diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6012 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999373158653545,
+  "eval_steps": 500,
+  "global_step": 997,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.6301,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.613,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 1.6397,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.5255,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.3498,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 1.2065,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 1.2296,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.1899,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 1.2451,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.1861,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.1463,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 1.1297,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 1.1189,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.0832,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 1.0713,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 1.0351,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.666666666666667e-05,
+      "loss": 1.0162,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6e-05,
+      "loss": 1.0099,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.333333333333333e-05,
+      "loss": 0.9961,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.9916,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7e-05,
+      "loss": 0.9658,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.9589,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.666666666666667e-05,
+      "loss": 0.9343,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8e-05,
+      "loss": 0.912,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 0.9183,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.9065,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9e-05,
+      "loss": 0.9022,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 0.8867,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.666666666666667e-05,
+      "loss": 0.8862,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0001,
+      "loss": 0.8906,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.999973613218314e-05,
+      "loss": 0.8628,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.999894453151759e-05,
+      "loss": 0.8386,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.99976252063585e-05,
+      "loss": 0.8607,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.99957781706309e-05,
+      "loss": 0.8601,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.999340344382979e-05,
+      "loss": 0.8406,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.999050105101971e-05,
+      "loss": 0.8355,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.998707102283458e-05,
+      "loss": 0.8266,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.998311339547734e-05,
+      "loss": 0.8367,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.997862821071965e-05,
+      "loss": 0.8311,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.997361551590132e-05,
+      "loss": 0.8274,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.996807536392989e-05,
+      "loss": 0.8355,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.996200781328011e-05,
+      "loss": 0.8169,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.99554129279932e-05,
+      "loss": 0.8185,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.994829077767631e-05,
+      "loss": 0.8062,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.994064143750167e-05,
+      "loss": 0.82,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.993246498820588e-05,
+      "loss": 0.7828,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.992376151608898e-05,
+      "loss": 0.7934,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.991453111301365e-05,
+      "loss": 0.7952,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.990477387640416e-05,
+      "loss": 0.7872,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.989448990924528e-05,
+      "loss": 0.7893,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.988367932008138e-05,
+      "loss": 0.8265,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.98723422230151e-05,
+      "loss": 0.7713,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.986047873770625e-05,
+      "loss": 0.791,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.98480889893705e-05,
+      "loss": 0.8084,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.983517310877811e-05,
+      "loss": 0.8018,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.982173123225243e-05,
+      "loss": 0.8045,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.980776350166867e-05,
+      "loss": 0.7996,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.979327006445218e-05,
+      "loss": 0.8036,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.977825107357703e-05,
+      "loss": 0.7881,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.976270668756434e-05,
+      "loss": 0.78,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.974663707048067e-05,
+      "loss": 0.7747,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.973004239193619e-05,
+      "loss": 0.7941,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.971292282708295e-05,
+      "loss": 0.7717,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.969527855661308e-05,
+      "loss": 0.7813,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.967710976675675e-05,
+      "loss": 0.7707,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.965841664928033e-05,
+      "loss": 0.7942,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.96391994014843e-05,
+      "loss": 0.777,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.961945822620119e-05,
+      "loss": 0.7814,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.959919333179344e-05,
+      "loss": 0.7627,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.957840493215117e-05,
+      "loss": 0.7617,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.955709324668999e-05,
+      "loss": 0.7801,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.953525850034857e-05,
+      "loss": 0.7707,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.951290092358644e-05,
+      "loss": 0.7683,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.949002075238139e-05,
+      "loss": 0.7533,
+      "step": 74
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.946661822822702e-05,
+      "loss": 0.7522,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.944269359813027e-05,
+      "loss": 0.7559,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.941824711460873e-05,
+      "loss": 0.7494,
+      "step": 77
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.939327903568802e-05,
+      "loss": 0.7425,
+      "step": 78
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.936778962489902e-05,
+      "loss": 0.7245,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.934177915127517e-05,
+      "loss": 0.7406,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.931524788934949e-05,
+      "loss": 0.7463,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.928819611915189e-05,
+      "loss": 0.742,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.926062412620601e-05,
+      "loss": 0.7425,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.923253220152629e-05,
+      "loss": 0.7476,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.920392064161492e-05,
+      "loss": 0.7305,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.917478974845874e-05,
+      "loss": 0.7398,
+      "step": 86
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.914513982952593e-05,
+      "loss": 0.7409,
+      "step": 87
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.911497119776286e-05,
+      "loss": 0.731,
+      "step": 88
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.908428417159077e-05,
+      "loss": 0.7361,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.905307907490242e-05,
+      "loss": 0.7273,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.902135623705864e-05,
+      "loss": 0.7208,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.898911599288485e-05,
+      "loss": 0.7311,
+      "step": 92
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.895635868266756e-05,
+      "loss": 0.7232,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.892308465215078e-05,
+      "loss": 0.7129,
+      "step": 94
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.888929425253236e-05,
+      "loss": 0.7192,
+      "step": 95
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.885498784046023e-05,
+      "loss": 0.7121,
+      "step": 96
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.882016577802873e-05,
+      "loss": 0.7137,
+      "step": 97
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.87848284327747e-05,
+      "loss": 0.7201,
+      "step": 98
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.874897617767369e-05,
+      "loss": 0.7196,
+      "step": 99
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.871260939113595e-05,
+      "loss": 0.7037,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.867572845700246e-05,
+      "loss": 0.7147,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.863833376454087e-05,
+      "loss": 0.7234,
+      "step": 102
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.860042570844142e-05,
+      "loss": 0.7442,
+      "step": 103
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.856200468881276e-05,
+      "loss": 0.7345,
+      "step": 104
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.852307111117772e-05,
+      "loss": 0.7261,
+      "step": 105
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.848362538646898e-05,
+      "loss": 0.7123,
+      "step": 106
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.844366793102488e-05,
+      "loss": 0.7019,
+      "step": 107
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.840319916658488e-05,
+      "loss": 0.7102,
+      "step": 108
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.836221952028512e-05,
+      "loss": 0.7051,
+      "step": 109
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.832072942465404e-05,
+      "loss": 0.7259,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.827872931760765e-05,
+      "loss": 0.7194,
+      "step": 111
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.8236219642445e-05,
+      "loss": 0.7213,
+      "step": 112
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.819320084784351e-05,
+      "loss": 0.7122,
+      "step": 113
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.814967338785424e-05,
+      "loss": 0.6899,
+      "step": 114
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.810563772189696e-05,
+      "loss": 0.6893,
+      "step": 115
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.806109431475549e-05,
+      "loss": 0.6745,
+      "step": 116
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.801604363657271e-05,
+      "loss": 0.6712,
+      "step": 117
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.797048616284559e-05,
+      "loss": 0.6738,
+      "step": 118
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.792442237442013e-05,
+      "loss": 0.6961,
+      "step": 119
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.787785275748644e-05,
+      "loss": 0.6838,
+      "step": 120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.78307778035734e-05,
+      "loss": 0.6789,
+      "step": 121
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.778319800954364e-05,
+      "loss": 0.6794,
+      "step": 122
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.773511387758823e-05,
+      "loss": 0.6669,
+      "step": 123
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.768652591522134e-05,
+      "loss": 0.6656,
+      "step": 124
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.763743463527497e-05,
+      "loss": 0.685,
+      "step": 125
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.758784055589346e-05,
+      "loss": 0.6697,
+      "step": 126
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.753774420052809e-05,
+      "loss": 0.6556,
+      "step": 127
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.748714609793148e-05,
+      "loss": 0.6776,
+      "step": 128
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.743604678215207e-05,
+      "loss": 0.6713,
+      "step": 129
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.738444679252845e-05,
+      "loss": 0.6724,
+      "step": 130
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.733234667368369e-05,
+      "loss": 0.6644,
+      "step": 131
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.72797469755196e-05,
+      "loss": 0.6703,
+      "step": 132
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.722664825321082e-05,
+      "loss": 0.6726,
+      "step": 133
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.717305106719916e-05,
+      "loss": 0.6643,
+      "step": 134
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.71189559831875e-05,
+      "loss": 0.6999,
+      "step": 135
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.706436357213391e-05,
+      "loss": 0.7048,
+      "step": 136
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.70092744102456e-05,
+      "loss": 0.6898,
+      "step": 137
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.695368907897287e-05,
+      "loss": 0.7053,
+      "step": 138
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.689760816500286e-05,
+      "loss": 0.6914,
+      "step": 139
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.684103226025357e-05,
+      "loss": 0.6732,
+      "step": 140
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.678396196186739e-05,
+      "loss": 0.6866,
+      "step": 141
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.67263978722049e-05,
+      "loss": 0.6882,
+      "step": 142
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.666834059883858e-05,
+      "loss": 0.6836,
+      "step": 143
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.660979075454622e-05,
