diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6544 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1086,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.7468,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.7543,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.7848,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.7554,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.7899,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.7236,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.6904,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.6955,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.6735,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.7221,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6643,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.67,
+      "step": 12
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.6701,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.6776,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.6976,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.6557,
+      "step": 16
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0303030303030304e-05,
+      "loss": 0.6567,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.6358,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1515151515151517e-05,
+      "loss": 0.6595,
+      "step": 19
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.6306,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.6498,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.646,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3939393939393942e-05,
+      "loss": 0.6837,
+      "step": 23
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.6296,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.6357,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.575757575757576e-05,
+      "loss": 0.6448,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.6318,
+      "step": 27
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6969696969696972e-05,
+      "loss": 0.6354,
+      "step": 28
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7575757575757576e-05,
+      "loss": 0.6259,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.6216,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8787878787878792e-05,
+      "loss": 0.6115,
+      "step": 31
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9393939393939395e-05,
+      "loss": 0.6146,
+      "step": 32
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.6415,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999955494602408e-05,
+      "loss": 0.6416,
+      "step": 34
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999982197880577e-05,
+      "loss": 0.6113,
+      "step": 35
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999599453798523e-05,
+      "loss": 0.6364,
+      "step": 36
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999287921561385e-05,
+      "loss": 0.6319,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998887384867323e-05,
+      "loss": 0.6109,
+      "step": 38
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998397847281548e-05,
+      "loss": 0.6396,
+      "step": 39
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997819313161476e-05,
+      "loss": 0.6393,
+      "step": 40
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997151787656678e-05,
+      "loss": 0.6061,
+      "step": 41
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996395276708856e-05,
+      "loss": 0.5879,
+      "step": 42
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995549787051772e-05,
+      "loss": 0.6099,
+      "step": 43
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994615326211203e-05,
+      "loss": 0.6371,
+      "step": 44
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993591902504854e-05,
+      "loss": 0.6422,
+      "step": 45
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992479525042305e-05,
+      "loss": 0.613,
+      "step": 46
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991278203724908e-05,
+      "loss": 0.6171,
+      "step": 47
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989987949245725e-05,
+      "loss": 0.6178,
+      "step": 48
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9988608773089413e-05,
+      "loss": 0.6064,
+      "step": 49
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998714068753213e-05,
+      "loss": 0.6083,
+      "step": 50
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9985583705641418e-05,
+      "loss": 0.6238,
+      "step": 51
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9983937841276103e-05,
+      "loss": 0.6101,
+      "step": 52
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982203109086153e-05,
+      "loss": 0.6307,
+      "step": 53
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998037952451255e-05,
+      "loss": 0.6024,
+      "step": 54
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978467103787176e-05,
+      "loss": 0.5762,
+      "step": 55
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976465863932632e-05,
+      "loss": 0.5982,
+      "step": 56
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9974375822762117e-05,
+      "loss": 0.5804,
+      "step": 57
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9972196998879254e-05,
+      "loss": 0.5854,
+      "step": 58
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.996992941167792e-05,
+      "loss": 0.6073,
+      "step": 59
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.6606,
+      "step": 60
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9965128028845676e-05,
+      "loss": 0.5868,
+      "step": 61
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9962594275952246e-05,
+      "loss": 0.6212,
+      "step": 62
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9959971845214953e-05,
+      "loss": 0.6308,
+      "step": 63
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995726075997626e-05,
+      "loss": 0.624,
+      "step": 64
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9954461044367752e-05,
+      "loss": 0.6143,
+      "step": 65
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9951572723309918e-05,
+      "loss": 0.6167,
+      "step": 66
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994859582251194e-05,
+      "loss": 0.5796,
+      "step": 67
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9945530368471444e-05,
+      "loss": 0.5787,
+      "step": 68
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9942376388474282e-05,
+      "loss": 0.6278,
+      "step": 69
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9939133910594276e-05,
+      "loss": 0.5963,
+      "step": 70
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9935802963692988e-05,
+      "loss": 0.6121,
+      "step": 71
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.6191,
+      "step": 72
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992887578220984e-05,
+      "loss": 0.5994,
+      "step": 73
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9925279609287384e-05,
+      "loss": 0.5929,
+      "step": 74
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9921595090661872e-05,
+      "loss": 0.6216,
+      "step": 75
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9917822259129508e-05,
+      "loss": 0.6072,
+      "step": 76
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991396114827256e-05,
+      "loss": 0.6063,
+      "step": 77
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9910011792459086e-05,
+      "loss": 0.6301,
+      "step": 78
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9905974226842614e-05,
+      "loss": 0.5881,
+      "step": 79
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9901848487361834e-05,
+      "loss": 0.5886,
+      "step": 80
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.989763461074029e-05,
+      "loss": 0.6074,
+      "step": 81
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.989333263448602e-05,
+      "loss": 0.606,
+      "step": 82
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888942596891267e-05,
+      "loss": 0.6321,
+      "step": 83
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9884464537032103e-05,
+      "loss": 0.5899,
+      "step": 84
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9879898494768093e-05,
+      "loss": 0.6051,
+      "step": 85
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987524451074194e-05,
+      "loss": 0.6085,
+      "step": 86
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.586,
+      "step": 87
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9865672883887553e-05,
+      "loss": 0.5848,
+      "step": 88
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9860755326257127e-05,
+      "loss": 0.5696,
+      "step": 89
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985574999725943e-05,
+      "loss": 0.6011,
+      "step": 90
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985065694144728e-05,
+      "loss": 0.6275,
+      "step": 91
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9845476204154387e-05,
+      "loss": 0.5999,
+      "step": 92
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9840207831494903e-05,
+      "loss": 0.6029,
+      "step": 93
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9834851870363024e-05,
+      "loss": 0.5926,
+      "step": 94
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9829408368432592e-05,
+      "loss": 0.6029,
+      "step": 95
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9823877374156647e-05,
+      "loss": 0.5967,
+      "step": 96
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9818258936767013e-05,
+      "loss": 0.588,
+      "step": 97
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9812553106273848e-05,
+      "loss": 0.5919,
+      "step": 98
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.98067599334652e-05,
+      "loss": 0.6277,
+      "step": 99
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.980087946990656e-05,
+      "loss": 0.6105,
+      "step": 100
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9794911767940405e-05,
+      "loss": 0.5777,
+      "step": 101
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978885688068572e-05,
+      "loss": 0.5907,
+      "step": 102
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9782714862037544e-05,
+      "loss": 0.6131,
+      "step": 103
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977648576666647e-05,
+      "loss": 0.5816,
+      "step": 104
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977016965001817e-05,
+      "loss": 0.6135,
+      "step": 105
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9763766568312906e-05,
+      "loss": 0.578,
+      "step": 106
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9757276578545013e-05,
+      "loss": 0.5997,
+      "step": 107
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9750699738482403e-05,
+      "loss": 0.5733,
+      "step": 108
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974403610666606e-05,
+      "loss": 0.6063,
+      "step": 109
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9737285742409495e-05,
+      "loss": 0.5773,
+      "step": 110
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.5972,
+      "step": 111
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.97235250576893e-05,
+      "loss": 0.6071,
+      "step": 112
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971651485971062e-05,
+      "loss": 0.5858,
+      "step": 113
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.6153,
+      "step": 114
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9702235064507175e-05,
+      "loss": 0.6004,
+      "step": 115
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9694965594388008e-05,
+      "loss": 0.608,
+      "step": 116
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9687609828609156e-05,
+      "loss": 0.6,
+      "step": 117
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9680167832644868e-05,
+      "loss": 0.634,
+      "step": 118
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9672639672736947e-05,
+      "loss": 0.6051,
+      "step": 119
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.966502541589414e-05,
+      "loss": 0.5959,
+      "step": 120
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9657325129891558e-05,
+      "loss": 0.5828,
+      "step": 121
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9649538883270053e-05,
+      "loss": 0.6068,
+      "step": 122
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9641666745335626e-05,
+      "loss": 0.6166,
+      "step": 123
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9633708786158803e-05,
+      "loss": 0.5812,
+      "step": 124
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962566507657402e-05,
+      "loss": 0.6173,
+      "step": 125
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.961753568817896e-05,
+      "loss": 0.5814,
+      "step": 126
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9609320693333967e-05,
+      "loss": 0.6002,
+      "step": 127
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.960102016516136e-05,
+      "loss": 0.6289,
+      "step": 128
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9592634177544803e-05,
+      "loss": 0.6091,
+      "step": 129
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584162805128636e-05,
+      "loss": 0.6092,
+      "step": 130
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9575606123317215e-05,
+      "loss": 0.601,
+      "step": 131
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9566964208274254e-05,
+      "loss": 0.5846,
+      "step": 132
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9558237136922117e-05,
+      "loss": 0.5726,
+      "step": 133
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.954942498694117e-05,
+      "loss": 0.6019,
+      "step": 134
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9540527836769047e-05,
+      "loss": 0.5931,
+      "step": 135
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.95315457656e-05,
+      "loss": 0.573,
+      "step": 136
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9522478853384154e-05,
+      "loss": 0.5933,
+      "step": 137
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.951332718082682e-05,
+      "loss": 0.6012,
+      "step": 138
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.950409082938776e-05,
+      "loss": 0.6078,
+      "step": 139
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.949476988128047e-05,
+      "loss": 0.6186,
+      "step": 140
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.5668,
+      "step": 141
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.947587452767948e-05,
+      "loss": 0.5752,
+      "step": 142
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.946630029037482e-05,
+      "loss": 0.5994,
+      "step": 143
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9456641792778527e-05,
+      "loss": 0.5723,
+      "step": 144
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9446899120861653e-05,
+      "loss": 0.5882,
+      "step": 145
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.94370723613445e-05,
+      "loss": 0.5831,
+      "step": 146
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9427161601695833e-05,
+      "loss": 0.5695,
+      "step": 147
