diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12324 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9997941210110945,
+  "eval_steps": 100.0,
+  "global_step": 2049,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.7507,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.7482,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 0.6573,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5637,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.56,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 0.5173,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.5959,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4862,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 0.5063,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5542,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.5165,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4e-05,
+      "loss": 0.6018,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.6962,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.7856,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 0.7155,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.997541789577188e-05,
+      "loss": 0.8667,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9950835791543757e-05,
+      "loss": 0.9359,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.992625368731564e-05,
+      "loss": 0.9198,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.990167158308752e-05,
+      "loss": 0.9953,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.987708947885939e-05,
+      "loss": 0.8008,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.985250737463127e-05,
+      "loss": 0.5827,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9827925270403145e-05,
+      "loss": 0.5492,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9803343166175026e-05,
+      "loss": 0.5417,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9778761061946906e-05,
+      "loss": 0.6181,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9754178957718786e-05,
+      "loss": 0.4864,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.972959685349066e-05,
+      "loss": 0.4988,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.970501474926254e-05,
+      "loss": 0.4282,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.968043264503442e-05,
+      "loss": 0.4931,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9655850540806295e-05,
+      "loss": 0.4636,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.963126843657817e-05,
+      "loss": 0.5745,
+      "step": 30
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.960668633235005e-05,
+      "loss": 0.5198,
+      "step": 31
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.958210422812193e-05,
+      "loss": 0.4511,
+      "step": 32
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.955752212389381e-05,
+      "loss": 0.4555,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.953294001966568e-05,
+      "loss": 0.5005,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9508357915437564e-05,
+      "loss": 0.4268,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9483775811209444e-05,
+      "loss": 0.4219,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9459193706981325e-05,
+      "loss": 0.4504,
+      "step": 37
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.94346116027532e-05,
+      "loss": 0.4064,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.941002949852507e-05,
+      "loss": 0.4123,
+      "step": 39
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.938544739429695e-05,
+      "loss": 0.4464,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.936086529006883e-05,
+      "loss": 0.4168,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9336283185840707e-05,
+      "loss": 0.4087,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.931170108161259e-05,
+      "loss": 0.3903,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.928711897738447e-05,
+      "loss": 0.4236,
+      "step": 44
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.926253687315635e-05,
+      "loss": 0.4084,
+      "step": 45
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.923795476892822e-05,
+      "loss": 0.377,
+      "step": 46
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.92133726647001e-05,
+      "loss": 0.4168,
+      "step": 47
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.918879056047198e-05,
+      "loss": 0.4255,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9164208456243856e-05,
+      "loss": 0.4473,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.913962635201573e-05,
+      "loss": 0.4385,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.911504424778761e-05,
+      "loss": 0.422,
+      "step": 51
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.909046214355949e-05,
+      "loss": 0.4401,
+      "step": 52
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.906588003933137e-05,
+      "loss": 0.412,
+      "step": 53
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9041297935103245e-05,
+      "loss": 0.4536,
+      "step": 54
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9016715830875125e-05,
+      "loss": 0.4466,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8992133726647006e-05,
+      "loss": 0.4278,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8967551622418886e-05,
+      "loss": 0.4027,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.894296951819076e-05,
+      "loss": 0.4224,
+      "step": 58
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.891838741396263e-05,
+      "loss": 0.4238,
+      "step": 59
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8893805309734514e-05,
+      "loss": 0.4074,
+      "step": 60
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8869223205506394e-05,
+      "loss": 0.4007,
+      "step": 61
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.884464110127827e-05,
+      "loss": 0.4057,
+      "step": 62
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.882005899705015e-05,
+      "loss": 0.4106,
+      "step": 63
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.879547689282203e-05,
+      "loss": 0.5015,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.877089478859391e-05,
+      "loss": 0.4244,
+      "step": 65
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.874631268436578e-05,
+      "loss": 0.408,
+      "step": 66
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.872173058013766e-05,
+      "loss": 0.4063,
+      "step": 67
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.869714847590954e-05,
+      "loss": 0.426,
+      "step": 68
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.867256637168142e-05,
+      "loss": 0.4208,
+      "step": 69
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.86479842674533e-05,
+      "loss": 0.4094,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.862340216322517e-05,
+      "loss": 0.4329,
+      "step": 71
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.859882005899705e-05,
+      "loss": 0.5189,
+      "step": 72
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.857423795476893e-05,
+      "loss": 0.4785,
+      "step": 73
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8549655850540806e-05,
+      "loss": 0.4019,
+      "step": 74
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8525073746312687e-05,
+      "loss": 0.4191,
+      "step": 75
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.850049164208457e-05,
+      "loss": 0.3913,
+      "step": 76
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.847590953785645e-05,
+      "loss": 0.3914,
+      "step": 77
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.845132743362832e-05,
+      "loss": 0.3801,
+      "step": 78
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8426745329400195e-05,
+      "loss": 0.4403,
+      "step": 79
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8402163225172075e-05,
+      "loss": 0.3919,
+      "step": 80
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8377581120943956e-05,
+      "loss": 0.413,
+      "step": 81
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8352999016715836e-05,
+      "loss": 0.4607,
+      "step": 82
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.832841691248771e-05,
+      "loss": 0.5239,
+      "step": 83
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.830383480825959e-05,
+      "loss": 0.7022,
+      "step": 84
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.827925270403147e-05,
+      "loss": 0.8226,
+      "step": 85
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.825467059980335e-05,
+      "loss": 0.8205,
+      "step": 86
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.823008849557522e-05,
+      "loss": 0.516,
+      "step": 87
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.82055063913471e-05,
+      "loss": 0.4735,
+      "step": 88
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.818092428711898e-05,
+      "loss": 0.4843,
+      "step": 89
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.815634218289086e-05,
+      "loss": 0.463,
+      "step": 90
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.813176007866273e-05,
+      "loss": 0.4591,
+      "step": 91
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.810717797443461e-05,
+      "loss": 0.541,
+      "step": 92
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8082595870206494e-05,
+      "loss": 0.4078,
+      "step": 93
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8058013765978374e-05,
+      "loss": 0.4228,
+      "step": 94
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.803343166175025e-05,
+      "loss": 0.4978,
+      "step": 95
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.800884955752213e-05,
+      "loss": 0.3989,
+      "step": 96
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7984267453294e-05,
+      "loss": 0.3833,
+      "step": 97
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.795968534906588e-05,
+      "loss": 0.4756,
+      "step": 98
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7935103244837756e-05,
+      "loss": 0.481,
+      "step": 99
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7910521140609637e-05,
+      "loss": 0.4679,
+      "step": 100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.788593903638152e-05,
+      "loss": 0.4033,
+      "step": 101
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.78613569321534e-05,
+      "loss": 0.4065,
+      "step": 102
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.783677482792527e-05,
+      "loss": 0.4853,
+      "step": 103
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.781219272369715e-05,
+      "loss": 0.406,
+      "step": 104
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.778761061946903e-05,
+      "loss": 0.4467,
+      "step": 105
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.776302851524091e-05,
+      "loss": 0.3774,
+      "step": 106
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7738446411012786e-05,
+      "loss": 0.3875,
+      "step": 107
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.771386430678466e-05,
+      "loss": 0.3843,
+      "step": 108
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.768928220255654e-05,
+      "loss": 0.3893,
+      "step": 109
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.766470009832842e-05,
+      "loss": 0.3808,
+      "step": 110
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7640117994100294e-05,
+      "loss": 0.3954,
+      "step": 111
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7615535889872175e-05,
+      "loss": 0.4638,
+      "step": 112
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7590953785644055e-05,
+      "loss": 0.3988,
+      "step": 113
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7566371681415936e-05,
+      "loss": 0.3952,
+      "step": 114
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.754178957718781e-05,
+      "loss": 0.4056,
+      "step": 115
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.751720747295968e-05,
+      "loss": 0.4063,
+      "step": 116
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.749262536873156e-05,
+      "loss": 0.3902,
+      "step": 117
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7468043264503444e-05,
+      "loss": 0.3822,
+      "step": 118
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7443461160275324e-05,
+      "loss": 0.4237,
+      "step": 119
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.74188790560472e-05,
+      "loss": 0.4644,
+      "step": 120
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.739429695181908e-05,
+      "loss": 0.3821,
+      "step": 121
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.736971484759096e-05,
+      "loss": 0.3835,
+      "step": 122
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.734513274336283e-05,
+      "loss": 0.3811,
+      "step": 123
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.732055063913471e-05,
+      "loss": 0.3866,
+      "step": 124
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.729596853490659e-05,
+      "loss": 0.463,
+      "step": 125
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.727138643067847e-05,
+      "loss": 0.4107,
+      "step": 126
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.724680432645035e-05,
+      "loss": 0.397,
+      "step": 127
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 0.3702,
+      "step": 128
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.71976401179941e-05,
+      "loss": 0.3876,
+      "step": 129
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.717305801376598e-05,
+      "loss": 0.3777,
+      "step": 130
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.714847590953786e-05,
+      "loss": 0.3933,
+      "step": 131
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7123893805309736e-05,
+      "loss": 0.4486,
+      "step": 132
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7099311701081617e-05,
+      "loss": 0.3806,
+      "step": 133
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.70747295968535e-05,
+      "loss": 0.4657,
+      "step": 134
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.705014749262537e-05,
+      "loss": 0.3876,
+      "step": 135
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.7025565388397244e-05,
+      "loss": 0.4651,
+      "step": 136
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.7000983284169125e-05,
+      "loss": 0.446,
+      "step": 137
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6976401179941005e-05,
+      "loss": 0.414,
+      "step": 138
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6951819075712886e-05,
+      "loss": 0.4414,
+      "step": 139
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.692723697148476e-05,
+      "loss": 0.3983,
+      "step": 140
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.690265486725664e-05,
+      "loss": 0.4141,
+      "step": 141
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.687807276302852e-05,
+      "loss": 0.3641,
+      "step": 142
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.68534906588004e-05,
+      "loss": 0.3995,
+      "step": 143
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6828908554572274e-05,
+      "loss": 0.3796,
+      "step": 144
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.680432645034415e-05,
+      "loss": 0.3743,
+      "step": 145
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.677974434611603e-05,
+      "loss": 0.3856,
+      "step": 146
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.675516224188791e-05,
+      "loss": 0.3929,
+      "step": 147
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.673058013765978e-05,
+      "loss": 0.4378,
+      "step": 148
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.670599803343166e-05,
+      "loss": 0.3868,
+      "step": 149
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.668141592920354e-05,
+      "loss": 0.3823,
+      "step": 150
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6656833824975424e-05,
+      "loss": 0.3861,
+      "step": 151
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.66322517207473e-05,
+      "loss": 0.3942,
+      "step": 152
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.660766961651918e-05,
+      "loss": 0.4449,
+      "step": 153
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.658308751229106e-05,
+      "loss": 0.3586,
+      "step": 154
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.655850540806293e-05,
+      "loss": 0.3815,
+      "step": 155
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6533923303834806e-05,
+      "loss": 0.3869,
+      "step": 156
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6509341199606686e-05,
+      "loss": 0.3691,
+      "step": 157
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6484759095378567e-05,