+      "loss": 0.6881,
+      "step": 144
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.655074895730462e-05,
+      "loss": 0.6988,
+      "step": 145
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.6491215830283e-05,
+      "loss": 0.6973,
+      "step": 146
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.643119200183639e-05,
+      "loss": 0.6976,
+      "step": 147
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.637067810549907e-05,
+      "loss": 0.6809,
+      "step": 148
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.630967477997781e-05,
+      "loss": 0.6832,
+      "step": 149
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.624818266914521e-05,
+      "loss": 0.684,
+      "step": 150
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.618620242203281e-05,
+      "loss": 0.691,
+      "step": 151
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.612373469282429e-05,
+      "loss": 0.6893,
+      "step": 152
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.606078014084862e-05,
+      "loss": 0.6835,
+      "step": 153
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.599733943057301e-05,
+      "loss": 0.7082,
+      "step": 154
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.59334132315959e-05,
+      "loss": 0.7115,
+      "step": 155
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.586900221863997e-05,
+      "loss": 0.7276,
+      "step": 156
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.580410707154494e-05,
+      "loss": 0.7335,
+      "step": 157
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.573872847526049e-05,
+      "loss": 0.7216,
+      "step": 158
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.567286711983885e-05,
+      "loss": 0.7112,
+      "step": 159
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.560652370042773e-05,
+      "loss": 0.7021,
+      "step": 160
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.553969891726289e-05,
+      "loss": 0.7004,
+      "step": 161
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.547239347566069e-05,
+      "loss": 0.7122,
+      "step": 162
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.54046080860107e-05,
+      "loss": 0.6966,
+      "step": 163
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.533634346376827e-05,
+      "loss": 0.7006,
+      "step": 164
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.526760032944687e-05,
+      "loss": 0.7121,
+      "step": 165
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.519837940861052e-05,
+      "loss": 0.7116,
+      "step": 166
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.512868143186615e-05,
+      "loss": 0.717,
+      "step": 167
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.505850713485589e-05,
+      "loss": 0.7099,
+      "step": 168
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.498785725824928e-05,
+      "loss": 0.7066,
+      "step": 169
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.491673254773545e-05,
+      "loss": 0.7165,
+      "step": 170
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.484513375401531e-05,
+      "loss": 0.716,
+      "step": 171
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.477306163279354e-05,
+      "loss": 0.7103,
+      "step": 172
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.470051694477067e-05,
+      "loss": 0.6948,
+      "step": 173
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.462750045563504e-05,
+      "loss": 0.6992,
+      "step": 174
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.455401293605471e-05,
+      "loss": 0.7054,
+      "step": 175
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.448005516166934e-05,
+      "loss": 0.695,
+      "step": 176
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.4405627913082e-05,
+      "loss": 0.6962,
+      "step": 177
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.433073197585089e-05,
+      "loss": 0.6829,
+      "step": 178
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.425536814048112e-05,
+      "loss": 0.7062,
+      "step": 179
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.417953720241633e-05,
+      "loss": 0.7001,
+      "step": 180
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.410323996203027e-05,
+      "loss": 0.6793,
+      "step": 181
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.40264772246184e-05,
+      "loss": 0.6935,
+      "step": 182
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.394924980038932e-05,
+      "loss": 0.6768,
+      "step": 183
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.387155850445634e-05,
+      "loss": 0.6954,
+      "step": 184
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.379340415682877e-05,
+      "loss": 0.6928,
+      "step": 185
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.371478758240328e-05,
+      "loss": 0.6856,
+      "step": 186
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.363570961095524e-05,
+      "loss": 0.6775,
+      "step": 187
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.355617107712989e-05,
+      "loss": 0.6858,
+      "step": 188
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.347617282043362e-05,
+      "loss": 0.6843,
+      "step": 189
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.339571568522503e-05,
+      "loss": 0.6834,
+      "step": 190
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.331480052070607e-05,
+      "loss": 0.6889,
+      "step": 191
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.323342818091307e-05,
+      "loss": 0.6793,
+      "step": 192
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.315159952470767e-05,
+      "loss": 0.6944,
+      "step": 193
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.306931541576784e-05,
+      "loss": 0.6899,
+      "step": 194
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.298657672257872e-05,
+      "loss": 0.691,
+      "step": 195
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.29033843184234e-05,
+      "loss": 0.7019,
+      "step": 196
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.281973908137385e-05,
+      "loss": 0.6926,
+      "step": 197
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.27356418942815e-05,
+      "loss": 0.6956,
+      "step": 198
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.265109364476798e-05,
+      "loss": 0.6944,
+      "step": 199
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.256609522521578e-05,
+      "loss": 0.7094,
+      "step": 200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.248064753275881e-05,
+      "loss": 0.7016,
+      "step": 201
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.23947514692729e-05,
+      "loss": 0.6918,
+      "step": 202
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.23084079413663e-05,
+      "loss": 0.6843,
+      "step": 203
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.222161786037018e-05,
+      "loss": 0.6985,
+      "step": 204
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.213438214232888e-05,
+      "loss": 0.6942,
+      "step": 205
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.204670170799034e-05,
+      "loss": 0.7057,
+      "step": 206
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.195857748279637e-05,
+      "loss": 0.7007,
+      "step": 207
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.187001039687284e-05,
+      "loss": 0.6914,
+      "step": 208
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.178100138501987e-05,
+      "loss": 0.6923,
+      "step": 209
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.169155138670203e-05,
+      "loss": 0.6707,
+      "step": 210
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.160166134603834e-05,
+      "loss": 0.6871,
+      "step": 211
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.151133221179236e-05,
+      "loss": 0.6924,
+      "step": 212
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.142056493736215e-05,
+      "loss": 0.6901,
+      "step": 213
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.132936048077021e-05,
+      "loss": 0.6847,
+      "step": 214
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.123771980465336e-05,
+      "loss": 0.6872,
+      "step": 215
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.114564387625262e-05,
+      "loss": 0.6693,
+      "step": 216
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.105313366740296e-05,
+      "loss": 0.6743,
+      "step": 217
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.096019015452304e-05,
+      "loss": 0.6903,
+      "step": 218
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.086681431860493e-05,
+      "loss": 0.6835,
+      "step": 219
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.077300714520379e-05,
+      "loss": 0.6684,
+      "step": 220
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.067876962442733e-05,
+      "loss": 0.6885,
+      "step": 221
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.058410275092554e-05,
+      "loss": 0.6779,
+      "step": 222
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.048900752388006e-05,
+      "loss": 0.6814,
+      "step": 223
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.039348494699367e-05,
+      "loss": 0.6755,
+      "step": 224
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.029753602847975e-05,
+      "loss": 0.6888,
+      "step": 225
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.020116178105154e-05,
+      "loss": 0.6759,
+      "step": 226
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.010436322191156e-05,
+      "loss": 0.6837,
+      "step": 227
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.000714137274078e-05,
+      "loss": 0.6797,
+      "step": 228
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.990949725968786e-05,
+      "loss": 0.6991,
+      "step": 229
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.981143191335839e-05,
+      "loss": 0.6731,
+      "step": 230
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.971294636880392e-05,
+      "loss": 0.686,
+      "step": 231
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.961404166551103e-05,
+      "loss": 0.6837,
+      "step": 232
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.95147188473905e-05,
+      "loss": 0.6758,
+      "step": 233
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.941497896276614e-05,
+      "loss": 0.6794,
+      "step": 234
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.931482306436374e-05,
+      "loss": 0.6683,
+      "step": 235
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.921425220930002e-05,
+      "loss": 0.6717,
+      "step": 236
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.911326745907152e-05,
+      "loss": 0.6794,
+      "step": 237
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.90118698795432e-05,
+      "loss": 0.6652,
+      "step": 238
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.89100605409374e-05,
+      "loss": 0.6784,
+      "step": 239
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.880784051782243e-05,
+      "loss": 0.6803,
+      "step": 240
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.870521088910129e-05,
+      "loss": 0.6814,
+      "step": 241
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.860217273800022e-05,
+      "loss": 0.6778,
+      "step": 242
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.849872715205725e-05,
+      "loss": 0.6921,
+      "step": 243
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.839487522311087e-05,
+      "loss": 0.6812,
+      "step": 244
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.829061804728834e-05,
+      "loss": 0.6767,
+      "step": 245
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.818595672499419e-05,
+      "loss": 0.6898,
+      "step": 246
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.808089236089858e-05,
+      "loss": 0.6683,
+      "step": 247
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.797542606392573e-05,
+      "loss": 0.6508,
+      "step": 248
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.786955894724207e-05,
+      "loss": 0.6482,
+      "step": 249
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.776329212824462e-05,
+      "loss": 0.6472,
+      "step": 250
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.765662672854908e-05,
+      "loss": 0.6587,
+      "step": 251
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.754956387397815e-05,
+      "loss": 0.6409,
+      "step": 252
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.744210469454946e-05,
+      "loss": 0.6479,
+      "step": 253
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.733425032446381e-05,
+      "loss": 0.6307,
+      "step": 254
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.722600190209303e-05,
+      "loss": 0.6273,
+      "step": 255
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.711736056996817e-05,
+      "loss": 0.6423,
+      "step": 256
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.700832747476726e-05,
+      "loss": 0.6407,
+      "step": 257
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.689890376730327e-05,
+      "loss": 0.6317,
+      "step": 258
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.678909060251201e-05,
+      "loss": 0.6315,
+      "step": 259
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.667888913943989e-05,
+      "loss": 0.6411,
+      "step": 260
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.656830054123169e-05,
+      "loss": 0.6323,
+      "step": 261
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.645732597511827e-05,
+      "loss": 0.6505,
+      "step": 262
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.634596661240428e-05,
+      "loss": 0.6474,
+      "step": 263
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.623422362845583e-05,
+      "loss": 0.6388,
+      "step": 264
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.612209820268799e-05,
+      "loss": 0.6355,
+      "step": 265
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.600959151855241e-05,
+      "loss": 0.6442,
+      "step": 266