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.941716693013211e-05,
+      "loss": 0.5691,
+      "step": 148
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.94070884356167e-05,
+      "loss": 0.5929,
+      "step": 149
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.5601,
+      "step": 150
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.938668033731406e-05,
+      "loss": 0.5864,
+      "step": 151
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9376350915180935e-05,
+      "loss": 0.588,
+      "step": 152
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9365938033402715e-05,
+      "loss": 0.591,
+      "step": 153
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9355441784665295e-05,
+      "loss": 0.6012,
+      "step": 154
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9344862262396612e-05,
+      "loss": 0.5935,
+      "step": 155
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933419956076584e-05,
+      "loss": 0.5993,
+      "step": 156
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932345377468253e-05,
+      "loss": 0.6078,
+      "step": 157
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9312624999795784e-05,
+      "loss": 0.585,
+      "step": 158
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9301713332493386e-05,
+      "loss": 0.5549,
+      "step": 159
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.929071886990095e-05,
+      "loss": 0.5647,
+      "step": 160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9279641709881067e-05,
+      "loss": 0.5995,
+      "step": 161
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.926848195103242e-05,
+      "loss": 0.5751,
+      "step": 162
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9257239692688907e-05,
+      "loss": 0.59,
+      "step": 163
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9245915034918763e-05,
+      "loss": 0.5716,
+      "step": 164
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.923450807852367e-05,
+      "loss": 0.5715,
+      "step": 165
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.922301892503785e-05,
+      "loss": 0.6003,
+      "step": 166
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9211447676727174e-05,
+      "loss": 0.572,
+      "step": 167
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.5934,
+      "step": 168
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9188059308347475e-05,
+      "loss": 0.6141,
+      "step": 169
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9176242396460184e-05,
+      "loss": 0.5905,
+      "step": 170
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.916434380610963e-05,
+      "loss": 0.5991,
+      "step": 171
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9152363643206126e-05,
+      "loss": 0.5836,
+      "step": 172
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9140302014386044e-05,
+      "loss": 0.5794,
+      "step": 173
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912815902701091e-05,
+      "loss": 0.6077,
+      "step": 174
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.911593478916641e-05,
+      "loss": 0.5837,
+      "step": 175
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9103629409661468e-05,
+      "loss": 0.579,
+      "step": 176
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.909124299802724e-05,
+      "loss": 0.5888,
+      "step": 177
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9078775664516157e-05,
+      "loss": 0.5875,
+      "step": 178
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.906622752010095e-05,
+      "loss": 0.6022,
+      "step": 179
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9053598676473656e-05,
+      "loss": 0.5567,
+      "step": 180
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.904088924604461e-05,
+      "loss": 0.584,
+      "step": 181
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9028099341941457e-05,
+      "loss": 0.5628,
+      "step": 182
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9015229078008163e-05,
+      "loss": 0.5586,
+      "step": 183
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.900227856880396e-05,
+      "loss": 0.5892,
+      "step": 184
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.898924792960237e-05,
+      "loss": 0.5874,
+      "step": 185
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8976137276390145e-05,
+      "loss": 0.6056,
+      "step": 186
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8962946725866246e-05,
+      "loss": 0.588,
+      "step": 187
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8949676395440818e-05,
+      "loss": 0.5679,
+      "step": 188
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.5755,
+      "step": 189
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.892289686807551e-05,
+      "loss": 0.5997,
+      "step": 190
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8909387909502335e-05,
+      "loss": 0.6096,
+      "step": 191
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8895799647758912e-05,
+      "loss": 0.5858,
+      "step": 192
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.888213220379544e-05,
+      "loss": 0.59,
+      "step": 193
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8868385699266928e-05,
+      "loss": 0.5967,
+      "step": 194
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.5666,
+      "step": 195
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8840655998652324e-05,
+      "loss": 0.5586,
+      "step": 196
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8826673049390508e-05,
+      "loss": 0.5788,
+      "step": 197
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.881261153320999e-05,
+      "loss": 0.5456,
+      "step": 198
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8798471575273445e-05,
+      "loss": 0.5741,
+      "step": 199
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8784253301441767e-05,
+      "loss": 0.5729,
+      "step": 200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8769956838272937e-05,
+      "loss": 0.5792,
+      "step": 201
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8755582313020912e-05,
+      "loss": 0.5823,
+      "step": 202
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8741129853634483e-05,
+      "loss": 0.6096,
+      "step": 203
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8726599588756144e-05,
+      "loss": 0.543,
+      "step": 204
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8711991647720936e-05,
+      "loss": 0.6138,
+      "step": 205
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8697306160555303e-05,
+      "loss": 0.5741,
+      "step": 206
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.868254325797594e-05,
+      "loss": 0.593,
+      "step": 207
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8667703071388607e-05,
+      "loss": 0.5786,
+      "step": 208
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8652785732886988e-05,
+      "loss": 0.563,
+      "step": 209
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8637791375251505e-05,
+      "loss": 0.6006,
+      "step": 210
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862272013194812e-05,
+      "loss": 0.5927,
+      "step": 211
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8607572137127175e-05,
+      "loss": 0.6052,
+      "step": 212
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.859234752562217e-05,
+      "loss": 0.5828,
+      "step": 213
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8577046432948586e-05,
+      "loss": 0.5507,
+      "step": 214
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8561668995302668e-05,
+      "loss": 0.5893,
+      "step": 215
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8546215349560204e-05,
+      "loss": 0.5711,
+      "step": 216
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.853068563327533e-05,
+      "loss": 0.5637,
+      "step": 217
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.851507998467929e-05,
+      "loss": 0.5787,
+      "step": 218
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.849939854267919e-05,
+      "loss": 0.5787,
+      "step": 219
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8483641446856798e-05,
+      "loss": 0.5654,
+      "step": 220
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8467808837467277e-05,
+      "loss": 0.5978,
+      "step": 221
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.5756,
+      "step": 222
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.843591764236702e-05,
+      "loss": 0.632,
+      "step": 223
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.841985934052234e-05,
+      "loss": 0.5594,
+      "step": 224
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.840372609284013e-05,
+      "loss": 0.5717,
+      "step": 225
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8387518042923715e-05,
+      "loss": 0.5942,
+      "step": 226
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8371235335042236e-05,
+      "loss": 0.5556,
+      "step": 227
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.5739,
+      "step": 228
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.833844652578203e-05,
+      "loss": 0.5416,
+      "step": 229
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.832194071625911e-05,
+      "loss": 0.5592,
+      "step": 230
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8305360832480118e-05,
+      "loss": 0.5863,
+      "step": 231
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8288707022023926e-05,
+      "loss": 0.5908,
+      "step": 232
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.827197943312742e-05,
+      "loss": 0.5676,
+      "step": 233
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.82551782146842e-05,
+      "loss": 0.5806,
+      "step": 234
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8238303516243253e-05,
+      "loss": 0.5638,
+      "step": 235
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8221355488007606e-05,
+      "loss": 0.5671,
+      "step": 236
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8204334280833005e-05,
+      "loss": 0.5803,
+      "step": 237
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8187240046226576e-05,
+      "loss": 0.5743,
+      "step": 238
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.817007293634545e-05,
+      "loss": 0.569,
+      "step": 239
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8152833103995443e-05,
+      "loss": 0.5674,
+      "step": 240
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8135520702629677e-05,
+      "loss": 0.5833,
+      "step": 241
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8118135886347207e-05,
+      "loss": 0.5816,
+      "step": 242
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8100678809891668e-05,
+      "loss": 0.5639,
+      "step": 243
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8083149628649887e-05,
+      "loss": 0.5853,
+      "step": 244
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8065548498650495e-05,
+      "loss": 0.601,
+      "step": 245
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8047875576562556e-05,
+      "loss": 0.5554,
+      "step": 246
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.803013101969415e-05,
+      "loss": 0.5508,
+      "step": 247
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.801231498599099e-05,
+      "loss": 0.5544,
+      "step": 248
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.5735,
+      "step": 249
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7976469123042955e-05,
+      "loss": 0.559,
+      "step": 250
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7958439612864954e-05,
+      "loss": 0.6069,
+      "step": 251
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7940339263983112e-05,
+      "loss": 0.5803,
+      "step": 252
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7922168237510076e-05,
+      "loss": 0.5735,
+      "step": 253
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7903926695187595e-05,
+      "loss": 0.6066,
+      "step": 254
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7885614799385086e-05,
+      "loss": 0.555,
+      "step": 255
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.78672327130982e-05,
+      "loss": 0.5682,
+      "step": 256
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7848780599947334e-05,
+      "loss": 0.5471,
+      "step": 257
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7830258624176224e-05,
+      "loss": 0.565,
+      "step": 258
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7811666950650445e-05,
+      "loss": 0.5991,
+      "step": 259
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7793005744855967e-05,
+      "loss": 0.5779,
+      "step": 260
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.777427517289766e-05,
+      "loss": 0.5721,
+      "step": 261
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.775547540149784e-05,
+      "loss": 0.5607,
+      "step": 262
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7736606597994763e-05,
+      "loss": 0.5671,
+      "step": 263
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7717668930341152e-05,
+      "loss": 0.5918,
+      "step": 264
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.769866256710269e-05,
+      "loss": 0.5635,
+      "step": 265
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.767958767745653e-05,
+      "loss": 0.5632,
+      "step": 266
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.5837,
+      "step": 267
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7641232998698e-05,
+      "loss": 0.5674,
+      "step": 268
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7621953550983677e-05,
+      "loss": 0.5711,
+      "step": 269
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7602606259654704e-05,
+      "loss": 0.5841,
+      "step": 270
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7583191296922866e-05,
+      "loss": 0.5553,
+      "step": 271
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7563708835602286e-05,
+      "loss": 0.5483,
+      "step": 272
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7544159049107902e-05,
+      "loss": 0.5745,
+      "step": 273
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7524542111453923e-05,
+      "loss": 0.5507,
+      "step": 274
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7504858197252263e-05,
+      "loss": 0.5859,
+      "step": 275
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.5756,
+      "step": 276
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.746529014063286e-05,
+      "loss": 0.58,
+      "step": 277
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7445406350413533e-05,
+      "loss": 0.5871,
+      "step": 278
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7425456288040236e-05,
+      "loss": 0.5763,
+      "step": 279
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.740544013109005e-05,
+      "loss": 0.5806,
+      "step": 280
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.738535805772838e-05,
+      "loss": 0.5669,
+      "step": 281
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.736521024670737e-05,
+      "loss": 0.6098,
+      "step": 282
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7344996877364282e-05,
+      "loss": 0.5752,
+      "step": 283