+      "loss": 0.3737,
+      "step": 158
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.646017699115045e-05,
+      "loss": 0.3723,
+      "step": 159
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.643559488692232e-05,
+      "loss": 0.3788,
+      "step": 160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.64110127826942e-05,
+      "loss": 0.4056,
+      "step": 161
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.638643067846608e-05,
+      "loss": 0.3851,
+      "step": 162
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.636184857423796e-05,
+      "loss": 0.375,
+      "step": 163
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6337266470009836e-05,
+      "loss": 0.4488,
+      "step": 164
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.631268436578171e-05,
+      "loss": 0.3593,
+      "step": 165
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.628810226155359e-05,
+      "loss": 0.5176,
+      "step": 166
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.626352015732547e-05,
+      "loss": 0.3966,
+      "step": 167
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6238938053097344e-05,
+      "loss": 0.4261,
+      "step": 168
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6214355948869224e-05,
+      "loss": 0.547,
+      "step": 169
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6189773844641105e-05,
+      "loss": 0.3607,
+      "step": 170
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6165191740412985e-05,
+      "loss": 0.3747,
+      "step": 171
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.614060963618486e-05,
+      "loss": 0.3777,
+      "step": 172
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.611602753195674e-05,
+      "loss": 0.4634,
+      "step": 173
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.609144542772861e-05,
+      "loss": 0.3612,
+      "step": 174
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.606686332350049e-05,
+      "loss": 0.3892,
+      "step": 175
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6042281219272374e-05,
+      "loss": 0.3882,
+      "step": 176
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.601769911504425e-05,
+      "loss": 0.372,
+      "step": 177
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.599311701081613e-05,
+      "loss": 0.3993,
+      "step": 178
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.596853490658801e-05,
+      "loss": 0.3963,
+      "step": 179
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.594395280235988e-05,
+      "loss": 0.4074,
+      "step": 180
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.591937069813176e-05,
+      "loss": 0.4225,
+      "step": 181
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.589478859390364e-05,
+      "loss": 0.4448,
+      "step": 182
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.5870206489675517e-05,
+      "loss": 0.3977,
+      "step": 183
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.58456243854474e-05,
+      "loss": 0.362,
+      "step": 184
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.582104228121927e-05,
+      "loss": 0.3984,
+      "step": 185
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.579646017699115e-05,
+      "loss": 0.3603,
+      "step": 186
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.577187807276303e-05,
+      "loss": 0.4703,
+      "step": 187
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.574729596853491e-05,
+      "loss": 0.3899,
+      "step": 188
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.5722713864306786e-05,
+      "loss": 0.3673,
+      "step": 189
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.5698131760078666e-05,
+      "loss": 0.3674,
+      "step": 190
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.5673549655850547e-05,
+      "loss": 0.3967,
+      "step": 191
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.564896755162242e-05,
+      "loss": 0.402,
+      "step": 192
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.5624385447394294e-05,
+      "loss": 0.3702,
+      "step": 193
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.5599803343166174e-05,
+      "loss": 0.3526,
+      "step": 194
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.5575221238938055e-05,
+      "loss": 0.3803,
+      "step": 195
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.5550639134709935e-05,
+      "loss": 0.3861,
+      "step": 196
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.552605703048181e-05,
+      "loss": 0.3941,
+      "step": 197
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.550147492625369e-05,
+      "loss": 0.4065,
+      "step": 198
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.547689282202557e-05,
+      "loss": 0.4138,
+      "step": 199
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.545231071779745e-05,
+      "loss": 0.3975,
+      "step": 200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.5427728613569324e-05,
+      "loss": 0.3879,
+      "step": 201
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5403146509341204e-05,
+      "loss": 0.4044,
+      "step": 202
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.537856440511308e-05,
+      "loss": 0.37,
+      "step": 203
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.535398230088496e-05,
+      "loss": 0.3684,
+      "step": 204
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.532940019665683e-05,
+      "loss": 0.4666,
+      "step": 205
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.530481809242871e-05,
+      "loss": 0.3814,
+      "step": 206
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.528023598820059e-05,
+      "loss": 0.3704,
+      "step": 207
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.525565388397247e-05,
+      "loss": 0.3783,
+      "step": 208
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.523107177974435e-05,
+      "loss": 0.411,
+      "step": 209
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.520648967551623e-05,
+      "loss": 0.4171,
+      "step": 210
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.518190757128811e-05,
+      "loss": 0.4678,
+      "step": 211
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.515732546705998e-05,
+      "loss": 0.4048,
+      "step": 212
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5132743362831855e-05,
+      "loss": 0.4288,
+      "step": 213
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5108161258603736e-05,
+      "loss": 0.381,
+      "step": 214
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5083579154375616e-05,
+      "loss": 0.4327,
+      "step": 215
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.5058997050147497e-05,
+      "loss": 0.371,
+      "step": 216
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.503441494591937e-05,
+      "loss": 0.4862,
+      "step": 217
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.500983284169125e-05,
+      "loss": 0.3967,
+      "step": 218
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.498525073746313e-05,
+      "loss": 0.3552,
+      "step": 219
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.496066863323501e-05,
+      "loss": 0.3657,
+      "step": 220
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.4936086529006885e-05,
+      "loss": 0.3642,
+      "step": 221
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.491150442477876e-05,
+      "loss": 0.3761,
+      "step": 222
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.488692232055064e-05,
+      "loss": 0.3635,
+      "step": 223
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.486234021632252e-05,
+      "loss": 0.4703,
+      "step": 224
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.48377581120944e-05,
+      "loss": 0.7381,
+      "step": 225
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.4813176007866274e-05,
+      "loss": 0.573,
+      "step": 226
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.4788593903638154e-05,
+      "loss": 0.3947,
+      "step": 227
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.4764011799410035e-05,
+      "loss": 0.4302,
+      "step": 228
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.473942969518191e-05,
+      "loss": 0.562,
+      "step": 229
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.471484759095379e-05,
+      "loss": 0.4153,
+      "step": 230
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.469026548672566e-05,
+      "loss": 0.5394,
+      "step": 231
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.466568338249754e-05,
+      "loss": 0.4339,
+      "step": 232
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.464110127826942e-05,
+      "loss": 0.4647,
+      "step": 233
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.46165191740413e-05,
+      "loss": 0.3934,
+      "step": 234
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.459193706981318e-05,
+      "loss": 0.4557,
+      "step": 235
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.456735496558506e-05,
+      "loss": 0.453,
+      "step": 236
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.454277286135694e-05,
+      "loss": 0.403,
+      "step": 237
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.451819075712881e-05,
+      "loss": 0.4653,
+      "step": 238
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.449360865290069e-05,
+      "loss": 0.3743,
+      "step": 239
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.446902654867257e-05,
+      "loss": 0.3652,
+      "step": 240
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.3609,
+      "step": 241
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.441986234021632e-05,
+      "loss": 0.3824,
+      "step": 242
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.43952802359882e-05,
+      "loss": 0.3617,
+      "step": 243
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.437069813176008e-05,
+      "loss": 0.3618,
+      "step": 244
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.434611602753196e-05,
+      "loss": 0.4493,
+      "step": 245
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4321533923303835e-05,
+      "loss": 0.3625,
+      "step": 246
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4296951819075716e-05,
+      "loss": 0.3644,
+      "step": 247
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4272369714847596e-05,
+      "loss": 0.356,
+      "step": 248
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4247787610619477e-05,
+      "loss": 0.4553,
+      "step": 249
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.422320550639135e-05,
+      "loss": 0.4031,
+      "step": 250
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.4198623402163224e-05,
+      "loss": 0.362,
+      "step": 251
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.4174041297935104e-05,
+      "loss": 0.406,
+      "step": 252
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.4149459193706985e-05,
+      "loss": 0.3762,
+      "step": 253
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.412487708947886e-05,
+      "loss": 0.3772,
+      "step": 254
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.410029498525074e-05,
+      "loss": 0.3525,
+      "step": 255
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.407571288102262e-05,
+      "loss": 0.3686,
+      "step": 256
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.40511307767945e-05,
+      "loss": 0.3649,
+      "step": 257
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4026548672566373e-05,
+      "loss": 0.3731,
+      "step": 258
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4001966568338254e-05,
+      "loss": 0.3788,
+      "step": 259
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.397738446411013e-05,
+      "loss": 0.3765,
+      "step": 260
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.395280235988201e-05,
+      "loss": 0.4245,
+      "step": 261
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.392822025565388e-05,
+      "loss": 0.366,
+      "step": 262
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.390363815142576e-05,
+      "loss": 0.3547,
+      "step": 263
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.387905604719764e-05,
+      "loss": 0.3663,
+      "step": 264
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.385447394296952e-05,
+      "loss": 0.3753,
+      "step": 265
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.3829891838741397e-05,
+      "loss": 0.3762,
+      "step": 266
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.380530973451328e-05,
+      "loss": 0.3682,
+      "step": 267
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.378072763028516e-05,
+      "loss": 0.3474,
+      "step": 268
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.375614552605704e-05,
+      "loss": 0.3694,
+      "step": 269
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.373156342182891e-05,
+      "loss": 0.4305,
+      "step": 270
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.3706981317600785e-05,
+      "loss": 0.3432,
+      "step": 271
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.3682399213372666e-05,
+      "loss": 0.3508,
+      "step": 272
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.3657817109144546e-05,
+      "loss": 0.349,
+      "step": 273
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.363323500491642e-05,
+      "loss": 0.451,
+      "step": 274
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.36086529006883e-05,
+      "loss": 0.3486,
+      "step": 275
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.358407079646018e-05,
+      "loss": 0.3716,
+      "step": 276
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.355948869223206e-05,
+      "loss": 0.3728,
+      "step": 277
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3534906588003935e-05,
+      "loss": 0.3883,
+      "step": 278
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.351032448377581e-05,
+      "loss": 0.3629,
+      "step": 279
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.348574237954769e-05,
+      "loss": 0.3626,
+      "step": 280
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.346116027531957e-05,
+      "loss": 0.3512,
+      "step": 281
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.343657817109145e-05,
+      "loss": 0.3594,
+      "step": 282
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3411996066863323e-05,
+      "loss": 0.3571,
+      "step": 283
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3387413962635204e-05,
+      "loss": 0.3783,
+      "step": 284
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3362831858407084e-05,
+      "loss": 0.3308,
+      "step": 285
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.333824975417896e-05,
+      "loss": 0.3349,
+      "step": 286
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.331366764995084e-05,
+      "loss": 0.5084,
+      "step": 287
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.328908554572272e-05,
+      "loss": 0.3579,
+      "step": 288
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.326450344149459e-05,
+      "loss": 0.3658,
+      "step": 289
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.323992133726647e-05,
+      "loss": 0.3596,
+      "step": 290
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.321533923303835e-05,
+      "loss": 0.3807,
+      "step": 291
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.319075712881023e-05,
+      "loss": 0.4011,
+      "step": 292
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.316617502458211e-05,
+      "loss": 0.3483,
+      "step": 293
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.314159292035399e-05,
+      "loss": 0.3528,
+      "step": 294
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.311701081612586e-05,
+      "loss": 0.3567,
+      "step": 295
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.309242871189774e-05,
+      "loss": 0.3677,
+      "step": 296
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.306784660766962e-05,
+      "loss": 0.3851,
+      "step": 297
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.3043264503441496e-05,
+      "loss": 0.4094,
+      "step": 298
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.301868239921337e-05,
+      "loss": 0.4575,
+      "step": 299
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.299410029498525e-05,
+      "loss": 0.3628,
+      "step": 300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.296951819075713e-05,
+      "loss": 0.3581,
+      "step": 301
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.294493608652901e-05,
+      "loss": 0.3753,
+      "step": 302
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.2920353982300885e-05,
+      "loss": 0.3548,
+      "step": 303
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2895771878072765e-05,
+      "loss": 0.3333,
