss": 1.4151,
+      "step": 772
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.982075242057025e-05,
+      "loss": 1.509,
+      "step": 773
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9820052291751993e-05,
+      "loss": 1.5051,
+      "step": 774
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9819350810683665e-05,
+      "loss": 1.4146,
+      "step": 775
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9818647977461864e-05,
+      "loss": 1.4569,
+      "step": 776
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9817943792183373e-05,
+      "loss": 1.4164,
+      "step": 777
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9817238254945158e-05,
+      "loss": 1.4247,
+      "step": 778
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9816531365844373e-05,
+      "loss": 1.4303,
+      "step": 779
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9815823124978357e-05,
+      "loss": 1.3933,
+      "step": 780
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9815113532444645e-05,
+      "loss": 1.4855,
+      "step": 781
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9814402588340944e-05,
+      "loss": 1.484,
+      "step": 782
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.981369029276515e-05,
+      "loss": 1.3155,
+      "step": 783
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9812976645815357e-05,
+      "loss": 1.3681,
+      "step": 784
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.981226164758983e-05,
+      "loss": 1.4086,
+      "step": 785
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9811545298187027e-05,
+      "loss": 1.4894,
+      "step": 786
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9810827597705595e-05,
+      "loss": 1.3771,
+      "step": 787
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9810108546244355e-05,
+      "loss": 1.3784,
+      "step": 788
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9809388143902334e-05,
+      "loss": 1.4298,
+      "step": 789
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9808666390778724e-05,
+      "loss": 1.4988,
+      "step": 790
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9807943286972918e-05,
+      "loss": 1.3983,
+      "step": 791
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9807218832584488e-05,
+      "loss": 1.4151,
+      "step": 792
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.980649302771319e-05,
+      "loss": 1.3332,
+      "step": 793
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9805765872458976e-05,
+      "loss": 1.343,
+      "step": 794
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9805037366921973e-05,
+      "loss": 1.3924,
+      "step": 795
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.98043075112025e-05,
+      "loss": 1.3664,
+      "step": 796
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.980357630540106e-05,
+      "loss": 1.3989,
+      "step": 797
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9802843749618338e-05,
+      "loss": 1.3577,
+      "step": 798
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.980210984395522e-05,
+      "loss": 1.3912,
+      "step": 799
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9801374588512754e-05,
+      "loss": 1.331,
+      "step": 800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9800637983392193e-05,
+      "loss": 1.4542,
+      "step": 801
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9799900028694972e-05,
+      "loss": 1.4107,
+      "step": 802
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.979916072452271e-05,
+      "loss": 1.3656,
+      "step": 803
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9798420070977205e-05,
+      "loss": 1.3943,
+      "step": 804
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9797678068160448e-05,
+      "loss": 1.3393,
+      "step": 805
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.979693471617462e-05,
+      "loss": 1.3929,
+      "step": 806
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.979619001512208e-05,
+      "loss": 1.38,
+      "step": 807
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9795443965105377e-05,
+      "loss": 1.4499,
+      "step": 808
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9794696566227242e-05,
+      "loss": 1.5925,
+      "step": 809
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9793947818590596e-05,
+      "loss": 1.347,
+      "step": 810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9793197722298544e-05,
+      "loss": 1.3907,
+      "step": 811
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9792446277454373e-05,
+      "loss": 1.346,
+      "step": 812
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9791693484161562e-05,
+      "loss": 1.4224,
+      "step": 813
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9790939342523773e-05,
+      "loss": 1.4361,
+      "step": 814
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.979018385264485e-05,
+      "loss": 1.3375,
+      "step": 815
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9789427014628836e-05,
+      "loss": 1.3869,
+      "step": 816
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9788668828579936e-05,
+      "loss": 1.4308,
+      "step": 817
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9787909294602564e-05,
+      "loss": 1.3518,
+      "step": 818
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9787148412801307e-05,
+      "loss": 1.3878,
+      "step": 819
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9786386183280937e-05,
+      "loss": 1.3596,
+      "step": 820
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9785622606146426e-05,
+      "loss": 1.3275,
+      "step": 821
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9784857681502912e-05,
+      "loss": 1.3859,
+      "step": 822
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9784091409455728e-05,
+      "loss": 1.3604,
+      "step": 823
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9783323790110396e-05,
+      "loss": 1.3671,
+      "step": 824
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9782554823572615e-05,
+      "loss": 1.388,
+      "step": 825
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9781784509948275e-05,
+      "loss": 1.3441,
+      "step": 826
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9781012849343452e-05,
+      "loss": 1.3543,
+      "step": 827
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9780239841864407e-05,
+      "loss": 1.3438,
+      "step": 828
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9779465487617583e-05,
+      "loss": 1.4472,
+      "step": 829
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.977868978670961e-05,
+      "loss": 1.3368,
+      "step": 830
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9777912739247305e-05,
+      "loss": 1.5205,
+      "step": 831
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.977713434533767e-05,
+      "loss": 1.4413,
+      "step": 832
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9776354605087896e-05,
+      "loss": 1.3639,
+      "step": 833
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9775573518605348e-05,
+      "loss": 1.4305,
+      "step": 834
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9774791085997593e-05,
+      "loss": 1.4453,
+      "step": 835
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9774007307372366e-05,
+      "loss": 1.3343,
+      "step": 836
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9773222182837597e-05,
+      "loss": 1.4079,
+      "step": 837
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9772435712501404e-05,
+      "loss": 1.4355,
+      "step": 838
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9771647896472083e-05,
+      "loss": 1.4027,
+      "step": 839
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9770858734858123e-05,
+      "loss": 1.3284,
+      "step": 840
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9770068227768188e-05,
+      "loss": 1.3645,
+      "step": 841
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9769276375311136e-05,
+      "loss": 1.3801,
+      "step": 842
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9768483177596008e-05,
+      "loss": 1.3811,
+      "step": 843
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9767688634732028e-05,
+      "loss": 1.3379,
+      "step": 844
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.976689274682861e-05,
+      "loss": 1.4047,
+      "step": 845
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9766095513995344e-05,
+      "loss": 1.413,
+      "step": 846
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.976529693634202e-05,
+      "loss": 1.3443,
+      "step": 847
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9764497013978594e-05,
+      "loss": 1.3464,
+      "step": 848
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9763695747015227e-05,
+      "loss": 1.4478,
+      "step": 849
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.976289313556225e-05,
+      "loss": 1.4345,
+      "step": 850
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9762089179730188e-05,
+      "loss": 1.4581,
+      "step": 851
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9761283879629743e-05,
+      "loss": 1.4257,
+      "step": 852
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9760477235371817e-05,
+      "loss": 1.3719,
+      "step": 853
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.975966924706748e-05,
+      "loss": 1.434,
+      "step": 854
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9758859914827997e-05,
+      "loss": 1.4238,
+      "step": 855
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9758049238764815e-05,
+      "loss": 1.3439,
+      "step": 856
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9757237218989563e-05,
+      "loss": 1.3232,
+      "step": 857
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.975642385561406e-05,
+      "loss": 1.3749,
+      "step": 858
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.975560914875031e-05,
+      "loss": 1.3542,
+      "step": 859
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9754793098510505e-05,
+      "loss": 1.4302,
+      "step": 860
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9753975705007012e-05,
+      "loss": 1.3722,
+      "step": 861
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9753156968352388e-05,
+      "loss": 1.3301,
+      "step": 862
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9752336888659374e-05,
+      "loss": 1.4458,
+      "step": 863
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9751515466040905e-05,
+      "loss": 1.4091,
+      "step": 864
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9750692700610087e-05,
+      "loss": 1.4248,
+      "step": 865
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.974986859248022e-05,
+      "loss": 1.307,
+      "step": 866
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.974904314176478e-05,
+      "loss": 1.3007,
+      "step": 867
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.974821634857744e-05,
+      "loss": 1.4258,
+      "step": 868
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9747388213032056e-05,
+      "loss": 1.4895,
+      "step": 869
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9746558735242657e-05,
+      "loss": 1.3732,
+      "step": 870
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9745727915323468e-05,
+      "loss": 1.3001,
+      "step": 871
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9744895753388892e-05,
+      "loss": 1.4223,
+      "step": 872
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.974406224955352e-05,
+      "loss": 1.4517,
+      "step": 873
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9743227403932135e-05,
+      "loss": 1.2469,
+      "step": 874
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9742391216639692e-05,
+      "loss": 1.4412,
+      "step": 875
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9741553687791338e-05,
+      "loss": 1.3634,
+      "step": 876
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.97407148175024e-05,
+      "loss": 1.4394,
+      "step": 877
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.97398746058884e-05,
+      "loss": 1.2815,
+      "step": 878
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9739033053065027e-05,
+      "loss": 1.3391,
+      "step": 879
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.973819015914818e-05,
+      "loss": 1.3141,
+      "step": 880
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9737345924253913e-05,
+      "loss": 1.3743,
+      "step": 881
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9736500348498485e-05,
+      "loss": 1.2785,
+      "step": 882
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.973565343199834e-05,
+      "loss": 1.4557,
+      "step": 883
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9734805174870094e-05,
+      "loss": 1.3249,
+      "step": 884
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.973395557723056e-05,
+      "loss": 1.4218,
+      "step": 885
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9733104639196724e-05,
+      "loss": 1.3612,
+      "step": 886
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9732252360885763e-05,
+      "loss": 1.3563,
+      "step": 887
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9731398742415044e-05,
+      "loss": 1.3738,
+      "step": 888
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.973054378390211e-05,
+      "loss": 1.2922,
+      "step": 889
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.972968748546469e-05,
+      "loss": 1.3431,
+      "step": 890
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.97288298472207e-05,
+      "loss": 1.2986,
+      "step": 891
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.972797086928824e-05,
+      "loss": 1.3357,
+      "step": 892
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.972711055178559e-05,
+      "loss": 1.3329,
+      "step": 893
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9726248894831222e-05,
+      "loss": 1.3913,
+      "step": 894
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9725385898543784e-05,
+      "loss": 1.3272,
+      "step": 895
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9724521563042124e-05,
+      "loss": 1.3687,
+      "step": 896
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.972365588844525e-05,
+      "loss": 1.379,
+      "step": 897
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.972278887487238e-05,
+      "loss": 1.4128,
+      "step": 898
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9721920522442898e-05,
+      "loss": 1.418,
+      "step": 899
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9721050831276372e-05,
+      "loss": 1.383,
+      "step": 900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9720179801492577e-05,
+      "loss": 1.431,
+      "step": 901
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9719307433211442e-05,
+      "loss": 1.375,
+      "step": 902
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9718433726553106e-05,
+      "loss": 1.3144,
+      "step": 903
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9717558681637873e-05,
+      "loss": 1.4161,
+      "step": 904
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.971668229858624e-05,
+      "loss": 1.3695,
+      "step": 905
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9715804577518887e-05,
+      "loss": 1.3729,
+      "step": 906
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9714925518556686e-05,
+      "loss": 1.3255,
+      "step": 907
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9714045121820676e-05,
+      "loss": 1.3136,
+      "step": 908
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.97131633874321e-05,
+      "loss": 1.2811,
+      "step": 909
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.971228031551237e-05,
+      "loss": 1.4467,
+      "step": 910
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9711395906183086e-05,
+      "loss": 1.3457,
+      "step": 911
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9710510159566033e-05,
+      "loss": 1.3551,
+      "step": 912
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.970962307578319e-05,
+      "loss": 1.3566,
+      "step": 913
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9708734654956703e-05,
+      "loss": 1.2815,
+      "step": 914
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9707844897208907e-05,
+      "loss": 1.3881,
+      "step": 915
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9706953802662333e-05,
+      "loss": 1.3823,
+      "step": 916
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9706061371439686e-05,
+      "loss": 1.3294,
+      "step": 917
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.970516760366385e-05,
+      "loss": 1.3079,
+      "step": 918
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9704272499457908e-05,
+      "loss": 1.3541,
+      "step": 919
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.970337605894511e-05,
+      "loss": 1.2949,
+      "step": 920
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9702478282248904e-05,
+      "loss": 1.1893,
+      "step": 921
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9701579169492916e-05,
+      "loss": 1.3554,
+      "step": 922
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.970067872080095e-05,
+      "loss": 1.4289,
+      "step": 923
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9699776936297015e-05,
+      "loss": 1.3353,
+      "step": 924
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9698873816105272e-05,
+      "loss": 1.3366,
+      "step": 925
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9697969360350098e-05,
+      "loss": 1.2836,
+      "step": 926
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9697063569156027e-05,
+      "loss": 1.3202,
+      "step": 927
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.96961564426478e-05,
+      "loss": 1.3361,
+      "step": 928
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9695247980950323e-05,
+      "loss": 1.3566,
+      "step": 929
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9694338184188696e-05,
+      "loss": 1.3508,
+      "step": 930
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.96934270524882e-05,
+      "loss": 1.3364,
+      "step": 931
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.9692514585974306e-05,
+      "loss": 1.3348,
+      "step": 932
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9691600784772657e-05,
+      "loss": 1.3796,
+      "step": 933
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.969068564900909e-05,
+      "loss": 1.4025,
+      "step": 934
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.968976917880962e-05,
+      "loss": 1.3524,
+      "step": 935
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.968885137430044e-05,
+      "loss": 1.342,
+      "step": 936
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9687932235607953e-05,
+      "loss": 1.2559,
+      "step": 937
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9687011762858713e-05,
+      "loss": 1.354,
+      "step": 938
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9686089956179473e-05,
+      "loss": 1.2324,
+      "step": 939
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.968516681569717e-05,
+      "loss": 1.3321,
+      "step": 940
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9684242341538926e-05,
+      "loss": 1.3775,
+      "step": 941
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.968331653383204e-05,
+      "loss": 1.3289,
+      "step": 942
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9682389392704003e-05,
+      "loss": 1.392,
+      "step": 943
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.968146091828248e-05,
+      "loss": 1.3239,
+      "step": 944
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9680531110695328e-05,
+      "loss": 1.4254,
+      "step": 945
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.967959997007058e-05,
+      "loss": 1.3219,
+      "step": 946
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9678667496536465e-05,
+      "loss": 1.3659,
+      "step": 947