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.732471812961992e-05,
+      "loss": 0.544,
+      "step": 284
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7304374183977032e-05,
+      "loss": 0.5334,
+      "step": 285
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.72839652215187e-05,
+      "loss": 0.5391,
+      "step": 286
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7263491423906716e-05,
+      "loss": 0.5541,
+      "step": 287
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7242952973379983e-05,
+      "loss": 0.5868,
+      "step": 288
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7222350052752883e-05,
+      "loss": 0.5443,
+      "step": 289
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.720168284541365e-05,
+      "loss": 0.5885,
+      "step": 290
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7180951535322742e-05,
+      "loss": 0.5516,
+      "step": 291
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7160156307011197e-05,
+      "loss": 0.5732,
+      "step": 292
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7139297345578992e-05,
+      "loss": 0.5804,
+      "step": 293
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7118374836693407e-05,
+      "loss": 0.5837,
+      "step": 294
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7097388966587355e-05,
+      "loss": 0.5721,
+      "step": 295
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7076339922057736e-05,
+      "loss": 0.5694,
+      "step": 296
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.705522789046377e-05,
+      "loss": 0.5661,
+      "step": 297
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7034053059725325e-05,
+      "loss": 0.5817,
+      "step": 298
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7012815618321256e-05,
+      "loss": 0.5695,
+      "step": 299
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6991515755287715e-05,
+      "loss": 0.583,
+      "step": 300
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.697015366021648e-05,
+      "loss": 0.598,
+      "step": 301
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.694872952325326e-05,
+      "loss": 0.542,
+      "step": 302
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.5527,
+      "step": 303
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.690569588699318e-05,
+      "loss": 0.5398,
+      "step": 304
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6884086770742138e-05,
+      "loss": 0.5721,
+      "step": 305
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.5677,
+      "step": 306
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6840684903718658e-05,
+      "loss": 0.5662,
+      "step": 307
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.681889253926969e-05,
+      "loss": 0.5613,
+      "step": 308
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6797039479315994e-05,
+      "loss": 0.5582,
+      "step": 309
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.67751259183734e-05,
+      "loss": 0.5778,
+      "step": 310
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.675315205149626e-05,
+      "loss": 0.5823,
+      "step": 311
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.67311180742757e-05,
+      "loss": 0.5737,
+      "step": 312
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6709024182837917e-05,
+      "loss": 0.5577,
+      "step": 313
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6686870573842388e-05,
+      "loss": 0.5767,
+      "step": 314
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6664657444480145e-05,
+      "loss": 0.5516,
+      "step": 315
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6642384992472026e-05,
+      "loss": 0.5617,
+      "step": 316
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6620053416066892e-05,
+      "loss": 0.5553,
+      "step": 317
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6597662914039885e-05,
+      "loss": 0.542,
+      "step": 318
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.657521368569064e-05,
+      "loss": 0.5736,
+      "step": 319
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6552705930841523e-05,
+      "loss": 0.5953,
+      "step": 320
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.653013984983585e-05,
+      "loss": 0.5712,
+      "step": 321
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6507515643536113e-05,
+      "loss": 0.5722,
+      "step": 322
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6484833513322155e-05,
+      "loss": 0.5616,
+      "step": 323
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6462093661089432e-05,
+      "loss": 0.5646,
+      "step": 324
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.643929628924717e-05,
+      "loss": 0.572,
+      "step": 325
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6416441600716593e-05,
+      "loss": 0.5912,
+      "step": 326
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6393529798929103e-05,
+      "loss": 0.5722,
+      "step": 327
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.637056108782446e-05,
+      "loss": 0.5787,
+      "step": 328
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6347535671848998e-05,
+      "loss": 0.5789,
+      "step": 329
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.5521,
+      "step": 330
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6301315545592753e-05,
+      "loss": 0.5546,
+      "step": 331
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.627812124672099e-05,
+      "loss": 0.5657,
+      "step": 332
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6254871065792776e-05,
+      "loss": 0.5581,
+      "step": 333
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.623156520975983e-05,
+      "loss": 0.5759,
+      "step": 334
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.620820388606942e-05,
+      "loss": 0.5368,
+      "step": 335
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.618478730266255e-05,
+      "loss": 0.5714,
+      "step": 336
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6161315667972095e-05,
+      "loss": 0.5804,
+      "step": 337
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6137789190920938e-05,
+      "loss": 0.5604,
+      "step": 338
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6114208080920125e-05,
+      "loss": 0.5542,
+      "step": 339
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6090572547866983e-05,
+      "loss": 0.5492,
+      "step": 340
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.606688280214328e-05,
+      "loss": 0.5766,
+      "step": 341
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6043139054613326e-05,
+      "loss": 0.5604,
+      "step": 342
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.60193415166221e-05,
+      "loss": 0.5355,
+      "step": 343
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.599549039999338e-05,
+      "loss": 0.5473,
+      "step": 344
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.5665,
+      "step": 345
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.594762828050124e-05,
+      "loss": 0.5532,
+      "step": 346
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5923617703662346e-05,
+      "loss": 0.5699,
+      "step": 347
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5899554400231233e-05,
+      "loss": 0.574,
+      "step": 348
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.587543858439727e-05,
+      "loss": 0.5763,
+      "step": 349
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.585127047081727e-05,
+      "loss": 0.5694,
+      "step": 350
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5827050274613512e-05,
+      "loss": 0.5707,
+      "step": 351
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.580277821137191e-05,
+      "loss": 0.5679,
+      "step": 352
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.577845449714001e-05,
+      "loss": 0.5722,
+      "step": 353
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5754079348425137e-05,
+      "loss": 0.5476,
+      "step": 354
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5729652982192428e-05,
+      "loss": 0.5583,
+      "step": 355
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5705175615862906e-05,
+      "loss": 0.5941,
+      "step": 356
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.5828,
+      "step": 357
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5656068754865388e-05,
+      "loss": 0.5647,
+      "step": 358
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5631439697301464e-05,
+      "loss": 0.5693,
+      "step": 359
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.560676051384499e-05,
+      "loss": 0.5682,
+      "step": 360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.558203142416734e-05,
+      "loss": 0.5573,
+      "step": 361
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5557252648384103e-05,
+      "loss": 0.5404,
+      "step": 362
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.553242440705314e-05,
+      "loss": 0.3841,
+      "step": 363
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5507546921172595e-05,
+      "loss": 0.3967,
+      "step": 364
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.548262041217895e-05,
+      "loss": 0.3613,
+      "step": 365
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5457645101945046e-05,
+      "loss": 0.3973,
+      "step": 366
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5432621212778105e-05,
+      "loss": 0.3661,
+      "step": 367
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5407548967417755e-05,
+      "loss": 0.3624,
+      "step": 368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.538242858903404e-05,
+      "loss": 0.3701,
+      "step": 369
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5357260301225448e-05,
+      "loss": 0.382,
+      "step": 370
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5332044328016916e-05,
+      "loss": 0.3836,
+      "step": 371
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.530678089385782e-05,
+      "loss": 0.3734,
+      "step": 372
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5281470223619995e-05,
+      "loss": 0.3927,
+      "step": 373
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.525611254259574e-05,
+      "loss": 0.3981,
+      "step": 374
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5230708076495777e-05,
+      "loss": 0.3578,
+      "step": 375
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5205257051447291e-05,
+      "loss": 0.3893,
+      "step": 376
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5179759693991869e-05,
+      "loss": 0.3726,
+      "step": 377
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5154216231083522e-05,
+      "loss": 0.367,
+      "step": 378
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5128626890086647e-05,
+      "loss": 0.3768,
+      "step": 379
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5102991898774e-05,
+      "loss": 0.3667,
+      "step": 380
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.507731148532468e-05,
+      "loss": 0.3372,
+      "step": 381
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.505158587832208e-05,
+      "loss": 0.3377,
+      "step": 382
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5025815306751888e-05,
+      "loss": 0.3908,
+      "step": 383
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.3642,
+      "step": 384
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.497414018785052e-05,
+      "loss": 0.3626,
+      "step": 385
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4948236100483688e-05,
+      "loss": 0.373,
+      "step": 386
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.492228796847385e-05,
+      "loss": 0.3765,
+      "step": 387
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4896296022787386e-05,
+      "loss": 0.3767,
+      "step": 388
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4870260494780679e-05,
+      "loss": 0.3589,
+      "step": 389
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4844181616198028e-05,
+      "loss": 0.3722,
+      "step": 390
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4818059619169606e-05,
+      "loss": 0.3806,
+      "step": 391
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.479189473620939e-05,
+      "loss": 0.3662,
+      "step": 392
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4765687200213079e-05,
+      "loss": 0.3588,
+      "step": 393
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4739437244456039e-05,
+      "loss": 0.3673,
+      "step": 394
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4713145102591209e-05,
+      "loss": 0.3733,
+      "step": 395
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4686811008647037e-05,
+      "loss": 0.3506,
+      "step": 396
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4660435197025391e-05,
+      "loss": 0.3537,
+      "step": 397
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.463401790249946e-05,
+      "loss": 0.3642,
+      "step": 398
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4607559360211688e-05,
+      "loss": 0.37,
+      "step": 399
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4581059805671662e-05,
+      "loss": 0.3789,
+      "step": 400
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4554519474754025e-05,
+      "loss": 0.3633,
+      "step": 401
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4527938603696376e-05,
+      "loss": 0.353,
+      "step": 402
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4501317429097155e-05,
+      "loss": 0.3657,
+      "step": 403
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4474656187913558e-05,
+      "loss": 0.3749,
+      "step": 404
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4447955117459414e-05,
+      "loss": 0.3729,
+      "step": 405
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4421214455403071e-05,
+      "loss": 0.368,
+      "step": 406
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4394434439765295e-05,
+      "loss": 0.3695,
+      "step": 407
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.436761530891713e-05,
+      "loss": 0.3527,
+      "step": 408
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4340757301577787e-05,
+      "loss": 0.3662,
+      "step": 409
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4313860656812537e-05,
+      "loss": 0.3699,
+      "step": 410
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4286925614030542e-05,
+      "loss": 0.3728,
+      "step": 411
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.425995241298277e-05,
+      "loss": 0.3927,
+      "step": 412
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.423294129375982e-05,
+      "loss": 0.3617,
+      "step": 413
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4205892496789816e-05,
+      "loss": 0.3785,
+      "step": 414
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4178806262836252e-05,
+      "loss": 0.378,
+      "step": 415
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4151682832995846e-05,
+      "loss": 0.3728,
+      "step": 416
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4124522448696407e-05,
+      "loss": 0.3805,
+      "step": 417
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4097325351694674e-05,
+      "loss": 0.3701,
+      "step": 418