+      "step": 304
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2871189773844646e-05,
+      "loss": 0.3546,
+      "step": 305
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2846607669616526e-05,
+      "loss": 0.3563,
+      "step": 306
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.28220255653884e-05,
+      "loss": 0.3833,
+      "step": 307
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2797443461160273e-05,
+      "loss": 0.3622,
+      "step": 308
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2772861356932154e-05,
+      "loss": 0.3649,
+      "step": 309
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.2748279252704034e-05,
+      "loss": 0.3329,
+      "step": 310
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.272369714847591e-05,
+      "loss": 0.4113,
+      "step": 311
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.269911504424779e-05,
+      "loss": 0.3501,
+      "step": 312
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.267453294001967e-05,
+      "loss": 0.3461,
+      "step": 313
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.264995083579155e-05,
+      "loss": 0.3643,
+      "step": 314
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.262536873156342e-05,
+      "loss": 0.356,
+      "step": 315
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2600786627335303e-05,
+      "loss": 0.4141,
+      "step": 316
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2576204523107184e-05,
+      "loss": 0.3736,
+      "step": 317
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.255162241887906e-05,
+      "loss": 0.4465,
+      "step": 318
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.252704031465093e-05,
+      "loss": 0.3881,
+      "step": 319
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.250245821042281e-05,
+      "loss": 0.3275,
+      "step": 320
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.247787610619469e-05,
+      "loss": 0.3418,
+      "step": 321
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.245329400196657e-05,
+      "loss": 0.3471,
+      "step": 322
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2428711897738446e-05,
+      "loss": 0.3657,
+      "step": 323
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2404129793510327e-05,
+      "loss": 0.3511,
+      "step": 324
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.237954768928221e-05,
+      "loss": 0.3856,
+      "step": 325
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.235496558505409e-05,
+      "loss": 0.3428,
+      "step": 326
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.233038348082596e-05,
+      "loss": 0.3708,
+      "step": 327
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.2305801376597835e-05,
+      "loss": 0.491,
+      "step": 328
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.2281219272369715e-05,
+      "loss": 0.4147,
+      "step": 329
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.2256637168141596e-05,
+      "loss": 0.3564,
+      "step": 330
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.223205506391347e-05,
+      "loss": 0.3592,
+      "step": 331
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.220747295968535e-05,
+      "loss": 0.354,
+      "step": 332
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.218289085545723e-05,
+      "loss": 0.3334,
+      "step": 333
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.215830875122911e-05,
+      "loss": 0.3622,
+      "step": 334
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.2133726647000984e-05,
+      "loss": 0.3559,
+      "step": 335
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.2109144542772865e-05,
+      "loss": 0.3412,
+      "step": 336
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.208456243854474e-05,
+      "loss": 0.4333,
+      "step": 337
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.205998033431662e-05,
+      "loss": 0.3296,
+      "step": 338
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.20353982300885e-05,
+      "loss": 0.3648,
+      "step": 339
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.201081612586037e-05,
+      "loss": 0.3583,
+      "step": 340
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.1986234021632253e-05,
+      "loss": 0.3444,
+      "step": 341
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.1961651917404134e-05,
+      "loss": 0.3364,
+      "step": 342
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.193706981317601e-05,
+      "loss": 0.3428,
+      "step": 343
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.191248770894789e-05,
+      "loss": 0.3629,
+      "step": 344
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.188790560471977e-05,
+      "loss": 0.4026,
+      "step": 345
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.186332350049165e-05,
+      "loss": 0.3264,
+      "step": 346
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.183874139626352e-05,
+      "loss": 0.3825,
+      "step": 347
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.1814159292035396e-05,
+      "loss": 0.3456,
+      "step": 348
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.178957718780728e-05,
+      "loss": 0.3541,
+      "step": 349
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.176499508357916e-05,
+      "loss": 0.3324,
+      "step": 350
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.174041297935104e-05,
+      "loss": 0.3348,
+      "step": 351
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.171583087512291e-05,
+      "loss": 0.4355,
+      "step": 352
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.169124877089479e-05,
+      "loss": 0.3524,
+      "step": 353
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.4303,
+      "step": 354
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.164208456243855e-05,
+      "loss": 0.3804,
+      "step": 355
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.161750245821042e-05,
+      "loss": 0.4189,
+      "step": 356
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.15929203539823e-05,
+      "loss": 0.3662,
+      "step": 357
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.156833824975418e-05,
+      "loss": 0.3571,
+      "step": 358
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.154375614552606e-05,
+      "loss": 0.3642,
+      "step": 359
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1519174041297934e-05,
+      "loss": 0.3465,
+      "step": 360
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1494591937069815e-05,
+      "loss": 0.367,
+      "step": 361
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1470009832841695e-05,
+      "loss": 0.3634,
+      "step": 362
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1445427728613576e-05,
+      "loss": 0.3657,
+      "step": 363
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.142084562438545e-05,
+      "loss": 0.3419,
+      "step": 364
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.139626352015733e-05,
+      "loss": 0.3558,
+      "step": 365
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1371681415929203e-05,
+      "loss": 0.3601,
+      "step": 366
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1347099311701084e-05,
+      "loss": 0.3379,
+      "step": 367
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.132251720747296e-05,
+      "loss": 0.3626,
+      "step": 368
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.129793510324484e-05,
+      "loss": 0.3738,
+      "step": 369
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.127335299901672e-05,
+      "loss": 0.3616,
+      "step": 370
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.12487708947886e-05,
+      "loss": 0.3772,
+      "step": 371
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.122418879056047e-05,
+      "loss": 0.3775,
+      "step": 372
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.119960668633235e-05,
+      "loss": 0.3295,
+      "step": 373
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1175024582104233e-05,
+      "loss": 0.3369,
+      "step": 374
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.115044247787611e-05,
+      "loss": 0.3544,
+      "step": 375
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.112586037364799e-05,
+      "loss": 0.3758,
+      "step": 376
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.110127826941986e-05,
+      "loss": 0.4212,
+      "step": 377
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.107669616519174e-05,
+      "loss": 0.436,
+      "step": 378
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.105211406096362e-05,
+      "loss": 0.3623,
+      "step": 379
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1027531956735496e-05,
+      "loss": 0.4006,
+      "step": 380
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1002949852507376e-05,
+      "loss": 0.3499,
+      "step": 381
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.097836774827926e-05,
+      "loss": 0.3504,
+      "step": 382
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.095378564405114e-05,
+      "loss": 0.3563,
+      "step": 383
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.092920353982301e-05,
+      "loss": 0.3663,
+      "step": 384
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.0904621435594884e-05,
+      "loss": 0.3773,
+      "step": 385
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0880039331366765e-05,
+      "loss": 0.3772,
+      "step": 386
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0855457227138645e-05,
+      "loss": 0.3592,
+      "step": 387
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0830875122910526e-05,
+      "loss": 0.424,
+      "step": 388
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.08062930186824e-05,
+      "loss": 0.3605,
+      "step": 389
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.078171091445428e-05,
+      "loss": 0.3457,
+      "step": 390
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.075712881022616e-05,
+      "loss": 0.4251,
+      "step": 391
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.0732546705998034e-05,
+      "loss": 0.4324,
+      "step": 392
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0707964601769914e-05,
+      "loss": 0.3755,
+      "step": 393
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.068338249754179e-05,
+      "loss": 0.3364,
+      "step": 394
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.065880039331367e-05,
+      "loss": 0.3429,
+      "step": 395
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.063421828908555e-05,
+      "loss": 0.368,
+      "step": 396
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.060963618485742e-05,
+      "loss": 0.3856,
+      "step": 397
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.05850540806293e-05,
+      "loss": 0.3586,
+      "step": 398
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0560471976401183e-05,
+      "loss": 0.3698,
+      "step": 399
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0535889872173064e-05,
+      "loss": 0.3833,
+      "step": 400
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.051130776794494e-05,
+      "loss": 0.3932,
+      "step": 401
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.048672566371682e-05,
+      "loss": 0.3442,
+      "step": 402
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.04621435594887e-05,
+      "loss": 0.3623,
+      "step": 403
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.043756145526057e-05,
+      "loss": 0.4418,
+      "step": 404
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0412979351032446e-05,
+      "loss": 0.3361,
+      "step": 405
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0388397246804326e-05,
+      "loss": 0.3652,
+      "step": 406
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.036381514257621e-05,
+      "loss": 0.3584,
+      "step": 407
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.033923303834809e-05,
+      "loss": 0.3512,
+      "step": 408
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.031465093411996e-05,
+      "loss": 0.3714,
+      "step": 409
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.029006882989184e-05,
+      "loss": 0.3445,
+      "step": 410
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.026548672566372e-05,
+      "loss": 0.361,
+      "step": 411
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.02409046214356e-05,
+      "loss": 0.3407,
+      "step": 412
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0216322517207476e-05,
+      "loss": 0.3365,
+      "step": 413
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.019174041297935e-05,
+      "loss": 0.3223,
+      "step": 414
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.016715830875123e-05,
+      "loss": 0.3629,
+      "step": 415
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.014257620452311e-05,
+      "loss": 0.3555,
+      "step": 416
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0117994100294984e-05,
+      "loss": 0.3738,
+      "step": 417
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0093411996066864e-05,
+      "loss": 0.3645,
+      "step": 418
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0068829891838745e-05,
+      "loss": 0.3469,
+      "step": 419
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0044247787610625e-05,
+      "loss": 0.3534,
+      "step": 420
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.00196656833825e-05,
+      "loss": 0.3556,
+      "step": 421
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.999508357915438e-05,
+      "loss": 0.3429,
+      "step": 422
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.997050147492625e-05,
+      "loss": 0.3464,
+      "step": 423
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.9945919370698133e-05,
+      "loss": 0.3509,
+      "step": 424
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.992133726647001e-05,
+      "loss": 0.3386,
+      "step": 425
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 3.989675516224189e-05,
+      "loss": 0.4153,
+      "step": 426
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.987217305801377e-05,
+      "loss": 0.4461,
+      "step": 427
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.984759095378565e-05,
+      "loss": 0.3577,
+      "step": 428
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.982300884955752e-05,
+      "loss": 0.4134,
+      "step": 429
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.97984267453294e-05,
+      "loss": 0.3611,
+      "step": 430
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.977384464110128e-05,
+      "loss": 0.3492,
+      "step": 431
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.9749262536873163e-05,
+      "loss": 0.3497,
+      "step": 432
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.972468043264504e-05,
+      "loss": 0.3594,
+      "step": 433
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.970009832841691e-05,
+      "loss": 0.3818,
+      "step": 434
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.967551622418879e-05,
+      "loss": 0.3808,
+      "step": 435
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.965093411996067e-05,
+      "loss": 0.3561,
+      "step": 436
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9626352015732545e-05,
+      "loss": 0.4329,
+      "step": 437
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9601769911504426e-05,
+      "loss": 0.3588,
+      "step": 438
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9577187807276306e-05,
+      "loss": 0.3555,
+      "step": 439
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.955260570304819e-05,
+      "loss": 0.3678,
+      "step": 440
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.952802359882006e-05,
+      "loss": 0.3843,
+      "step": 441
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9503441494591934e-05,
+      "loss": 0.3418,
+      "step": 442
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9478859390363814e-05,
+      "loss": 0.3477,
+      "step": 443
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9454277286135695e-05,
+      "loss": 0.3385,
+      "step": 444
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9429695181907575e-05,
+      "loss": 0.3696,
+      "step": 445
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.940511307767945e-05,
+      "loss": 0.3351,
+      "step": 446
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.938053097345133e-05,
+      "loss": 0.326,
+      "step": 447
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.935594886922321e-05,
+      "loss": 0.3673,
+      "step": 448
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9331366764995083e-05,
+      "loss": 0.3494,
+      "step": 449
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9306784660766964e-05,
+      "loss": 0.3654,