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.589670476352485e-05,
+      "loss": 0.6309,
+      "step": 267
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.578343912909252e-05,
+      "loss": 0.6438,
+      "step": 268
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.566979581074168e-05,
+      "loss": 0.645,
+      "step": 269
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.555577600794489e-05,
+      "loss": 0.6322,
+      "step": 270
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.544138092414842e-05,
+      "loss": 0.6283,
+      "step": 271
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.532661176675951e-05,
+      "loss": 0.6294,
+      "step": 272
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.521146974713363e-05,
+      "loss": 0.6446,
+      "step": 273
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.509595608056171e-05,
+      "loss": 0.6236,
+      "step": 274
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.498007198625733e-05,
+      "loss": 0.6406,
+      "step": 275
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.486381868734379e-05,
+      "loss": 0.6341,
+      "step": 276
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.474719741084128e-05,
+      "loss": 0.6208,
+      "step": 277
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.463020938765384e-05,
+      "loss": 0.6352,
+      "step": 278
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.451285585255646e-05,
+      "loss": 0.633,
+      "step": 279
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.439513804418197e-05,
+      "loss": 0.6259,
+      "step": 280
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.4277057205008e-05,
+      "loss": 0.6272,
+      "step": 281
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.415861458134394e-05,
+      "loss": 0.6166,
+      "step": 282
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.403981142331759e-05,
+      "loss": 0.6316,
+      "step": 283
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.392064898486216e-05,
+      "loss": 0.6368,
+      "step": 284
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.380112852370297e-05,
+      "loss": 0.63,
+      "step": 285
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.368125130134414e-05,
+      "loss": 0.6435,
+      "step": 286
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.35610185830553e-05,
+      "loss": 0.6235,
+      "step": 287
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.344043163785824e-05,
+      "loss": 0.626,
+      "step": 288
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.331949173851355e-05,
+      "loss": 0.6258,
+      "step": 289
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.319820016150706e-05,
+      "loss": 0.6421,
+      "step": 290
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.307655818703658e-05,
+      "loss": 0.6284,
+      "step": 291
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.295456709899816e-05,
+      "loss": 0.6321,
+      "step": 292
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.28322281849727e-05,
+      "loss": 0.6232,
+      "step": 293
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.270954273621229e-05,
+      "loss": 0.6185,
+      "step": 294
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.258651204762657e-05,
+      "loss": 0.6315,
+      "step": 295
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.246313741776911e-05,
+      "loss": 0.6429,
+      "step": 296
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.23394201488237e-05,
+      "loss": 0.6334,
+      "step": 297
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.221536154659056e-05,
+      "loss": 0.6392,
+      "step": 298
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.209096292047257e-05,
+      "loss": 0.622,
+      "step": 299
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.196622558346152e-05,
+      "loss": 0.6114,
+      "step": 300
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.184115085212413e-05,
+      "loss": 0.6422,
+      "step": 301
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.171574004658829e-05,
+      "loss": 0.633,
+      "step": 302
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.158999449052898e-05,
+      "loss": 0.6259,
+      "step": 303
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.146391551115444e-05,
+      "loss": 0.6351,
+      "step": 304
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.133750443919206e-05,
+      "loss": 0.626,
+      "step": 305
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.121076260887437e-05,
+      "loss": 0.624,
+      "step": 306
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.108369135792499e-05,
+      "loss": 0.6371,
+      "step": 307
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.095629202754448e-05,
+      "loss": 0.6424,
+      "step": 308
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.082856596239614e-05,
+      "loss": 0.631,
+      "step": 309
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.070051451059188e-05,
+      "loss": 0.6446,
+      "step": 310
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.057213902367801e-05,
+      "loss": 0.6417,
+      "step": 311
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.044344085662092e-05,
+      "loss": 0.6504,
+      "step": 312
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.031442136779273e-05,
+      "loss": 0.6408,
+      "step": 313
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.018508191895712e-05,
+      "loss": 0.6318,
+      "step": 314
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.005542387525479e-05,
+      "loss": 0.6375,
+      "step": 315
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.992544860518916e-05,
+      "loss": 0.6298,
+      "step": 316
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.979515748061182e-05,
+      "loss": 0.6422,
+      "step": 317
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.966455187670819e-05,
+      "loss": 0.6487,
+      "step": 318
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.953363317198287e-05,
+      "loss": 0.6522,
+      "step": 319
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.940240274824519e-05,
+      "loss": 0.6446,
+      "step": 320
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.927086199059456e-05,
+      "loss": 0.6435,
+      "step": 321
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.91390122874059e-05,
+      "loss": 0.6322,
+      "step": 322
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.900685503031491e-05,
+      "loss": 0.6343,
+      "step": 323
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.887439161420347e-05,
+      "loss": 0.6372,
+      "step": 324
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.87416234371849e-05,
+      "loss": 0.6418,
+      "step": 325
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.860855190058913e-05,
+      "loss": 0.64,
+      "step": 326
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.847517840894805e-05,
+      "loss": 0.6504,
+      "step": 327
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.834150436998047e-05,
+      "loss": 0.6411,
+      "step": 328
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.82075311945775e-05,
+      "loss": 0.6241,
+      "step": 329
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.807326029678754e-05,
+      "loss": 0.5894,
+      "step": 330
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.793869309380129e-05,
+      "loss": 0.6057,
+      "step": 331
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.780383100593693e-05,
+      "loss": 0.5981,
+      "step": 332
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.766867545662506e-05,
+      "loss": 0.5975,
+      "step": 333
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.753322787239365e-05,
+      "loss": 0.5886,
+      "step": 334
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.739748968285305e-05,
+      "loss": 0.6032,
+      "step": 335
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.726146232068083e-05,
+      "loss": 0.5985,
+      "step": 336
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.712514722160674e-05,
+      "loss": 0.6043,
+      "step": 337
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.698854582439746e-05,
+      "loss": 0.5909,
+      "step": 338
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.685165957084148e-05,
+      "loss": 0.614,
+      "step": 339
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.671448990573393e-05,
+      "loss": 0.6095,
+      "step": 340
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.657703827686115e-05,
+      "loss": 0.598,
+      "step": 341
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.643930613498563e-05,
+      "loss": 0.6007,
+      "step": 342
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.630129493383052e-05,
+      "loss": 0.5979,
+      "step": 343
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.616300613006443e-05,
+      "loss": 0.5995,
+      "step": 344
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.602444118328594e-05,
+      "loss": 0.6011,
+      "step": 345
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.588560155600824e-05,
+      "loss": 0.6084,
+      "step": 346
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.57464887136437e-05,
+      "loss": 0.5883,
+      "step": 347
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.560710412448839e-05,
+      "loss": 0.6031,
+      "step": 348
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.546744925970664e-05,
+      "loss": 0.5964,
+      "step": 349
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.532752559331539e-05,
+      "loss": 0.5996,
+      "step": 350
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.518733460216876e-05,
+      "loss": 0.5872,
+      "step": 351
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.504687776594234e-05,
+      "loss": 0.576,
+      "step": 352
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.490615656711773e-05,
+      "loss": 0.5794,
+      "step": 353
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.476517249096671e-05,
+      "loss": 0.5889,
+      "step": 354
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.462392702553572e-05,
+      "loss": 0.5805,
+      "step": 355
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.448242166163005e-05,
+      "loss": 0.5822,
+      "step": 356
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.434065789279816e-05,
+      "loss": 0.571,
+      "step": 357
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.41986372153159e-05,
+      "loss": 0.5816,
+      "step": 358
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.405636112817073e-05,
+      "loss": 0.5807,
+      "step": 359
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.391383113304583e-05,
+      "loss": 0.5744,
+      "step": 360
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.377104873430439e-05,
+      "loss": 0.588,
+      "step": 361
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.362801543897358e-05,
+      "loss": 0.5941,
+      "step": 362
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.348473275672874e-05,
+      "loss": 0.585,
+      "step": 363
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.334120219987741e-05,
+      "loss": 0.5858,
+      "step": 364
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.31974252833434e-05,
+      "loss": 0.5865,
+      "step": 365
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.305340352465071e-05,
+      "loss": 0.5847,
+      "step": 366
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.290913844390765e-05,
+      "loss": 0.5798,
+      "step": 367
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.27646315637907e-05,
+      "loss": 0.5761,
+      "step": 368
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.261988440952844e-05,
+      "loss": 0.58,
+      "step": 369
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.247489850888551e-05,
+      "loss": 0.5828,
+      "step": 370
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.232967539214643e-05,
+      "loss": 0.6166,
+      "step": 371
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.218421659209948e-05,
+      "loss": 0.6039,
+      "step": 372
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.203852364402048e-05,
+      "loss": 0.6079,
+      "step": 373
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.189259808565665e-05,
+      "loss": 0.6137,
+      "step": 374
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.17464414572103e-05,
+      "loss": 0.5941,
+      "step": 375
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.16000553013227e-05,
+      "loss": 0.6014,
+      "step": 376
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.145344116305763e-05,
+      "loss": 0.6202,
+      "step": 377
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.130660058988521e-05,
+      "loss": 0.611,
+      "step": 378
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.115953513166549e-05,
+      "loss": 0.6072,
+      "step": 379
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.101224634063213e-05,
+      "loss": 0.6153,
+      "step": 380
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.086473577137597e-05,
+      "loss": 0.6034,
+      "step": 381
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.071700498082873e-05,
+      "loss": 0.5943,
+      "step": 382
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.056905552824645e-05,
+      "loss": 0.6201,
+      "step": 383
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.042088897519307e-05,
+      "loss": 0.605,
+      "step": 384
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.027250688552399e-05,
+      "loss": 0.5969,
+      "step": 385
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.012391082536956e-05,
+      "loss": 0.5917,
+      "step": 386
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.997510236311846e-05,
+      "loss": 0.5935,
+      "step": 387
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.982608306940128e-05,
+      "loss": 0.6003,
+      "step": 388
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.967685451707384e-05,