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.407009178407417e-05,
+      "loss": 0.3797,
+      "step": 419
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.404282198824305e-05,
+      "loss": 0.3732,
+      "step": 420
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4015516206931932e-05,
+      "loss": 0.3713,
+      "step": 421
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3988174683191744e-05,
+      "loss": 0.3615,
+      "step": 422
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.3651,
+      "step": 423
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.393338538221646e-05,
+      "loss": 0.3725,
+      "step": 424
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3905938092665283e-05,
+      "loss": 0.3889,
+      "step": 425
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.387845603604855e-05,
+      "loss": 0.3567,
+      "step": 426
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.385093945698623e-05,
+      "loss": 0.3966,
+      "step": 427
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.382338860040558e-05,
+      "loss": 0.3757,
+      "step": 428
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3795803711538966e-05,
+      "loss": 0.3588,
+      "step": 429
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3768185035921677e-05,
+      "loss": 0.3555,
+      "step": 430
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.374053281938974e-05,
+      "loss": 0.3723,
+      "step": 431
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3712847308077737e-05,
+      "loss": 0.3679,
+      "step": 432
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3685128748416603e-05,
+      "loss": 0.3496,
+      "step": 433
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3657377387131443e-05,
+      "loss": 0.3468,
+      "step": 434
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3629593471239328e-05,
+      "loss": 0.3759,
+      "step": 435
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3601777248047105e-05,
+      "loss": 0.3776,
+      "step": 436
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3573928965149188e-05,
+      "loss": 0.3797,
+      "step": 437
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.3842,
+      "step": 438
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3518137212038554e-05,
+      "loss": 0.3429,
+      "step": 439
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3490194238432665e-05,
+      "loss": 0.3625,
+      "step": 440
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.346222019833033e-05,
+      "loss": 0.3915,
+      "step": 441
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3434215340730692e-05,
+      "loss": 0.3763,
+      "step": 442
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.340617991490722e-05,
+      "loss": 0.3835,
+      "step": 443
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3378114170405473e-05,
+      "loss": 0.3642,
+      "step": 444
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.335001835704087e-05,
+      "loss": 0.3768,
+      "step": 445
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3321892724896483e-05,
+      "loss": 0.3579,
+      "step": 446
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3293737524320798e-05,
+      "loss": 0.3911,
+      "step": 447
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3265553005925494e-05,
+      "loss": 0.3634,
+      "step": 448
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3237339420583213e-05,
+      "loss": 0.3706,
+      "step": 449
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3209097019425317e-05,
+      "loss": 0.3438,
+      "step": 450
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3180826053839668e-05,
+      "loss": 0.3613,
+      "step": 451
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.315252677546838e-05,
+      "loss": 0.3611,
+      "step": 452
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3124199436205575e-05,
+      "loss": 0.3731,
+      "step": 453
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.309584428819516e-05,
+      "loss": 0.3602,
+      "step": 454
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3067461583828553e-05,
+      "loss": 0.3541,
+      "step": 455
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.303905157574247e-05,
+      "loss": 0.3681,
+      "step": 456
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3010614516816652e-05,
+      "loss": 0.3884,
+      "step": 457
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2982150660171613e-05,
+      "loss": 0.3739,
+      "step": 458
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2953660259166413e-05,
+      "loss": 0.3745,
+      "step": 459
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2925143567396374e-05,
+      "loss": 0.3525,
+      "step": 460
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2896600838690838e-05,
+      "loss": 0.3991,
+      "step": 461
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.3769,
+      "step": 462
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2839438286947163e-05,
+      "loss": 0.3871,
+      "step": 463
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2810818972717438e-05,
+      "loss": 0.3711,
+      "step": 464
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2782174639164528e-05,
+      "loss": 0.366,
+      "step": 465
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2753505541253917e-05,
+      "loss": 0.3688,
+      "step": 466
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.272481193417153e-05,
+      "loss": 0.3839,
+      "step": 467
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.269609407332144e-05,
+      "loss": 0.3831,
+      "step": 468
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2667352214323614e-05,
+      "loss": 0.3902,
+      "step": 469
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2638586613011625e-05,
+      "loss": 0.3719,
+      "step": 470
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2609797525430374e-05,
+      "loss": 0.3711,
+      "step": 471
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.258098520783382e-05,
+      "loss": 0.3709,
+      "step": 472
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2552149916682695e-05,
+      "loss": 0.3957,
+      "step": 473
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2523291908642219e-05,
+      "loss": 0.3679,
+      "step": 474
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2494411440579814e-05,
+      "loss": 0.3723,
+      "step": 475
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2465508769562824e-05,
+      "loss": 0.3823,
+      "step": 476
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.243658415285622e-05,
+      "loss": 0.3715,
+      "step": 477
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.240763784792032e-05,
+      "loss": 0.3638,
+      "step": 478
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2378670112408482e-05,
+      "loss": 0.3618,
+      "step": 479
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2349681204164823e-05,
+      "loss": 0.3665,
+      "step": 480
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2320671381221923e-05,
+      "loss": 0.3723,
+      "step": 481
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.229164090179852e-05,
+      "loss": 0.3785,
+      "step": 482
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2262590024297226e-05,
+      "loss": 0.3561,
+      "step": 483
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2233519007302201e-05,
+      "loss": 0.3552,
+      "step": 484
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2204428109576888e-05,
+      "loss": 0.3798,
+      "step": 485
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2175317590061676e-05,
+      "loss": 0.3712,
+      "step": 486
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2146187707871617e-05,
+      "loss": 0.3617,
+      "step": 487
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.211703872229411e-05,
+      "loss": 0.3623,
+      "step": 488
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2087870892786588e-05,
+      "loss": 0.3896,
+      "step": 489
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2058684478974226e-05,
+      "loss": 0.3755,
+      "step": 490
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2029479740647613e-05,
+      "loss": 0.378,
+      "step": 491
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2000256937760446e-05,
+      "loss": 0.3861,
+      "step": 492
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1971016330427217e-05,
+      "loss": 0.3755,
+      "step": 493
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1941758178920899e-05,
+      "loss": 0.3651,
+      "step": 494
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1912482743670624e-05,
+      "loss": 0.3527,
+      "step": 495
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.188319028525937e-05,
+      "loss": 0.3736,
+      "step": 496
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1853881064421634e-05,
+      "loss": 0.3706,
+      "step": 497
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1824555342041129e-05,
+      "loss": 0.3642,
+      "step": 498
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1795213379148436e-05,
+      "loss": 0.3746,
+      "step": 499
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.17658554369187e-05,
+      "loss": 0.3695,
+      "step": 500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.3776,
+      "step": 501
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1707092659857531e-05,
+      "loss": 0.3835,
+      "step": 502
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1677688348078244e-05,
+      "loss": 0.3601,
+      "step": 503
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1648269103061567e-05,
+      "loss": 0.3816,
+      "step": 504
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1618835186670532e-05,
+      "loss": 0.3655,
+      "step": 505
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1589386860898773e-05,
+      "loss": 0.3838,
+      "step": 506
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.155992438786818e-05,
+      "loss": 0.3729,
+      "step": 507
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1530448029826566e-05,
+      "loss": 0.3459,
+      "step": 508
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1500958049145342e-05,
+      "loss": 0.3848,
+      "step": 509
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1471454708317163e-05,
+      "loss": 0.3574,
+      "step": 510
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1441938269953619e-05,
+      "loss": 0.3762,
+      "step": 511
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1412408996782871e-05,
+      "loss": 0.36,
+      "step": 512
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1382867151647333e-05,
+      "loss": 0.3817,
+      "step": 513
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1353312997501313e-05,
+      "loss": 0.3557,
+      "step": 514
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1323746797408687e-05,
+      "loss": 0.3541,
+      "step": 515
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1294168814540554e-05,
+      "loss": 0.3618,
+      "step": 516
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1264579312172896e-05,
+      "loss": 0.3815,
+      "step": 517
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.123497855368422e-05,
+      "loss": 0.3815,
+      "step": 518
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.3814,
+      "step": 519
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1175744322356488e-05,
+      "loss": 0.3881,
+      "step": 520
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1146111376766033e-05,
+      "loss": 0.3601,
+      "step": 521
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1116468229547079e-05,
+      "loss": 0.3655,
+      "step": 522
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1086815144555633e-05,
+      "loss": 0.3659,
+      "step": 523
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.105715238573616e-05,
+      "loss": 0.3724,
+      "step": 524
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1027480217119245e-05,
+      "loss": 0.3842,
+      "step": 525
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0997798902819208e-05,
+      "loss": 0.3529,
+      "step": 526
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0968108707031792e-05,
+      "loss": 0.3669,
+      "step": 527
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0938409894031793e-05,
+      "loss": 0.3791,
+      "step": 528
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0908702728170706e-05,
+      "loss": 0.3616,
+      "step": 529
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0878987473874381e-05,
+      "loss": 0.3622,
+      "step": 530
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.084926439564065e-05,
+      "loss": 0.359,
+      "step": 531
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0819533758037002e-05,
+      "loss": 0.3835,
+      "step": 532
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0789795825698206e-05,
+      "loss": 0.3724,
+      "step": 533
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0760050863323961e-05,
+      "loss": 0.3876,
+      "step": 534
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0730299135676545e-05,
+      "loss": 0.3643,
+      "step": 535
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0700540907578447e-05,
+      "loss": 0.3728,
+      "step": 536
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0670776443910024e-05,
+      "loss": 0.3879,
+      "step": 537
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0641006009607137e-05,
+      "loss": 0.3653,
+      "step": 538
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0611229869658785e-05,
+      "loss": 0.3573,
+      "step": 539
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.3778,
+      "step": 540
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0551661533033274e-05,
+      "loss": 0.3645,
+      "step": 541
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.052186986657862e-05,
+      "loss": 0.3689,
+      "step": 542
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0492073554918782e-05,
+      "loss": 0.3755,
+      "step": 543
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0462272863273105e-05,
+      "loss": 0.3854,
+      "step": 544
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0432468056899909e-05,
+      "loss": 0.3588,
+      "step": 545
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0402659401094154e-05,
+      "loss": 0.3766,
+      "step": 546
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0372847161185047e-05,
+      "loss": 0.366,
+      "step": 547
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0343031602533713e-05,
+      "loss": 0.3857,
+      "step": 548
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0313212990530804e-05,
+      "loss": 0.3709,
+      "step": 549
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.028339159059416e-05,
+      "loss": 0.3844,
+      "step": 550
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0253567668166436e-05,
+      "loss": 0.3667,
+      "step": 551
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0223741488712732e-05,
+      "loss": 0.3599,
+      "step": 552
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0193913317718245e-05,