+      "step": 450
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9282202556538844e-05,
+      "loss": 0.3434,
+      "step": 451
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.925762045231072e-05,
+      "loss": 0.3821,
+      "step": 452
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.92330383480826e-05,
+      "loss": 0.3814,
+      "step": 453
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.920845624385447e-05,
+      "loss": 0.3419,
+      "step": 454
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.918387413962635e-05,
+      "loss": 0.3345,
+      "step": 455
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.915929203539823e-05,
+      "loss": 0.3707,
+      "step": 456
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9134709931170113e-05,
+      "loss": 0.3592,
+      "step": 457
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.911012782694199e-05,
+      "loss": 0.3471,
+      "step": 458
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.908554572271387e-05,
+      "loss": 0.3431,
+      "step": 459
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.906096361848575e-05,
+      "loss": 0.3935,
+      "step": 460
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.903638151425762e-05,
+      "loss": 0.4194,
+      "step": 461
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.9011799410029495e-05,
+      "loss": 0.3418,
+      "step": 462
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.8987217305801376e-05,
+      "loss": 0.319,
+      "step": 463
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.8962635201573256e-05,
+      "loss": 0.3752,
+      "step": 464
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.893805309734514e-05,
+      "loss": 0.3415,
+      "step": 465
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.891347099311701e-05,
+      "loss": 0.3431,
+      "step": 466
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.3483,
+      "step": 467
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.886430678466077e-05,
+      "loss": 0.3736,
+      "step": 468
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.883972468043265e-05,
+      "loss": 0.3487,
+      "step": 469
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8815142576204525e-05,
+      "loss": 0.327,
+      "step": 470
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.87905604719764e-05,
+      "loss": 0.3352,
+      "step": 471
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.876597836774828e-05,
+      "loss": 0.3553,
+      "step": 472
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.874139626352016e-05,
+      "loss": 0.3569,
+      "step": 473
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.8716814159292034e-05,
+      "loss": 0.3422,
+      "step": 474
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8692232055063914e-05,
+      "loss": 0.3339,
+      "step": 475
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8667649950835794e-05,
+      "loss": 0.3193,
+      "step": 476
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8643067846607675e-05,
+      "loss": 0.3422,
+      "step": 477
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.861848574237955e-05,
+      "loss": 0.3663,
+      "step": 478
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.859390363815143e-05,
+      "loss": 0.3407,
+      "step": 479
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.856932153392331e-05,
+      "loss": 0.3395,
+      "step": 480
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.854473942969518e-05,
+      "loss": 0.3851,
+      "step": 481
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.852015732546706e-05,
+      "loss": 0.3656,
+      "step": 482
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.849557522123894e-05,
+      "loss": 0.3277,
+      "step": 483
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.847099311701082e-05,
+      "loss": 0.3532,
+      "step": 484
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.84464110127827e-05,
+      "loss": 0.33,
+      "step": 485
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.842182890855457e-05,
+      "loss": 0.3237,
+      "step": 486
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.839724680432645e-05,
+      "loss": 0.3516,
+      "step": 487
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.837266470009833e-05,
+      "loss": 0.3981,
+      "step": 488
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.834808259587021e-05,
+      "loss": 0.3579,
+      "step": 489
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.832350049164209e-05,
+      "loss": 0.3393,
+      "step": 490
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.829891838741396e-05,
+      "loss": 0.3223,
+      "step": 491
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.827433628318584e-05,
+      "loss": 0.3237,
+      "step": 492
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.824975417895772e-05,
+      "loss": 0.3914,
+      "step": 493
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.82251720747296e-05,
+      "loss": 0.3469,
+      "step": 494
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8200589970501475e-05,
+      "loss": 0.3212,
+      "step": 495
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8176007866273356e-05,
+      "loss": 0.349,
+      "step": 496
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8151425762045236e-05,
+      "loss": 0.3212,
+      "step": 497
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.812684365781711e-05,
+      "loss": 0.3624,
+      "step": 498
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.810226155358899e-05,
+      "loss": 0.3232,
+      "step": 499
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8077679449360864e-05,
+      "loss": 0.3302,
+      "step": 500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8053097345132744e-05,
+      "loss": 0.3419,
+      "step": 501
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8028515240904625e-05,
+      "loss": 0.3468,
+      "step": 502
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.80039331366765e-05,
+      "loss": 0.3413,
+      "step": 503
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.797935103244838e-05,
+      "loss": 0.3466,
+      "step": 504
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.795476892822026e-05,
+      "loss": 0.3879,
+      "step": 505
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.793018682399214e-05,
+      "loss": 0.3454,
+      "step": 506
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.7905604719764013e-05,
+      "loss": 0.3549,
+      "step": 507
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.7881022615535894e-05,
+      "loss": 0.3397,
+      "step": 508
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.7856440511307774e-05,
+      "loss": 0.3208,
+      "step": 509
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.783185840707965e-05,
+      "loss": 0.4078,
+      "step": 510
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.780727630285152e-05,
+      "loss": 0.3148,
+      "step": 511
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.77826941986234e-05,
+      "loss": 0.3173,
+      "step": 512
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.775811209439528e-05,
+      "loss": 0.3393,
+      "step": 513
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.773352999016716e-05,
+      "loss": 0.4234,
+      "step": 514
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.770894788593904e-05,
+      "loss": 0.3265,
+      "step": 515
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.768436578171092e-05,
+      "loss": 0.328,
+      "step": 516
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.76597836774828e-05,
+      "loss": 0.3423,
+      "step": 517
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.763520157325468e-05,
+      "loss": 0.3376,
+      "step": 518
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7610619469026545e-05,
+      "loss": 0.3342,
+      "step": 519
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7586037364798425e-05,
+      "loss": 0.3834,
+      "step": 520
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7561455260570306e-05,
+      "loss": 0.3453,
+      "step": 521
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7536873156342186e-05,
+      "loss": 0.3382,
+      "step": 522
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.751229105211406e-05,
+      "loss": 0.3282,
+      "step": 523
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.748770894788594e-05,
+      "loss": 0.3587,
+      "step": 524
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.746312684365782e-05,
+      "loss": 0.3534,
+      "step": 525
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.74385447394297e-05,
+      "loss": 0.3542,
+      "step": 526
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7413962635201575e-05,
+      "loss": 0.3534,
+      "step": 527
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7389380530973455e-05,
+      "loss": 0.3616,
+      "step": 528
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.736479842674533e-05,
+      "loss": 0.358,
+      "step": 529
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.734021632251721e-05,
+      "loss": 0.3253,
+      "step": 530
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.731563421828908e-05,
+      "loss": 0.3355,
+      "step": 531
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7291052114060964e-05,
+      "loss": 0.3393,
+      "step": 532
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7266470009832844e-05,
+      "loss": 0.3476,
+      "step": 533
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7241887905604724e-05,
+      "loss": 0.355,
+      "step": 534
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.72173058013766e-05,
+      "loss": 0.363,
+      "step": 535
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.719272369714848e-05,
+      "loss": 0.3715,
+      "step": 536
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.716814159292036e-05,
+      "loss": 0.3965,
+      "step": 537
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.714355948869223e-05,
+      "loss": 0.3182,
+      "step": 538
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.711897738446411e-05,
+      "loss": 0.3236,
+      "step": 539
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.709439528023599e-05,
+      "loss": 0.3216,
+      "step": 540
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.706981317600787e-05,
+      "loss": 0.3405,
+      "step": 541
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.704523107177975e-05,
+      "loss": 0.3322,
+      "step": 542
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.702064896755162e-05,
+      "loss": 0.3932,
+      "step": 543
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.69960668633235e-05,
+      "loss": 0.3074,
+      "step": 544
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.697148475909538e-05,
+      "loss": 0.3343,
+      "step": 545
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.694690265486726e-05,
+      "loss": 0.3387,
+      "step": 546
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.6922320550639136e-05,
+      "loss": 0.3795,
+      "step": 547
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.689773844641101e-05,
+      "loss": 0.3625,
+      "step": 548
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.687315634218289e-05,
+      "loss": 0.3637,
+      "step": 549
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.684857423795477e-05,
+      "loss": 0.3925,
+      "step": 550
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.682399213372665e-05,
+      "loss": 0.352,
+      "step": 551
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6799410029498525e-05,
+      "loss": 0.3763,
+      "step": 552
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6774827925270405e-05,
+      "loss": 0.3499,
+      "step": 553
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6750245821042286e-05,
+      "loss": 0.3368,
+      "step": 554
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.672566371681416e-05,
+      "loss": 0.3371,
+      "step": 555
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.670108161258604e-05,
+      "loss": 0.3299,
+      "step": 556
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.667649950835792e-05,
+      "loss": 0.3222,
+      "step": 557
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6651917404129794e-05,
+      "loss": 0.349,
+      "step": 558
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6627335299901674e-05,
+      "loss": 0.3301,
+      "step": 559
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.660275319567355e-05,
+      "loss": 0.3474,
+      "step": 560
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.657817109144543e-05,
+      "loss": 0.4259,
+      "step": 561
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.655358898721731e-05,
+      "loss": 0.3387,
+      "step": 562
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.652900688298919e-05,
+      "loss": 0.417,
+      "step": 563
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.650442477876106e-05,
+      "loss": 0.3326,
+      "step": 564
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6479842674532944e-05,
+      "loss": 0.3405,
+      "step": 565
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6455260570304824e-05,
+      "loss": 0.3241,
+      "step": 566
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.64306784660767e-05,
+      "loss": 0.3499,
+      "step": 567
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.640609636184857e-05,
+      "loss": 0.3316,
+      "step": 568
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.638151425762045e-05,
+      "loss": 0.3171,
+      "step": 569
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.635693215339233e-05,
+      "loss": 0.3439,
+      "step": 570
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.633235004916421e-05,
+      "loss": 0.4409,
+      "step": 571
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6307767944936086e-05,
+      "loss": 0.3153,
+      "step": 572
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.628318584070797e-05,
+      "loss": 0.328,
+      "step": 573
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.625860373647985e-05,
+      "loss": 0.3644,
+      "step": 574
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.623402163225173e-05,
+      "loss": 0.3227,
+      "step": 575
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.62094395280236e-05,
+      "loss": 0.3641,
+      "step": 576
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6184857423795475e-05,
+      "loss": 0.34,
+      "step": 577
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.6160275319567355e-05,
+      "loss": 0.3985,
+      "step": 578
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.6135693215339236e-05,
+      "loss": 0.3324,
+      "step": 579
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 0.3937,
+      "step": 580
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.608652900688299e-05,
+      "loss": 0.3352,
+      "step": 581
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.606194690265487e-05,
+      "loss": 0.3483,
+      "step": 582
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.603736479842675e-05,
+      "loss": 0.3297,
+      "step": 583
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.6012782694198624e-05,
+      "loss": 0.3645,
+      "step": 584
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5988200589970505e-05,
+      "loss": 0.3508,
+      "step": 585
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.596361848574238e-05,
+      "loss": 0.329,
+      "step": 586
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.593903638151426e-05,
+      "loss": 0.3619,
+      "step": 587
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.591445427728613e-05,
+      "loss": 0.3796,
+      "step": 588
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.588987217305801e-05,
+      "loss": 0.3524,
+      "step": 589
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.5865290068829894e-05,
+      "loss": 0.3495,
+      "step": 590
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.5840707964601774e-05,
+      "loss": 0.3618,
+      "step": 591
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.581612586037365e-05,
+      "loss": 0.3478,
+      "step": 592
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.579154375614553e-05,
+      "loss": 0.3181,
+      "step": 593
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.576696165191741e-05,
+      "loss": 0.3171,
+      "step": 594
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.574237954768929e-05,
+      "loss": 0.3438,
+      "step": 595
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.571779744346116e-05,
+      "loss": 0.3332,
+      "step": 596