+      "loss": 0.6101,
+      "step": 389
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.952741828120062e-05,
+      "loss": 0.6083,
+      "step": 390
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.937777593903818e-05,
+      "loss": 0.5976,
+      "step": 391
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.922792907001842e-05,
+      "loss": 0.5976,
+      "step": 392
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 6.9077879255732e-05,
+      "loss": 0.6051,
+      "step": 393
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.892762807991159e-05,
+      "loss": 0.5825,
+      "step": 394
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.87771771284152e-05,
+      "loss": 0.5876,
+      "step": 395
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.862652798920939e-05,
+      "loss": 0.5921,
+      "step": 396
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.84756822523525e-05,
+      "loss": 0.5942,
+      "step": 397
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.8324641509978e-05,
+      "loss": 0.5922,
+      "step": 398
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.817340735627747e-05,
+      "loss": 0.5858,
+      "step": 399
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.802198138748398e-05,
+      "loss": 0.5785,
+      "step": 400
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.787036520185511e-05,
+      "loss": 0.5951,
+      "step": 401
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.771856039965615e-05,
+      "loss": 0.587,
+      "step": 402
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.756656858314318e-05,
+      "loss": 0.5922,
+      "step": 403
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.741439135654613e-05,
+      "loss": 0.5884,
+      "step": 404
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.726203032605191e-05,
+      "loss": 0.5864,
+      "step": 405
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.710948709978741e-05,
+      "loss": 0.601,
+      "step": 406
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.695676328780257e-05,
+      "loss": 0.5912,
+      "step": 407
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.680386050205332e-05,
+      "loss": 0.6002,
+      "step": 408
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.665078035638466e-05,
+      "loss": 0.6011,
+      "step": 409
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.649752446651352e-05,
+      "loss": 0.6005,
+      "step": 410
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.63440944500118e-05,
+      "loss": 0.5906,
+      "step": 411
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.619049192628924e-05,
+      "loss": 0.5822,
+      "step": 412
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.603671851657634e-05,
+      "loss": 0.5928,
+      "step": 413
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.588277584390726e-05,
+      "loss": 0.5949,
+      "step": 414
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.572866553310266e-05,
+      "loss": 0.5793,
+      "step": 415
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.557438921075258e-05,
+      "loss": 0.5953,
+      "step": 416
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.541994850519934e-05,
+      "loss": 0.5858,
+      "step": 417
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.526534504652014e-05,
+      "loss": 0.5986,
+      "step": 418
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.511058046651011e-05,
+      "loss": 0.579,
+      "step": 419
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.49556563986649e-05,
+      "loss": 0.591,
+      "step": 420
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.480057447816356e-05,
+      "loss": 0.5856,
+      "step": 421
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.464533634185117e-05,
+      "loss": 0.5881,
+      "step": 422
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.448994362822168e-05,
+      "loss": 0.5859,
+      "step": 423
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.433439797740049e-05,
+      "loss": 0.5736,
+      "step": 424
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.417870103112732e-05,
+      "loss": 0.5977,
+      "step": 425
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.402285443273865e-05,
+      "loss": 0.5927,
+      "step": 426
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.386685982715057e-05,
+      "loss": 0.5908,
+      "step": 427
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.371071886084132e-05,
+      "loss": 0.5823,
+      "step": 428
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.355443318183394e-05,
+      "loss": 0.597,
+      "step": 429
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.339800443967884e-05,
+      "loss": 0.5977,
+      "step": 430
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.324143428543647e-05,
+      "loss": 0.5826,
+      "step": 431
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.308472437165982e-05,
+      "loss": 0.5909,
+      "step": 432
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.292787635237701e-05,
+      "loss": 0.5981,
+      "step": 433
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.27708918830738e-05,
+      "loss": 0.5735,
+      "step": 434
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.261377262067614e-05,
+      "loss": 0.5804,
+      "step": 435
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.245652022353276e-05,
+      "loss": 0.5884,
+      "step": 436
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.229913635139749e-05,
+      "loss": 0.583,
+      "step": 437
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.214162266541188e-05,
+      "loss": 0.5801,
+      "step": 438
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.198398082808765e-05,
+      "loss": 0.5868,
+      "step": 439
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.182621250328905e-05,
+      "loss": 0.5784,
+      "step": 440
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.166831935621547e-05,
+      "loss": 0.5871,
+      "step": 441
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.151030305338368e-05,
+      "loss": 0.5824,
+      "step": 442
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.135216526261036e-05,
+      "loss": 0.5928,
+      "step": 443
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.119390765299448e-05,
+      "loss": 0.6104,
+      "step": 444
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.10355318948996e-05,
+      "loss": 0.5927,
+      "step": 445
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.0877039659936366e-05,
+      "loss": 0.5937,
+      "step": 446
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.071843262094477e-05,
+      "loss": 0.5949,
+      "step": 447
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.0559712451976525e-05,
+      "loss": 0.5977,
+      "step": 448
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.040088082827744e-05,
+      "loss": 0.5865,
+      "step": 449
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.0241939426269625e-05,
+      "loss": 0.6103,
+      "step": 450
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.008288992353396e-05,
+      "loss": 0.6038,
+      "step": 451
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.992373399879221e-05,
+      "loss": 0.6117,
+      "step": 452
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.976447333188945e-05,
+      "loss": 0.5909,
+      "step": 453
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.9605109603776267e-05,
+      "loss": 0.6105,
+      "step": 454
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.944564449649099e-05,
+      "loss": 0.596,
+      "step": 455
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.928607969314201e-05,
+      "loss": 0.5953,
+      "step": 456
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.9126416877890026e-05,
+      "loss": 0.5806,
+      "step": 457
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.896665773593013e-05,
+      "loss": 0.6111,
+      "step": 458
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8806803953474186e-05,
+      "loss": 0.5902,
+      "step": 459
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8646857217732934e-05,
+      "loss": 0.6033,
+      "step": 460
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8486819216898205e-05,
+      "loss": 0.5999,
+      "step": 461
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.832669164012513e-05,
+      "loss": 0.5926,
+      "step": 462
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8166476177514254e-05,
+      "loss": 0.6135,
+      "step": 463
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.800617452009376e-05,
+      "loss": 0.6171,
+      "step": 464
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.784578835980158e-05,
+      "loss": 0.6211,
+      "step": 465
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7685319389467565e-05,
+      "loss": 0.601,
+      "step": 466
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.752476930279558e-05,
+      "loss": 0.6164,
+      "step": 467
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7364139794345663e-05,
+      "loss": 0.595,
+      "step": 468
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7203432559516114e-05,
+      "loss": 0.6095,
+      "step": 469
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.704264929452563e-05,
+      "loss": 0.5944,
+      "step": 470
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.688179169639538e-05,
+      "loss": 0.6036,
+      "step": 471
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.6720861462931086e-05,
+      "loss": 0.6211,
+      "step": 472
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.6559860292705114e-05,
+      "loss": 0.6038,
+      "step": 473
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.6398789885038587e-05,
+      "loss": 0.6079,
+      "step": 474
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.623765193998334e-05,
+      "loss": 0.61,
+      "step": 475
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.607644815830412e-05,
+      "loss": 0.6111,
+      "step": 476
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.59151802414605e-05,
+      "loss": 0.6076,
+      "step": 477
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5753849891589e-05,
+      "loss": 0.6059,
+      "step": 478
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5592458811485135e-05,
+      "loss": 0.6104,
+      "step": 479
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.543100870458538e-05,
+      "loss": 0.5947,
+      "step": 480
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.5269501274949184e-05,
+      "loss": 0.599,
+      "step": 481
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.510793822724111e-05,
+      "loss": 0.5925,
+      "step": 482
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.494632126671269e-05,
+      "loss": 0.5961,
+      "step": 483
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.4784652099184495e-05,
+      "loss": 0.6004,
+      "step": 484
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.462293243102816e-05,
+      "loss": 0.5953,
+      "step": 485
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.4461163969148296e-05,
+      "loss": 0.6023,
+      "step": 486
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.4299348420964534e-05,
+      "loss": 0.5999,
+      "step": 487
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.413748749439351e-05,
+      "loss": 0.5961,
+      "step": 488
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.397558289783079e-05,
+      "loss": 0.6103,
+      "step": 489
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.381363634013286e-05,
+      "loss": 0.6091,
+      "step": 490
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.365164953059911e-05,
+      "loss": 0.5985,
+      "step": 491
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.348962417895379e-05,
+      "loss": 0.6126,
+      "step": 492
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.332756199532791e-05,
+      "loss": 0.6097,
+      "step": 493
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.3165464690241287e-05,
+      "loss": 0.6032,
+      "step": 494
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.300333397458436e-05,
+      "loss": 0.5986,
+      "step": 495
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.284117155960026e-05,
+      "loss": 0.606,
+      "step": 496
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.2678979156866684e-05,
+      "loss": 0.6082,
+      "step": 497
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.251675847827785e-05,
+      "loss": 0.6025,
+      "step": 498
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.2354511236026413e-05,
+      "loss": 0.606,
+      "step": 499
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.2192239142585384e-05,
+      "loss": 0.585,
+      "step": 500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.2029943910690095e-05,
+      "loss": 0.5995,
+      "step": 501
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.186762725332008e-05,
+      "loss": 0.6013,
+      "step": 502
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.1705290883681034e-05,
+      "loss": 0.6027,
+      "step": 503
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.154293651518667e-05,
+      "loss": 0.6076,
+      "step": 504
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.138056586144072e-05,
+      "loss": 0.6251,
+      "step": 505
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.121818063621877e-05,
+      "loss": 0.6343,
+      "step": 506
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.1055782553450214e-05,
+      "loss": 0.6204,
+      "step": 507
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.089337332720017e-05,
+      "loss": 0.6261,
+      "step": 508
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.073095467165134e-05,
+      "loss": 0.624,
+      "step": 509
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.0568528301085984e-05,
+      "loss": 0.6282,
+      "step": 510
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.040609592986776e-05,