+      "loss": 0.4021,
+      "step": 553
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0164083420685898e-05,
+      "loss": 0.3895,
+      "step": 554
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0134252063133976e-05,
+      "loss": 0.3554,
+      "step": 555
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0104419510593764e-05,
+      "loss": 0.3687,
+      "step": 556
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0074586028607184e-05,
+      "loss": 0.3595,
+      "step": 557
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0044751882724436e-05,
+      "loss": 0.3596,
+      "step": 558
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0014917338501618e-05,
+      "loss": 0.376,
+      "step": 559
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.985082661498384e-06,
+      "loss": 0.3869,
+      "step": 560
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.955248117275566e-06,
+      "loss": 0.3608,
+      "step": 561
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.925413971392816e-06,
+      "loss": 0.3611,
+      "step": 562
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.89558048940624e-06,
+      "loss": 0.3865,
+      "step": 563
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.865747936866027e-06,
+      "loss": 0.345,
+      "step": 564
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.835916579314105e-06,
+      "loss": 0.3651,
+      "step": 565
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.806086682281759e-06,
+      "loss": 0.3669,
+      "step": 566
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.776258511287271e-06,
+      "loss": 0.3631,
+      "step": 567
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.746432331833568e-06,
+      "loss": 0.3675,
+      "step": 568
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.716608409405843e-06,
+      "loss": 0.3678,
+      "step": 569
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.6867870094692e-06,
+      "loss": 0.3642,
+      "step": 570
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.65696839746629e-06,
+      "loss": 0.3542,
+      "step": 571
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.627152838814954e-06,
+      "loss": 0.3592,
+      "step": 572
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.597340598905851e-06,
+      "loss": 0.3719,
+      "step": 573
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.567531943100093e-06,
+      "loss": 0.3701,
+      "step": 574
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.537727136726899e-06,
+      "loss": 0.3874,
+      "step": 575
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.50792644508122e-06,
+      "loss": 0.3636,
+      "step": 576
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.478130133421381e-06,
+      "loss": 0.3749,
+      "step": 577
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.448338466966727e-06,
+      "loss": 0.351,
+      "step": 578
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.3623,
+      "step": 579
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.388770130341217e-06,
+      "loss": 0.3531,
+      "step": 580
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.358993990392864e-06,
+      "loss": 0.3644,
+      "step": 581
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.329223556089976e-06,
+      "loss": 0.3581,
+      "step": 582
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.299459092421558e-06,
+      "loss": 0.3524,
+      "step": 583
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.269700864323462e-06,
+      "loss": 0.3537,
+      "step": 584
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.239949136676042e-06,
+      "loss": 0.3397,
+      "step": 585
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.210204174301797e-06,
+      "loss": 0.3755,
+      "step": 586
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.180466241963e-06,
+      "loss": 0.3735,
+      "step": 587
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.15073560435935e-06,
+      "loss": 0.3342,
+      "step": 588
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.121012526125625e-06,
+      "loss": 0.3677,
+      "step": 589
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.091297271829296e-06,
+      "loss": 0.3452,
+      "step": 590
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.061590105968208e-06,
+      "loss": 0.3471,
+      "step": 591
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.03189129296821e-06,
+      "loss": 0.3572,
+      "step": 592
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.002201097180796e-06,
+      "loss": 0.3559,
+      "step": 593
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.97251978288076e-06,
+      "loss": 0.3595,
+      "step": 594
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.942847614263842e-06,
+      "loss": 0.3808,
+      "step": 595
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.91318485544437e-06,
+      "loss": 0.3558,
+      "step": 596
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.883531770452924e-06,
+      "loss": 0.3632,
+      "step": 597
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.853888623233967e-06,
+      "loss": 0.3592,
+      "step": 598
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.824255677643518e-06,
+      "loss": 0.3785,
+      "step": 599
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.3601,
+      "step": 600
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.765021446315784e-06,
+      "loss": 0.3609,
+      "step": 601
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.735420687827107e-06,
+      "loss": 0.336,
+      "step": 602
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.705831185459446e-06,
+      "loss": 0.3749,
+      "step": 603
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.676253202591318e-06,
+      "loss": 0.3816,
+      "step": 604
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.646687002498692e-06,
+      "loss": 0.3731,
+      "step": 605
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.617132848352672e-06,
+      "loss": 0.3894,
+      "step": 606
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.58759100321713e-06,
+      "loss": 0.3695,
+      "step": 607
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.558061730046385e-06,
+      "loss": 0.38,
+      "step": 608
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.528545291682839e-06,
+      "loss": 0.3665,
+      "step": 609
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.499041950854665e-06,
+      "loss": 0.3779,
+      "step": 610
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.469551970173437e-06,
+      "loss": 0.3664,
+      "step": 611
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.440075612131823e-06,
+      "loss": 0.3528,
+      "step": 612
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.410613139101229e-06,
+      "loss": 0.3897,
+      "step": 613
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.38116481332947e-06,
+      "loss": 0.3561,
+      "step": 614
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.351730896938438e-06,
+      "loss": 0.3549,
+      "step": 615
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.322311651921759e-06,
+      "loss": 0.3662,
+      "step": 616
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.292907340142472e-06,
+      "loss": 0.3593,
+      "step": 617
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.3719,
+      "step": 618
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.2341445630813e-06,
+      "loss": 0.3702,
+      "step": 619
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.204786620851569e-06,
+      "loss": 0.3527,
+      "step": 620
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.175444657958875e-06,
+      "loss": 0.36,
+      "step": 621
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.146118935578367e-06,
+      "loss": 0.3684,
+      "step": 622
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.116809714740634e-06,
+      "loss": 0.3731,
+      "step": 623
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.087517256329376e-06,
+      "loss": 0.3632,
+      "step": 624
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.058241821079106e-06,
+      "loss": 0.3903,
+      "step": 625
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.028983669572786e-06,
+      "loss": 0.3502,
+      "step": 626
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.999743062239557e-06,
+      "loss": 0.3556,
+      "step": 627
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.97052025935239e-06,
+      "loss": 0.3632,
+      "step": 628
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.941315521025776e-06,
+      "loss": 0.3486,
+      "step": 629
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.912129107213417e-06,
+      "loss": 0.3796,
+      "step": 630
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.882961277705897e-06,
+      "loss": 0.3785,
+      "step": 631
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.853812292128386e-06,
+      "loss": 0.3523,
+      "step": 632
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.824682409938328e-06,
+      "loss": 0.3475,
+      "step": 633
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.795571890423116e-06,
+      "loss": 0.3608,
+      "step": 634
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.766480992697802e-06,
+      "loss": 0.3741,
+      "step": 635
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.73740997570278e-06,
+      "loss": 0.3508,
+      "step": 636
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.708359098201483e-06,
+      "loss": 0.358,
+      "step": 637
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.67932861877808e-06,
+      "loss": 0.3719,
+      "step": 638
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.650318795835179e-06,
+      "loss": 0.4074,
+      "step": 639
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.621329887591519e-06,
+      "loss": 0.3753,
+      "step": 640
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.592362152079684e-06,
+      "loss": 0.3655,
+      "step": 641
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.563415847143782e-06,
+      "loss": 0.36,
+      "step": 642
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.5344912304371785e-06,
+      "loss": 0.377,
+      "step": 643
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.505588559420188e-06,
+      "loss": 0.3507,
+      "step": 644
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.476708091357783e-06,
+      "loss": 0.3477,
+      "step": 645
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.447850083317308e-06,
+      "loss": 0.3449,
+      "step": 646
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.419014792166182e-06,
+      "loss": 0.3621,
+      "step": 647
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.39020247456963e-06,
+      "loss": 0.3528,
+      "step": 648
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.361413386988379e-06,
+      "loss": 0.3636,
+      "step": 649
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.332647785676388e-06,
+      "loss": 0.382,
+      "step": 650
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.303905926678565e-06,
+      "loss": 0.3727,
+      "step": 651
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.275188065828476e-06,
+      "loss": 0.3747,
+      "step": 652
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.246494458746085e-06,
+      "loss": 0.3647,
+      "step": 653
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.3822,
+      "step": 654
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.189181027282561e-06,
+      "loss": 0.3632,
+      "step": 655
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.160561713052843e-06,
+      "loss": 0.3586,
+      "step": 656
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.374,
+      "step": 657
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.103399161309165e-06,
+      "loss": 0.3574,
+      "step": 658
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.074856432603628e-06,
+      "loss": 0.3624,
+      "step": 659
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.04633974083359e-06,
+      "loss": 0.3573,
+      "step": 660
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.017849339828388e-06,
+      "loss": 0.3515,
+      "step": 661
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.989385483183355e-06,
+      "loss": 0.3655,
+      "step": 662
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.960948424257532e-06,
+      "loss": 0.3645,
+      "step": 663
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9325384161714485e-06,
+      "loss": 0.358,
+      "step": 664
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.904155711804843e-06,
+      "loss": 0.3885,
+      "step": 665
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.8758005637944245e-06,
+      "loss": 0.3531,
+      "step": 666
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.8474732245316245e-06,
+      "loss": 0.3602,
+      "step": 667
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.819173946160336e-06,
+      "loss": 0.3647,
+      "step": 668
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.7909029805746855e-06,
+      "loss": 0.3486,
+      "step": 669
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.762660579416791e-06,
+      "loss": 0.3629,
+      "step": 670
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.734446994074507e-06,
+      "loss": 0.3513,
+      "step": 671
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.706262475679205e-06,
+      "loss": 0.3663,
+      "step": 672
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.678107275103519e-06,
+      "loss": 0.3781,
+      "step": 673
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.649981642959133e-06,
+      "loss": 0.3477,
+      "step": 674
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.62188582959453e-06,
+      "loss": 0.3521,
+      "step": 675
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.593820085092782e-06,
+      "loss": 0.374,
+      "step": 676
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.565784659269314e-06,
+      "loss": 0.3693,
+      "step": 677
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.537779801669677e-06,
+      "loss": 0.3716,
+      "step": 678
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.509805761567336e-06,
+      "loss": 0.3757,
+      "step": 679
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.481862787961448e-06,
+      "loss": 0.3731,
+      "step": 680
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.3586,
+      "step": 681
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.4260710348508115e-06,
+      "loss": 0.3442,
+      "step": 682
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.3982227519528986e-06,
+      "loss": 0.3531,
+      "step": 683
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.370406528760675e-06,
+      "loss": 0.347,
+      "step": 684
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.34262261286856e-06,
+      "loss": 0.3709,
+      "step": 685
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.3148712515833985e-06,
+      "loss": 0.3621,
+      "step": 686
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.287152691922264e-06,
+      "loss": 0.3637,
+      "step": 687
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.259467180610262e-06,