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.5693215339233036e-05,
+      "loss": 0.3142,
+      "step": 597
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.566863323500492e-05,
+      "loss": 0.31,
+      "step": 598
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.56440511307768e-05,
+      "loss": 0.3903,
+      "step": 599
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.561946902654867e-05,
+      "loss": 0.3371,
+      "step": 600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.559488692232055e-05,
+      "loss": 0.3461,
+      "step": 601
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.557030481809243e-05,
+      "loss": 0.3765,
+      "step": 602
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.554572271386431e-05,
+      "loss": 0.3329,
+      "step": 603
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.5521140609636186e-05,
+      "loss": 0.4012,
+      "step": 604
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5496558505408066e-05,
+      "loss": 0.3498,
+      "step": 605
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.547197640117994e-05,
+      "loss": 0.3532,
+      "step": 606
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.544739429695182e-05,
+      "loss": 0.3158,
+      "step": 607
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.54228121927237e-05,
+      "loss": 0.366,
+      "step": 608
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5398230088495574e-05,
+      "loss": 0.3398,
+      "step": 609
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5373647984267455e-05,
+      "loss": 0.3233,
+      "step": 610
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5349065880039335e-05,
+      "loss": 0.3958,
+      "step": 611
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.532448377581121e-05,
+      "loss": 0.3286,
+      "step": 612
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.529990167158309e-05,
+      "loss": 0.3326,
+      "step": 613
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.527531956735497e-05,
+      "loss": 0.3855,
+      "step": 614
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5250737463126844e-05,
+      "loss": 0.3303,
+      "step": 615
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5226155358898724e-05,
+      "loss": 0.3136,
+      "step": 616
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.52015732546706e-05,
+      "loss": 0.325,
+      "step": 617
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.517699115044248e-05,
+      "loss": 0.3363,
+      "step": 618
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.515240904621436e-05,
+      "loss": 0.324,
+      "step": 619
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.512782694198624e-05,
+      "loss": 0.4002,
+      "step": 620
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.510324483775811e-05,
+      "loss": 0.3403,
+      "step": 621
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.507866273352999e-05,
+      "loss": 0.3224,
+      "step": 622
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5054080629301874e-05,
+      "loss": 0.3018,
+      "step": 623
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5029498525073754e-05,
+      "loss": 0.3498,
+      "step": 624
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.500491642084562e-05,
+      "loss": 0.3263,
+      "step": 625
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.49803343166175e-05,
+      "loss": 0.3239,
+      "step": 626
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.495575221238938e-05,
+      "loss": 0.3313,
+      "step": 627
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.493117010816126e-05,
+      "loss": 0.3355,
+      "step": 628
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4906588003933136e-05,
+      "loss": 0.3309,
+      "step": 629
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4882005899705016e-05,
+      "loss": 0.3505,
+      "step": 630
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.48574237954769e-05,
+      "loss": 0.3167,
+      "step": 631
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.483284169124878e-05,
+      "loss": 0.3182,
+      "step": 632
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.480825958702065e-05,
+      "loss": 0.3319,
+      "step": 633
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4783677482792524e-05,
+      "loss": 0.3275,
+      "step": 634
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4759095378564405e-05,
+      "loss": 0.3244,
+      "step": 635
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4734513274336285e-05,
+      "loss": 0.3291,
+      "step": 636
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.470993117010816e-05,
+      "loss": 0.3915,
+      "step": 637
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.468534906588004e-05,
+      "loss": 0.379,
+      "step": 638
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.466076696165192e-05,
+      "loss": 0.3806,
+      "step": 639
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.46361848574238e-05,
+      "loss": 0.3381,
+      "step": 640
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4611602753195674e-05,
+      "loss": 0.3644,
+      "step": 641
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4587020648967554e-05,
+      "loss": 0.3533,
+      "step": 642
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4562438544739435e-05,
+      "loss": 0.3178,
+      "step": 643
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.453785644051131e-05,
+      "loss": 0.3378,
+      "step": 644
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.451327433628319e-05,
+      "loss": 0.3154,
+      "step": 645
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.448869223205506e-05,
+      "loss": 0.3295,
+      "step": 646
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.446411012782694e-05,
+      "loss": 0.3298,
+      "step": 647
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4439528023598824e-05,
+      "loss": 0.3114,
+      "step": 648
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.44149459193707e-05,
+      "loss": 0.3214,
+      "step": 649
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.439036381514258e-05,
+      "loss": 0.3747,
+      "step": 650
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.436578171091446e-05,
+      "loss": 0.3205,
+      "step": 651
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.434119960668634e-05,
+      "loss": 0.382,
+      "step": 652
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.431661750245821e-05,
+      "loss": 0.3296,
+      "step": 653
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4292035398230086e-05,
+      "loss": 0.3143,
+      "step": 654
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4267453294001966e-05,
+      "loss": 0.3228,
+      "step": 655
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.424287118977385e-05,
+      "loss": 0.3292,
+      "step": 656
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.421828908554573e-05,
+      "loss": 0.3103,
+      "step": 657
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.41937069813176e-05,
+      "loss": 0.3447,
+      "step": 658
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.416912487708948e-05,
+      "loss": 0.3492,
+      "step": 659
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.414454277286136e-05,
+      "loss": 0.3785,
+      "step": 660
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.4119960668633235e-05,
+      "loss": 0.3353,
+      "step": 661
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.4095378564405116e-05,
+      "loss": 0.3092,
+      "step": 662
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.407079646017699e-05,
+      "loss": 0.3102,
+      "step": 663
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.404621435594887e-05,
+      "loss": 0.3231,
+      "step": 664
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.402163225172075e-05,
+      "loss": 0.3819,
+      "step": 665
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.3997050147492624e-05,
+      "loss": 0.3143,
+      "step": 666
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.3972468043264504e-05,
+      "loss": 0.3088,
+      "step": 667
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.3947885939036385e-05,
+      "loss": 0.3334,
+      "step": 668
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.3923303834808265e-05,
+      "loss": 0.326,
+      "step": 669
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.389872173058014e-05,
+      "loss": 0.4287,
+      "step": 670
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.387413962635202e-05,
+      "loss": 0.3311,
+      "step": 671
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.38495575221239e-05,
+      "loss": 0.318,
+      "step": 672
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.3824975417895774e-05,
+      "loss": 0.2865,
+      "step": 673
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.380039331366765e-05,
+      "loss": 0.4216,
+      "step": 674
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.377581120943953e-05,
+      "loss": 0.3467,
+      "step": 675
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.375122910521141e-05,
+      "loss": 0.3622,
+      "step": 676
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.372664700098329e-05,
+      "loss": 0.3573,
+      "step": 677
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.370206489675516e-05,
+      "loss": 0.3179,
+      "step": 678
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.367748279252704e-05,
+      "loss": 0.3148,
+      "step": 679
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365290068829892e-05,
+      "loss": 0.3474,
+      "step": 680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3628318584070804e-05,
+      "loss": 0.3406,
+      "step": 681
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360373647984267e-05,
+      "loss": 0.3167,
+      "step": 682
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357915437561455e-05,
+      "loss": 0.3029,
+      "step": 683
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355457227138643e-05,
+      "loss": 0.3191,
+      "step": 684
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.352999016715831e-05,
+      "loss": 0.3358,
+      "step": 685
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3505408062930185e-05,
+      "loss": 0.3094,
+      "step": 686
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3480825958702066e-05,
+      "loss": 0.3355,
+      "step": 687
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3456243854473946e-05,
+      "loss": 0.3147,
+      "step": 688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343166175024583e-05,
+      "loss": 0.2923,
+      "step": 689
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34070796460177e-05,
+      "loss": 0.3172,
+      "step": 690
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338249754178958e-05,
+      "loss": 0.3327,
+      "step": 691
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3357915437561454e-05,
+      "loss": 0.288,
+      "step": 692
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.3017,
+      "step": 693
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.330875122910521e-05,
+      "loss": 0.2912,
+      "step": 694
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.328416912487709e-05,
+      "loss": 0.3003,
+      "step": 695
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.325958702064897e-05,
+      "loss": 0.3866,
+      "step": 696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.323500491642085e-05,
+      "loss": 0.3078,
+      "step": 697
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3210422812192724e-05,
+      "loss": 0.3079,
+      "step": 698
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3185840707964604e-05,
+      "loss": 0.3044,
+      "step": 699
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3161258603736484e-05,
+      "loss": 0.312,
+      "step": 700
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.313667649950836e-05,
+      "loss": 0.3027,
+      "step": 701
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.311209439528024e-05,
+      "loss": 0.3305,
+      "step": 702
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.308751229105211e-05,
+      "loss": 0.303,
+      "step": 703
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.306293018682399e-05,
+      "loss": 0.3323,
+      "step": 704
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.303834808259587e-05,
+      "loss": 0.316,
+      "step": 705
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.301376597836775e-05,
+      "loss": 0.3074,
+      "step": 706
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.298918387413963e-05,
+      "loss": 0.3111,
+      "step": 707
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.296460176991151e-05,
+      "loss": 0.3138,
+      "step": 708
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.294001966568339e-05,
+      "loss": 0.3005,
+      "step": 709
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.291543756145526e-05,
+      "loss": 0.3002,
+      "step": 710
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.2890855457227135e-05,
+      "loss": 0.3061,
+      "step": 711
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.2866273352999016e-05,
+      "loss": 0.3249,
+      "step": 712
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.2841691248770896e-05,
+      "loss": 0.3161,
+      "step": 713
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.281710914454278e-05,
+      "loss": 0.2885,
+      "step": 714
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.279252704031465e-05,
+      "loss": 0.2935,
+      "step": 715
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.276794493608653e-05,
+      "loss": 0.3099,
+      "step": 716
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.274336283185841e-05,
+      "loss": 0.2843,
+      "step": 717
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.2718780727630285e-05,
+      "loss": 0.2949,
+      "step": 718
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.2694198623402165e-05,
+      "loss": 0.2926,
+      "step": 719
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.2669616519174046e-05,
+      "loss": 0.3162,
+      "step": 720
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.264503441494592e-05,
+      "loss": 0.301,
+      "step": 721
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.26204523107178e-05,
+      "loss": 0.2866,
+      "step": 722
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2595870206489674e-05,
+      "loss": 0.2969,
+      "step": 723
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2571288102261554e-05,
+      "loss": 0.295,
+      "step": 724
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2546705998033434e-05,
+      "loss": 0.2844,
+      "step": 725
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.2522123893805315e-05,
+      "loss": 0.3013,
+      "step": 726
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.249754178957719e-05,
+      "loss": 0.3233,
+      "step": 727
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.247295968534907e-05,
+      "loss": 0.325,
+      "step": 728
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.244837758112095e-05,
+      "loss": 0.3093,
+      "step": 729
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.242379547689282e-05,
+      "loss": 0.4103,
+      "step": 730
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.23992133726647e-05,
+      "loss": 0.3032,
+      "step": 731
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.237463126843658e-05,
+      "loss": 0.2897,
+      "step": 732
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.235004916420846e-05,
+      "loss": 0.2908,
+      "step": 733
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.232546705998034e-05,
+      "loss": 0.3145,
+      "step": 734
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.230088495575221e-05,
+      "loss": 0.3033,
+      "step": 735
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.227630285152409e-05,
+      "loss": 0.3365,
+      "step": 736
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.225172074729597e-05,
+      "loss": 0.2896,
+      "step": 737
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.222713864306785e-05,
+      "loss": 0.316,
+      "step": 738
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.220255653883973e-05,
+      "loss": 0.3086,
+      "step": 739
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.21779744346116e-05,
+      "loss": 0.3185,
+      "step": 740
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.215339233038348e-05,
+      "loss": 0.3074,
+      "step": 741
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.212881022615536e-05,
+      "loss": 0.3575,
+      "step": 742
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.2104228121927235e-05,