+      "loss": 0.6265,
+      "step": 511
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.0243659272423674e-05,
+      "loss": 0.6348,
+      "step": 512
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.0081220043225974e-05,
+      "loss": 0.624,
+      "step": 513
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.9918779956774044e-05,
+      "loss": 0.621,
+      "step": 514
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.9756340727576344e-05,
+      "loss": 0.6347,
+      "step": 515
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.959390407013226e-05,
+      "loss": 0.6264,
+      "step": 516
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.943147169891403e-05,
+      "loss": 0.6273,
+      "step": 517
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.9269045328348665e-05,
+      "loss": 0.6314,
+      "step": 518
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.9106626672799835e-05,
+      "loss": 0.6198,
+      "step": 519
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.89442174465498e-05,
+      "loss": 0.6271,
+      "step": 520
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.878181936378124e-05,
+      "loss": 0.6404,
+      "step": 521
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8619434138559286e-05,
+      "loss": 0.6152,
+      "step": 522
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.8457063484813334e-05,
+      "loss": 0.613,
+      "step": 523
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8294709116318984e-05,
+      "loss": 0.6137,
+      "step": 524
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.813237274667993e-05,
+      "loss": 0.645,
+      "step": 525
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.797005608930991e-05,
+      "loss": 0.6602,
+      "step": 526
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.780776085741463e-05,
+      "loss": 0.6545,
+      "step": 527
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.76454887639736e-05,
+      "loss": 0.6569,
+      "step": 528
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.748324152172216e-05,
+      "loss": 0.6597,
+      "step": 529
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.732102084313333e-05,
+      "loss": 0.6759,
+      "step": 530
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.715882844039975e-05,
+      "loss": 0.6644,
+      "step": 531
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6996666025415654e-05,
+      "loss": 0.6666,
+      "step": 532
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.6834535309758725e-05,
+      "loss": 0.6646,
+      "step": 533
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.667243800467208e-05,
+      "loss": 0.6676,
+      "step": 534
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6510375821046206e-05,
+      "loss": 0.6472,
+      "step": 535
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.634835046940089e-05,
+      "loss": 0.6617,
+      "step": 536
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.618636365986714e-05,
+      "loss": 0.6683,
+      "step": 537
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.602441710216922e-05,
+      "loss": 0.6477,
+      "step": 538
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.5862512505606485e-05,
+      "loss": 0.6561,
+      "step": 539
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.570065157903546e-05,
+      "loss": 0.6547,
+      "step": 540
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.553883603085171e-05,
+      "loss": 0.652,
+      "step": 541
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.5377067568971844e-05,
+      "loss": 0.657,
+      "step": 542
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.5215347900815496e-05,
+      "loss": 0.6568,
+      "step": 543
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.505367873328731e-05,
+      "loss": 0.6546,
+      "step": 544
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4892061772758895e-05,
+      "loss": 0.6574,
+      "step": 545
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.473049872505082e-05,
+      "loss": 0.6732,
+      "step": 546
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.456899129541464e-05,
+      "loss": 0.6759,
+      "step": 547
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4407541188514863e-05,
+      "loss": 0.7034,
+      "step": 548
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4246150108410995e-05,
+      "loss": 0.6739,
+      "step": 549
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4084819758539506e-05,
+      "loss": 0.6815,
+      "step": 550
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.392355184169589e-05,
+      "loss": 0.6665,
+      "step": 551
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.376234806001666e-05,
+      "loss": 0.6723,
+      "step": 552
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.360121011496142e-05,
+      "loss": 0.6701,
+      "step": 553
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.34401397072949e-05,
+      "loss": 0.6613,
+      "step": 554
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.327913853706894e-05,
+      "loss": 0.6919,
+      "step": 555
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.311820830360464e-05,
+      "loss": 0.6544,
+      "step": 556
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.295735070547439e-05,
+      "loss": 0.6717,
+      "step": 557
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.279656744048391e-05,
+      "loss": 0.6647,
+      "step": 558
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.263586020565437e-05,
+      "loss": 0.6608,
+      "step": 559
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.247523069720445e-05,
+      "loss": 0.6778,
+      "step": 560
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2314680610532446e-05,
+      "loss": 0.6679,
+      "step": 561
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.215421164019843e-05,
+      "loss": 0.676,
+      "step": 562
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1993825479906255e-05,
+      "loss": 0.6682,
+      "step": 563
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.183352382248577e-05,
+      "loss": 0.6787,
+      "step": 564
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.16733083598749e-05,
+      "loss": 0.6691,
+      "step": 565
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1513180783101814e-05,
+      "loss": 0.6867,
+      "step": 566
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1353142782267084e-05,
+      "loss": 0.6761,
+      "step": 567
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.119319604652583e-05,
+      "loss": 0.6702,
+      "step": 568
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.103334226406989e-05,
+      "loss": 0.6644,
+      "step": 569
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0873583122109985e-05,
+      "loss": 0.6607,
+      "step": 570
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.071392030685799e-05,
+      "loss": 0.6653,
+      "step": 571
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0554355503509035e-05,
+      "loss": 0.663,
+      "step": 572
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.039489039622376e-05,
+      "loss": 0.6844,
+      "step": 573
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.023552666811056e-05,
+      "loss": 0.659,
+      "step": 574
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.00762660012078e-05,
+      "loss": 0.6736,
+      "step": 575
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.991711007646606e-05,
+      "loss": 0.67,
+      "step": 576
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.975806057373038e-05,
+      "loss": 0.6485,
+      "step": 577
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.959911917172258e-05,
+      "loss": 0.6715,
+      "step": 578
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.944028754802349e-05,
+      "loss": 0.6829,
+      "step": 579
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.928156737905525e-05,
+      "loss": 0.6709,
+      "step": 580
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.912296034006365e-05,
+      "loss": 0.6543,
+      "step": 581
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.896446810510041e-05,
+      "loss": 0.647,
+      "step": 582
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 3.880609234700554e-05,
+      "loss": 0.6559,
+      "step": 583
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.864783473738964e-05,
+      "loss": 0.6519,
+      "step": 584
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.8489696946616335e-05,
+      "loss": 0.6656,
+      "step": 585
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.833168064378455e-05,
+      "loss": 0.6747,
+      "step": 586
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.817378749671096e-05,
+      "loss": 0.6846,
+      "step": 587
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.801601917191238e-05,
+      "loss": 0.6918,
+      "step": 588
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.785837733458813e-05,
+      "loss": 0.684,
+      "step": 589
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.770086364860252e-05,
+      "loss": 0.6925,
+      "step": 590
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.754347977646725e-05,
+      "loss": 0.6762,
+      "step": 591
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.7386227379323856e-05,
+      "loss": 0.6957,
+      "step": 592
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.7229108116926225e-05,
+      "loss": 0.6901,
+      "step": 593
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.707212364762301e-05,
+      "loss": 0.6879,
+      "step": 594
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.691527562834019e-05,
+      "loss": 0.6843,
+      "step": 595
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6758565714563535e-05,
+      "loss": 0.7032,
+      "step": 596
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6601995560321164e-05,
+      "loss": 0.6897,
+      "step": 597
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.6445566818166074e-05,
+      "loss": 0.6991,
+      "step": 598
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.628928113915869e-05,
+      "loss": 0.6987,
+      "step": 599
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.613314017284943e-05,
+      "loss": 0.6878,
+      "step": 600
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5977145567261354e-05,
+      "loss": 0.7091,
+      "step": 601
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.58212989688727e-05,
+      "loss": 0.6993,
+      "step": 602
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.5665602022599515e-05,
+      "loss": 0.6764,
+      "step": 603
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.5510056371778336e-05,
+      "loss": 0.694,
+      "step": 604
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.535466365814884e-05,
+      "loss": 0.6984,
+      "step": 605
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.519942552183645e-05,
+      "loss": 0.6873,
+      "step": 606
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.504434360133511e-05,
+      "loss": 0.6899,
+      "step": 607
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.4889419533489896e-05,
+      "loss": 0.6933,
+      "step": 608
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.473465495347987e-05,
+      "loss": 0.6914,
+      "step": 609
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.458005149480068e-05,
+      "loss": 0.6896,
+      "step": 610
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.442561078924742e-05,
+      "loss": 0.6941,
+      "step": 611
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.427133446689735e-05,
+      "loss": 0.6842,
+      "step": 612
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.4117224156092754e-05,
+      "loss": 0.7012,
+      "step": 613
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.396328148342366e-05,
+      "loss": 0.6817,
+      "step": 614
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3809508073710756e-05,
+      "loss": 0.671,
+      "step": 615
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3655905549988194e-05,
+      "loss": 0.7013,
+      "step": 616
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3502475533486475e-05,
+      "loss": 0.6839,
+      "step": 617
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.334921964361535e-05,
+      "loss": 0.6964,
+      "step": 618
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.319613949794668e-05,
+      "loss": 0.6918,
+      "step": 619
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.3043236712197446e-05,
+      "loss": 0.6728,
+      "step": 620
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.289051290021259e-05,
+      "loss": 0.6864,
+      "step": 621
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.273796967394809e-05,
+      "loss": 0.6774,
+      "step": 622
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.258560864345387e-05,
+      "loss": 0.6917,
+      "step": 623
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.243343141685682e-05,
+      "loss": 0.675,
+      "step": 624
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.2281439600343836e-05,
+      "loss": 0.6832,
+      "step": 625
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.212963479814489e-05,
+      "loss": 0.6865,
+      "step": 626
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1978018612516026e-05,
+      "loss": 0.6682,
+      "step": 627
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.182659264372254e-05,
+      "loss": 0.6795,
+      "step": 628
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1675358490022014e-05,
+      "loss": 0.7023,
+      "step": 629
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1524317747647494e-05,
+      "loss": 0.6796,
+      "step": 630
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.1373472010790614e-05,
+      "loss": 0.6785,
+      "step": 631
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.122282287158479e-05,
+      "loss": 0.6948,
+      "step": 632
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.107237192008839e-05,
+      "loss": 0.6742,