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9677733690221375e-05,
+      "loss": 1.3002,
+      "step": 948
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.967679855125391e-05,
+      "loss": 1.3564,
+      "step": 949
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9675862079762838e-05,
+      "loss": 1.3605,
+      "step": 950
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.967492427587711e-05,
+      "loss": 1.4094,
+      "step": 951
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9673985139725864e-05,
+      "loss": 1.3075,
+      "step": 952
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.967304467143843e-05,
+      "loss": 1.2459,
+      "step": 953
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9672102871144302e-05,
+      "loss": 1.3331,
+      "step": 954
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9671159738973177e-05,
+      "loss": 1.2849,
+      "step": 955
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9670215275054923e-05,
+      "loss": 1.4564,
+      "step": 956
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.96692694795196e-05,
+      "loss": 1.3958,
+      "step": 957
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.966832235249744e-05,
+      "loss": 1.3472,
+      "step": 958
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9667373894118872e-05,
+      "loss": 1.3303,
+      "step": 959
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9666424104514495e-05,
+      "loss": 1.3579,
+      "step": 960
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.96654729838151e-05,
+      "loss": 1.3249,
+      "step": 961
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9664520532151664e-05,
+      "loss": 1.2811,
+      "step": 962
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9663566749655342e-05,
+      "loss": 1.3018,
+      "step": 963
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9662611636457464e-05,
+      "loss": 1.3029,
+      "step": 964
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.966165519268956e-05,
+      "loss": 1.4054,
+      "step": 965
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9660697418483327e-05,
+      "loss": 1.3445,
+      "step": 966
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9659738313970663e-05,
+      "loss": 1.3252,
+      "step": 967
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9658777879283638e-05,
+      "loss": 1.3056,
+      "step": 968
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.96578161145545e-05,
+      "loss": 1.2888,
+      "step": 969
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9656853019915692e-05,
+      "loss": 1.424,
+      "step": 970
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9655888595499832e-05,
+      "loss": 1.3487,
+      "step": 971
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.965492284143973e-05,
+      "loss": 1.3815,
+      "step": 972
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9653955757868363e-05,
+      "loss": 1.3278,
+      "step": 973
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9652987344918912e-05,
+      "loss": 1.3339,
+      "step": 974
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9652017602724722e-05,
+      "loss": 1.2851,
+      "step": 975
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9651046531419335e-05,
+      "loss": 1.2601,
+      "step": 976
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9650074131136468e-05,
+      "loss": 1.3356,
+      "step": 977
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9649100402010024e-05,
+      "loss": 1.3739,
+      "step": 978
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9648125344174085e-05,
+      "loss": 1.3845,
+      "step": 979
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9647148957762927e-05,
+      "loss": 1.3083,
+      "step": 980
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9646171242910993e-05,
+      "loss": 1.3582,
+      "step": 981
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.964519219975292e-05,
+      "loss": 1.3434,
+      "step": 982
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.964421182842353e-05,
+      "loss": 1.2415,
+      "step": 983
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9643230129057817e-05,
+      "loss": 1.2638,
+      "step": 984
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9642247101790966e-05,
+      "loss": 1.3154,
+      "step": 985
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.964126274675834e-05,
+      "loss": 1.3972,
+      "step": 986
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.9640277064095494e-05,
+      "loss": 1.2991,
+      "step": 987
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9639290053938153e-05,
+      "loss": 1.4007,
+      "step": 988
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9638301716422236e-05,
+      "loss": 1.4292,
+      "step": 989
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9637312051683835e-05,
+      "loss": 1.2494,
+      "step": 990
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9636321059859233e-05,
+      "loss": 1.2995,
+      "step": 991
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9635328741084892e-05,
+      "loss": 1.3466,
+      "step": 992
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9634335095497458e-05,
+      "loss": 1.3113,
+      "step": 993
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.963334012323376e-05,
+      "loss": 1.3734,
+      "step": 994
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9632343824430804e-05,
+      "loss": 1.3161,
+      "step": 995
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9631346199225783e-05,
+      "loss": 1.2256,
+      "step": 996
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9630347247756083e-05,
+      "loss": 1.4183,
+      "step": 997
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9629346970159253e-05,
+      "loss": 1.2894,
+      "step": 998
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9628345366573033e-05,
+      "loss": 1.2815,
+      "step": 999
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9627342437135356e-05,
+      "loss": 1.4156,
+      "step": 1000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9626338181984317e-05,
+      "loss": 1.2577,
+      "step": 1001
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9625332601258214e-05,
+      "loss": 1.3291,
+      "step": 1002
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9624325695095514e-05,
+      "loss": 1.3199,
+      "step": 1003
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.962331746363487e-05,
+      "loss": 1.2249,
+      "step": 1004
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9622307907015124e-05,
+      "loss": 1.3726,
+      "step": 1005
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.962129702537529e-05,
+      "loss": 1.2416,
+      "step": 1006
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.9620284818854572e-05,
+      "loss": 1.3003,
+      "step": 1007
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.961927128759235e-05,
+      "loss": 1.3782,
+      "step": 1008
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.961825643172819e-05,
+      "loss": 1.36,
+      "step": 1009
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.961724025140185e-05,
+      "loss": 1.2441,
+      "step": 1010
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.961622274675325e-05,
+      "loss": 1.278,
+      "step": 1011
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.961520391792251e-05,
+      "loss": 1.3865,
+      "step": 1012
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9614183765049922e-05,
+      "loss": 1.3593,
+      "step": 1013
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9613162288275968e-05,
+      "loss": 1.2566,
+      "step": 1014
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.96121394877413e-05,
+      "loss": 1.3574,
+      "step": 1015
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9611115363586772e-05,
+      "loss": 1.1906,
+      "step": 1016
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9610089915953403e-05,
+      "loss": 1.2625,
+      "step": 1017
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.9609063144982398e-05,
+      "loss": 1.4244,
+      "step": 1018
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.960803505081515e-05,
+      "loss": 1.3523,
+      "step": 1019
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.960700563359323e-05,
+      "loss": 1.3195,
+      "step": 1020
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.960597489345839e-05,
+      "loss": 1.312,
+      "step": 1021
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.960494283055257e-05,
+      "loss": 1.2375,
+      "step": 1022
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9603909445017885e-05,
+      "loss": 1.247,
+      "step": 1023
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9602874736996632e-05,
+      "loss": 1.2304,
+      "step": 1024
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.96018387066313e-05,
+      "loss": 1.3232,
+      "step": 1025
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9600801354064552e-05,
+      "loss": 1.2657,
+      "step": 1026
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.959976267943923e-05,
+      "loss": 1.4391,
+      "step": 1027
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9598722682898365e-05,
+      "loss": 1.324,
+      "step": 1028
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.959768136458517e-05,
+      "loss": 1.3632,
+      "step": 1029
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9596638724643034e-05,
+      "loss": 1.3265,
+      "step": 1030
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9595594763215537e-05,
+      "loss": 1.2332,
+      "step": 1031
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9594549480446426e-05,
+      "loss": 1.3078,
+      "step": 1032
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9593502876479648e-05,
+      "loss": 1.2327,
+      "step": 1033
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.959245495145932e-05,
+      "loss": 1.3178,
+      "step": 1034
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9591405705529745e-05,
+      "loss": 1.3165,
+      "step": 1035
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.959035513883541e-05,
+      "loss": 1.3267,
+      "step": 1036
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.9589303251520976e-05,
+      "loss": 1.2779,
+      "step": 1037
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.958825004373129e-05,
+      "loss": 1.3419,
+      "step": 1038
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.958719551561139e-05,
+      "loss": 1.3855,
+      "step": 1039
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.958613966730648e-05,
+      "loss": 1.2705,
+      "step": 1040
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.958508249896196e-05,
+      "loss": 1.2984,
+      "step": 1041
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.95840240107234e-05,
+      "loss": 1.2531,
+      "step": 1042
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9582964202736553e-05,
+      "loss": 1.2395,
+      "step": 1043
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9581903075147372e-05,
+      "loss": 1.2887,
+      "step": 1044
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.958084062810196e-05,
+      "loss": 1.2481,
+      "step": 1045
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9579776861746634e-05,
+      "loss": 1.3415,
+      "step": 1046
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9578711776227867e-05,
+      "loss": 1.2857,
+      "step": 1047
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.957764537169233e-05,
+      "loss": 1.1847,
+      "step": 1048
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.957657764828687e-05,
+      "loss": 1.3067,
+      "step": 1049
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9575508606158513e-05,
+      "loss": 1.2688,
+      "step": 1050
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9574438245454472e-05,
+      "loss": 1.2828,
+      "step": 1051
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9573366566322137e-05,
+      "loss": 1.177,
+      "step": 1052
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.957229356890908e-05,
+      "loss": 1.2537,
+      "step": 1053
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9571219253363057e-05,
+      "loss": 1.3694,
+      "step": 1054
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.957014361983201e-05,
+      "loss": 1.2553,
+      "step": 1055
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.956906666846405e-05,
+      "loss": 1.2578,
+      "step": 1056
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9567988399407473e-05,
+      "loss": 1.4287,
+      "step": 1057
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9566908812810772e-05,
+      "loss": 1.3261,
+      "step": 1058
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.95658279088226e-05,
+      "loss": 1.2616,
+      "step": 1059
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9564745687591804e-05,
+      "loss": 1.2552,
+      "step": 1060
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9563662149267405e-05,
+      "loss": 1.3762,
+      "step": 1061
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9562577293998616e-05,
+      "loss": 1.3171,
+      "step": 1062
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.956149112193482e-05,
+      "loss": 1.2897,
+      "step": 1063
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9560403633225587e-05,
+      "loss": 1.267,
+      "step": 1064
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.955931482802067e-05,
+      "loss": 1.2468,
+      "step": 1065
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9558224706469996e-05,
+      "loss": 1.2468,
+      "step": 1066
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.955713326872368e-05,
+      "loss": 1.1662,
+      "step": 1067
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9556040514932016e-05,
+      "loss": 1.3034,
+      "step": 1068
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9554946445245485e-05,
+      "loss": 1.2663,
+      "step": 1069
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.955385105981473e-05,
+      "loss": 1.2861,
+      "step": 1070
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9552754358790605e-05,
+      "loss": 1.2632,
+      "step": 1071
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9551656342324118e-05,
+      "loss": 1.3138,
+      "step": 1072
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9550557010566473e-05,
+      "loss": 1.3163,
+      "step": 1073
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9549456363669052e-05,
+      "loss": 1.3597,
+      "step": 1074
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9548354401783416e-05,
+      "loss": 1.3039,
+      "step": 1075
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9547251125061305e-05,
+      "loss": 1.2443,
+      "step": 1076
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.954614653365465e-05,
+      "loss": 1.3092,
+      "step": 1077
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9545040627715554e-05,
+      "loss": 1.3061,
+      "step": 1078
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.95439334073963e-05,
+      "loss": 1.282,
+      "step": 1079
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.954282487284936e-05,
+      "loss": 1.2258,
+      "step": 1080
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9541715024227382e-05,
+      "loss": 1.1984,
+      "step": 1081
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.954060386168319e-05,
+      "loss": 1.2602,
+      "step": 1082
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9539491385369803e-05,
+      "loss": 1.3139,
+      "step": 1083
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9538377595440404e-05,
+      "loss": 1.3163,
+      "step": 1084
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.953726249204837e-05,
+      "loss": 1.2,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9536146075347257e-05,
+      "loss": 1.2997,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9535028345490792e-05,
+      "loss": 1.1989,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.953390930263289e-05,
+      "loss": 1.1874,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9532788946927653e-05,
+      "loss": 1.2893,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.953166727852935e-05,
+      "loss": 1.3093,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.953054429759244e-05,
+      "loss": 1.275,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9529420004271568e-05,
+      "loss": 1.1615,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9528294398721543e-05,
+      "loss": 1.2582,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9527167481097368e-05,
+      "loss": 1.1617,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9526039251554222e-05,
+      "loss": 1.2342,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9524909710247465e-05,
+      "loss": 1.253,
+      "step": 1096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9523778857332642e-05,
+      "loss": 1.1803,
+      "step": 1097
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.952264669296547e-05,
+      "loss": 1.2095,
+      "step": 1098
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.952151321730186e-05,
+      "loss": 1.2354,
+      "step": 1099
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9520378430497882e-05,
+      "loss": 1.1911,
+      "step": 1100
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9519242332709804e-05,
+      "loss": 1.2879,
+      "step": 1101
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9518104924094077e-05,
+      "loss": 1.2155,
+      "step": 1102
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951696620480732e-05,
+      "loss": 1.2096,
+      "step": 1103
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9515826175006333e-05,
+      "loss": 1.2402,
+      "step": 1104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951468483484811e-05,
+      "loss": 1.209,
+      "step": 1105
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.9513542184489814e-05,
+      "loss": 1.3071,
+      "step": 1106
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.951239822408879e-05,
+      "loss": 1.1456,
+      "step": 1107
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9511252953802568e-05,
+      "loss": 1.2644,
+      "step": 1108
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.951010637378885e-05,
+      "loss": 1.2433,
+      "step": 1109
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.950895848420553e-05,
+      "loss": 1.2146,
+      "step": 1110
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9507809285210672e-05,
+      "loss": 1.3293,
+      "step": 1111
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9506658776962522e-05,
+      "loss": 1.2426,
+      "step": 1112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9505506959619507e-05,
+      "loss": 1.2697,
+      "step": 1113
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9504353833340244e-05,
+      "loss": 1.2571,
+      "step": 1114
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9503199398283515e-05,
+      "loss": 1.2641,
+      "step": 1115
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9502043654608293e-05,
+      "loss": 1.277,
+      "step": 1116
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9500886602473725e-05,
+      "loss": 1.2176,
+      "step": 1117
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.949972824203914e-05,
+      "loss": 1.2307,
+      "step": 1118
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.949856857346405e-05,
+      "loss": 1.2565,
+      "step": 1119
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.949740759690814e-05,
+      "loss": 1.2309,
+      "step": 1120
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9496245312531286e-05,
+      "loss": 1.2726,
+      "step": 1121
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9495081720493534e-05,
+      "loss": 1.1973,
+      "step": 1122
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9493916820955114e-05,
+      "loss": 1.1321,
+      "step": 1123
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.949275061407644e-05,