+      "loss": 0.2889,
+      "step": 743
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.2079646017699115e-05,
+      "loss": 0.326,
+      "step": 744
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.2055063913470996e-05,
+      "loss": 0.343,
+      "step": 745
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.2030481809242876e-05,
+      "loss": 0.3846,
+      "step": 746
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.200589970501475e-05,
+      "loss": 0.3189,
+      "step": 747
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.198131760078663e-05,
+      "loss": 0.3002,
+      "step": 748
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.1956735496558504e-05,
+      "loss": 0.2817,
+      "step": 749
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.1932153392330385e-05,
+      "loss": 0.3126,
+      "step": 750
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.190757128810226e-05,
+      "loss": 0.2958,
+      "step": 751
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.188298918387414e-05,
+      "loss": 0.3348,
+      "step": 752
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.185840707964602e-05,
+      "loss": 0.2949,
+      "step": 753
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.18338249754179e-05,
+      "loss": 0.291,
+      "step": 754
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.180924287118977e-05,
+      "loss": 0.2844,
+      "step": 755
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.1784660766961654e-05,
+      "loss": 0.3242,
+      "step": 756
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.1760078662733534e-05,
+      "loss": 0.3134,
+      "step": 757
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.1735496558505414e-05,
+      "loss": 0.2966,
+      "step": 758
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.171091445427729e-05,
+      "loss": 0.3636,
+      "step": 759
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.168633235004916e-05,
+      "loss": 0.3684,
+      "step": 760
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.166175024582104e-05,
+      "loss": 0.3387,
+      "step": 761
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.163716814159292e-05,
+      "loss": 0.3129,
+      "step": 762
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.16125860373648e-05,
+      "loss": 0.2969,
+      "step": 763
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.158800393313668e-05,
+      "loss": 0.2894,
+      "step": 764
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.156342182890856e-05,
+      "loss": 0.3617,
+      "step": 765
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.153883972468044e-05,
+      "loss": 0.3631,
+      "step": 766
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.151425762045231e-05,
+      "loss": 0.3058,
+      "step": 767
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.148967551622419e-05,
+      "loss": 0.3017,
+      "step": 768
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.1465093411996065e-05,
+      "loss": 0.3305,
+      "step": 769
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.1440511307767946e-05,
+      "loss": 0.3093,
+      "step": 770
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.1415929203539826e-05,
+      "loss": 0.3006,
+      "step": 771
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.13913470993117e-05,
+      "loss": 0.2822,
+      "step": 772
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.136676499508358e-05,
+      "loss": 0.3018,
+      "step": 773
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.134218289085546e-05,
+      "loss": 0.2783,
+      "step": 774
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.131760078662734e-05,
+      "loss": 0.3666,
+      "step": 775
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.1293018682399215e-05,
+      "loss": 0.319,
+      "step": 776
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.1268436578171095e-05,
+      "loss": 0.2886,
+      "step": 777
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.124385447394297e-05,
+      "loss": 0.2919,
+      "step": 778
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.121927236971485e-05,
+      "loss": 0.297,
+      "step": 779
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.119469026548672e-05,
+      "loss": 0.3108,
+      "step": 780
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.1170108161258604e-05,
+      "loss": 0.2977,
+      "step": 781
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.1145526057030484e-05,
+      "loss": 0.3955,
+      "step": 782
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.1120943952802364e-05,
+      "loss": 0.3241,
+      "step": 783
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.109636184857424e-05,
+      "loss": 0.2913,
+      "step": 784
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.107177974434612e-05,
+      "loss": 0.3097,
+      "step": 785
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.1047197640118e-05,
+      "loss": 0.3132,
+      "step": 786
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.102261553588988e-05,
+      "loss": 0.2978,
+      "step": 787
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.0998033431661746e-05,
+      "loss": 0.3044,
+      "step": 788
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.097345132743363e-05,
+      "loss": 0.3085,
+      "step": 789
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.094886922320551e-05,
+      "loss": 0.2924,
+      "step": 790
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.092428711897739e-05,
+      "loss": 0.2884,
+      "step": 791
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.089970501474926e-05,
+      "loss": 0.3207,
+      "step": 792
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.087512291052114e-05,
+      "loss": 0.2721,
+      "step": 793
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.085054080629302e-05,
+      "loss": 0.2939,
+      "step": 794
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.08259587020649e-05,
+      "loss": 0.3105,
+      "step": 795
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.0801376597836776e-05,
+      "loss": 0.2978,
+      "step": 796
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.077679449360865e-05,
+      "loss": 0.2797,
+      "step": 797
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.075221238938053e-05,
+      "loss": 0.3522,
+      "step": 798
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.072763028515241e-05,
+      "loss": 0.2938,
+      "step": 799
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.0703048180924285e-05,
+      "loss": 0.3143,
+      "step": 800
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.0678466076696165e-05,
+      "loss": 0.3004,
+      "step": 801
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.0653883972468045e-05,
+      "loss": 0.3034,
+      "step": 802
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.0629301868239926e-05,
+      "loss": 0.3109,
+      "step": 803
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.06047197640118e-05,
+      "loss": 0.2998,
+      "step": 804
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.058013765978368e-05,
+      "loss": 0.2802,
+      "step": 805
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 0.3133,
+      "step": 806
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.0530973451327434e-05,
+      "loss": 0.3037,
+      "step": 807
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.050639134709931e-05,
+      "loss": 0.2846,
+      "step": 808
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.048180924287119e-05,
+      "loss": 0.3199,
+      "step": 809
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.045722713864307e-05,
+      "loss": 0.3126,
+      "step": 810
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0432645034414946e-05,
+      "loss": 0.3009,
+      "step": 811
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0408062930186826e-05,
+      "loss": 0.2917,
+      "step": 812
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0383480825958703e-05,
+      "loss": 0.2844,
+      "step": 813
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.0358898721730584e-05,
+      "loss": 0.3186,
+      "step": 814
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.033431661750246e-05,
+      "loss": 0.2698,
+      "step": 815
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.030973451327434e-05,
+      "loss": 0.3166,
+      "step": 816
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.0285152409046215e-05,
+      "loss": 0.2963,
+      "step": 817
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.0260570304818092e-05,
+      "loss": 0.3185,
+      "step": 818
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.0235988200589972e-05,
+      "loss": 0.321,
+      "step": 819
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.021140609636185e-05,
+      "loss": 0.2887,
+      "step": 820
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.018682399213373e-05,
+      "loss": 0.3225,
+      "step": 821
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.0162241887905607e-05,
+      "loss": 0.2947,
+      "step": 822
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.0137659783677484e-05,
+      "loss": 0.2963,
+      "step": 823
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0113077679449364e-05,
+      "loss": 0.3068,
+      "step": 824
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.008849557522124e-05,
+      "loss": 0.291,
+      "step": 825
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0063913470993115e-05,
+      "loss": 0.3097,
+      "step": 826
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0039331366764995e-05,
+      "loss": 0.3145,
+      "step": 827
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0014749262536872e-05,
+      "loss": 0.2956,
+      "step": 828
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.9990167158308753e-05,
+      "loss": 0.2935,
+      "step": 829
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.996558505408063e-05,
+      "loss": 0.2907,
+      "step": 830
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.994100294985251e-05,
+      "loss": 0.2649,
+      "step": 831
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9916420845624387e-05,
+      "loss": 0.298,
+      "step": 832
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9891838741396268e-05,
+      "loss": 0.3357,
+      "step": 833
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9867256637168145e-05,
+      "loss": 0.3105,
+      "step": 834
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9842674532940022e-05,
+      "loss": 0.2985,
+      "step": 835
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9818092428711896e-05,
+      "loss": 0.2898,
+      "step": 836
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.9793510324483776e-05,
+      "loss": 0.2925,
+      "step": 837
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.9768928220255653e-05,
+      "loss": 0.3084,
+      "step": 838
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.9744346116027534e-05,
+      "loss": 0.2988,
+      "step": 839
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.971976401179941e-05,
+      "loss": 0.2962,
+      "step": 840
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.969518190757129e-05,
+      "loss": 0.318,
+      "step": 841
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.9670599803343168e-05,
+      "loss": 0.2849,
+      "step": 842
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.964601769911505e-05,
+      "loss": 0.2938,
+      "step": 843
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9621435594886926e-05,
+      "loss": 0.3065,
+      "step": 844
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.95968534906588e-05,
+      "loss": 0.2967,
+      "step": 845
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9572271386430676e-05,
+      "loss": 0.3233,
+      "step": 846
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9547689282202557e-05,
+      "loss": 0.2833,
+      "step": 847
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9523107177974434e-05,
+      "loss": 0.2966,
+      "step": 848
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.9498525073746314e-05,
+      "loss": 0.3037,
+      "step": 849
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.947394296951819e-05,
+      "loss": 0.3164,
+      "step": 850
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.9449360865290072e-05,
+      "loss": 0.2943,
+      "step": 851
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.942477876106195e-05,
+      "loss": 0.2873,
+      "step": 852
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.940019665683383e-05,
+      "loss": 0.3508,
+      "step": 853
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.9375614552605706e-05,
+      "loss": 0.2979,
+      "step": 854
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.935103244837758e-05,
+      "loss": 0.2944,
+      "step": 855
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.9326450344149457e-05,
+      "loss": 0.3358,
+      "step": 856
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.9301868239921338e-05,
+      "loss": 0.3009,
+      "step": 857
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9277286135693215e-05,
+      "loss": 0.2965,
+      "step": 858
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9252704031465095e-05,
+      "loss": 0.3672,
+      "step": 859
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9228121927236972e-05,
+      "loss": 0.344,
+      "step": 860
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9203539823008852e-05,
+      "loss": 0.3071,
+      "step": 861
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.917895771878073e-05,
+      "loss": 0.2939,
+      "step": 862
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.915437561455261e-05,
+      "loss": 0.3213,
+      "step": 863
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.9129793510324487e-05,
+      "loss": 0.3004,
+      "step": 864
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.910521140609636e-05,
+      "loss": 0.3016,
+      "step": 865
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.908062930186824e-05,
+      "loss": 0.3187,
+      "step": 866
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9056047197640118e-05,
+      "loss": 0.2964,
+      "step": 867
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9031465093412e-05,
+      "loss": 0.2804,
+      "step": 868
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9006882989183876e-05,
+      "loss": 0.2951,
+      "step": 869
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8982300884955753e-05,
+      "loss": 0.2823,
+      "step": 870
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8957718780727633e-05,
+      "loss": 0.2965,
+      "step": 871
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.893313667649951e-05,
+      "loss": 0.2858,
+      "step": 872
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.890855457227139e-05,
+      "loss": 0.3111,
+      "step": 873
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8883972468043264e-05,
+      "loss": 0.2859,
+      "step": 874
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.885939036381514e-05,
+      "loss": 0.3111,
+      "step": 875
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8834808259587022e-05,
+      "loss": 0.3922,
+      "step": 876
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.88102261553589e-05,
+      "loss": 0.3584,
+      "step": 877
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.878564405113078e-05,
+      "loss": 0.2942,
+      "step": 878
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8761061946902656e-05,
+      "loss": 0.3197,
+      "step": 879
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8736479842674537e-05,
+      "loss": 0.2841,
+      "step": 880
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8711897738446414e-05,
+      "loss": 0.2952,
+      "step": 881
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.868731563421829e-05,
+      "loss": 0.3051,
+      "step": 882
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.866273352999017e-05,
+      "loss": 0.3377,
+      "step": 883
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8638151425762045e-05,
+      "loss": 0.2891,
+      "step": 884
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8613569321533922e-05,
+      "loss": 0.2816,
+      "step": 885
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8588987217305803e-05,
+      "loss": 0.295,
+      "step": 886
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.856440511307768e-05,
+      "loss": 0.3086,
+      "step": 887
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.853982300884956e-05,
+      "loss": 0.2995,
+      "step": 888
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8515240904621437e-05,