+      "loss": 0.3662,
+      "step": 688
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.231814964078327e-06,
+      "loss": 0.3664,
+      "step": 689
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.204196288461037e-06,
+      "loss": 0.357,
+      "step": 690
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.176611399594422e-06,
+      "loss": 0.3548,
+      "step": 691
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.149060543013772e-06,
+      "loss": 0.3525,
+      "step": 692
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.121543963951453e-06,
+      "loss": 0.3715,
+      "step": 693
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.094061907334718e-06,
+      "loss": 0.374,
+      "step": 694
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.066614617783542e-06,
+      "loss": 0.3788,
+      "step": 695
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.3699,
+      "step": 696
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.0118253168082555e-06,
+      "loss": 0.3657,
+      "step": 697
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.984483793068072e-06,
+      "loss": 0.3583,
+      "step": 698
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.957178011756952e-06,
+      "loss": 0.3758,
+      "step": 699
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.92990821592583e-06,
+      "loss": 0.3538,
+      "step": 700
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.902674648305329e-06,
+      "loss": 0.3587,
+      "step": 701
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.875477551303596e-06,
+      "loss": 0.3666,
+      "step": 702
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.848317167004159e-06,
+      "loss": 0.3548,
+      "step": 703
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8211937371637525e-06,
+      "loss": 0.375,
+      "step": 704
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.794107503210187e-06,
+      "loss": 0.3574,
+      "step": 705
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.767058706240183e-06,
+      "loss": 0.3744,
+      "step": 706
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.740047587017232e-06,
+      "loss": 0.3681,
+      "step": 707
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.713074385969457e-06,
+      "loss": 0.3551,
+      "step": 708
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.686139343187468e-06,
+      "loss": 0.3486,
+      "step": 709
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.659242698422213e-06,
+      "loss": 0.3761,
+      "step": 710
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.632384691082874e-06,
+      "loss": 0.3496,
+      "step": 711
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.605565560234707e-06,
+      "loss": 0.3514,
+      "step": 712
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.578785544596928e-06,
+      "loss": 0.3481,
+      "step": 713
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.55204488254059e-06,
+      "loss": 0.3694,
+      "step": 714
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.525343812086445e-06,
+      "loss": 0.3715,
+      "step": 715
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.498682570902849e-06,
+      "loss": 0.3674,
+      "step": 716
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.47206139630363e-06,
+      "loss": 0.3527,
+      "step": 717
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.445480525245976e-06,
+      "loss": 0.3754,
+      "step": 718
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.418940194328344e-06,
+      "loss": 0.366,
+      "step": 719
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.3924406397883174e-06,
+      "loss": 0.3668,
+      "step": 720
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.365982097500545e-06,
+      "loss": 0.354,
+      "step": 721
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.339564802974615e-06,
+      "loss": 0.3738,
+      "step": 722
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.313188991352964e-06,
+      "loss": 0.3561,
+      "step": 723
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.286854897408793e-06,
+      "loss": 0.3586,
+      "step": 724
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.2605627555439635e-06,
+      "loss": 0.2377,
+      "step": 725
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.234312799786921e-06,
+      "loss": 0.2292,
+      "step": 726
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.208105263790611e-06,
+      "loss": 0.2117,
+      "step": 727
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.181940380830393e-06,
+      "loss": 0.2141,
+      "step": 728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.155818383801976e-06,
+      "loss": 0.2237,
+      "step": 729
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.129739505219325e-06,
+      "loss": 0.2236,
+      "step": 730
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.103703977212615e-06,
+      "loss": 0.2053,
+      "step": 731
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.077712031526153e-06,
+      "loss": 0.2234,
+      "step": 732
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.051763899516313e-06,
+      "loss": 0.2308,
+      "step": 733
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.025859812149481e-06,
+      "loss": 0.2188,
+      "step": 734
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.2197,
+      "step": 735
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.9741846932481154e-06,
+      "loss": 0.2091,
+      "step": 736
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.94841412167792e-06,
+      "loss": 0.224,
+      "step": 737
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.922688514675325e-06,
+      "loss": 0.2023,
+      "step": 738
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.8970081012260014e-06,
+      "loss": 0.2062,
+      "step": 739
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.8713731099133576e-06,
+      "loss": 0.2181,
+      "step": 740
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.845783768916482e-06,
+      "loss": 0.223,
+      "step": 741
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.820240306008136e-06,
+      "loss": 0.2088,
+      "step": 742
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.794742948552716e-06,
+      "loss": 0.2118,
+      "step": 743
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.769291923504226e-06,
+      "loss": 0.1877,
+      "step": 744
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.743887457404268e-06,
+      "loss": 0.2117,
+      "step": 745
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.718529776380009e-06,
+      "loss": 0.2046,
+      "step": 746
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.693219106142186e-06,
+      "loss": 0.2045,
+      "step": 747
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.66795567198309e-06,
+      "loss": 0.2031,
+      "step": 748
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.642739698774555e-06,
+      "loss": 0.2075,
+      "step": 749
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.617571410965964e-06,
+      "loss": 0.2085,
+      "step": 750
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.59245103258225e-06,
+      "loss": 0.2185,
+      "step": 751
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.567378787221896e-06,
+      "loss": 0.2095,
+      "step": 752
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.542354898054953e-06,
+      "loss": 0.2002,
+      "step": 753
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.517379587821049e-06,
+      "loss": 0.2133,
+      "step": 754
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.492453078827409e-06,
+      "loss": 0.2211,
+      "step": 755
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.467575592946865e-06,
+      "loss": 0.2032,
+      "step": 756
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.442747351615899e-06,
+      "loss": 0.2272,
+      "step": 757
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.417968575832664e-06,
+      "loss": 0.2108,
+      "step": 758
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.393239486155011e-06,
+      "loss": 0.2128,
+      "step": 759
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3685603026985356e-06,
+      "loss": 0.2137,
+      "step": 760
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.343931245134616e-06,
+      "loss": 0.2043,
+      "step": 761
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.2334,
+      "step": 762
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.294824384137096e-06,
+      "loss": 0.2026,
+      "step": 763
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.270347017807575e-06,
+      "loss": 0.2238,
+      "step": 764
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.245920651574864e-06,
+      "loss": 0.2203,
+      "step": 765
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.221545502859994e-06,
+      "loss": 0.2158,
+      "step": 766
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.197221788628096e-06,
+      "loss": 0.2122,
+      "step": 767
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.172949725386488e-06,
+      "loss": 0.2159,
+      "step": 768
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.148729529182736e-06,
+      "loss": 0.2009,
+      "step": 769
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.124561415602729e-06,
+      "loss": 0.2157,
+      "step": 770
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.100445599768774e-06,
+      "loss": 0.2038,
+      "step": 771
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.0763822963376585e-06,
+      "loss": 0.2241,
+      "step": 772
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0523717194987634e-06,
+      "loss": 0.204,
+      "step": 773
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.2032,
+      "step": 774
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.004509600006618e-06,
+      "loss": 0.2168,
+      "step": 775
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.980658483377903e-06,
+      "loss": 0.2236,
+      "step": 776
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.956860945386677e-06,
+      "loss": 0.2009,
+      "step": 777
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9331171978567206e-06,
+      "loss": 0.1957,
+      "step": 778
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.909427452133017e-06,
+      "loss": 0.2151,
+      "step": 779
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.885791919079878e-06,
+      "loss": 0.209,
+      "step": 780
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.862210809079061e-06,
+      "loss": 0.2104,
+      "step": 781
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.838684332027908e-06,
+      "loss": 0.2049,
+      "step": 782
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.815212697337451e-06,
+      "loss": 0.2045,
+      "step": 783
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7917961139305835e-06,
+      "loss": 0.2135,
+      "step": 784
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7684347902401753e-06,
+      "loss": 0.2193,
+      "step": 785
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.745128934207225e-06,
+      "loss": 0.2092,
+      "step": 786
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7218787532790167e-06,
+      "loss": 0.2122,
+      "step": 787
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6986844544072496e-06,
+      "loss": 0.2306,
+      "step": 788
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6755462440462288e-06,
+      "loss": 0.2156,
+      "step": 789
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6524643281510018e-06,
+      "loss": 0.2118,
+      "step": 790
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6294389121755404e-06,
+      "loss": 0.2066,
+      "step": 791
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.606470201070904e-06,
+      "loss": 0.2072,
+      "step": 792
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.58355839928341e-06,
+      "loss": 0.2063,
+      "step": 793
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.560703710752833e-06,
+      "loss": 0.218,
+      "step": 794
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5379063389105727e-06,
+      "loss": 0.2176,
+      "step": 795
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.515166486677848e-06,
+      "loss": 0.213,
+      "step": 796
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4924843564638946e-06,
+      "loss": 0.2186,
+      "step": 797
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4698601501641517e-06,
+      "loss": 0.2201,
+      "step": 798
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.447294069158481e-06,
+      "loss": 0.2075,
+      "step": 799
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.424786314309365e-06,
+      "loss": 0.2041,
+      "step": 800
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4023370859601192e-06,
+      "loss": 0.2091,
+      "step": 801
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3799465839331103e-06,
+      "loss": 0.2237,
+      "step": 802
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3576150075279757e-06,
+      "loss": 0.2098,
+      "step": 803
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.335342555519855e-06,
+      "loss": 0.1901,
+      "step": 804
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.313129426157613e-06,
+      "loss": 0.1949,
+      "step": 805
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.290975817162082e-06,
+      "loss": 0.2193,
+      "step": 806
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2688819257242963e-06,
+      "loss": 0.2073,
+      "step": 807
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.246847948503744e-06,
+      "loss": 0.195,
+      "step": 808
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2248740816266012e-06,
+      "loss": 0.2088,
+      "step": 809
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2029605206840088e-06,
+      "loss": 0.2114,
+      "step": 810
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.181107460730314e-06,
+      "loss": 0.2148,
+      "step": 811
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1593150962813425e-06,
+      "loss": 0.217,
+      "step": 812
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 0.2092,
+      "step": 813
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.115913229257864e-06,
+      "loss": 0.2159,
+      "step": 814
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0943041130068243e-06,
+      "loss": 0.216,
+      "step": 815
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0727564649040066e-06,
+      "loss": 0.2258,
+      "step": 816
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0512704767467417e-06,
+      "loss": 0.1863,
+      "step": 817
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0298463397835223e-06,
+      "loss": 0.2179,
+      "step": 818
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.008484244712286e-06,
+      "loss": 0.1884,
+      "step": 819
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.987184381678747e-06,
+      "loss": 0.214,
+      "step": 820
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.965946940274678e-06,
+      "loss": 0.1999,
+      "step": 821
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9447721095362325e-06,
+      "loss": 0.2018,