+      "loss": 1.1666,
+      "step": 1124
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.94915831000181e-05,
+      "loss": 1.2235,
+      "step": 1125
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.949041427894086e-05,
+      "loss": 1.2529,
+      "step": 1126
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9489244151005673e-05,
+      "loss": 1.264,
+      "step": 1127
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9488072716373674e-05,
+      "loss": 1.1907,
+      "step": 1128
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9486899975206165e-05,
+      "loss": 1.2014,
+      "step": 1129
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9485725927664643e-05,
+      "loss": 1.236,
+      "step": 1130
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.948455057391077e-05,
+      "loss": 1.252,
+      "step": 1131
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9483373914106407e-05,
+      "loss": 1.213,
+      "step": 1132
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.948219594841357e-05,
+      "loss": 1.2364,
+      "step": 1133
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9481016676994477e-05,
+      "loss": 1.2715,
+      "step": 1134
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9479836100011512e-05,
+      "loss": 1.3033,
+      "step": 1135
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.947865421762725e-05,
+      "loss": 1.287,
+      "step": 1136
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.947747103000444e-05,
+      "loss": 1.2207,
+      "step": 1137
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9476286537306002e-05,
+      "loss": 1.2085,
+      "step": 1138
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.947510073969505e-05,
+      "loss": 1.225,
+      "step": 1139
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9473913637334873e-05,
+      "loss": 1.2314,
+      "step": 1140
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9472725230388936e-05,
+      "loss": 1.2293,
+      "step": 1141
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.947153551902089e-05,
+      "loss": 1.3197,
+      "step": 1142
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9470344503394553e-05,
+      "loss": 1.3448,
+      "step": 1143
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9469152183673936e-05,
+      "loss": 1.2603,
+      "step": 1144
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.946795856002323e-05,
+      "loss": 1.2173,
+      "step": 1145
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.946676363260679e-05,
+      "loss": 1.2239,
+      "step": 1146
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.946556740158917e-05,
+      "loss": 1.2422,
+      "step": 1147
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9464369867135097e-05,
+      "loss": 1.3027,
+      "step": 1148
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9463171029409463e-05,
+      "loss": 1.2071,
+      "step": 1149
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.946197088857736e-05,
+      "loss": 1.2291,
+      "step": 1150
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9460769444804052e-05,
+      "loss": 1.2431,
+      "step": 1151
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9459566698254974e-05,
+      "loss": 1.1603,
+      "step": 1152
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9458362649095753e-05,
+      "loss": 1.1997,
+      "step": 1153
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9457157297492194e-05,
+      "loss": 1.1859,
+      "step": 1154
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9455950643610268e-05,
+      "loss": 1.22,
+      "step": 1155
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.945474268761614e-05,
+      "loss": 1.2252,
+      "step": 1156
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9453533429676148e-05,
+      "loss": 1.2542,
+      "step": 1157
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9452322869956813e-05,
+      "loss": 1.2087,
+      "step": 1158
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.945111100862483e-05,
+      "loss": 1.2335,
+      "step": 1159
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.9449897845847077e-05,
+      "loss": 1.1906,
+      "step": 1160
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.944868338179061e-05,
+      "loss": 1.246,
+      "step": 1161
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.944746761662266e-05,
+      "loss": 1.1977,
+      "step": 1162
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.944625055051065e-05,
+      "loss": 1.231,
+      "step": 1163
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9445032183622166e-05,
+      "loss": 1.2078,
+      "step": 1164
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9443812516124986e-05,
+      "loss": 1.2563,
+      "step": 1165
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9442591548187056e-05,
+      "loss": 1.329,
+      "step": 1166
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.944136927997651e-05,
+      "loss": 1.2954,
+      "step": 1167
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9440145711661663e-05,
+      "loss": 1.1927,
+      "step": 1168
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9438920843410996e-05,
+      "loss": 1.1735,
+      "step": 1169
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.943769467539318e-05,
+      "loss": 1.1881,
+      "step": 1170
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.9436467207777062e-05,
+      "loss": 1.2686,
+      "step": 1171
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.943523844073167e-05,
+      "loss": 1.262,
+      "step": 1172
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9434008374426204e-05,
+      "loss": 1.2237,
+      "step": 1173
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9432777009030054e-05,
+      "loss": 1.2057,
+      "step": 1174
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9431544344712776e-05,
+      "loss": 1.2236,
+      "step": 1175
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9430310381644115e-05,
+      "loss": 1.163,
+      "step": 1176
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9429075119993992e-05,
+      "loss": 1.2202,
+      "step": 1177
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9427838559932507e-05,
+      "loss": 1.2265,
+      "step": 1178
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.942660070162993e-05,
+      "loss": 1.2292,
+      "step": 1179
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.942536154525673e-05,
+      "loss": 1.1912,
+      "step": 1180
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9424121090983532e-05,
+      "loss": 1.1617,
+      "step": 1181
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9422879338981158e-05,
+      "loss": 1.2249,
+      "step": 1182
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9421636289420593e-05,
+      "loss": 1.2025,
+      "step": 1183
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9420391942473018e-05,
+      "loss": 1.1637,
+      "step": 1184
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9419146298309775e-05,
+      "loss": 1.1963,
+      "step": 1185
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9417899357102395e-05,
+      "loss": 1.1254,
+      "step": 1186
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.941665111902259e-05,
+      "loss": 1.1694,
+      "step": 1187
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.941540158424224e-05,
+      "loss": 1.1892,
+      "step": 1188
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9414150752933416e-05,
+      "loss": 1.2337,
+      "step": 1189
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9412898625268354e-05,
+      "loss": 1.0893,
+      "step": 1190
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9411645201419477e-05,
+      "loss": 1.2168,
+      "step": 1191
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.941039048155939e-05,
+      "loss": 1.2219,
+      "step": 1192
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.940913446586087e-05,
+      "loss": 1.1314,
+      "step": 1193
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.940787715449687e-05,
+      "loss": 1.2281,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.940661854764053e-05,
+      "loss": 1.3266,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.940535864546516e-05,
+      "loss": 1.2355,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9404097448144256e-05,
+      "loss": 1.3345,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.940283495585149e-05,
+      "loss": 1.1812,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9401571168760705e-05,
+      "loss": 1.2545,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.940030608704593e-05,
+      "loss": 1.1972,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9399039710881375e-05,
+      "loss": 1.1985,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9397772040441415e-05,
+      "loss": 1.1798,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.939650307590062e-05,
+      "loss": 1.2915,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9395232817433728e-05,
+      "loss": 1.2199,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9393961265215656e-05,
+      "loss": 1.2923,
+      "step": 1205
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.93926884194215e-05,
+      "loss": 1.1621,
+      "step": 1206
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.939141428022654e-05,
+      "loss": 1.2277,
+      "step": 1207
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.939013884780622e-05,
+      "loss": 1.2065,
+      "step": 1208
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.938886212233618e-05,
+      "loss": 1.1748,
+      "step": 1209
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9387584103992217e-05,
+      "loss": 1.1481,
+      "step": 1210
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.938630479295033e-05,
+      "loss": 1.1996,
+      "step": 1211
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9385024189386678e-05,
+      "loss": 1.1591,
+      "step": 1212
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9383742293477604e-05,
+      "loss": 1.2891,
+      "step": 1213
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9382459105399634e-05,
+      "loss": 1.2562,
+      "step": 1214
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9381174625329455e-05,
+      "loss": 1.1626,
+      "step": 1215
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9379888853443954e-05,
+      "loss": 1.1854,
+      "step": 1216
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9378601789920186e-05,
+      "loss": 1.1246,
+      "step": 1217
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9377313434935376e-05,
+      "loss": 1.1588,
+      "step": 1218
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.937602378866694e-05,
+      "loss": 1.2171,
+      "step": 1219
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9374732851292466e-05,
+      "loss": 1.2302,
+      "step": 1220
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9373440622989717e-05,
+      "loss": 1.1389,
+      "step": 1221
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9372147103936637e-05,
+      "loss": 1.2006,
+      "step": 1222
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.937085229431135e-05,
+      "loss": 1.2835,
+      "step": 1223
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.936955619429215e-05,
+      "loss": 1.213,
+      "step": 1224
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.936825880405752e-05,
+      "loss": 1.2475,
+      "step": 1225
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9366960123786113e-05,
+      "loss": 1.0803,
+      "step": 1226
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9365660153656754e-05,
+      "loss": 1.1079,
+      "step": 1227
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9364358893848464e-05,
+      "loss": 1.1723,
+      "step": 1228
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9363056344540423e-05,
+      "loss": 1.2025,
+      "step": 1229
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9361752505911997e-05,
+      "loss": 1.2077,
+      "step": 1230
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.936044737814273e-05,
+      "loss": 1.2278,
+      "step": 1231
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9359140961412338e-05,
+      "loss": 1.1292,
+      "step": 1232
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9357833255900725e-05,
+      "loss": 1.2578,
+      "step": 1233
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9356524261787962e-05,
+      "loss": 1.181,
+      "step": 1234
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9355213979254302e-05,
+      "loss": 1.2243,
+      "step": 1235
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9353902408480172e-05,
+      "loss": 1.1647,
+      "step": 1236
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9352589549646183e-05,
+      "loss": 1.1948,
+      "step": 1237
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9351275402933116e-05,
+      "loss": 1.2079,
+      "step": 1238
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9349959968521935e-05,
+      "loss": 1.1982,
+      "step": 1239
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9348643246593782e-05,
+      "loss": 1.2071,
+      "step": 1240
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.934732523732997e-05,
+      "loss": 1.2419,
+      "step": 1241
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9346005940911992e-05,
+      "loss": 1.2162,
+      "step": 1242
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.934468535752152e-05,
+      "loss": 1.2517,
+      "step": 1243
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9343363487340403e-05,
+      "loss": 1.0877,
+      "step": 1244
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9342040330550665e-05,
+      "loss": 1.0515,
+      "step": 1245
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.934071588733451e-05,
+      "loss": 1.2234,
+      "step": 1246
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.933939015787432e-05,
+      "loss": 1.1071,
+      "step": 1247
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9338063142352644e-05,
+      "loss": 1.1721,
+      "step": 1248
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9336734840952223e-05,
+      "loss": 1.2409,
+      "step": 1249
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.933540525385597e-05,
+      "loss": 1.1999,
+      "step": 1250
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9334074381246965e-05,
+      "loss": 1.1333,
+      "step": 1251
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9332742223308477e-05,
+      "loss": 1.287,
+      "step": 1252
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9331408780223953e-05,
+      "loss": 1.1557,
+      "step": 1253
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9330074052177e-05,
+      "loss": 1.1372,
+      "step": 1254
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9328738039351427e-05,
+      "loss": 1.189,
+      "step": 1255
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9327400741931196e-05,
+      "loss": 1.1725,
+      "step": 1256
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9326062160100466e-05,
+      "loss": 1.1818,
+      "step": 1257
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 1.2252,
+      "step": 1258
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.932338114394498e-05,
+      "loss": 1.1922,
+      "step": 1259
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.9322038709989406e-05,
+      "loss": 1.1891,
+      "step": 1260
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9320694992361697e-05,
+      "loss": 1.2364,
+      "step": 1261
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9319349991246887e-05,
+      "loss": 1.169,
+      "step": 1262
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9318003706830186e-05,
+      "loss": 1.1702,
+      "step": 1263
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.931665613929698e-05,
+      "loss": 1.1049,
+      "step": 1264
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9315307288832835e-05,
+      "loss": 1.2261,
+      "step": 1265
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.931395715562349e-05,
+      "loss": 1.193,
+      "step": 1266
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9312605739854862e-05,
+      "loss": 1.1614,
+      "step": 1267
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9311253041713053e-05,
+      "loss": 1.1379,
+      "step": 1268
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9309899061384322e-05,
+      "loss": 1.1663,
+      "step": 1269
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9308543799055118e-05,
+      "loss": 1.184,
+      "step": 1270
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.930718725491207e-05,
+      "loss": 1.1689,
+      "step": 1271
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.930582942914198e-05,
+      "loss": 1.1865,
+      "step": 1272
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.930447032193181e-05,
+      "loss": 1.2186,
+      "step": 1273
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9303109933468732e-05,
+      "loss": 1.1837,
+      "step": 1274
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9301748263940064e-05,
+      "loss": 1.2505,
+      "step": 1275
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9300385313533313e-05,
+      "loss": 1.1087,
+      "step": 1276
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9299021082436164e-05,
+      "loss": 1.1514,
+      "step": 1277
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9297655570836476e-05,
+      "loss": 1.1912,
+      "step": 1278
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.929628877892228e-05,
+      "loss": 1.2198,
+      "step": 1279
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9294920706881787e-05,
+      "loss": 1.1869,
+      "step": 1280
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9293551354903393e-05,
+      "loss": 1.1148,
+      "step": 1281
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9292180723175656e-05,
+      "loss": 1.1564,
+      "step": 1282
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.929080881188731e-05,
+      "loss": 1.1286,
+      "step": 1283
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9289435621227283e-05,
+      "loss": 1.1962,
+      "step": 1284
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9288061151384658e-05,
+      "loss": 1.1856,
+      "step": 1285
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9286685402548707e-05,
+      "loss": 1.1712,
+      "step": 1286
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9285308374908876e-05,
+      "loss": 1.126,
+      "step": 1287
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9283930068654787e-05,
+      "loss": 1.3294,
+      "step": 1288
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9282550483976236e-05,
+      "loss": 1.1895,
+      "step": 1289
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9281169621063188e-05,
+      "loss": 1.1615,
+      "step": 1290
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9279787480105802e-05,
+      "loss": 1.1351,
+      "step": 1291
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9278404061294397e-05,
+      "loss": 1.1722,
+      "step": 1292
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9277019364819476e-05,
+      "loss": 1.1763,
+      "step": 1293
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9275633390871718e-05,
+      "loss": 1.2654,
+      "step": 1294
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9274246139641972e-05,
+      "loss": 1.2165,
+      "step": 1295
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9272857611321268e-05,
+      "loss": 1.1931,
+      "step": 1296
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.927146780610081e-05,
+      "loss": 1.1248,
+      "step": 1297
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.927007672417198e-05,
+      "loss": 1.2842,
+      "step": 1298
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9268684365726326e-05,
+      "loss": 1.1177,
+      "step": 1299
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9267290730955593e-05,