+      "loss": 0.2909,
+      "step": 889
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8490658800393317e-05,
+      "loss": 0.292,
+      "step": 890
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.8466076696165195e-05,
+      "loss": 0.2969,
+      "step": 891
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8441494591937075e-05,
+      "loss": 0.2983,
+      "step": 892
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8416912487708945e-05,
+      "loss": 0.3035,
+      "step": 893
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8392330383480826e-05,
+      "loss": 0.3534,
+      "step": 894
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8367748279252703e-05,
+      "loss": 0.296,
+      "step": 895
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8343166175024583e-05,
+      "loss": 0.3702,
+      "step": 896
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.831858407079646e-05,
+      "loss": 0.3138,
+      "step": 897
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.829400196656834e-05,
+      "loss": 0.2954,
+      "step": 898
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8269419862340218e-05,
+      "loss": 0.2796,
+      "step": 899
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8244837758112098e-05,
+      "loss": 0.3277,
+      "step": 900
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8220255653883975e-05,
+      "loss": 0.3148,
+      "step": 901
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8195673549655856e-05,
+      "loss": 0.2898,
+      "step": 902
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8171091445427726e-05,
+      "loss": 0.3014,
+      "step": 903
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8146509341199606e-05,
+      "loss": 0.2989,
+      "step": 904
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8121927236971483e-05,
+      "loss": 0.3177,
+      "step": 905
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8097345132743364e-05,
+      "loss": 0.3565,
+      "step": 906
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.807276302851524e-05,
+      "loss": 0.3209,
+      "step": 907
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.804818092428712e-05,
+      "loss": 0.2939,
+      "step": 908
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8023598820059e-05,
+      "loss": 0.2953,
+      "step": 909
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.799901671583088e-05,
+      "loss": 0.2917,
+      "step": 910
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.7974434611602756e-05,
+      "loss": 0.2907,
+      "step": 911
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.7949852507374636e-05,
+      "loss": 0.2974,
+      "step": 912
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.792527040314651e-05,
+      "loss": 0.3723,
+      "step": 913
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.7900688298918387e-05,
+      "loss": 0.3482,
+      "step": 914
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.7876106194690264e-05,
+      "loss": 0.289,
+      "step": 915
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.7851524090462145e-05,
+      "loss": 0.3052,
+      "step": 916
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.782694198623402e-05,
+      "loss": 0.2952,
+      "step": 917
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.7802359882005902e-05,
+      "loss": 0.2905,
+      "step": 918
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.2839,
+      "step": 919
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.775319567354966e-05,
+      "loss": 0.2846,
+      "step": 920
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.7728613569321537e-05,
+      "loss": 0.3014,
+      "step": 921
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.770403146509341e-05,
+      "loss": 0.2743,
+      "step": 922
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.767944936086529e-05,
+      "loss": 0.2974,
+      "step": 923
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.7654867256637168e-05,
+      "loss": 0.3299,
+      "step": 924
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.7630285152409048e-05,
+      "loss": 0.2794,
+      "step": 925
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7605703048180925e-05,
+      "loss": 0.2794,
+      "step": 926
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7581120943952802e-05,
+      "loss": 0.3323,
+      "step": 927
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7556538839724683e-05,
+      "loss": 0.2798,
+      "step": 928
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.753195673549656e-05,
+      "loss": 0.314,
+      "step": 929
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.750737463126844e-05,
+      "loss": 0.3441,
+      "step": 930
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.7482792527040317e-05,
+      "loss": 0.28,
+      "step": 931
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.745821042281219e-05,
+      "loss": 0.2925,
+      "step": 932
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.743362831858407e-05,
+      "loss": 0.2865,
+      "step": 933
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.740904621435595e-05,
+      "loss": 0.2811,
+      "step": 934
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.738446411012783e-05,
+      "loss": 0.3375,
+      "step": 935
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.7359882005899706e-05,
+      "loss": 0.2769,
+      "step": 936
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.7335299901671586e-05,
+      "loss": 0.2932,
+      "step": 937
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.7310717797443463e-05,
+      "loss": 0.2911,
+      "step": 938
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.7286135693215344e-05,
+      "loss": 0.286,
+      "step": 939
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.726155358898722e-05,
+      "loss": 0.3251,
+      "step": 940
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.7236971484759095e-05,
+      "loss": 0.291,
+      "step": 941
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.721238938053097e-05,
+      "loss": 0.3066,
+      "step": 942
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.7187807276302852e-05,
+      "loss": 0.3113,
+      "step": 943
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.716322517207473e-05,
+      "loss": 0.2967,
+      "step": 944
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.713864306784661e-05,
+      "loss": 0.2897,
+      "step": 945
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.7114060963618487e-05,
+      "loss": 0.3487,
+      "step": 946
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7089478859390367e-05,
+      "loss": 0.2998,
+      "step": 947
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7064896755162244e-05,
+      "loss": 0.338,
+      "step": 948
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7040314650934125e-05,
+      "loss": 0.2962,
+      "step": 949
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7015732546706e-05,
+      "loss": 0.3128,
+      "step": 950
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6991150442477875e-05,
+      "loss": 0.3041,
+      "step": 951
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.6966568338249752e-05,
+      "loss": 0.3116,
+      "step": 952
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6941986234021633e-05,
+      "loss": 0.275,
+      "step": 953
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.691740412979351e-05,
+      "loss": 0.2822,
+      "step": 954
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.689282202556539e-05,
+      "loss": 0.2969,
+      "step": 955
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6868239921337267e-05,
+      "loss": 0.2904,
+      "step": 956
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6843657817109148e-05,
+      "loss": 0.307,
+      "step": 957
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6819075712881025e-05,
+      "loss": 0.2898,
+      "step": 958
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.6794493608652905e-05,
+      "loss": 0.2956,
+      "step": 959
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6769911504424782e-05,
+      "loss": 0.2998,
+      "step": 960
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6745329400196656e-05,
+      "loss": 0.2864,
+      "step": 961
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6720747295968533e-05,
+      "loss": 0.2914,
+      "step": 962
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6696165191740413e-05,
+      "loss": 0.3156,
+      "step": 963
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.667158308751229e-05,
+      "loss": 0.3011,
+      "step": 964
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.664700098328417e-05,
+      "loss": 0.2854,
+      "step": 965
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.6622418879056048e-05,
+      "loss": 0.3267,
+      "step": 966
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.659783677482793e-05,
+      "loss": 0.2874,
+      "step": 967
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6573254670599805e-05,
+      "loss": 0.2962,
+      "step": 968
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6548672566371686e-05,
+      "loss": 0.3296,
+      "step": 969
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.652409046214356e-05,
+      "loss": 0.2804,
+      "step": 970
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6499508357915437e-05,
+      "loss": 0.294,
+      "step": 971
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6474926253687317e-05,
+      "loss": 0.3005,
+      "step": 972
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.6450344149459194e-05,
+      "loss": 0.2793,
+      "step": 973
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.642576204523107e-05,
+      "loss": 0.2827,
+      "step": 974
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.640117994100295e-05,
+      "loss": 0.279,
+      "step": 975
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.637659783677483e-05,
+      "loss": 0.274,
+      "step": 976
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.635201573254671e-05,
+      "loss": 0.321,
+      "step": 977
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6327433628318586e-05,
+      "loss": 0.2913,
+      "step": 978
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6302851524090467e-05,
+      "loss": 0.3504,
+      "step": 979
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.627826941986234e-05,
+      "loss": 0.2956,
+      "step": 980
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6253687315634217e-05,
+      "loss": 0.3051,
+      "step": 981
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6229105211406098e-05,
+      "loss": 0.2791,
+      "step": 982
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6204523107177975e-05,
+      "loss": 0.2827,
+      "step": 983
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6179941002949855e-05,
+      "loss": 0.2581,
+      "step": 984
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6155358898721732e-05,
+      "loss": 0.2962,
+      "step": 985
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.613077679449361e-05,
+      "loss": 0.2872,
+      "step": 986
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.610619469026549e-05,
+      "loss": 0.3034,
+      "step": 987
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6081612586037367e-05,
+      "loss": 0.2858,
+      "step": 988
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.605703048180924e-05,
+      "loss": 0.2973,
+      "step": 989
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.603244837758112e-05,
+      "loss": 0.2842,
+      "step": 990
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6007866273352998e-05,
+      "loss": 0.2752,
+      "step": 991
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.598328416912488e-05,
+      "loss": 0.517,
+      "step": 992
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.5958702064896756e-05,
+      "loss": 0.2802,
+      "step": 993
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5934119960668636e-05,
+      "loss": 0.3019,
+      "step": 994
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5909537856440513e-05,
+      "loss": 0.2745,
+      "step": 995
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5884955752212393e-05,
+      "loss": 0.3152,
+      "step": 996
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.586037364798427e-05,
+      "loss": 0.2672,
+      "step": 997
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.583579154375615e-05,
+      "loss": 0.2936,
+      "step": 998
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.581120943952802e-05,
+      "loss": 0.2958,
+      "step": 999
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.57866273352999e-05,
+      "loss": 0.2896,
+      "step": 1000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.576204523107178e-05,
+      "loss": 0.2843,
+      "step": 1001
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.573746312684366e-05,
+      "loss": 0.286,
+      "step": 1002
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5712881022615536e-05,
+      "loss": 0.2963,
+      "step": 1003
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5688298918387417e-05,
+      "loss": 0.299,
+      "step": 1004
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5663716814159294e-05,
+      "loss": 0.2914,
+      "step": 1005
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5639134709931174e-05,
+      "loss": 0.4149,
+      "step": 1006
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.561455260570305e-05,
+      "loss": 0.2713,
+      "step": 1007
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.558997050147493e-05,
+      "loss": 0.2909,
+      "step": 1008
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5565388397246802e-05,
+      "loss": 0.291,
+      "step": 1009
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5540806293018682e-05,
+      "loss": 0.3736,
+      "step": 1010
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.551622418879056e-05,
+      "loss": 0.2879,
+      "step": 1011
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.549164208456244e-05,
+      "loss": 0.3583,
+      "step": 1012
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5467059980334317e-05,
+      "loss": 0.2763,
+      "step": 1013
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.5442477876106197e-05,
+      "loss": 0.3326,
+      "step": 1014
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5417895771878074e-05,
+      "loss": 0.3435,
+      "step": 1015
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5393313667649955e-05,
+      "loss": 0.2758,
+      "step": 1016
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5368731563421832e-05,
+      "loss": 0.2679,
+      "step": 1017
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5344149459193706e-05,
+      "loss": 0.273,
+      "step": 1018
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5319567354965586e-05,
+      "loss": 0.2829,
+      "step": 1019
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.5294985250737463e-05,
+      "loss": 0.3186,
+      "step": 1020
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.527040314650934e-05,
+      "loss": 0.295,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.524582104228122e-05,
+      "loss": 0.2967,
+      "step": 1022
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5221238938053098e-05,
+      "loss": 0.2818,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5196656833824978e-05,
+      "loss": 0.3018,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5172074729596855e-05,
+      "loss": 0.2912,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5147492625368735e-05,
+      "loss": 0.3628,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5122910521140613e-05,
+      "loss": 0.2968,
+      "step": 1027
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5098328416912486e-05,
+      "loss": 0.2989,
+      "step": 1028
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5073746312684367e-05,
+      "loss": 0.3141,
+      "step": 1029
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5049164208456244e-05,
+      "loss": 0.2712,
+      "step": 1030
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5024582104228124e-05,
+      "loss": 0.3049,
+      "step": 1031
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5e-05,
+      "loss": 0.2694,
+      "step": 1032
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.4975417895771878e-05,
+      "loss": 0.3519,
+      "step": 1033
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.495083579154376e-05,
+      "loss": 0.2809,
+      "step": 1034
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4926253687315636e-05,
+      "loss": 0.2968,
+      "step": 1035