+      "step": 822
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9236600779422674e-06,
+      "loss": 0.2079,
+      "step": 823
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.902611033412648e-06,
+      "loss": 0.2203,
+      "step": 824
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8816251633065963e-06,
+      "loss": 0.2093,
+      "step": 825
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8607026544210115e-06,
+      "loss": 0.2058,
+      "step": 826
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8398436929888085e-06,
+      "loss": 0.2076,
+      "step": 827
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.819048464677261e-06,
+      "loss": 0.208,
+      "step": 828
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.798317154586352e-06,
+      "loss": 0.1956,
+      "step": 829
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7776499472471184e-06,
+      "loss": 0.2107,
+      "step": 830
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7570470266200177e-06,
+      "loss": 0.2141,
+      "step": 831
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.736508576093285e-06,
+      "loss": 0.2141,
+      "step": 832
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.716034778481301e-06,
+      "loss": 0.2197,
+      "step": 833
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.69562581602297e-06,
+      "loss": 0.2077,
+      "step": 834
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.675281870380082e-06,
+      "loss": 0.2023,
+      "step": 835
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.65500312263572e-06,
+      "loss": 0.203,
+      "step": 836
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6347897532926293e-06,
+      "loss": 0.2251,
+      "step": 837
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6146419422716174e-06,
+      "loss": 0.2172,
+      "step": 838
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.594559868909956e-06,
+      "loss": 0.2092,
+      "step": 839
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5745437119597704e-06,
+      "loss": 0.2122,
+      "step": 840
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.554593649586469e-06,
+      "loss": 0.205,
+      "step": 841
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5347098593671417e-06,
+      "loss": 0.1979,
+      "step": 842
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.2135,
+      "step": 843
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.49514180274774e-06,
+      "loss": 0.2108,
+      "step": 844
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4754578885460813e-06,
+      "loss": 0.2084,
+      "step": 845
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.455840950892099e-06,
+      "loss": 0.2069,
+      "step": 846
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.436291164397715e-06,
+      "loss": 0.2034,
+      "step": 847
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.416808703077135e-06,
+      "loss": 0.2168,
+      "step": 848
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3973937403452983e-06,
+      "loss": 0.2091,
+      "step": 849
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3780464490163267e-06,
+      "loss": 0.2042,
+      "step": 850
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3587670013020026e-06,
+      "loss": 0.2047,
+      "step": 851
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.2086,
+      "step": 852
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3204123225434714e-06,
+      "loss": 0.2015,
+      "step": 853
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3013374328973113e-06,
+      "loss": 0.2081,
+      "step": 854
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.28233106965885e-06,
+      "loss": 0.2055,
+      "step": 855
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2633934020052383e-06,
+      "loss": 0.2047,
+      "step": 856
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2445245985021613e-06,
+      "loss": 0.2157,
+      "step": 857
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2257248271023424e-06,
+      "loss": 0.2085,
+      "step": 858
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.206994255144036e-06,
+      "loss": 0.2021,
+      "step": 859
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.188333049349556e-06,
+      "loss": 0.2264,
+      "step": 860
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1697413758237785e-06,
+      "loss": 0.218,
+      "step": 861
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1512194000526676e-06,
+      "loss": 0.2069,
+      "step": 862
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1327672869018036e-06,
+      "loss": 0.1998,
+      "step": 863
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.114385200614912e-06,
+      "loss": 0.2074,
+      "step": 864
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0960733048124082e-06,
+      "loss": 0.2001,
+      "step": 865
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.077831762489927e-06,
+      "loss": 0.2143,
+      "step": 866
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0596607360168897e-06,
+      "loss": 0.2186,
+      "step": 867
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0415603871350476e-06,
+      "loss": 0.2068,
+      "step": 868
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.023530876957045e-06,
+      "loss": 0.211,
+      "step": 869
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0055723659649907e-06,
+      "loss": 0.204,
+      "step": 870
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.987685014009011e-06,
+      "loss": 0.2119,
+      "step": 871
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9698689803058523e-06,
+      "loss": 0.1991,
+      "step": 872
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.952124423437447e-06,
+      "loss": 0.2133,
+      "step": 873
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.934451501349507e-06,
+      "loss": 0.215,
+      "step": 874
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9168503713501184e-06,
+      "loss": 0.2058,
+      "step": 875
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8993211901083353e-06,
+      "loss": 0.203,
+      "step": 876
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8818641136527959e-06,
+      "loss": 0.2132,
+      "step": 877
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8644792973703252e-06,
+      "loss": 0.2112,
+      "step": 878
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8471668960045575e-06,
+      "loss": 0.2036,
+      "step": 879
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8299270636545518e-06,
+      "loss": 0.2042,
+      "step": 880
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8127599537734297e-06,
+      "loss": 0.1876,
+      "step": 881
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7956657191669969e-06,
+      "loss": 0.2145,
+      "step": 882
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7786445119923967e-06,
+      "loss": 0.203,
+      "step": 883
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7616964837567497e-06,
+      "loss": 0.2054,
+      "step": 884
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7448217853158e-06,
+      "loss": 0.2125,
+      "step": 885
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7280205668725814e-06,
+      "loss": 0.2167,
+      "step": 886
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7112929779760768e-06,
+      "loss": 0.2062,
+      "step": 887
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6946391675198838e-06,
+      "loss": 0.2024,
+      "step": 888
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6780592837408926e-06,
+      "loss": 0.2113,
+      "step": 889
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6615534742179684e-06,
+      "loss": 0.206,
+      "step": 890
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 0.2166,
+      "step": 891
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6287646649577672e-06,
+      "loss": 0.2073,
+      "step": 892
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6124819570762862e-06,
+      "loss": 0.2009,
+      "step": 893
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5962739071598709e-06,
+      "loss": 0.2003,
+      "step": 894
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5801406594776625e-06,
+      "loss": 0.2164,
+      "step": 895
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5640823576329844e-06,
+      "loss": 0.218,
+      "step": 896
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5480991445620541e-06,
+      "loss": 0.2209,
+      "step": 897
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5321911625327224e-06,
+      "loss": 0.1961,
+      "step": 898
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5163585531432046e-06,
+      "loss": 0.2148,
+      "step": 899
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.500601457320814e-06,
+      "loss": 0.2102,
+      "step": 900
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4849200153207176e-06,
+      "loss": 0.2132,
+      "step": 901
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4693143667246713e-06,
+      "loss": 0.2135,
+      "step": 902
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.453784650439798e-06,
+      "loss": 0.2001,
+      "step": 903
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4383310046973365e-06,
+      "loss": 0.2126,
+      "step": 904
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4229535670514162e-06,
+      "loss": 0.1928,
+      "step": 905
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.407652474377832e-06,
+      "loss": 0.2168,
+      "step": 906
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3924278628728305e-06,
+      "loss": 0.2055,
+      "step": 907
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3772798680518828e-06,
+      "loss": 0.2018,
+      "step": 908
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3622086247484989e-06,
+      "loss": 0.2123,
+      "step": 909
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3472142671130139e-06,
+      "loss": 0.1973,
+      "step": 910
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3322969286113973e-06,
+      "loss": 0.2237,
+      "step": 911
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3174567420240647e-06,
+      "loss": 0.2202,
+      "step": 912
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3026938394446976e-06,
+      "loss": 0.2175,
+      "step": 913
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2880083522790654e-06,
+      "loss": 0.2101,
+      "step": 914
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.273400411243857e-06,
+      "loss": 0.2067,
+      "step": 915
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2588701463655172e-06,
+      "loss": 0.2163,
+      "step": 916
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2444176869790925e-06,
+      "loss": 0.2074,
+      "step": 917
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2300431617270669e-06,
+      "loss": 0.2046,
+      "step": 918
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2157466985582367e-06,
+      "loss": 0.2107,
+      "step": 919
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2015284247265567e-06,
+      "loss": 0.2218,
+      "step": 920
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1873884667900125e-06,
+      "loss": 0.205,
+      "step": 921
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1733269506094957e-06,
+      "loss": 0.21,
+      "step": 922
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1593440013476775e-06,
+      "loss": 0.2147,
+      "step": 923
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.2157,
+      "step": 924
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1316143007330739e-06,
+      "loss": 0.2109,
+      "step": 925
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1178677962045604e-06,
+      "loss": 0.2158,
+      "step": 926
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1042003522410882e-06,
+      "loss": 0.2149,
+      "step": 927
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.090612090497668e-06,
+      "loss": 0.2054,
+      "step": 928
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.077103131924493e-06,
+      "loss": 0.2122,
+      "step": 929
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 0.213,
+      "step": 930
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0503236045591857e-06,
+      "loss": 0.2051,
+      "step": 931
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.037053274133758e-06,
+      "loss": 0.198,
+      "step": 932
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0238627236098619e-06,
+      "loss": 0.2066,
+      "step": 933
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0107520703976325e-06,
+      "loss": 0.212,
+      "step": 934
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.977214311960404e-07,
+      "loss": 0.2104,
+      "step": 935
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.8477092199184e-07,
+      "loss": 0.1995,
+      "step": 936
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.719006580585444e-07,
+      "loss": 0.2121,
+      "step": 937
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.591107539553945e-07,
+      "loss": 0.2092,
+      "step": 938
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.464013235263458e-07,
+      "loss": 0.2011,
+      "step": 939
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.337724798990489e-07,
+      "loss": 0.1923,
+      "step": 940
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.212243354838435e-07,
+      "loss": 0.2009,
+      "step": 941
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.08757001972762e-07,
+      "loss": 0.2085,
+      "step": 942
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.963705903385344e-07,
+      "loss": 0.2063,
+      "step": 943
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.8406521083359e-07,
+      "loss": 0.2023,
+      "step": 944
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.71840972989092e-07,
+      "loss": 0.201,
+      "step": 945
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.596979856139553e-07,
+      "loss": 0.2025,
+      "step": 946
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.476363567938751e-07,
+      "loss": 0.2015,
+      "step": 947
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.356561938903707e-07,
+      "loss": 0.2111,
+      "step": 948
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.237576035398198e-07,
+      "loss": 0.1969,
+      "step": 949
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.119406916525252e-07,
+      "loss": 0.224,
+      "step": 950
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.002055634117578e-07,
+      "loss": 0.211,
+      "step": 951
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.885523232728287e-07,
+      "loss": 0.2029,
+      "step": 952
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.769810749621532e-07,
+      "loss": 0.1962,
+      "step": 953
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.654919214763357e-07,
+      "loss": 0.2137,
+      "step": 954
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.540849650812409e-07,
+      "loss": 0.2189,
+      "step": 955
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.427603073110967e-07,
+      "loss": 0.199,
+      "step": 956
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.315180489675822e-07,