+      "loss": 1.1422,
+      "step": 1300
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9265895820051683e-05,
+      "loss": 1.1189,
+      "step": 1301
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9264499633206675e-05,
+      "loss": 1.238,
+      "step": 1302
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9263102170612833e-05,
+      "loss": 1.2069,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9261703432462592e-05,
+      "loss": 1.1823,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9260303418948554e-05,
+      "loss": 1.1402,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9258902130263518e-05,
+      "loss": 1.1701,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.925749956660044e-05,
+      "loss": 1.2124,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9256095728152452e-05,
+      "loss": 1.1811,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.925469061511287e-05,
+      "loss": 1.0657,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.925328422767519e-05,
+      "loss": 1.1101,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9251876566033062e-05,
+      "loss": 1.2998,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9250467630380335e-05,
+      "loss": 1.1224,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9249057420911018e-05,
+      "loss": 1.1064,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9247645937819306e-05,
+      "loss": 1.1783,
+      "step": 1314
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.924623318129956e-05,
+      "loss": 1.1586,
+      "step": 1315
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9244819151546325e-05,
+      "loss": 1.1707,
+      "step": 1316
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9243403848754312e-05,
+      "loss": 1.1678,
+      "step": 1317
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9241987273118416e-05,
+      "loss": 1.245,
+      "step": 1318
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9240569424833703e-05,
+      "loss": 1.1762,
+      "step": 1319
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9239150304095416e-05,
+      "loss": 1.2003,
+      "step": 1320
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9237729911098972e-05,
+      "loss": 1.1058,
+      "step": 1321
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9236308246039957e-05,
+      "loss": 1.1858,
+      "step": 1322
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9234885309114144e-05,
+      "loss": 1.1276,
+      "step": 1323
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.923346110051748e-05,
+      "loss": 1.2029,
+      "step": 1324
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.923203562044607e-05,
+      "loss": 1.0934,
+      "step": 1325
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9230608869096218e-05,
+      "loss": 1.1768,
+      "step": 1326
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.922918084666439e-05,
+      "loss": 1.1753,
+      "step": 1327
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9227751553347226e-05,
+      "loss": 1.163,
+      "step": 1328
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9226320989341547e-05,
+      "loss": 1.1092,
+      "step": 1329
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.922488915484434e-05,
+      "loss": 1.1622,
+      "step": 1330
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.922345605005278e-05,
+      "loss": 1.1744,
+      "step": 1331
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.922202167516421e-05,
+      "loss": 1.1039,
+      "step": 1332
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9220586030376135e-05,
+      "loss": 1.1021,
+      "step": 1333
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9219149115886265e-05,
+      "loss": 1.1465,
+      "step": 1334
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9217710931892453e-05,
+      "loss": 1.1217,
+      "step": 1335
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9216271478592753e-05,
+      "loss": 1.2125,
+      "step": 1336
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9214830756185374e-05,
+      "loss": 1.1196,
+      "step": 1337
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9213388764868707e-05,
+      "loss": 1.208,
+      "step": 1338
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9211945504841324e-05,
+      "loss": 1.1574,
+      "step": 1339
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.921050097630196e-05,
+      "loss": 1.1611,
+      "step": 1340
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.920905517944954e-05,
+      "loss": 1.0815,
+      "step": 1341
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9207608114483144e-05,
+      "loss": 1.2017,
+      "step": 1342
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9206159781602044e-05,
+      "loss": 1.134,
+      "step": 1343
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9204710181005676e-05,
+      "loss": 1.1553,
+      "step": 1344
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.920325931289366e-05,
+      "loss": 1.218,
+      "step": 1345
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9201807177465775e-05,
+      "loss": 1.152,
+      "step": 1346
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9200353774921995e-05,
+      "loss": 1.0675,
+      "step": 1347
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9198899105462447e-05,
+      "loss": 1.1446,
+      "step": 1348
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9197443169287454e-05,
+      "loss": 1.2236,
+      "step": 1349
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9195985966597495e-05,
+      "loss": 1.0899,
+      "step": 1350
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9194527497593233e-05,
+      "loss": 1.1854,
+      "step": 1351
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.919306776247551e-05,
+      "loss": 1.2317,
+      "step": 1352
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9191606761445324e-05,
+      "loss": 1.1981,
+      "step": 1353
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9190144494703865e-05,
+      "loss": 1.0638,
+      "step": 1354
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9188680962452496e-05,
+      "loss": 1.2912,
+      "step": 1355
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9187216164892743e-05,
+      "loss": 1.162,
+      "step": 1356
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9185750102226318e-05,
+      "loss": 1.1589,
+      "step": 1357
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9184282774655096e-05,
+      "loss": 1.1295,
+      "step": 1358
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9182814182381136e-05,
+      "loss": 1.1716,
+      "step": 1359
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9181344325606667e-05,
+      "loss": 1.1345,
+      "step": 1360
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9179873204534095e-05,
+      "loss": 1.1449,
+      "step": 1361
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9178400819365994e-05,
+      "loss": 1.1294,
+      "step": 1362
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.917692717030512e-05,
+      "loss": 1.1242,
+      "step": 1363
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9175452257554393e-05,
+      "loss": 1.1287,
+      "step": 1364
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9173976081316915e-05,
+      "loss": 1.1908,
+      "step": 1365
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9172498641795964e-05,
+      "loss": 1.2946,
+      "step": 1366
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9171019939194982e-05,
+      "loss": 1.1122,
+      "step": 1367
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9169539973717594e-05,
+      "loss": 1.2165,
+      "step": 1368
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.916805874556759e-05,
+      "loss": 1.2032,
+      "step": 1369
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9166576254948953e-05,
+      "loss": 1.1235,
+      "step": 1370
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9165092502065815e-05,
+      "loss": 1.0782,
+      "step": 1371
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.916360748712249e-05,
+      "loss": 1.2162,
+      "step": 1372
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.916212121032348e-05,
+      "loss": 1.1069,
+      "step": 1373
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9160633671873447e-05,
+      "loss": 1.131,
+      "step": 1374
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.915914487197722e-05,
+      "loss": 1.1476,
+      "step": 1375
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.915765481083982e-05,
+      "loss": 1.1218,
+      "step": 1376
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9156163488666433e-05,
+      "loss": 1.1787,
+      "step": 1377
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9154670905662412e-05,
+      "loss": 1.1262,
+      "step": 1378
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9153177062033297e-05,
+      "loss": 1.2267,
+      "step": 1379
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9151681957984793e-05,
+      "loss": 1.1435,
+      "step": 1380
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9150185593722778e-05,
+      "loss": 1.1497,
+      "step": 1381
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9148687969453304e-05,
+      "loss": 1.1293,
+      "step": 1382
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9147189085382603e-05,
+      "loss": 1.1393,
+      "step": 1383
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9145688941717074e-05,
+      "loss": 1.1798,
+      "step": 1384
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.914418753866329e-05,
+      "loss": 1.1738,
+      "step": 1385
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9142684876428e-05,
+      "loss": 1.0535,
+      "step": 1386
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9141180955218127e-05,
+      "loss": 1.1763,
+      "step": 1387
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9139675775240763e-05,
+      "loss": 1.1819,
+      "step": 1388
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9138169336703173e-05,
+      "loss": 1.2148,
+      "step": 1389
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.91366616398128e-05,
+      "loss": 1.0504,
+      "step": 1390
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9135152684777262e-05,
+      "loss": 1.1111,
+      "step": 1391
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9133642471804342e-05,
+      "loss": 1.1279,
+      "step": 1392
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9132131001102e-05,
+      "loss": 1.0798,
+      "step": 1393
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9130618272878374e-05,
+      "loss": 1.1879,
+      "step": 1394
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9129104287341772e-05,
+      "loss": 1.2112,
+      "step": 1395
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.912758904470067e-05,
+      "loss": 1.1534,
+      "step": 1396
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.912607254516372e-05,
+      "loss": 1.0871,
+      "step": 1397
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.912455478893976e-05,
+      "loss": 1.1842,
+      "step": 1398
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.912303577623777e-05,
+      "loss": 1.0489,
+      "step": 1399
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.912151550726694e-05,
+      "loss": 1.1086,
+      "step": 1400
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9119993982236608e-05,
+      "loss": 1.1133,
+      "step": 1401
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9118471201356293e-05,
+      "loss": 1.1494,
+      "step": 1402
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9116947164835686e-05,
+      "loss": 1.2392,
+      "step": 1403
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.911542187288465e-05,
+      "loss": 1.1789,
+      "step": 1404
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.911389532571323e-05,
+      "loss": 1.198,
+      "step": 1405
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9112367523531627e-05,
+      "loss": 1.1681,
+      "step": 1406
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9110838466550227e-05,
+      "loss": 1.2425,
+      "step": 1407
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9109308154979588e-05,
+      "loss": 1.1262,
+      "step": 1408
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9107776589030436e-05,
+      "loss": 0.9683,
+      "step": 1409
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.910624376891367e-05,
+      "loss": 1.228,
+      "step": 1410
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.910470969484037e-05,
+      "loss": 1.1503,
+      "step": 1411
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.9103174367021777e-05,
+      "loss": 1.0929,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9101637785669308e-05,
+      "loss": 1.1794,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9100099950994564e-05,
+      "loss": 1.2562,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9098560863209302e-05,
+      "loss": 1.2559,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9097020522525462e-05,
+      "loss": 1.147,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.909547892915515e-05,
+      "loss": 1.1385,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9093936083310653e-05,
+      "loss": 1.0959,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.909239198520442e-05,
+      "loss": 1.133,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9090846635049084e-05,
+      "loss": 1.0868,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.908930003305744e-05,
+      "loss": 1.1409,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.908775217944246e-05,
+      "loss": 1.117,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9086203074417288e-05,
+      "loss": 1.1447,
+      "step": 1423
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9084652718195237e-05,
+      "loss": 1.156,
+      "step": 1424
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9083101110989807e-05,
+      "loss": 1.1893,
+      "step": 1425
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9081548253014645e-05,
+      "loss": 1.1558,
+      "step": 1426
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.907999414448359e-05,
+      "loss": 1.1835,
+      "step": 1427
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9078438785610652e-05,
+      "loss": 1.1366,
+      "step": 1428
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9076882176610003e-05,
+      "loss": 1.1204,
+      "step": 1429
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9075324317695996e-05,
+      "loss": 1.1411,
+      "step": 1430
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9073765209083145e-05,
+      "loss": 1.1518,
+      "step": 1431
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.9072204850986153e-05,
+      "loss": 1.1336,
+      "step": 1432
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.907064324361988e-05,
+      "loss": 1.1449,
+      "step": 1433
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.906908038719937e-05,
+      "loss": 1.1208,
+      "step": 1434
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9067516281939826e-05,
+      "loss": 1.132,
+      "step": 1435
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9065950928056636e-05,
+      "loss": 1.0531,
+      "step": 1436
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9064384325765347e-05,
+      "loss": 1.1867,
+      "step": 1437
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9062816475281693e-05,
+      "loss": 1.0506,
+      "step": 1438
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9061247376821565e-05,
+      "loss": 1.1595,
+      "step": 1439
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9059677030601037e-05,
+      "loss": 1.1645,
+      "step": 1440
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9058105436836345e-05,
+      "loss": 1.1718,
+      "step": 1441
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9056532595743903e-05,
+      "loss": 1.1079,
+      "step": 1442
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9054958507540304e-05,
+      "loss": 1.1588,
+      "step": 1443
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.905338317244229e-05,
+      "loss": 1.1483,
+      "step": 1444
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.9051806590666807e-05,
+      "loss": 1.1094,
+      "step": 1445
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.905022876243094e-05,
+      "loss": 1.1528,
+      "step": 1446
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.904864968795197e-05,
+      "loss": 1.1829,
+      "step": 1447
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9047069367447336e-05,
+      "loss": 1.1866,
+      "step": 1448
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.904548780113465e-05,
+      "loss": 1.1485,
+      "step": 1449
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9043904989231703e-05,
+      "loss": 1.161,
+      "step": 1450
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.904232093195645e-05,
+      "loss": 1.1999,
+      "step": 1451
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9040735629527027e-05,
+      "loss": 1.1871,
+      "step": 1452
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9039149082161724e-05,
+      "loss": 1.0714,
+      "step": 1453
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9037561290079022e-05,
+      "loss": 1.1575,
+      "step": 1454
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.903597225349756e-05,
+      "loss": 1.0895,
+      "step": 1455
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.903438197263616e-05,
+      "loss": 1.1583,
+      "step": 1456
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9032790447713797e-05,
+      "loss": 1.1596,
+      "step": 1457
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9031197678949633e-05,
+      "loss": 1.0433,
+      "step": 1458
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9029603666563003e-05,
+      "loss": 1.1374,
+      "step": 1459
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9028008410773398e-05,
+      "loss": 1.062,
+      "step": 1460
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9026411911800496e-05,
+      "loss": 0.997,
+      "step": 1461
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.902481416986414e-05,
+      "loss": 1.1099,
+      "step": 1462
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9023215185184342e-05,
+      "loss": 1.2044,
+      "step": 1463
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9021614957981283e-05,
+      "loss": 1.0899,
+      "step": 1464
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.902001348847532e-05,
+      "loss": 1.1238,
+      "step": 1465
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.901841077688699e-05,
+      "loss": 1.0159,
+      "step": 1466
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9016806823436978e-05,
+      "loss": 1.1297,
+      "step": 1467
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.901520162834616e-05,
+      "loss": 1.1819,
+      "step": 1468
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9013595191835575e-05,
+      "loss": 1.1348,
+      "step": 1469
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9011987514126432e-05,
+      "loss": 1.0526,
+      "step": 1470
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9010378595440116e-05,
+      "loss": 1.2138,
+      "step": 1471
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.900876843599818e-05,
+      "loss": 1.0901,
+      "step": 1472
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9007157036022345e-05,
+      "loss": 1.0599,
+      "step": 1473
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9005544395734504e-05,
+      "loss": 1.1204,
+      "step": 1474
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9003930515356727e-05,
+      "loss": 1.0567,
+      "step": 1475
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.900231539511125e-05,
+      "loss": 1.1028,
+      "step": 1476
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.9000699035220473e-05,
+      "loss": 1.0965,
+      "step": 1477
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.8999081435906984e-05,