+      "step": 633
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.092212074426799e-05,
+      "loss": 0.6783,
+      "step": 634
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.077207092998159e-05,
+      "loss": 0.6806,
+      "step": 635
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0622224060961834e-05,
+      "loss": 0.6814,
+      "step": 636
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.047258171879939e-05,
+      "loss": 0.6926,
+      "step": 637
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0323145482926186e-05,
+      "loss": 0.6665,
+      "step": 638
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0173916930598744e-05,
+      "loss": 0.6781,
+      "step": 639
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.0024897636881562e-05,
+      "loss": 0.6655,
+      "step": 640
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.987608917463047e-05,
+      "loss": 0.6804,
+      "step": 641
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9727493114476022e-05,
+      "loss": 0.6601,
+      "step": 642
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9579111024806945e-05,
+      "loss": 0.6729,
+      "step": 643
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9430944471753563e-05,
+      "loss": 0.6666,
+      "step": 644
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9282995019171276e-05,
+      "loss": 0.6646,
+      "step": 645
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.913526422862404e-05,
+      "loss": 0.685,
+      "step": 646
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.898775365936789e-05,
+      "loss": 0.6848,
+      "step": 647
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8840464868334533e-05,
+      "loss": 0.6817,
+      "step": 648
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8693399410114796e-05,
+      "loss": 0.6816,
+      "step": 649
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8546558836942383e-05,
+      "loss": 0.6712,
+      "step": 650
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8399944698677304e-05,
+      "loss": 0.6531,
+      "step": 651
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8253558542789706e-05,
+      "loss": 0.6805,
+      "step": 652
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.8107401914343366e-05,
+      "loss": 0.6648,
+      "step": 653
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7961476355979543e-05,
+      "loss": 0.6533,
+      "step": 654
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.781578340790053e-05,
+      "loss": 0.6692,
+      "step": 655
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.767032460785356e-05,
+      "loss": 0.6602,
+      "step": 656
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.752510149111449e-05,
+      "loss": 0.6658,
+      "step": 657
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7380115590471557e-05,
+      "loss": 0.6636,
+      "step": 658
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.7235368436209312e-05,
+      "loss": 0.6954,
+      "step": 659
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.709086155609235e-05,
+      "loss": 0.6848,
+      "step": 660
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6946596475349306e-05,
+      "loss": 0.7165,
+      "step": 661
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6802574716656615e-05,
+      "loss": 0.7032,
+      "step": 662
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6658797800122604e-05,
+      "loss": 0.6971,
+      "step": 663
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6515267243271267e-05,
+      "loss": 0.6968,
+      "step": 664
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.637198456102642e-05,
+      "loss": 0.7056,
+      "step": 665
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6228951265695623e-05,
+      "loss": 0.6844,
+      "step": 666
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6086168866954176e-05,
+      "loss": 0.7023,
+      "step": 667
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5943638871829302e-05,
+      "loss": 0.6841,
+      "step": 668
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.580136278468411e-05,
+      "loss": 0.6796,
+      "step": 669
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.565934210720186e-05,
+      "loss": 0.6852,
+      "step": 670
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.551757833836996e-05,
+      "loss": 0.6878,
+      "step": 671
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5376072974464283e-05,
+      "loss": 0.6814,
+      "step": 672
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.5234827509033298e-05,
+      "loss": 0.689,
+      "step": 673
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.5093843432882274e-05,
+      "loss": 0.6961,
+      "step": 674
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4953122234057663e-05,
+      "loss": 0.6685,
+      "step": 675
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4812665397831248e-05,
+      "loss": 0.6853,
+      "step": 676
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.467247440668462e-05,
+      "loss": 0.6787,
+      "step": 677
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4532550740293365e-05,
+      "loss": 0.6798,
+      "step": 678
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.439289587551162e-05,
+      "loss": 0.7063,
+      "step": 679
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.4253511286356324e-05,
+      "loss": 0.6917,
+      "step": 680
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.411439844399177e-05,
+      "loss": 0.6824,
+      "step": 681
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.3975558816714077e-05,
+      "loss": 0.6712,
+      "step": 682
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.383699386993557e-05,
+      "loss": 0.6748,
+      "step": 683
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3698705066169486e-05,
+      "loss": 0.6766,
+      "step": 684
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3560693865014383e-05,
+      "loss": 0.6574,
+      "step": 685
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3422961723138864e-05,
+      "loss": 0.6836,
+      "step": 686
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.328551009426609e-05,
+      "loss": 0.6881,
+      "step": 687
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.3148340429158526e-05,
+      "loss": 0.6727,
+      "step": 688
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.301145417560256e-05,
+      "loss": 0.68,
+      "step": 689
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.287485277839327e-05,
+      "loss": 0.6591,
+      "step": 690
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2738537679319184e-05,
+      "loss": 0.6738,
+      "step": 691
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.2602510317146957e-05,
+      "loss": 0.6732,
+      "step": 692
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.246677212760636e-05,
+      "loss": 0.672,
+      "step": 693
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2331324543374942e-05,
+      "loss": 0.6788,
+      "step": 694
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2196168994063076e-05,
+      "loss": 0.675,
+      "step": 695
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.2061306906198708e-05,
+      "loss": 0.6741,
+      "step": 696
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1926739703212473e-05,
+      "loss": 0.6745,
+      "step": 697
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.179246880542249e-05,
+      "loss": 0.6805,
+      "step": 698
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.165849563001952e-05,
+      "loss": 0.6597,
+      "step": 699
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1524821591051962e-05,
+      "loss": 0.6863,
+      "step": 700
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.1391448099410854e-05,
+      "loss": 0.6685,
+      "step": 701
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.125837656281511e-05,
+      "loss": 0.6793,
+      "step": 702
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.112560838579653e-05,
+      "loss": 0.685,
+      "step": 703
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0993144969685108e-05,
+      "loss": 0.6743,
+      "step": 704
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0860987712594104e-05,
+      "loss": 0.7037,
+      "step": 705
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.072913800940542e-05,
+      "loss": 0.6881,
+      "step": 706
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0597597251754803e-05,
+      "loss": 0.6795,
+      "step": 707
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0466366828017113e-05,
+      "loss": 0.6932,
+      "step": 708
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.033544812329181e-05,
+      "loss": 0.6745,
+      "step": 709
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0204842519388174e-05,
+      "loss": 0.6764,
+      "step": 710
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.0074551394810854e-05,
+      "loss": 0.6931,
+      "step": 711
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.994457612474521e-05,
+      "loss": 0.6799,
+      "step": 712
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9814918081042887e-05,
+      "loss": 0.6898,
+      "step": 713
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.968557863220727e-05,
+      "loss": 0.6798,
+      "step": 714
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.95565591433791e-05,
+      "loss": 0.6855,
+      "step": 715
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9427860976321998e-05,
+      "loss": 0.6864,
+      "step": 716
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9299485489408125e-05,
+      "loss": 0.6875,
+      "step": 717
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9171434037603886e-05,
+      "loss": 0.6747,
+      "step": 718
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.904370797245554e-05,
+      "loss": 0.6721,
+      "step": 719
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.891630864207501e-05,
+      "loss": 0.6757,
+      "step": 720
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8789237391125647e-05,
+      "loss": 0.6513,
+      "step": 721
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.8662495560807957e-05,
+      "loss": 0.6772,
+      "step": 722
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8536084488845584e-05,
+      "loss": 0.6699,
+      "step": 723
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8410005509471028e-05,
+      "loss": 0.6726,
+      "step": 724
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8284259953411732e-05,
+      "loss": 0.6748,
+      "step": 725
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.8158849147875873e-05,
+      "loss": 0.6621,
+      "step": 726
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.80337744165385e-05,
+      "loss": 0.6734,
+      "step": 727
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7909037079527434e-05,
+      "loss": 0.6593,
+      "step": 728
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7784638453409452e-05,
+      "loss": 0.6574,
+      "step": 729
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7660579851176318e-05,
+      "loss": 0.6759,
+      "step": 730
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7536862582230896e-05,
+      "loss": 0.6587,
+      "step": 731
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7413487952373457e-05,
+      "loss": 0.6726,
+      "step": 732
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.729045726378773e-05,
+      "loss": 0.6689,
+      "step": 733
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.716777181502732e-05,
+      "loss": 0.6744,
+      "step": 734
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7045432901001847e-05,
+      "loss": 0.6692,
+      "step": 735
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6923441812963436e-05,
+      "loss": 0.6705,
+      "step": 736
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6801799838492944e-05,
+      "loss": 0.6664,
+      "step": 737
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6680508261486466e-05,
+      "loss": 0.6837,
+      "step": 738
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.655956836214177e-05,
+      "loss": 0.6657,
+      "step": 739
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.643898141694471e-05,
+      "loss": 0.671,
+      "step": 740
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6318748698655873e-05,
+      "loss": 0.6677,
+      "step": 741
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.6198871476297034e-05,
+      "loss": 0.6401,
+      "step": 742
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.607935101513785e-05,
+      "loss": 0.6422,
+      "step": 743
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5960188576682423e-05,
+      "loss": 0.6501,
+      "step": 744
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.584138541865607e-05,
+      "loss": 0.655,
+      "step": 745
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5722942794991995e-05,
+      "loss": 0.6467,
+      "step": 746
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.560486195581804e-05,
+      "loss": 0.6566,
+      "step": 747
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.548714414744356e-05,
+      "loss": 0.6463,
+      "step": 748
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5369790612346168e-05,
+      "loss": 0.6463,
+      "step": 749
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5252802589158738e-05,
+      "loss": 0.6331,
+      "step": 750
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.513618131265621e-05,
+      "loss": 0.6443,
+      "step": 751
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.5019928013742684e-05,
+      "loss": 0.6499,
+      "step": 752
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4904043919438294e-05,
+      "loss": 0.6505,
+      "step": 753
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4788530252866373e-05,
+      "loss": 0.6444,
+      "step": 754
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4673388233240503e-05,
+      "loss": 0.6304,
+      "step": 755