+      "loss": 0.2082,
+      "step": 957
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.203582901189332e-07,
+      "loss": 0.1972,
+      "step": 958
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.092811300990521e-07,
+      "loss": 0.2113,
+      "step": 959
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.98286667506618e-07,
+      "loss": 0.2164,
+      "step": 960
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.87375000204219e-07,
+      "loss": 0.2203,
+      "step": 961
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.765462253174715e-07,
+      "loss": 0.2231,
+      "step": 962
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.658004392341633e-07,
+      "loss": 0.1949,
+      "step": 963
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.551377376033896e-07,
+      "loss": 0.1995,
+      "step": 964
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.445582153347074e-07,
+      "loss": 0.2068,
+      "step": 965
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.340619665972847e-07,
+      "loss": 0.2132,
+      "step": 966
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.236490848190657e-07,
+      "loss": 0.2218,
+      "step": 967
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.133196626859406e-07,
+      "loss": 0.1993,
+      "step": 968
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.2027,
+      "step": 969
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.929115643833005e-07,
+      "loss": 0.2108,
+      "step": 970
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.828330698678908e-07,
+      "loss": 0.2094,
+      "step": 971
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.728383983041696e-07,
+      "loss": 0.2108,
+      "step": 972
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.629276386555016e-07,
+      "loss": 0.2078,
+      "step": 973
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.531008791383485e-07,
+      "loss": 0.2077,
+      "step": 974
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.43358207221476e-07,
+      "loss": 0.1993,
+      "step": 975
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.336997096251816e-07,
+      "loss": 0.1934,
+      "step": 976
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.241254723205225e-07,
+      "loss": 0.2026,
+      "step": 977
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.146355805285452e-07,
+      "loss": 0.2125,
+      "step": 978
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.052301187195296e-07,
+      "loss": 0.2046,
+      "step": 979
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.959091706122431e-07,
+      "loss": 0.2029,
+      "step": 980
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.866728191731829e-07,
+      "loss": 0.1976,
+      "step": 981
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.775211466158469e-07,
+      "loss": 0.2035,
+      "step": 982
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.6845423440000315e-07,
+      "loss": 0.1887,
+      "step": 983
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.594721632309551e-07,
+      "loss": 0.2134,
+      "step": 984
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.505750130588371e-07,
+      "loss": 0.2085,
+      "step": 985
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.4176286307788475e-07,
+      "loss": 0.2059,
+      "step": 986
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3303579172574884e-07,
+      "loss": 0.208,
+      "step": 987
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.243938766827849e-07,
+      "loss": 0.2023,
+      "step": 988
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1583719487136575e-07,
+      "loss": 0.1943,
+      "step": 989
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.0736582245519795e-07,
+      "loss": 0.199,
+      "step": 990
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9897983483863866e-07,
+      "loss": 0.2023,
+      "step": 991
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9067930666603304e-07,
+      "loss": 0.2193,
+      "step": 992
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.824643118210403e-07,
+      "loss": 0.2184,
+      "step": 993
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.743349234259841e-07,
+      "loss": 0.1987,
+      "step": 994
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.662912138411967e-07,
+      "loss": 0.2166,
+      "step": 995
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5833325466437697e-07,
+      "loss": 0.2128,
+      "step": 996
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5046111672995097e-07,
+      "loss": 0.2143,
+      "step": 997
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.426748701084448e-07,
+      "loss": 0.2111,
+      "step": 998
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.349745841058605e-07,
+      "loss": 0.2187,
+      "step": 999
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2736032726305546e-07,
+      "loss": 0.2042,
+      "step": 1000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.198321673551341e-07,
+      "loss": 0.2043,
+      "step": 1001
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1239017139084725e-07,
+      "loss": 0.196,
+      "step": 1002
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.050344056119925e-07,
+      "loss": 0.2067,
+      "step": 1003
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.977649354928258e-07,
+      "loss": 0.2104,
+      "step": 1004
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.207,
+      "step": 1005
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.834851402893857e-07,
+      "loss": 0.1951,
+      "step": 1006
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.764749423107027e-07,
+      "loss": 0.1985,
+      "step": 1007
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 0.2044,
+      "step": 1008
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.627142575905062e-07,
+      "loss": 0.2083,
+      "step": 1009
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.559638933339414e-07,
+      "loss": 0.2098,
+      "step": 1010
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.493002615175977e-07,
+      "loss": 0.2041,
+      "step": 1011
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.4272342145499006e-07,
+      "loss": 0.2159,
+      "step": 1012
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3623343168709624e-07,
+      "loss": 0.1974,
+      "step": 1013
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2983034998182997e-07,
+      "loss": 0.2098,
+      "step": 1014
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.235142333335316e-07,
+      "loss": 0.193,
+      "step": 1015
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1728513796245855e-07,
+      "loss": 0.2025,
+      "step": 1016
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.11143119314281e-07,
+      "loss": 0.2063,
+      "step": 1017
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0508823205959815e-07,
+      "loss": 0.2191,
+      "step": 1018
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.991205300934429e-07,
+      "loss": 0.2038,
+      "step": 1019
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9324006653480332e-07,
+      "loss": 0.2147,
+      "step": 1020
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.874468937261531e-07,
+      "loss": 0.2003,
+      "step": 1021
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8174106323298634e-07,
+      "loss": 0.2085,
+      "step": 1022
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.761226258433524e-07,
+      "loss": 0.2071,
+      "step": 1023
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7059163156740943e-07,
+      "loss": 0.2045,
+      "step": 1024
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6514812963697723e-07,
+      "loss": 0.1973,
+      "step": 1025
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5979216850509848e-07,
+      "loss": 0.2033,
+      "step": 1026
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.545237958456125e-07,
+      "loss": 0.2017,
+      "step": 1027
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4934305855271892e-07,
+      "loss": 0.2188,
+      "step": 1028
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4425000274057577e-07,
+      "loss": 0.2063,
+      "step": 1029
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3924467374287432e-07,
+      "loss": 0.2151,
+      "step": 1030
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.343271161124493e-07,
+      "loss": 0.2166,
+      "step": 1031
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2949737362087156e-07,
+      "loss": 0.2095,
+      "step": 1032
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.247554892580616e-07,
+      "loss": 0.2165,
+      "step": 1033
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.201015052319099e-07,
+      "loss": 0.2222,
+      "step": 1034
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1553546296789952e-07,
+      "loss": 0.22,
+      "step": 1035
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1105740310873414e-07,
+      "loss": 0.2047,
+      "step": 1036
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.066673655139816e-07,
+      "loss": 0.2112,
+      "step": 1037
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0236538925971429e-07,
+      "loss": 0.2146,
+      "step": 1038
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.815151263816714e-08,
+      "loss": 0.214,
+      "step": 1039
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.402577315738904e-08,
+      "loss": 0.2114,
+      "step": 1040
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.99882075409153e-08,
+      "loss": 0.2129,
+      "step": 1041
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.603885172744131e-08,
+      "loss": 0.1971,
+      "step": 1042
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.217774087049268e-08,
+      "loss": 0.2054,
+      "step": 1043
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.840490933812783e-08,
+      "loss": 0.197,
+      "step": 1044
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.472039071261927e-08,
+      "loss": 0.2009,
+      "step": 1045
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.112421779015944e-08,
+      "loss": 0.22,
+      "step": 1046
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 0.2093,
+      "step": 1047
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.419703630701546e-08,
+      "loss": 0.2009,
+      "step": 1048
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.086608940572447e-08,
+      "loss": 0.2121,
+      "step": 1049
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.7623611525721155e-08,
+      "loss": 0.2113,
+      "step": 1050
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.446963152855644e-08,
+      "loss": 0.2117,
+      "step": 1051
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.140417748806026e-08,
+      "loss": 0.1995,
+      "step": 1052
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.8427276690081735e-08,
+      "loss": 0.1929,
+      "step": 1053
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.553895563225053e-08,
+      "loss": 0.2024,
+      "step": 1054
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2739240023742526e-08,
+      "loss": 0.1983,
+      "step": 1055
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.002815478505007e-08,
+      "loss": 0.2105,
+      "step": 1056
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.7405724047756554e-08,
+      "loss": 0.1979,
+      "step": 1057
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.487197115432883e-08,
+      "loss": 0.2052,
+      "step": 1058
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.242691865790071e-08,
+      "loss": 0.2053,
+      "step": 1059
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.0070588322079765e-08,
+      "loss": 0.2026,
+      "step": 1060
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.780300112074974e-08,
+      "loss": 0.2049,
+      "step": 1061
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.5624177237884017e-08,
+      "loss": 0.2184,
+      "step": 1062
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.3534136067369094e-08,
+      "loss": 0.2051,
+      "step": 1063
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1532896212825837e-08,
+      "loss": 0.2046,
+      "step": 1064
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.962047548744961e-08,
+      "loss": 0.2134,
+      "step": 1065
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.7796890913850395e-08,
+      "loss": 0.1919,
+      "step": 1066
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.606215872389738e-08,
+      "loss": 0.2097,
+      "step": 1067
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4416294358582383e-08,
+      "loss": 0.2103,
+      "step": 1068
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2859312467872197e-08,
+      "loss": 0.2087,
+      "step": 1069
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.1391226910588693e-08,
+      "loss": 0.2145,
+      "step": 1070
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0012050754277802e-08,
+      "loss": 0.209,
+      "step": 1071
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.721796275095173e-09,
+      "loss": 0.2114,
+      "step": 1072
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.520474957699586e-09,
+      "loss": 0.2224,
+      "step": 1073
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.40809749514637e-09,
+      "loss": 0.2072,
+      "step": 1074
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.384673788797479e-09,
+      "loss": 0.2242,
+      "step": 1075
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.450212948227117e-09,
+      "loss": 0.2183,
+      "step": 1076
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.6047232911462506e-09,
+      "loss": 0.2179,
+      "step": 1077
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.8482123433248853e-09,
+      "loss": 0.2012,
+      "step": 1078
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.180686838527679e-09,
+      "loss": 0.1969,
+      "step": 1079
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6021527184528761e-09,
+      "loss": 0.1997,
+      "step": 1080
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.1126151326779077e-09,
+      "loss": 0.2175,
+      "step": 1081
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.120784386160928e-10,
+      "loss": 0.2035,
+      "step": 1082
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.005462014766703e-10,
+      "loss": 0.2045,
+      "step": 1083
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.7802119423149244e-10,
+      "loss": 0.2136,
+      "step": 1084
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.450539759393024e-11,
+      "loss": 0.2111,
+      "step": 1085
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.2032,
+      "step": 1086
+    },
+    {
+      "epoch": 3.0,
+      "step": 1086,
+      "total_flos": 2.877522978659107e+18,
+      "train_loss": 0.38975011184125535,
+      "train_runtime": 36926.9258,
+      "train_samples_per_second": 3.764,
+      "train_steps_per_second": 0.029
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1086,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "total_flos": 2.877522978659107e+18,
+  "trial_name": null,
+  "trial_params": null
+}