+      "loss": 1.1954,
+      "step": 1478
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8997462597393522e-05,
+      "loss": 1.1743,
+      "step": 1479
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8995842519903013e-05,
+      "loss": 1.0921,
+      "step": 1480
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8994221203658538e-05,
+      "loss": 1.0657,
+      "step": 1481
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.899259864888336e-05,
+      "loss": 1.1269,
+      "step": 1482
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8990974855800918e-05,
+      "loss": 1.0935,
+      "step": 1483
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8989349824634803e-05,
+      "loss": 1.0404,
+      "step": 1484
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8987723555608793e-05,
+      "loss": 1.1727,
+      "step": 1485
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8986096048946826e-05,
+      "loss": 1.072,
+      "step": 1486
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.8984467304873013e-05,
+      "loss": 1.0975,
+      "step": 1487
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.898283732361164e-05,
+      "loss": 1.1157,
+      "step": 1488
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.898120610538716e-05,
+      "loss": 1.0937,
+      "step": 1489
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8979573650424196e-05,
+      "loss": 1.1376,
+      "step": 1490
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.897793995894754e-05,
+      "loss": 1.1592,
+      "step": 1491
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.897630503118216e-05,
+      "loss": 1.1541,
+      "step": 1492
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8974668867353192e-05,
+      "loss": 1.1521,
+      "step": 1493
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8973031467685933e-05,
+      "loss": 0.9926,
+      "step": 1494
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.897139283240586e-05,
+      "loss": 1.0548,
+      "step": 1495
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8969752961738625e-05,
+      "loss": 1.0404,
+      "step": 1496
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.896811185591003e-05,
+      "loss": 1.0795,
+      "step": 1497
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8966469515146076e-05,
+      "loss": 1.1357,
+      "step": 1498
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.8964825939672907e-05,
+      "loss": 1.0871,
+      "step": 1499
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.896318112971685e-05,
+      "loss": 1.1266,
+      "step": 1500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8961535085504402e-05,
+      "loss": 1.0666,
+      "step": 1501
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8959887807262225e-05,
+      "loss": 1.1487,
+      "step": 1502
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.895823929521716e-05,
+      "loss": 1.0132,
+      "step": 1503
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8956589549596207e-05,
+      "loss": 1.0727,
+      "step": 1504
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.895493857062654e-05,
+      "loss": 1.1194,
+      "step": 1505
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.895328635853551e-05,
+      "loss": 1.1079,
+      "step": 1506
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8951632913550625e-05,
+      "loss": 1.1305,
+      "step": 1507
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8949978235899573e-05,
+      "loss": 1.1214,
+      "step": 1508
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8948322325810203e-05,
+      "loss": 1.0909,
+      "step": 1509
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8946665183510546e-05,
+      "loss": 1.0459,
+      "step": 1510
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8945006809228788e-05,
+      "loss": 1.156,
+      "step": 1511
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8943347203193298e-05,
+      "loss": 1.0555,
+      "step": 1512
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8941686365632607e-05,
+      "loss": 1.1013,
+      "step": 1513
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8940024296775415e-05,
+      "loss": 1.1143,
+      "step": 1514
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8938360996850597e-05,
+      "loss": 1.0271,
+      "step": 1515
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.893669646608719e-05,
+      "loss": 1.0549,
+      "step": 1516
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8935030704714407e-05,
+      "loss": 1.1158,
+      "step": 1517
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.893336371296163e-05,
+      "loss": 1.1146,
+      "step": 1518
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.893169549105841e-05,
+      "loss": 1.1158,
+      "step": 1519
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.893002603923446e-05,
+      "loss": 1.1547,
+      "step": 1520
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8928355357719673e-05,
+      "loss": 1.0478,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8926683446744102e-05,
+      "loss": 1.1296,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.892501030653798e-05,
+      "loss": 1.0397,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8923335937331702e-05,
+      "loss": 1.072,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8921660339355828e-05,
+      "loss": 1.1788,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8919983512841104e-05,
+      "loss": 1.1236,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.891830545801842e-05,
+      "loss": 1.1227,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8916626175118863e-05,
+      "loss": 1.0749,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8914945664373664e-05,
+      "loss": 1.1497,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.891326392601424e-05,
+      "loss": 1.0794,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8911580960272168e-05,
+      "loss": 1.0516,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.89098967673792e-05,
+      "loss": 1.1126,
+      "step": 1532
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8908211347567257e-05,
+      "loss": 1.168,
+      "step": 1533
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.890652470106842e-05,
+      "loss": 1.049,
+      "step": 1534
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.890483682811495e-05,
+      "loss": 1.0999,
+      "step": 1535
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8903147728939265e-05,
+      "loss": 1.2091,
+      "step": 1536
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.890145740377397e-05,
+      "loss": 1.081,
+      "step": 1537
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8899765852851815e-05,
+      "loss": 1.0866,
+      "step": 1538
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8898073076405744e-05,
+      "loss": 1.0974,
+      "step": 1539
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8896379074668848e-05,
+      "loss": 1.0412,
+      "step": 1540
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.88946838478744e-05,
+      "loss": 1.0583,
+      "step": 1541
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.889298739625584e-05,
+      "loss": 1.1289,
+      "step": 1542
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.8891289720046774e-05,
+      "loss": 1.1125,
+      "step": 1543
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8889590819480973e-05,
+      "loss": 1.1149,
+      "step": 1544
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8887890694792377e-05,
+      "loss": 1.07,
+      "step": 1545
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8886189346215107e-05,
+      "loss": 1.1284,
+      "step": 1546
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.888448677398344e-05,
+      "loss": 1.1074,
+      "step": 1547
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8882782978331825e-05,
+      "loss": 1.1337,
+      "step": 1548
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8881077959494882e-05,
+      "loss": 1.134,
+      "step": 1549
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8879371717707396e-05,
+      "loss": 1.0886,
+      "step": 1550
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8877664253204317e-05,
+      "loss": 1.1419,
+      "step": 1551
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.887595556622077e-05,
+      "loss": 1.1775,
+      "step": 1552
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8874245656992052e-05,
+      "loss": 1.072,
+      "step": 1553
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.8872534525753617e-05,
+      "loss": 1.0956,
+      "step": 1554
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.8870822172741093e-05,
+      "loss": 1.0428,
+      "step": 1555
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.886910859819028e-05,
+      "loss": 1.0905,
+      "step": 1556
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.886739380233714e-05,
+      "loss": 1.0985,
+      "step": 1557
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.88656777854178e-05,
+      "loss": 1.0633,
+      "step": 1558
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.886396054766857e-05,
+      "loss": 1.0977,
+      "step": 1559
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.886224208932591e-05,
+      "loss": 1.1041,
+      "step": 1560
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.8860522410626462e-05,
+      "loss": 1.0499,
+      "step": 1561
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.885880151180703e-05,
+      "loss": 1.0715,
+      "step": 1562
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.8857079393104588e-05,
+      "loss": 1.064,
+      "step": 1563
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.8855356054756274e-05,
+      "loss": 1.0545,
+      "step": 1564
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.88536314969994e-05,
+      "loss": 1.1982,
+      "step": 1565
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8851905720071438e-05,
+      "loss": 1.2264,
+      "step": 1566
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8850178724210037e-05,
+      "loss": 1.0826,
+      "step": 1567
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8848450509653006e-05,
+      "loss": 1.0639,
+      "step": 1568
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.884672107663833e-05,
+      "loss": 1.1534,
+      "step": 1569
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8844990425404147e-05,
+      "loss": 1.0153,
+      "step": 1570
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8843258556188787e-05,
+      "loss": 1.0947,
+      "step": 1571
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8841525469230724e-05,
+      "loss": 1.1044,
+      "step": 1572
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.883979116476861e-05,
+      "loss": 1.0563,
+      "step": 1573
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.883805564304127e-05,
+      "loss": 1.0363,
+      "step": 1574
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.883631890428768e-05,
+      "loss": 1.0461,
+      "step": 1575
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8834580948746997e-05,
+      "loss": 1.087,
+      "step": 1576
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8832841776658552e-05,
+      "loss": 1.1264,
+      "step": 1577
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8831101388261827e-05,
+      "loss": 1.0127,
+      "step": 1578
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8829359783796474e-05,
+      "loss": 1.1267,
+      "step": 1579
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8827616963502325e-05,
+      "loss": 1.1266,
+      "step": 1580
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.882587292761937e-05,
+      "loss": 1.1806,
+      "step": 1581
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8824127676387764e-05,
+      "loss": 1.1441,
+      "step": 1582
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8822381210047836e-05,
+      "loss": 1.0549,
+      "step": 1583
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.882063352884008e-05,
+      "loss": 1.1639,
+      "step": 1584
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8818884633005153e-05,
+      "loss": 1.1438,
+      "step": 1585
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.881713452278389e-05,
+      "loss": 1.0208,
+      "step": 1586
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.8815383198417285e-05,
+      "loss": 1.1031,
+      "step": 1587
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.881363066014649e-05,
+      "loss": 1.1324,
+      "step": 1588
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8811876908212846e-05,
+      "loss": 1.0524,
+      "step": 1589
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8810121942857848e-05,
+      "loss": 1.1315,
+      "step": 1590
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8808365764323157e-05,
+      "loss": 1.0621,
+      "step": 1591
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8806608372850604e-05,
+      "loss": 1.0474,
+      "step": 1592
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.880484976868219e-05,
+      "loss": 1.0648,
+      "step": 1593
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8803089952060075e-05,
+      "loss": 1.1173,
+      "step": 1594
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8801328923226596e-05,
+      "loss": 1.0619,
+      "step": 1595
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.879956668242425e-05,
+      "loss": 1.184,
+      "step": 1596
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8797803229895702e-05,
+      "loss": 0.9693,
+      "step": 1597
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.8796038565883788e-05,
+      "loss": 0.9898,
+      "step": 1598
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8794272690631502e-05,
+      "loss": 1.1352,
+      "step": 1599
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8792505604382013e-05,
+      "loss": 1.0811,
+      "step": 1600
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8790737307378655e-05,
+      "loss": 1.1143,
+      "step": 1601
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8788967799864927e-05,
+      "loss": 1.1554,
+      "step": 1602
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8787197082084493e-05,
+      "loss": 1.0439,
+      "step": 1603
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.878542515428119e-05,
+      "loss": 1.1099,
+      "step": 1604
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8783652016699014e-05,
+      "loss": 1.1092,
+      "step": 1605
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.878187766958213e-05,
+      "loss": 1.077,
+      "step": 1606
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.878010211317488e-05,
+      "loss": 1.1618,
+      "step": 1607
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.8778325347721754e-05,
+      "loss": 1.1167,
+      "step": 1608
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.877654737346742e-05,
+      "loss": 1.102,
+      "step": 1609
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8774768190656713e-05,
+      "loss": 1.0507,
+      "step": 1610
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.877298779953463e-05,
+      "loss": 1.0674,
+      "step": 1611
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8771206200346332e-05,
+      "loss": 1.0967,
+      "step": 1612
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8769423393337157e-05,
+      "loss": 1.0431,
+      "step": 1613
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.87676393787526e-05,
+      "loss": 0.9751,
+      "step": 1614
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.876585415683833e-05,
+      "loss": 1.0969,
+      "step": 1615
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8764067727840163e-05,
+      "loss": 1.0872,
+      "step": 1616
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8762280092004112e-05,
+      "loss": 1.0289,
+      "step": 1617
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.876049124957633e-05,
+      "loss": 1.0348,
+      "step": 1618
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.875870120080315e-05,
+      "loss": 1.0626,
+      "step": 1619
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8756909945931065e-05,
+      "loss": 1.0189,
+      "step": 1620
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8755117485206736e-05,
+      "loss": 1.1173,
+      "step": 1621
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.875332381887699e-05,
+      "loss": 1.0546,
+      "step": 1622
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8751528947188825e-05,
+      "loss": 1.0799,
+      "step": 1623
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.874973287038939e-05,
+      "loss": 1.1238,
+      "step": 1624
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.874793558872602e-05,
+      "loss": 1.0454,
+      "step": 1625
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8746137102446203e-05,
+      "loss": 1.119,
+      "step": 1626
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8744337411797594e-05,
+      "loss": 1.039,
+      "step": 1627
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8742536517028017e-05,
+      "loss": 1.0116,
+      "step": 1628
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.874073441838546e-05,
+      "loss": 1.1256,
+      "step": 1629
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8738931116118076e-05,
+      "loss": 1.1917,
+      "step": 1630
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8737126610474187e-05,
+      "loss": 1.164,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.8735320901702278e-05,
+      "loss": 1.1178,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.8733513990051e-05,
+      "loss": 1.1046,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.873170587576917e-05,
+      "loss": 1.0198,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.872989655910577e-05,
+      "loss": 1.0233,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.8728086040309952e-05,
+      "loss": 1.0183,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.872627431963103e-05,
+      "loss": 1.1879,
+      "step": 1637
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.8724461397318476e-05,
+      "loss": 1.1,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.872264727362194e-05,
+      "loss": 0.9294,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.872083194879123e-05,
+      "loss": 1.1345,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.8719015423076324e-05,
+      "loss": 1.1866,
+      "step": 1641
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8717197696727364e-05,
+      "loss": 1.0849,
+      "step": 1642
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.871537876999465e-05,
+      "loss": 1.0255,
+      "step": 1643
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8713558643128662e-05,
+      "loss": 1.1276,
+      "step": 1644
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8711737316380033e-05,
+      "loss": 1.052,
+      "step": 1645
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.870991478999956e-05,
+      "loss": 1.147,
+      "step": 1646
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.870809106423822e-05,
+      "loss": 1.0409,
+      "step": 1647
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8706266139347136e-05,
+      "loss": 1.1414,
+      "step": 1648
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8704440015577613e-05,
+      "loss": 1.0253,
+      "step": 1649
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.8702612693181107e-05,
+      "loss": 0.9661,
+      "step": 1650
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.870078417240925e-05,
+      "loss": 0.9967,
+      "step": 1651
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.869895445351383e-05,
+      "loss": 1.105,
+      "step": 1652
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8697123536746815e-05,
+      "loss": 1.0191,
+      "step": 1653
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8695291422360317e-05,
+      "loss": 1.0636,
+      "step": 1654
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8693458110606628e-05,