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4558619075851582e-05,
+      "loss": 0.656,
+      "step": 756
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4444223992055118e-05,
+      "loss": 0.6552,
+      "step": 757
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4330204189258328e-05,
+      "loss": 0.6613,
+      "step": 758
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4216560870907497e-05,
+      "loss": 0.6418,
+      "step": 759
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4103295236475167e-05,
+      "loss": 0.65,
+      "step": 760
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3990408481447597e-05,
+      "loss": 0.649,
+      "step": 761
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.3877901797312021e-05,
+      "loss": 0.618,
+      "step": 762
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3765776371544176e-05,
+      "loss": 0.6216,
+      "step": 763
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3654033387595733e-05,
+      "loss": 0.6126,
+      "step": 764
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3542674024881746e-05,
+      "loss": 0.6177,
+      "step": 765
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3431699458768333e-05,
+      "loss": 0.6138,
+      "step": 766
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.332111086056011e-05,
+      "loss": 0.6139,
+      "step": 767
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3210909397487997e-05,
+      "loss": 0.6113,
+      "step": 768
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.3101096232696735e-05,
+      "loss": 0.6127,
+      "step": 769
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2991672525232757e-05,
+      "loss": 0.6036,
+      "step": 770
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2882639430031835e-05,
+      "loss": 0.6271,
+      "step": 771
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.2773998097906964e-05,
+      "loss": 0.616,
+      "step": 772
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.266574967553621e-05,
+      "loss": 0.6214,
+      "step": 773
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2557895305450535e-05,
+      "loss": 0.6255,
+      "step": 774
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2450436126021864e-05,
+      "loss": 0.615,
+      "step": 775
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.234337327145092e-05,
+      "loss": 0.6122,
+      "step": 776
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2236707871755405e-05,
+      "loss": 0.6224,
+      "step": 777
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.213044105275794e-05,
+      "loss": 0.605,
+      "step": 778
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.2024573936074273e-05,
+      "loss": 0.612,
+      "step": 779
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1919107639101424e-05,
+      "loss": 0.6216,
+      "step": 780
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1814043275005821e-05,
+      "loss": 0.6274,
+      "step": 781
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.1709381952711668e-05,
+      "loss": 0.5762,
+      "step": 782
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1605124776889126e-05,
+      "loss": 0.5662,
+      "step": 783
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1501272847942752e-05,
+      "loss": 0.5713,
+      "step": 784
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1397827261999795e-05,
+      "loss": 0.5601,
+      "step": 785
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1294789110898712e-05,
+      "loss": 0.5692,
+      "step": 786
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1192159482177561e-05,
+      "loss": 0.5685,
+      "step": 787
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.1089939459062605e-05,
+      "loss": 0.5835,
+      "step": 788
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0988130120456814e-05,
+      "loss": 0.5719,
+      "step": 789
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.088673254092849e-05,
+      "loss": 0.5545,
+      "step": 790
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0785747790699979e-05,
+      "loss": 0.5669,
+      "step": 791
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.0685176935636265e-05,
+      "loss": 0.5768,
+      "step": 792
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0585021037233873e-05,
+      "loss": 0.5704,
+      "step": 793
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0485281152609483e-05,
+      "loss": 0.5713,
+      "step": 794
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0385958334488966e-05,
+      "loss": 0.5556,
+      "step": 795
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0287053631196108e-05,
+      "loss": 0.5661,
+      "step": 796
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0188568086641615e-05,
+      "loss": 0.5621,
+      "step": 797
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.0090502740312152e-05,
+      "loss": 0.5667,
+      "step": 798
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.992858627259238e-06,
+      "loss": 0.5646,
+      "step": 799
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.895636778088457e-06,
+      "loss": 0.5748,
+      "step": 800
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.79883821894847e-06,
+      "loss": 0.5678,
+      "step": 801
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.702463971520264e-06,
+      "loss": 0.5728,
+      "step": 802
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.606515053006349e-06,
+      "loss": 0.5551,
+      "step": 803
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.510992476119961e-06,
+      "loss": 0.5569,
+      "step": 804
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.41589724907448e-06,
+      "loss": 0.562,
+      "step": 805
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.321230375572681e-06,
+      "loss": 0.5542,
+      "step": 806
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.226992854796234e-06,
+      "loss": 0.5669,
+      "step": 807
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.133185681395073e-06,
+      "loss": 0.5571,
+      "step": 808
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.03980984547697e-06,
+      "loss": 0.5604,
+      "step": 809
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.946866332597064e-06,
+      "loss": 0.5645,
+      "step": 810
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.854356123747393e-06,
+      "loss": 0.5693,
+      "step": 811
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.762280195346656e-06,
+      "loss": 0.565,
+      "step": 812
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.6706395192298e-06,
+      "loss": 0.5711,
+      "step": 813
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.579435062637863e-06,
+      "loss": 0.5602,
+      "step": 814
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.488667788207644e-06,
+      "loss": 0.5674,
+      "step": 815
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.398338653961673e-06,
+      "loss": 0.5726,
+      "step": 816
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.308448613297981e-06,
+      "loss": 0.5624,
+      "step": 817
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.218998614980133e-06,
+      "loss": 0.551,
+      "step": 818
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.129989603127181e-06,
+      "loss": 0.5739,
+      "step": 819
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.041422517203628e-06,
+      "loss": 0.5525,
+      "step": 820
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.953298292009659e-06,
+      "loss": 0.5651,
+      "step": 821
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 7.86561785767112e-06,
+      "loss": 0.5611,
+      "step": 822
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.778382139629831e-06,
+      "loss": 0.553,
+      "step": 823
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.691592058633695e-06,
+      "loss": 0.5777,
+      "step": 824
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.605248530727116e-06,
+      "loss": 0.5673,
+      "step": 825
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.519352467241198e-06,
+      "loss": 0.5584,
+      "step": 826
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.433904774784217e-06,
+      "loss": 0.5693,
+      "step": 827
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.348906355232027e-06,
+      "loss": 0.5709,
+      "step": 828
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.264358105718505e-06,
+      "loss": 0.5616,
+      "step": 829
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.180260918626153e-06,
+      "loss": 0.5557,
+      "step": 830
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.096615681576596e-06,
+      "loss": 0.5565,
+      "step": 831
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 7.013423277421299e-06,
+      "loss": 0.5623,
+      "step": 832
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.930684584232161e-06,
+      "loss": 0.5661,
+      "step": 833
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.848400475292344e-06,
+      "loss": 0.582,
+      "step": 834
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.766571819086942e-06,
+      "loss": 0.5725,
+      "step": 835
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.68519947929393e-06,
+      "loss": 0.5747,
+      "step": 836
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.6042843147749836e-06,
+      "loss": 0.5768,
+      "step": 837
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.523827179566394e-06,
+      "loss": 0.5907,
+      "step": 838
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.443828922870126e-06,
+      "loss": 0.5669,
+      "step": 839
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.364290389044769e-06,
+      "loss": 0.5863,
+      "step": 840
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.2852124175967195e-06,
+      "loss": 0.5807,
+      "step": 841
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 6.2065958431712254e-06,
+      "loss": 0.5773,
+      "step": 842
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.128441495543647e-06,
+      "loss": 0.5722,
+      "step": 843
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 6.050750199610683e-06,
+      "loss": 0.5734,
+      "step": 844
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.973522775381618e-06,
+      "loss": 0.5865,
+      "step": 845
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.89676003796974e-06,
+      "loss": 0.6016,
+      "step": 846
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.82046279758367e-06,
+      "loss": 0.5808,
+      "step": 847
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.7446318595188785e-06,
+      "loss": 0.5662,
+      "step": 848
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.66926802414911e-06,
+      "loss": 0.5887,
+      "step": 849
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.594372086918009e-06,
+      "loss": 0.5884,
+      "step": 850
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.519944838330659e-06,
+      "loss": 0.5719,
+      "step": 851
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.445987063945291e-06,
+      "loss": 0.5731,
+      "step": 852
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.372499544364973e-06,
+      "loss": 0.5918,
+      "step": 853
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.2994830552293374e-06,
+      "loss": 0.5832,
+      "step": 854
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.226938367206474e-06,
+      "loss": 0.5526,
+      "step": 855
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.154866245984696e-06,
+      "loss": 0.543,
+      "step": 856
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.0832674522645565e-06,
+      "loss": 0.539,
+      "step": 857
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.012142741750725e-06,
+      "loss": 0.5432,
+      "step": 858
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.941492865144115e-06,
+      "loss": 0.5409,
+      "step": 859
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.871318568133848e-06,
+      "loss": 0.5339,
+      "step": 860
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.801620591389477e-06,
+      "loss": 0.5443,
+      "step": 861
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.732399670553133e-06,
+      "loss": 0.5556,
+      "step": 862
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.663656536231731e-06,
+      "loss": 0.5373,
+      "step": 863
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.595391913989311e-06,
+      "loss": 0.5458,
+      "step": 864
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.527606524339328e-06,
+      "loss": 0.5513,
+      "step": 865
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.460301082737123e-06,
+      "loss": 0.5403,
+      "step": 866
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.393476299572263e-06,
+      "loss": 0.5598,
+      "step": 867
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.327132880161161e-06,
+      "loss": 0.5516,
+      "step": 868
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.261271524739524e-06,
+      "loss": 0.5497,
+      "step": 869
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.195892928455048e-06,
+      "loss": 0.5438,
+      "step": 870
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.130997781360035e-06,
+      "loss": 0.5496,
+      "step": 871
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.066586768404101e-06,
+      "loss": 0.5503,
+      "step": 872
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.002660569426997e-06,
+      "loss": 0.5545,
+      "step": 873
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.939219859151377e-06,
+      "loss": 0.5395,
+      "step": 874
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.876265307175714e-06,
+      "loss": 0.5381,
+      "step": 875
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.8137975779672098e-06,
+      "loss": 0.5622,
+      "step": 876
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.7518173308548064e-06,
+      "loss": 0.5506,
+      "step": 877