+      "loss": 1.0433,
+      "step": 1655
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.86916236017382e-05,
+      "loss": 1.0277,
+      "step": 1656
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8689787896007653e-05,
+      "loss": 1.0363,
+      "step": 1657
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8687950993667762e-05,
+      "loss": 0.9986,
+      "step": 1658
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.868611289497148e-05,
+      "loss": 0.9658,
+      "step": 1659
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.868427360017192e-05,
+      "loss": 1.0418,
+      "step": 1660
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.868243310952235e-05,
+      "loss": 1.0804,
+      "step": 1661
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.868059142327622e-05,
+      "loss": 1.085,
+      "step": 1662
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.8678748541687127e-05,
+      "loss": 1.0315,
+      "step": 1663
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.867690446500885e-05,
+      "loss": 1.0275,
+      "step": 1664
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8675059193495317e-05,
+      "loss": 0.9368,
+      "step": 1665
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.867321272740063e-05,
+      "loss": 0.9655,
+      "step": 1666
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8671365066979056e-05,
+      "loss": 1.0331,
+      "step": 1667
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8669516212485013e-05,
+      "loss": 0.989,
+      "step": 1668
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8667666164173106e-05,
+      "loss": 1.0792,
+      "step": 1669
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8665814922298088e-05,
+      "loss": 1.0648,
+      "step": 1670
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8663962487114874e-05,
+      "loss": 0.9678,
+      "step": 1671
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8662108858878558e-05,
+      "loss": 1.0339,
+      "step": 1672
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 1.0072,
+      "step": 1673
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8658398024267777e-05,
+      "loss": 1.1634,
+      "step": 1674
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8656540818404304e-05,
+      "loss": 1.1848,
+      "step": 1675
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.865468242050971e-05,
+      "loss": 1.079,
+      "step": 1676
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8652822830839902e-05,
+      "loss": 1.1298,
+      "step": 1677
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8650962049650955e-05,
+      "loss": 1.0278,
+      "step": 1678
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8649100077199102e-05,
+      "loss": 1.1182,
+      "step": 1679
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.864723691374074e-05,
+      "loss": 1.083,
+      "step": 1680
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8645372559532434e-05,
+      "loss": 1.0991,
+      "step": 1681
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.864350701483091e-05,
+      "loss": 1.1252,
+      "step": 1682
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8641640279893063e-05,
+      "loss": 1.0921,
+      "step": 1683
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.863977235497594e-05,
+      "loss": 0.946,
+      "step": 1684
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8637903240336767e-05,
+      "loss": 1.143,
+      "step": 1685
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8636032936232923e-05,
+      "loss": 1.0929,
+      "step": 1686
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8634161442921956e-05,
+      "loss": 0.9434,
+      "step": 1687
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8632288760661575e-05,
+      "loss": 1.0626,
+      "step": 1688
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.863041488970965e-05,
+      "loss": 0.9347,
+      "step": 1689
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.862853983032423e-05,
+      "loss": 1.0217,
+      "step": 1690
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8626663582763504e-05,
+      "loss": 1.061,
+      "step": 1691
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8624786147285846e-05,
+      "loss": 1.0797,
+      "step": 1692
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8622907524149778e-05,
+      "loss": 1.0539,
+      "step": 1693
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8621027713613993e-05,
+      "loss": 1.0735,
+      "step": 1694
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.8619146715937345e-05,
+      "loss": 1.018,
+      "step": 1695
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.861726453137886e-05,
+      "loss": 1.0287,
+      "step": 1696
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8615381160197718e-05,
+      "loss": 0.994,
+      "step": 1697
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.861349660265326e-05,
+      "loss": 0.9841,
+      "step": 1698
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8611610859004993e-05,
+      "loss": 1.0788,
+      "step": 1699
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8609723929512596e-05,
+      "loss": 1.0762,
+      "step": 1700
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8607835814435905e-05,
+      "loss": 0.9672,
+      "step": 1701
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8605946514034917e-05,
+      "loss": 1.0326,
+      "step": 1702
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.860405602856979e-05,
+      "loss": 0.9353,
+      "step": 1703
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.860216435830086e-05,
+      "loss": 1.1272,
+      "step": 1704
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8600271503488608e-05,
+      "loss": 1.0652,
+      "step": 1705
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8598377464393682e-05,
+      "loss": 0.9799,
+      "step": 1706
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8596482241276903e-05,
+      "loss": 1.0168,
+      "step": 1707
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.859458583439925e-05,
+      "loss": 1.0748,
+      "step": 1708
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.859268824402186e-05,
+      "loss": 0.9333,
+      "step": 1709
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8590789470406034e-05,
+      "loss": 1.1091,
+      "step": 1710
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8588889513813246e-05,
+      "loss": 0.9804,
+      "step": 1711
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8586988374505122e-05,
+      "loss": 1.0879,
+      "step": 1712
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8585086052743452e-05,
+      "loss": 0.9525,
+      "step": 1713
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8583182548790196e-05,
+      "loss": 1.0382,
+      "step": 1714
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.858127786290747e-05,
+      "loss": 0.9941,
+      "step": 1715
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8579371995357555e-05,
+      "loss": 0.9982,
+      "step": 1716
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8577464946402893e-05,
+      "loss": 1.0199,
+      "step": 1717
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8575556716306092e-05,
+      "loss": 1.0239,
+      "step": 1718
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8573647305329922e-05,
+      "loss": 1.0967,
+      "step": 1719
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8571736713737313e-05,
+      "loss": 0.9835,
+      "step": 1720
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.856982494179136e-05,
+      "loss": 1.022,
+      "step": 1721
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8567911989755318e-05,
+      "loss": 0.9622,
+      "step": 1722
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.85659978578926e-05,
+      "loss": 1.0056,
+      "step": 1723
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8564082546466804e-05,
+      "loss": 0.9953,
+      "step": 1724
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.856216605574166e-05,
+      "loss": 1.0946,
+      "step": 1725
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.856024838598108e-05,
+      "loss": 1.033,
+      "step": 1726
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8558329537449128e-05,
+      "loss": 1.0297,
+      "step": 1727
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8556409510410045e-05,
+      "loss": 1.051,
+      "step": 1728
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8554488305128214e-05,
+      "loss": 1.0101,
+      "step": 1729
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8552565921868196e-05,
+      "loss": 0.929,
+      "step": 1730
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8550642360894708e-05,
+      "loss": 1.1175,
+      "step": 1731
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8548717622472628e-05,
+      "loss": 1.0181,
+      "step": 1732
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8546791706867e-05,
+      "loss": 1.0102,
+      "step": 1733
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8544864614343026e-05,
+      "loss": 1.015,
+      "step": 1734
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8542936345166075e-05,
+      "loss": 1.0897,
+      "step": 1735
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8541006899601676e-05,
+      "loss": 1.0263,
+      "step": 1736
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.853907627791552e-05,
+      "loss": 1.0184,
+      "step": 1737
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8537144480373455e-05,
+      "loss": 1.139,
+      "step": 1738
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8535211507241495e-05,
+      "loss": 1.0843,
+      "step": 1739
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.853327735878582e-05,
+      "loss": 1.0011,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8531342035272768e-05,
+      "loss": 1.0897,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8529405536968835e-05,
+      "loss": 0.9688,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8527467864140687e-05,
+      "loss": 0.9998,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8525529017055144e-05,
+      "loss": 1.0251,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8523588995979192e-05,
+      "loss": 1.0499,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8521647801179975e-05,
+      "loss": 1.0438,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8519705432924806e-05,
+      "loss": 1.0112,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.851776189148115e-05,
+      "loss": 0.9959,
+      "step": 1748
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8515817177116646e-05,
+      "loss": 1.0892,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8513871290099074e-05,
+      "loss": 1.1214,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.85119242306964e-05,
+      "loss": 0.9799,
+      "step": 1751
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8509975999176736e-05,
+      "loss": 0.9969,
+      "step": 1752
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.850802659580836e-05,
+      "loss": 1.1077,
+      "step": 1753
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8506076020859708e-05,
+      "loss": 0.9774,
+      "step": 1754
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.850412427459938e-05,
+      "loss": 0.9452,
+      "step": 1755
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 0.914,
+      "step": 1756
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.850021726921891e-05,
+      "loss": 0.9721,
+      "step": 1757
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8498262010636777e-05,
+      "loss": 0.9081,
+      "step": 1758
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8496305581818978e-05,
+      "loss": 0.9713,
+      "step": 1759
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.849434798303492e-05,
+      "loss": 0.9928,
+      "step": 1760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.8492389214554177e-05,
+      "loss": 1.0305,
+      "step": 1761
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.849042927664647e-05,
+      "loss": 0.976,
+      "step": 1762
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.848846816958169e-05,
+      "loss": 1.0208,
+      "step": 1763
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8486505893629894e-05,
+      "loss": 0.9992,
+      "step": 1764
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8484542449061284e-05,
+      "loss": 1.0472,
+      "step": 1765
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8482577836146238e-05,
+      "loss": 0.8241,
+      "step": 1766
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8480612055155282e-05,
+      "loss": 1.0847,
+      "step": 1767
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8478645106359118e-05,
+      "loss": 1.0335,
+      "step": 1768
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.84766769900286e-05,
+      "loss": 0.9795,
+      "step": 1769
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8474707706434735e-05,
+      "loss": 1.0243,
+      "step": 1770
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.8472737255848705e-05,
+      "loss": 0.9797,
+      "step": 1771
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.847076563854185e-05,
+      "loss": 1.0168,
+      "step": 1772
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.846879285478566e-05,
+      "loss": 0.9565,
+      "step": 1773
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.84668189048518e-05,
+      "loss": 1.0406,
+      "step": 1774
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8464843789012086e-05,
+      "loss": 1.0606,
+      "step": 1775
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8462867507538495e-05,
+      "loss": 0.9712,
+      "step": 1776
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.846089006070317e-05,
+      "loss": 1.021,
+      "step": 1777
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.845891144877841e-05,
+      "loss": 0.9731,
+      "step": 1778
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8456931672036676e-05,
+      "loss": 1.0159,
+      "step": 1779
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8454950730750586e-05,
+      "loss": 0.9556,
+      "step": 1780
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8452968625192928e-05,
+      "loss": 0.9843,
+      "step": 1781
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8450985355636638e-05,
+      "loss": 1.0157,
+      "step": 1782
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8449000922354823e-05,
+      "loss": 1.0707,
+      "step": 1783
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.844701532562074e-05,
+      "loss": 0.9894,
+      "step": 1784
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8445028565707814e-05,
+      "loss": 1.0393,
+      "step": 1785
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.844304064288963e-05,
+      "loss": 0.9424,
+      "step": 1786
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8441051557439923e-05,
+      "loss": 1.0221,
+      "step": 1787
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8439061309632607e-05,
+      "loss": 1.0127,
+      "step": 1788
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8437069899741736e-05,
+      "loss": 1.0172,
+      "step": 1789
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.843507732804154e-05,
+      "loss": 0.9948,
+      "step": 1790
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.843308359480639e-05,
+      "loss": 0.9728,
+      "step": 1791
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8431088700310846e-05,
+      "loss": 1.1283,
+      "step": 1792
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.84290926448296e-05,
+      "loss": 0.89,
+      "step": 1793
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.8427095428637514e-05,
+      "loss": 0.9053,
+      "step": 1794
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8425097052009615e-05,
+      "loss": 1.0198,
+      "step": 1795
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8423097515221083e-05,
+      "loss": 0.9826,
+      "step": 1796
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.842109681854726e-05,
+      "loss": 1.023,
+      "step": 1797
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.841909496226365e-05,
+      "loss": 1.0604,
+      "step": 1798
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.841709194664591e-05,
+      "loss": 0.996,
+      "step": 1799
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.841508777196987e-05,
+      "loss": 1.0378,
+      "step": 1800
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8413082438511504e-05,
+      "loss": 0.9667,
+      "step": 1801
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.841107594654695e-05,
+      "loss": 1.0854,
+      "step": 1802
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.840906829635251e-05,
+      "loss": 0.9547,
+      "step": 1803
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8407059488204647e-05,
+      "loss": 0.9728,
+      "step": 1804
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.8405049522379975e-05,
+      "loss": 0.9738,
+      "step": 1805
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8403038399155278e-05,
+      "loss": 0.9428,
+      "step": 1806
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8401026118807487e-05,
+      "loss": 0.9814,
+      "step": 1807
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8399012681613705e-05,
+      "loss": 1.099,
+      "step": 1808
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8396998087851182e-05,
+      "loss": 1.033,
+      "step": 1809
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.839498233779734e-05,
+      "loss": 1.0182,
+      "step": 1810
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8392965431729745e-05,
+      "loss": 0.9367,
+      "step": 1811
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8390947369926138e-05,
+      "loss": 0.98,
+      "step": 1812
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.838892815266441e-05,
+      "loss": 0.973,
+      "step": 1813
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8386907780222615e-05,
+      "loss": 0.978,
+      "step": 1814
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.838488625287896e-05,
+      "loss": 1.0251,
+      "step": 1815
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.8382863570911823e-05,
+      "loss": 0.9282,
+      "step": 1816
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8380839734599724e-05,
+      "loss": 0.9992,
+      "step": 1817
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8378814744221358e-05,
+      "loss": 0.9215,
+      "step": 1818
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8376788600055565e-05,
+      "loss": 1.0054,
+      "step": 1819
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8374761302381362e-05,
+      "loss": 1.0559,
+      "step": 1820
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.83727328514779e-05,
+      "loss": 0.9821,
+      "step": 1821
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8370703247624514e-05,
+      "loss": 0.9922,
+      "step": 1822
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8368672491100686e-05,
+      "loss": 1.0174,
+      "step": 1823
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8366640582186047e-05,
+      "loss": 1.0455,
+      "step": 1824
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.836460752116041e-05,
+      "loss": 0.9801,
+      "step": 1825
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.836257330830372e-05,
+      "loss": 0.9501,
+      "step": 1826
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8360537943896104e-05,
+      "loss": 1.1,
+      "step": 1827
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8358501428217835e-05,
+      "loss": 1.0811,
+      "step": 1828
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8356463761549347e-05,
+      "loss": 0.9717,
+      "step": 1829
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.835442494417123e-05,
+      "loss": 1.0408,
+      "step": 1830
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8352384976364236e-05,
+      "loss": 0.9249,
+      "step": 1831
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.835034385840928e-05,
+      "loss": 1.0018,
+      "step": 1832
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.834830159058743e-05,
+      "loss": 0.9921,
+      "step": 1833
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8346258173179904e-05,
+      "loss": 1.0141,
+      "step": 1834
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8344213606468087e-05,
+      "loss": 0.9248,
+      "step": 1835
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.834216789073353e-05,
+      "loss": 0.947,
+      "step": 1836
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.834012102625793e-05,
+      "loss": 0.8883,
+      "step": 1837
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8338073013323147e-05,
+      "loss": 0.9459,
+      "step": 1838
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8336023852211197e-05,
+      "loss": 0.9587,
+      "step": 1839
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8333973543204256e-05,
+      "loss": 1.0141,
+      "step": 1840
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8331922086584656e-05,
+      "loss": 0.9947,
+      "step": 1841
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.832986948263489e-05,
+      "loss": 0.9597,
+      "step": 1842
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8327815731637612e-05,
+      "loss": 0.9614,
+      "step": 1843
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8325760833875622e-05,
+      "loss": 0.9183,
+      "step": 1844
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.832370478963189e-05,
+      "loss": 0.9765,
+      "step": 1845
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8321647599189538e-05,
+      "loss": 0.9215,
+      "step": 1846
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8319589262831847e-05,
+      "loss": 0.9865,
+      "step": 1847
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8317529780842257e-05,
+      "loss": 1.0435,
+      "step": 1848
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.8315469153504365e-05,
+      "loss": 0.9532,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.831340738110192e-05,
+      "loss": 0.9776,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8311344463918846e-05,
+      "loss": 0.9239,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.83092804022392e-05,
+      "loss": 0.9452,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8307215196347215e-05,
+      "loss": 0.954,
+      "step": 1853
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8305148846527275e-05,
+      "loss": 1.012,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8303081353063928e-05,
+      "loss": 1.0368,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8301012716241864e-05,
+      "loss": 1.0116,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8298942936345948e-05,
+      "loss": 1.0141,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8296872013661192e-05,
+      "loss": 1.0231,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8294799948472766e-05,
+      "loss": 0.9315,
+      "step": 1859
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8292726741066008e-05,
+      "loss": 0.8733,
+      "step": 1860
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.829065239172639e-05,
+      "loss": 1.0438,
+      "step": 1861
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8288576900739573e-05,
+      "loss": 1.0406,
+      "step": 1862
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8286500268391343e-05,
+      "loss": 0.9611,
+      "step": 1863
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8284422494967668e-05,
+      "loss": 0.9557,
+      "step": 1864
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8282343580754663e-05,
+      "loss": 1.1226,
+      "step": 1865
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8280263526038594e-05,
+      "loss": 0.9959,
+      "step": 1866
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.82781823311059e-05,
+      "loss": 1.0031,
+      "step": 1867
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8276099996243158e-05,
+      "loss": 0.9388,
+      "step": 1868
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8274016521737115e-05,
+      "loss": 0.9389,
+      "step": 1869
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8271931907874675e-05,
+      "loss": 1.013,
+      "step": 1870
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8269846154942897e-05,
+      "loss": 0.9616,
+      "step": 1871
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.826775926322899e-05,
+      "loss": 0.9883,
+      "step": 1872
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.826567123302032e-05,
+      "loss": 1.0638,
+      "step": 1873
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.826358206460443e-05,
+      "loss": 1.0381,
+      "step": 1874
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8261491758268993e-05,
+      "loss": 0.9363,
+      "step": 1875
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.825940031430185e-05,
+      "loss": 0.9811,
+      "step": 1876
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8257307732991008e-05,
+      "loss": 0.9214,
+      "step": 1877
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8255214014624613e-05,
+      "loss": 0.9189,
+      "step": 1878
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8253119159490983e-05,
+      "loss": 1.0511,
+      "step": 1879
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8251023167878577e-05,
+      "loss": 0.9525,
+      "step": 1880
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8248926040076024e-05,
+      "loss": 0.8798,
+      "step": 1881
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.824682777637211e-05,
+      "loss": 1.0389,
+      "step": 1882
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8244728377055763e-05,
+      "loss": 0.996,
+      "step": 1883
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8242627842416083e-05,
+      "loss": 1.0639,
+      "step": 1884
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8240526172742314e-05,
+      "loss": 0.9506,
+      "step": 1885
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.823842336832387e-05,
+      "loss": 1.0,
+      "step": 1886
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8236319429450304e-05,
+      "loss": 1.0104,
+      "step": 1887
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8234214356411343e-05,
+      "loss": 1.037,
+      "step": 1888
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8232108149496853e-05,
+      "loss": 0.9772,
+      "step": 1889
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8230000808996875e-05,
+      "loss": 1.0406,
+      "step": 1890
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8227892335201586e-05,
+      "loss": 0.925,
+      "step": 1891
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8225782728401336e-05,
+      "loss": 1.0083,
+      "step": 1892
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8223671988886623e-05,
+      "loss": 0.9304,
+      "step": 1893
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8221560116948103e-05,
+      "loss": 1.0321,
+      "step": 1894
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8219447112876582e-05,
+      "loss": 1.0154,
+      "step": 1895
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8217332976963028e-05,
+      "loss": 0.9419,
+      "step": 1896
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.821521770949857e-05,
+      "loss": 1.0422,
+      "step": 1897
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.821310131077448e-05,
+      "loss": 1.1018,
+      "step": 1898
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8210983781082192e-05,
+      "loss": 1.0314,
+      "step": 1899
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.82088651207133e-05,
+      "loss": 0.8826,
+      "step": 1900
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.820674532995955e-05,
+      "loss": 1.0398,
+      "step": 1901
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8204624409112842e-05,
+      "loss": 1.0778,
+      "step": 1902
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.820250235846523e-05,
+      "loss": 1.0378,
+      "step": 1903
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8200379178308928e-05,
+      "loss": 0.894,
+      "step": 1904
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.819825486893631e-05,
+      "loss": 0.9508,
+      "step": 1905
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8196129430639895e-05,
+      "loss": 0.9668,
+      "step": 1906
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.819400286371236e-05,
+      "loss": 1.0533,
+      "step": 1907
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.819187516844654e-05,
+      "loss": 1.007,
+      "step": 1908
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.818974634513543e-05,
+      "loss": 0.8703,
+      "step": 1909
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.818761639407217e-05,
+      "loss": 0.9351,
+      "step": 1910
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8185485315550062e-05,
+      "loss": 0.9956,
+      "step": 1911
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8183353109862563e-05,
+      "loss": 0.9243,
+      "step": 1912
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.8181219777303284e-05,
+      "loss": 1.0618,
+      "step": 1913
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.817908531816599e-05,
+      "loss": 0.9143,
+      "step": 1914
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.81769497327446e-05,
+      "loss": 0.9552,
+      "step": 1915
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.81748130213332e-05,
+      "loss": 0.9176,
+      "step": 1916
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8172675184226008e-05,
+      "loss": 0.8657,
+      "step": 1917
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8170536221717416e-05,
+      "loss": 0.9084,
+      "step": 1918
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8168396134101968e-05,
+      "loss": 1.0629,
+      "step": 1919
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.816625492167436e-05,
+      "loss": 1.0389,
+      "step": 1920
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.816411258472944e-05,
+      "loss": 0.9789,
+      "step": 1921
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.816196912356222e-05,
+      "loss": 1.0879,
+      "step": 1922
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.815982453846785e-05,
+      "loss": 0.9791,
+      "step": 1923
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.815767882974166e-05,
+      "loss": 0.9215,
+      "step": 1924
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.8155531997679106e-05,
+      "loss": 0.8816,
+      "step": 1925
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8153384042575827e-05,
+      "loss": 0.9796,
+      "step": 1926
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8151234964727592e-05,
+      "loss": 0.9242,
+      "step": 1927
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.814908476443034e-05,
+      "loss": 0.9544,
+      "step": 1928
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.814693344198016e-05,
+      "loss": 0.92,
+      "step": 1929
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8144780997673293e-05,
+      "loss": 1.0826,
+      "step": 1930
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.814262743180614e-05,
+      "loss": 0.9498,
+      "step": 1931
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8140472744675247e-05,
+      "loss": 0.9419,
+      "step": 1932
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8138316936577333e-05,
+      "loss": 0.9014,
+      "step": 1933
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8136160007809246e-05,
+      "loss": 0.8847,
+      "step": 1934
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8134001958668006e-05,
+      "loss": 0.9213,
+      "step": 1935
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8131842789450787e-05,
+      "loss": 1.0158,
+      "step": 1936
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8129682500454907e-05,
+      "loss": 1.007,
+      "step": 1937
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.812752109197785e-05,
+      "loss": 1.0474,
+      "step": 1938
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8125358564317242e-05,
+      "loss": 0.994,
+      "step": 1939
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8123194917770878e-05,
+      "loss": 0.9002,
+      "step": 1940
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.812103015263669e-05,
+      "loss": 0.9653,
+      "step": 1941
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8118864269212775e-05,
+      "loss": 0.9159,
+      "step": 1942
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.811669726779738e-05,
+      "loss": 0.9998,
+      "step": 1943
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8114529148688912e-05,
+      "loss": 0.9328,
+      "step": 1944
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8112359912185923e-05,
+      "loss": 0.9775,
+      "step": 1945
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8110189558587127e-05,
+      "loss": 0.9898,
+      "step": 1946
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8108018088191386e-05,
+      "loss": 0.9934,
+      "step": 1947
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8105845501297722e-05,
+      "loss": 0.9836,
+      "step": 1948
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8103671798205294e-05,
+      "loss": 0.9772,
+      "step": 1949
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8101496979213443e-05,
+      "loss": 0.9176,
+      "step": 1950
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8099321044621637e-05,
+      "loss": 1.011,
+      "step": 1951
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8097143994729517e-05,
+      "loss": 0.9289,
+      "step": 1952
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.809496582983686e-05,
+      "loss": 0.9182,
+      "step": 1953
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8092786550243613e-05,
+      "loss": 0.9149,
+      "step": 1954
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8090606156249866e-05,
+      "loss": 0.8381,
+      "step": 1955
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8088424648155866e-05,
+      "loss": 0.9281,
+      "step": 1956
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8086242026262014e-05,
+      "loss": 0.998,
+      "step": 1957
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.808405829086886e-05,
+      "loss": 0.8813,
+      "step": 1958
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8081873442277122e-05,
+      "loss": 0.9911,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.807968748078764e-05,
+      "loss": 0.8838,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8077500406701448e-05,
+      "loss": 0.922,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.80753122203197e-05,
+      "loss": 0.9198,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.807312292194372e-05,
+      "loss": 0.876,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.807093251187498e-05,
+      "loss": 0.9984,
+      "step": 1964
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8068740990415105e-05,
+      "loss": 0.9383,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8066548357865876e-05,
+      "loss": 1.0399,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8064354614529224e-05,
+      "loss": 0.8699,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8062159760707235e-05,
+      "loss": 0.9899,
+      "step": 1968
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8059963796702143e-05,
+      "loss": 1.013,
+      "step": 1969
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.805776672281634e-05,
+      "loss": 0.8208,
+      "step": 1970
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8055568539352376e-05,
+      "loss": 0.8284,
+      "step": 1971
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.805336924661294e-05,
+      "loss": 0.9743,
+      "step": 1972
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.805116884490088e-05,
+      "loss": 0.9732,
+      "step": 1973
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8048967334519202e-05,
+      "loss": 0.9752,
+      "step": 1974
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.804676471577106e-05,
+      "loss": 0.9115,
+      "step": 1975
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8044560988959766e-05,
+      "loss": 0.9534,
+      "step": 1976
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8042356154388768e-05,
+      "loss": 0.9311,
+      "step": 1977
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8040150212361687e-05,
+      "loss": 0.9604,
+      "step": 1978
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8037943163182283e-05,
+      "loss": 0.7873,
+      "step": 1979
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8035735007154477e-05,
+      "loss": 0.8921,
+      "step": 1980
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8033525744582335e-05,
+      "loss": 0.9304,
+      "step": 1981
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8031315375770086e-05,
+      "loss": 1.019,
+      "step": 1982
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.802910390102209e-05,
+      "loss": 0.9813,
+      "step": 1983
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.802689132064289e-05,
+      "loss": 0.8787,
+      "step": 1984
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8024677634937155e-05,
+      "loss": 0.9512,
+      "step": 1985
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.802246284420972e-05,
+      "loss": 0.9154,
+      "step": 1986
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8020246948765565e-05,
+      "loss": 0.9043,
+      "step": 1987
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8018029948909827e-05,
+      "loss": 0.9743,
+      "step": 1988
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8015811844947794e-05,
+      "loss": 0.9596,
+      "step": 1989
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8013592637184903e-05,
+      "loss": 0.9233,
+      "step": 1990
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.8011372325926753e-05,
+      "loss": 0.9704,
+      "step": 1991
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8009150911479077e-05,
+      "loss": 0.9385,
+      "step": 1992
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8006928394147777e-05,
+      "loss": 0.8599,
+      "step": 1993
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8004704774238893e-05,
+      "loss": 0.9348,
+      "step": 1994
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8002480052058636e-05,
+      "loss": 0.8847,
+      "step": 1995
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8000254227913346e-05,
+      "loss": 0.8752,
+      "step": 1996
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.799802730210953e-05,
+      "loss": 0.8546,
+      "step": 1997
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7995799274953842e-05,
+      "loss": 0.9547,
+      "step": 1998
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7993570146753088e-05,
+      "loss": 0.9099,
+      "step": 1999
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7991339917814227e-05,
+      "loss": 0.8577,
+      "step": 2000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.7989108588444365e-05,
+      "loss": 0.8483,
+      "step": 2001
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.798687615895076e-05,
+      "loss": 0.9054,
+      "step": 2002
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7984642629640832e-05,
+      "loss": 0.8814,
+      "step": 2003
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7982408000822138e-05,
+      "loss": 0.9838,
+      "step": 2004
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7980172272802398e-05,
+      "loss": 0.8573,
+      "step": 2005
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7977935445889475e-05,
+      "loss": 0.8391,
+      "step": 2006
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7975697520391384e-05,
+      "loss": 0.929,
+      "step": 2007
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.79734584966163e-05,
+      "loss": 0.9438,
+      "step": 2008