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6903252200222004e-06,
+      "loss": 0.5514,
+      "step": 878
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6293218945009367e-06,
+      "loss": 0.5351,
+      "step": 879
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.56880799816362e-06,
+      "loss": 0.5446,
+      "step": 880
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.508784169717011e-06,
+      "loss": 0.5439,
+      "step": 881
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.4492510426953783e-06,
+      "loss": 0.5613,
+      "step": 882
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.3902092454537904e-06,
+      "loss": 0.5366,
+      "step": 883
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.3316594011614354e-06,
+      "loss": 0.5426,
+      "step": 884
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.2736021277951057e-06,
+      "loss": 0.5368,
+      "step": 885
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.2160380381326237e-06,
+      "loss": 0.5352,
+      "step": 886
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.1589677397464433e-06,
+      "loss": 0.5338,
+      "step": 887
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.1023918349971424e-06,
+      "loss": 0.5476,
+      "step": 888
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.046310921027157e-06,
+      "loss": 0.5389,
+      "step": 889
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.990725589754406e-06,
+      "loss": 0.549,
+      "step": 890
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.9356364278660954e-06,
+      "loss": 0.5507,
+      "step": 891
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.8810440168125062e-06,
+      "loss": 0.5503,
+      "step": 892
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.8269489328008434e-06,
+      "loss": 0.5453,
+      "step": 893
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.7733517467891824e-06,
+      "loss": 0.5511,
+      "step": 894
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.720253024480418e-06,
+      "loss": 0.5409,
+      "step": 895
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.6676533263163107e-06,
+      "loss": 0.536,
+      "step": 896
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.615553207471555e-06,
+      "loss": 0.5335,
+      "step": 897
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.5639532178479418e-06,
+      "loss": 0.5253,
+      "step": 898
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.512853902068529e-06,
+      "loss": 0.5415,
+      "step": 899
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.462255799471913e-06,
+      "loss": 0.5447,
+      "step": 900
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.412159444106543e-06,
+      "loss": 0.5332,
+      "step": 901
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.362565364725039e-06,
+      "loss": 0.5407,
+      "step": 902
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.313474084778672e-06,
+      "loss": 0.5355,
+      "step": 903
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.2648861224117855e-06,
+      "loss": 0.5483,
+      "step": 904
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.2168019904563686e-06,
+      "loss": 0.531,
+      "step": 905
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.1692221964266123e-06,
+      "loss": 0.537,
+      "step": 906
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.122147242513578e-06,
+      "loss": 0.5425,
+      "step": 907
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.075577625579872e-06,
+      "loss": 0.5354,
+      "step": 908
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.029513837154423e-06,
+      "loss": 0.5295,
+      "step": 909
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9839563634272974e-06,
+      "loss": 0.5182,
+      "step": 910
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.938905685244513e-06,
+      "loss": 0.5311,
+      "step": 911
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.8943622781030568e-06,
+      "loss": 0.5398,
+      "step": 912
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.8503266121457753e-06,
+      "loss": 0.5309,
+      "step": 913
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.8067991521564854e-06,
+      "loss": 0.5365,
+      "step": 914
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7637803575550115e-06,
+      "loss": 0.5336,
+      "step": 915
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.7212706823923674e-06,
+      "loss": 0.5506,
+      "step": 916
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.679270575345976e-06,
+      "loss": 0.533,
+      "step": 917
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.637780479714879e-06,
+      "loss": 0.5451,
+      "step": 918
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5968008334151351e-06,
+      "loss": 0.544,
+      "step": 919
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5563320689751193e-06,
+      "loss": 0.547,
+      "step": 920
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.5163746135310186e-06,
+      "loss": 0.5346,
+      "step": 921
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4769288888222987e-06,
+      "loss": 0.5481,
+      "step": 922
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4379953111872458e-06,
+      "loss": 0.546,
+      "step": 923
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3995742915585807e-06,
+      "loss": 0.5401,
+      "step": 924
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.361666235459136e-06,
+      "loss": 0.5531,
+      "step": 925
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.3242715429975516e-06,
+      "loss": 0.5487,
+      "step": 926
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.2873906088640475e-06,
+      "loss": 0.5443,
+      "step": 927
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.2510238223263082e-06,
+      "loss": 0.5442,
+      "step": 928
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.2151715672252984e-06,
+      "loss": 0.5266,
+      "step": 929
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1798342219712822e-06,
+      "loss": 0.5391,
+      "step": 930
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1450121595397711e-06,
+      "loss": 0.5492,
+      "step": 931
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.1107057474676441e-06,
+      "loss": 0.5374,
+      "step": 932
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0769153478492112e-06,
+      "loss": 0.5457,
+      "step": 933
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.043641317332439e-06,
+      "loss": 0.537,
+      "step": 934
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.0108840071151648e-06,
+      "loss": 0.5384,
+      "step": 935
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.78643762941367e-07,
+      "loss": 0.5248,
+      "step": 936
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.469209250975774e-07,
+      "loss": 0.5177,
+      "step": 937
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.157158284092249e-07,
+      "loss": 0.5159,
+      "step": 938
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.850288022371478e-07,
+      "loss": 0.507,
+      "step": 939
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.548601704740756e-07,
+      "loss": 0.5173,
+      "step": 940
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.252102515412641e-07,
+      "loss": 0.51,
+      "step": 941
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.960793583850767e-07,
+      "loss": 0.526,
+      "step": 942
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.674677984737256e-07,
+      "loss": 0.5294,
+      "step": 943
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.393758737940126e-07,
+      "loss": 0.5226,
+      "step": 944
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.118038808481164e-07,
+      "loss": 0.5146,
+      "step": 945
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 6.847521106505106e-07,
+      "loss": 0.5153,
+      "step": 946
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 6.582208487248497e-07,
+      "loss": 0.5236,
+      "step": 947
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 6.322103751009834e-07,
+      "loss": 0.5047,
+      "step": 948
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 6.067209643119909e-07,
+      "loss": 0.5237,
+      "step": 949
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.817528853912735e-07,
+      "loss": 0.5211,
+      "step": 950
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.573064018697393e-07,
+      "loss": 0.5162,
+      "step": 951
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.333817717729895e-07,
+      "loss": 0.4983,
+      "step": 952
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.099792476186249e-07,
+      "loss": 0.51,
+      "step": 953
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.870990764135553e-07,
+      "loss": 0.5019,
+      "step": 954
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.647414996514277e-07,
+      "loss": 0.5122,
+      "step": 955
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.4290675331002953e-07,
+      "loss": 0.5158,
+      "step": 956
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.2159506784884005e-07,
+      "loss": 0.5099,
+      "step": 957
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.008066682065714e-07,
+      "loss": 0.5258,
+      "step": 958
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.805417737988148e-07,
+      "loss": 0.4972,
+      "step": 959
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.608005985157037e-07,
+      "loss": 0.5108,
+      "step": 960
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.415833507196764e-07,
+      "loss": 0.5313,
+      "step": 961
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.2289023324325595e-07,
+      "loss": 0.5273,
+      "step": 962
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.047214433869239e-07,
+      "loss": 0.5297,
+      "step": 963
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.8707717291704404e-07,
+      "loss": 0.5152,
+      "step": 964
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.6995760806382e-07,
+      "loss": 0.514,
+      "step": 965
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.5336292951933514e-07,
+      "loss": 0.5228,
+      "step": 966
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3729331243566023e-07,
+      "loss": 0.5186,
+      "step": 967
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2174892642298216e-07,
+      "loss": 0.526,
+      "step": 968
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.067299355478336e-07,
+      "loss": 0.5271,
+      "step": 969
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.922364983313385e-07,
+      "loss": 0.5225,
+      "step": 970
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.7826876774757473e-07,
+      "loss": 0.5182,
+      "step": 971
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.648268912219142e-07,
+      "loss": 0.5243,
+      "step": 972
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5191101062950186e-07,
+      "loss": 0.5231,
+      "step": 973
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3952126229375695e-07,
+      "loss": 0.5196,
+      "step": 974
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2765777698490188e-07,
+      "loss": 0.5246,
+      "step": 975
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.163206799186245e-07,
+      "loss": 0.5236,
+      "step": 976
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.0551009075471796e-07,
+      "loss": 0.5412,
+      "step": 977
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.522612359585403e-08,
+      "loss": 0.5388,
+      "step": 978
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.546888698634514e-08,
+      "loss": 0.538,
+      "step": 979
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.623848391102306e-08,
+      "loss": 0.5336,
+      "step": 980
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.753501179413423e-08,
+      "loss": 0.5416,
+      "step": 981
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.935856249833505e-08,
+      "loss": 0.5505,
+      "step": 982
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1709222323692577e-08,
+      "loss": 0.5365,
+      "step": 983
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.458707200679646e-08,
+      "loss": 0.5521,
+      "step": 984
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.799218671989291e-08,
+      "loss": 0.5363,
+      "step": 985
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.192463607010754e-08,
+      "loss": 0.5497,
+      "step": 986
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.638448409869043e-08,
+      "loss": 0.5477,
+      "step": 987
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.1371789280355547e-08,
+      "loss": 0.5456,
+      "step": 988
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6886604522659e-08,
+      "loss": 0.5463,
+      "step": 989
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.2928977165427292e-08,
+      "loss": 0.5398,
+      "step": 990
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.498948980291022e-09,
+      "loss": 0.5414,
+      "step": 991
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.596556170207491e-09,
+      "loss": 0.5382,
+      "step": 992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.221829369094321e-09,
+      "loss": 0.5537,
+      "step": 993
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.374793641518602e-09,
+      "loss": 0.5544,
+      "step": 994
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0554684824137795e-09,
+      "loss": 0.5401,
+      "step": 995
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6386781686316144e-10,
+      "loss": 0.546,
+      "step": 996
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.5491,
+      "step": 997
+    },
+    {
+      "epoch": 1.0,
+      "step": 997,
+      "total_flos": 0.0,
+      "train_loss": 0.6507606941276233,
+      "train_runtime": 58049.3541,
+      "train_samples_per_second": 35.176,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 997,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}