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.797121837487254e-05,
+      "loss": 0.9225,
+      "step": 2009
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7968977155468575e-05,
+      "loss": 0.9409,
+      "step": 2010
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7966734838713032e-05,
+      "loss": 1.0027,
+      "step": 2011
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7964491424914677e-05,
+      "loss": 0.8812,
+      "step": 2012
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.796224691438244e-05,
+      "loss": 0.8803,
+      "step": 2013
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7960001307425394e-05,
+      "loss": 0.9029,
+      "step": 2014
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.795775460435277e-05,
+      "loss": 0.9082,
+      "step": 2015
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.795550680547394e-05,
+      "loss": 0.9225,
+      "step": 2016
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7953257911098435e-05,
+      "loss": 0.9122,
+      "step": 2017
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7951007921535936e-05,
+      "loss": 0.9218,
+      "step": 2018
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.794875683709627e-05,
+      "loss": 0.8602,
+      "step": 2019
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.794650465808942e-05,
+      "loss": 0.8817,
+      "step": 2020
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7944251384825516e-05,
+      "loss": 0.9792,
+      "step": 2021
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7941997017614843e-05,
+      "loss": 0.9091,
+      "step": 2022
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.7939741556767837e-05,
+      "loss": 0.9061,
+      "step": 2023
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7937485002595073e-05,
+      "loss": 0.9004,
+      "step": 2024
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7935227355407287e-05,
+      "loss": 1.0017,
+      "step": 2025
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7932968615515368e-05,
+      "loss": 0.9717,
+      "step": 2026
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.793070878323035e-05,
+      "loss": 0.8404,
+      "step": 2027
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7928447858863417e-05,
+      "loss": 0.8637,
+      "step": 2028
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7926185842725904e-05,
+      "loss": 0.8858,
+      "step": 2029
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7923922735129303e-05,
+      "loss": 0.9284,
+      "step": 2030
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7921658536385244e-05,
+      "loss": 0.9525,
+      "step": 2031
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7919393246805513e-05,
+      "loss": 0.8238,
+      "step": 2032
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.791712686670205e-05,
+      "loss": 0.9748,
+      "step": 2033
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7914859396386948e-05,
+      "loss": 0.8981,
+      "step": 2034
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7912590836172438e-05,
+      "loss": 0.8308,
+      "step": 2035
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7910321186370905e-05,
+      "loss": 1.0003,
+      "step": 2036
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7908050447294893e-05,
+      "loss": 0.8374,
+      "step": 2037
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7905778619257087e-05,
+      "loss": 0.9177,
+      "step": 2038
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.790350570257032e-05,
+      "loss": 0.954,
+      "step": 2039
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7901231697547582e-05,
+      "loss": 0.8346,
+      "step": 2040
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7898956604502012e-05,
+      "loss": 0.8183,
+      "step": 2041
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.78966804237469e-05,
+      "loss": 0.8313,
+      "step": 2042
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7894403155595674e-05,
+      "loss": 0.9324,
+      "step": 2043
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7892124800361927e-05,
+      "loss": 0.8987,
+      "step": 2044
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.7889845358359395e-05,
+      "loss": 0.7885,
+      "step": 2045
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.788756482990196e-05,
+      "loss": 0.7914,
+      "step": 2046
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.788528321530366e-05,
+      "loss": 0.9062,
+      "step": 2047
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.788300051487868e-05,
+      "loss": 0.8645,
+      "step": 2048
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7880716728941355e-05,
+      "loss": 0.9251,
+      "step": 2049
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.787843185780617e-05,
+      "loss": 0.8862,
+      "step": 2050
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7876145901787755e-05,
+      "loss": 0.8671,
+      "step": 2051
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7873858861200895e-05,
+      "loss": 0.9266,
+      "step": 2052
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.787157073636052e-05,
+      "loss": 0.9066,
+      "step": 2053
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7869281527581718e-05,
+      "loss": 0.8695,
+      "step": 2054
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7866991235179713e-05,
+      "loss": 0.9501,
+      "step": 2055
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.7864699859469887e-05,
+      "loss": 0.9024,
+      "step": 2056
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.786240740076777e-05,
+      "loss": 0.8546,
+      "step": 2057
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7860113859389038e-05,
+      "loss": 0.9723,
+      "step": 2058
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7857819235649525e-05,
+      "loss": 0.9326,
+      "step": 2059
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.78555235298652e-05,
+      "loss": 0.9433,
+      "step": 2060
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.785322674235219e-05,
+      "loss": 0.9556,
+      "step": 2061
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.785092887342677e-05,
+      "loss": 0.937,
+      "step": 2062
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7848629923405364e-05,
+      "loss": 0.8323,
+      "step": 2063
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7846329892604548e-05,
+      "loss": 0.9927,
+      "step": 2064
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7844028781341033e-05,
+      "loss": 0.8884,
+      "step": 2065
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.78417265899317e-05,
+      "loss": 0.9143,
+      "step": 2066
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7839423318693558e-05,
+      "loss": 0.855,
+      "step": 2067
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.783711896794378e-05,
+      "loss": 0.8441,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.783481353799968e-05,
+      "loss": 0.8962,
+      "step": 2069
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.7832507029178725e-05,
+      "loss": 0.9057,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.7830199441798526e-05,
+      "loss": 0.9364,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.782789077617684e-05,
+      "loss": 0.8468,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.7825581032631584e-05,
+      "loss": 0.8377,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.7823270211480817e-05,
+      "loss": 0.8986,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.782095831304274e-05,
+      "loss": 0.877,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.781864533763571e-05,
+      "loss": 0.9103,
+      "step": 2076
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.781633128557823e-05,
+      "loss": 0.8807,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.781401615718896e-05,
+      "loss": 0.8921,
+      "step": 2078
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7811699952786686e-05,
+      "loss": 0.8901,
+      "step": 2079
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7809382672690368e-05,
+      "loss": 0.8847,
+      "step": 2080
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7807064317219096e-05,
+      "loss": 0.9248,
+      "step": 2081
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7804744886692117e-05,
+      "loss": 0.979,
+      "step": 2082
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7802424381428823e-05,
+      "loss": 0.8715,
+      "step": 2083
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7800102801748755e-05,
+      "loss": 0.8945,
+      "step": 2084
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7797780147971594e-05,
+      "loss": 0.8222,
+      "step": 2085
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.779545642041719e-05,
+      "loss": 0.9387,
+      "step": 2086
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7793131619405513e-05,
+      "loss": 0.8754,
+      "step": 2087
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7790805745256703e-05,
+      "loss": 0.8896,
+      "step": 2088
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7788478798291043e-05,
+      "loss": 0.8677,
+      "step": 2089
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7786150778828952e-05,
+      "loss": 0.8542,
+      "step": 2090
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.778382168719101e-05,
+      "loss": 0.8893,
+      "step": 2091
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7781491523697938e-05,
+      "loss": 0.8638,
+      "step": 2092
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.777916028867061e-05,
+      "loss": 0.9414,
+      "step": 2093
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7776827982430035e-05,
+      "loss": 0.9157,
+      "step": 2094
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.777449460529739e-05,
+      "loss": 0.8753,
+      "step": 2095
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7772160157593976e-05,
+      "loss": 0.8422,
+      "step": 2096
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7769824639641266e-05,
+      "loss": 0.816,
+      "step": 2097
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7767488051760858e-05,
+      "loss": 0.9129,
+      "step": 2098
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.776515039427451e-05,
+      "loss": 0.9119,
+      "step": 2099
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.7762811667504123e-05,
+      "loss": 1.0246,
+      "step": 2100
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7760471871771752e-05,
+      "loss": 0.9712,
+      "step": 2101
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7758131007399585e-05,
+      "loss": 0.8582,
+      "step": 2102
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7755789074709973e-05,
+      "loss": 0.9302,
+      "step": 2103
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.775344607402541e-05,
+      "loss": 0.9105,
+      "step": 2104
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7751102005668523e-05,
+      "loss": 0.9099,
+      "step": 2105
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7748756869962104e-05,
+      "loss": 0.827,
+      "step": 2106
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7746410667229083e-05,
+      "loss": 0.8609,
+      "step": 2107
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.774406339779254e-05,
+      "loss": 0.8626,
+      "step": 2108
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.77417150619757e-05,
+      "loss": 0.8799,
+      "step": 2109
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.773936566010194e-05,
+      "loss": 0.9663,
+      "step": 2110
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7737015192494777e-05,
+      "loss": 0.9592,
+      "step": 2111
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7734663659477874e-05,
+      "loss": 0.9038,
+      "step": 2112
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7732311061375048e-05,
+      "loss": 0.8506,
+      "step": 2113
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7729957398510252e-05,
+      "loss": 0.7953,
+      "step": 2114
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7727602671207605e-05,
+      "loss": 0.8044,
+      "step": 2115
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.772524687979135e-05,
+      "loss": 0.8019,
+      "step": 2116
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.772289002458589e-05,
+      "loss": 0.9181,
+      "step": 2117
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7720532105915763e-05,
+      "loss": 0.9567,
+      "step": 2118
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7718173124105674e-05,
+      "loss": 0.8947,
+      "step": 2119
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7715813079480452e-05,
+      "loss": 0.8373,
+      "step": 2120
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7713451972365087e-05,
+      "loss": 0.7996,
+      "step": 2121
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7711089803084714e-05,
+      "loss": 0.9533,
+      "step": 2122
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7708726571964598e-05,
+      "loss": 0.9316,
+      "step": 2123
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7706362279330175e-05,
+      "loss": 0.9515,
+      "step": 2124
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7703996925507014e-05,
+      "loss": 0.8636,
+      "step": 2125
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7701630510820826e-05,
+      "loss": 0.8765,
+      "step": 2126
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.769926303559747e-05,
+      "loss": 0.8771,
+      "step": 2127
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7696894500162964e-05,
+      "loss": 0.8576,
+      "step": 2128
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7694524904843455e-05,
+      "loss": 0.8175,
+      "step": 2129
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7692154249965248e-05,
+      "loss": 0.8092,
+      "step": 2130
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7689782535854786e-05,
+      "loss": 0.9438,
+      "step": 2131
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7687409762838666e-05,
+      "loss": 0.8456,
+      "step": 2132
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7685035931243618e-05,
+      "loss": 0.8612,
+      "step": 2133
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7682661041396532e-05,
+      "loss": 0.863,
+      "step": 2134
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7680285093624435e-05,
+      "loss": 0.9051,
+      "step": 2135
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7677908088254498e-05,
+      "loss": 0.8545,
+      "step": 2136
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.767553002561405e-05,
+      "loss": 0.8201,
+      "step": 2137
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.767315090603055e-05,
+      "loss": 0.8261,
+      "step": 2138
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7670770729831616e-05,
+      "loss": 0.8459,
+      "step": 2139
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7668389497345e-05,
+      "loss": 0.834,
+      "step": 2140
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7666007208898602e-05,
+      "loss": 0.852,
+      "step": 2141
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7663623864820478e-05,
+      "loss": 0.9381,
+      "step": 2142
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.766123946543882e-05,
+      "loss": 0.8279,
+      "step": 2143
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.765885401108196e-05,
+      "loss": 0.8006,
+      "step": 2144
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7656467502078392e-05,
+      "loss": 0.9437,
+      "step": 2145
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.765407993875674e-05,
+      "loss": 0.884,
+      "step": 2146
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.765169132144578e-05,
+      "loss": 0.8347,
+      "step": 2147
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7649301650474425e-05,
+      "loss": 0.8997,
+      "step": 2148
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7646910926171747e-05,
+      "loss": 0.9134,
+      "step": 2149
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7644519148866957e-05,
+      "loss": 0.8403,
+      "step": 2150
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.76421263188894e-05,
+      "loss": 0.8959,
+      "step": 2151
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7639732436568588e-05,
+      "loss": 0.8872,
+      "step": 2152
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7637337502234162e-05,
+      "loss": 0.8836,
+      "step": 2153
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7634941516215903e-05,
+      "loss": 0.8661,
+      "step": 2154
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7632544478843756e-05,
+      "loss": 0.8718,
+      "step": 2155
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7630146390447794e-05,
+      "loss": 0.8948,
+      "step": 2156
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7627747251358246e-05,
+      "loss": 0.9169,
+      "step": 2157
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7625347061905475e-05,
+      "loss": 0.9021,
+      "step": 2158
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7622945822419996e-05,
+      "loss": 0.8384,
+      "step": 2159
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7620543533232468e-05,
+      "loss": 0.8998,
+      "step": 2160
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.761814019467369e-05,
+      "loss": 0.8563,
+      "step": 2161
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7615735807074616e-05,
+      "loss": 0.8896,
+      "step": 2162
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7613330370766326e-05,
+      "loss": 0.8887,
+      "step": 2163
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7610923886080066e-05,
+      "loss": 0.8478,
+      "step": 2164
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.760851635334721e-05,
+      "loss": 0.8368,
+      "step": 2165
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.760610777289929e-05,
+      "loss": 0.8106,
+      "step": 2166
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7603698145067962e-05,
+      "loss": 0.9666,
+      "step": 2167
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.760128747018505e-05,
+      "loss": 0.8488,
+      "step": 2168
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7598875748582504e-05,
+      "loss": 0.81,
+      "step": 2169
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.759646298059243e-05,
+      "loss": 0.8019,
+      "step": 2170
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7594049166547073e-05,
+      "loss": 0.7785,
+      "step": 2171
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.759163430677882e-05,
+      "loss": 0.8108,
+      "step": 2172
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.75892184016202e-05,
+      "loss": 0.868,
+      "step": 2173
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.75868014514039e-05,
+      "loss": 0.8381,
+      "step": 2174
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7584383456462736e-05,
+      "loss": 0.7827,
+      "step": 2175
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7581964417129673e-05,
+      "loss": 0.8427,
+      "step": 2176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7579544333737815e-05,
+      "loss": 0.7852,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7577123206620427e-05,
+      "loss": 0.7756,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7574701036110894e-05,
+      "loss": 0.8981,
+      "step": 2179
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7572277822542763e-05,
+      "loss": 0.8653,
+      "step": 2180
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.756985356624971e-05,
+      "loss": 0.7999,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.7567428267565572e-05,
+      "loss": 0.8061,
+      "step": 2182
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8728,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 500,
+  "total_flos": 5.489446229316207e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}