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4901671583087513e-05,
+      "loss": 0.345,
+      "step": 1036
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4877089478859393e-05,
+      "loss": 0.2867,
+      "step": 1037
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.485250737463127e-05,
+      "loss": 0.2962,
+      "step": 1038
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4827925270403147e-05,
+      "loss": 0.2846,
+      "step": 1039
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4803343166175024e-05,
+      "loss": 0.2901,
+      "step": 1040
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.4778761061946905e-05,
+      "loss": 0.3379,
+      "step": 1041
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4754178957718782e-05,
+      "loss": 0.288,
+      "step": 1042
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4729596853490662e-05,
+      "loss": 0.2932,
+      "step": 1043
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4705014749262536e-05,
+      "loss": 0.2673,
+      "step": 1044
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4680432645034416e-05,
+      "loss": 0.2871,
+      "step": 1045
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4655850540806293e-05,
+      "loss": 0.2917,
+      "step": 1046
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.4631268436578174e-05,
+      "loss": 0.3019,
+      "step": 1047
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.460668633235005e-05,
+      "loss": 0.2757,
+      "step": 1048
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4582104228121928e-05,
+      "loss": 0.2893,
+      "step": 1049
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4557522123893805e-05,
+      "loss": 0.2888,
+      "step": 1050
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4532940019665686e-05,
+      "loss": 0.2785,
+      "step": 1051
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4508357915437563e-05,
+      "loss": 0.3076,
+      "step": 1052
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4483775811209443e-05,
+      "loss": 0.2821,
+      "step": 1053
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4459193706981317e-05,
+      "loss": 0.3288,
+      "step": 1054
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.4434611602753197e-05,
+      "loss": 0.2755,
+      "step": 1055
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4410029498525074e-05,
+      "loss": 0.2994,
+      "step": 1056
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4385447394296955e-05,
+      "loss": 0.3179,
+      "step": 1057
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.436086529006883e-05,
+      "loss": 0.295,
+      "step": 1058
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.433628318584071e-05,
+      "loss": 0.2903,
+      "step": 1059
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4311701081612586e-05,
+      "loss": 0.2813,
+      "step": 1060
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4287118977384466e-05,
+      "loss": 0.2764,
+      "step": 1061
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.4262536873156343e-05,
+      "loss": 0.2711,
+      "step": 1062
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4237954768928224e-05,
+      "loss": 0.2649,
+      "step": 1063
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4213372664700097e-05,
+      "loss": 0.2789,
+      "step": 1064
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4188790560471978e-05,
+      "loss": 0.2917,
+      "step": 1065
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4164208456243855e-05,
+      "loss": 0.2861,
+      "step": 1066
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4139626352015735e-05,
+      "loss": 0.3272,
+      "step": 1067
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.411504424778761e-05,
+      "loss": 0.2916,
+      "step": 1068
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.409046214355949e-05,
+      "loss": 0.3188,
+      "step": 1069
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4065880039331366e-05,
+      "loss": 0.2815,
+      "step": 1070
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4041297935103247e-05,
+      "loss": 0.3558,
+      "step": 1071
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4016715830875124e-05,
+      "loss": 0.3027,
+      "step": 1072
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3992133726647e-05,
+      "loss": 0.2703,
+      "step": 1073
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.3967551622418878e-05,
+      "loss": 0.2787,
+      "step": 1074
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.394296951819076e-05,
+      "loss": 0.2878,
+      "step": 1075
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3918387413962636e-05,
+      "loss": 0.2894,
+      "step": 1076
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3893805309734516e-05,
+      "loss": 0.2857,
+      "step": 1077
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3869223205506393e-05,
+      "loss": 0.2949,
+      "step": 1078
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.384464110127827e-05,
+      "loss": 0.2785,
+      "step": 1079
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3820058997050147e-05,
+      "loss": 0.2919,
+      "step": 1080
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3795476892822028e-05,
+      "loss": 0.2676,
+      "step": 1081
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.3770894788593905e-05,
+      "loss": 0.2714,
+      "step": 1082
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.374631268436578e-05,
+      "loss": 0.2727,
+      "step": 1083
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3721730580137662e-05,
+      "loss": 0.2735,
+      "step": 1084
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.369714847590954e-05,
+      "loss": 0.303,
+      "step": 1085
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3672566371681416e-05,
+      "loss": 0.2813,
+      "step": 1086
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3647984267453297e-05,
+      "loss": 0.2707,
+      "step": 1087
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.3623402163225174e-05,
+      "loss": 0.3067,
+      "step": 1088
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.359882005899705e-05,
+      "loss": 0.2788,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.357423795476893e-05,
+      "loss": 0.2907,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3549655850540808e-05,
+      "loss": 0.2624,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3525073746312685e-05,
+      "loss": 0.3881,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3500491642084562e-05,
+      "loss": 0.2786,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.3475909537856443e-05,
+      "loss": 0.3459,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.345132743362832e-05,
+      "loss": 0.2901,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.34267453294002e-05,
+      "loss": 0.2613,
+      "step": 1096
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3402163225172074e-05,
+      "loss": 0.2815,
+      "step": 1097
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3377581120943954e-05,
+      "loss": 0.2854,
+      "step": 1098
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.335299901671583e-05,
+      "loss": 0.2716,
+      "step": 1099
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3328416912487712e-05,
+      "loss": 0.3004,
+      "step": 1100
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.330383480825959e-05,
+      "loss": 0.2846,
+      "step": 1101
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3279252704031466e-05,
+      "loss": 0.3095,
+      "step": 1102
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3254670599803343e-05,
+      "loss": 0.2768,
+      "step": 1103
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3230088495575223e-05,
+      "loss": 0.3289,
+      "step": 1104
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.32055063913471e-05,
+      "loss": 0.3008,
+      "step": 1105
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.318092428711898e-05,
+      "loss": 0.2971,
+      "step": 1106
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3156342182890855e-05,
+      "loss": 0.2722,
+      "step": 1107
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3131760078662735e-05,
+      "loss": 0.2711,
+      "step": 1108
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3107177974434612e-05,
+      "loss": 0.2808,
+      "step": 1109
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3082595870206493e-05,
+      "loss": 0.2809,
+      "step": 1110
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.305801376597837e-05,
+      "loss": 0.2727,
+      "step": 1111
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3033431661750247e-05,
+      "loss": 0.2829,
+      "step": 1112
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3008849557522124e-05,
+      "loss": 0.2959,
+      "step": 1113
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2984267453294004e-05,
+      "loss": 0.2902,
+      "step": 1114
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.295968534906588e-05,
+      "loss": 0.2761,
+      "step": 1115
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2935103244837758e-05,
+      "loss": 0.2722,
+      "step": 1116
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2910521140609635e-05,
+      "loss": 0.2667,
+      "step": 1117
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2885939036381516e-05,
+      "loss": 0.2913,
+      "step": 1118
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2861356932153393e-05,
+      "loss": 0.3134,
+      "step": 1119
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2836774827925273e-05,
+      "loss": 0.2774,
+      "step": 1120
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2812192723697147e-05,
+      "loss": 0.2878,
+      "step": 1121
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2787610619469027e-05,
+      "loss": 0.2832,
+      "step": 1122
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2763028515240904e-05,
+      "loss": 0.3183,
+      "step": 1123
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2738446411012785e-05,
+      "loss": 0.2815,
+      "step": 1124
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2713864306784662e-05,
+      "loss": 0.3022,
+      "step": 1125
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.268928220255654e-05,
+      "loss": 0.3047,
+      "step": 1126
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2664700098328416e-05,
+      "loss": 0.2859,
+      "step": 1127
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2640117994100296e-05,
+      "loss": 0.2716,
+      "step": 1128
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2615535889872174e-05,
+      "loss": 0.2953,
+      "step": 1129
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.2590953785644054e-05,
+      "loss": 0.2798,
+      "step": 1130
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2566371681415928e-05,
+      "loss": 0.2813,
+      "step": 1131
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2541789577187808e-05,
+      "loss": 0.3547,
+      "step": 1132
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2517207472959685e-05,
+      "loss": 0.2953,
+      "step": 1133
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2492625368731566e-05,
+      "loss": 0.2748,
+      "step": 1134
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.2468043264503443e-05,
+      "loss": 0.3208,
+      "step": 1135
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.244346116027532e-05,
+      "loss": 0.2731,
+      "step": 1136
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.24188790560472e-05,
+      "loss": 0.2959,
+      "step": 1137
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2394296951819077e-05,
+      "loss": 0.3298,
+      "step": 1138
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2369714847590954e-05,
+      "loss": 0.2713,
+      "step": 1139
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.234513274336283e-05,
+      "loss": 0.2714,
+      "step": 1140
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.232055063913471e-05,
+      "loss": 0.2947,
+      "step": 1141
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.229596853490659e-05,
+      "loss": 0.2815,
+      "step": 1142
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.227138643067847e-05,
+      "loss": 0.3121,
+      "step": 1143
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.2246804326450346e-05,
+      "loss": 0.2936,
+      "step": 1144
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.3547,
+      "step": 1145
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.21976401179941e-05,
+      "loss": 0.3367,
+      "step": 1146
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.217305801376598e-05,
+      "loss": 0.2737,
+      "step": 1147
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2148475909537858e-05,
+      "loss": 0.2822,
+      "step": 1148
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2123893805309738e-05,
+      "loss": 0.2901,
+      "step": 1149
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2099311701081612e-05,
+      "loss": 0.2669,
+      "step": 1150
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2074729596853492e-05,
+      "loss": 0.285,
+      "step": 1151
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.205014749262537e-05,
+      "loss": 0.2646,
+      "step": 1152
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.202556538839725e-05,
+      "loss": 0.2712,
+      "step": 1153
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2000983284169127e-05,
+      "loss": 0.2914,
+      "step": 1154
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.1976401179941004e-05,
+      "loss": 0.2789,
+      "step": 1155
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.195181907571288e-05,
+      "loss": 0.3383,
+      "step": 1156
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.192723697148476e-05,
+      "loss": 0.2736,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.190265486725664e-05,
+      "loss": 0.2691,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.187807276302852e-05,
+      "loss": 0.2643,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1853490658800393e-05,
+      "loss": 0.2573,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1828908554572273e-05,
+      "loss": 0.2639,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.180432645034415e-05,
+      "loss": 0.2669,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.177974434611603e-05,
+      "loss": 0.2727,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.1755162241887904e-05,
+      "loss": 0.2775,
+      "step": 1164
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1730580137659785e-05,
+      "loss": 0.2982,
+      "step": 1165
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1705998033431662e-05,
+      "loss": 0.2591,
+      "step": 1166
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1681415929203542e-05,
+      "loss": 0.3374,
+      "step": 1167
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.165683382497542e-05,
+      "loss": 0.2807,
+      "step": 1168
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1632251720747296e-05,
+      "loss": 0.3213,
+      "step": 1169
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1607669616519173e-05,
+      "loss": 0.2653,
+      "step": 1170
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.1583087512291054e-05,
+      "loss": 0.3323,
+      "step": 1171
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.155850540806293e-05,
+      "loss": 0.2862,
+      "step": 1172
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.153392330383481e-05,
+      "loss": 0.2745,
+      "step": 1173
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1509341199606685e-05,
+      "loss": 0.3133,
+      "step": 1174
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1484759095378565e-05,
+      "loss": 0.2887,
+      "step": 1175
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1460176991150442e-05,
+      "loss": 0.3313,
+      "step": 1176
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.1435594886922323e-05,
+      "loss": 0.2843,
+      "step": 1177
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.14110127826942e-05,
+      "loss": 0.2531,
+      "step": 1178
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1386430678466077e-05,
+      "loss": 0.2807,
+      "step": 1179
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1361848574237954e-05,
+      "loss": 0.2713,
+      "step": 1180
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1337266470009834e-05,
+      "loss": 0.2849,
+      "step": 1181
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.131268436578171e-05,
+      "loss": 0.3321,
+      "step": 1182
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1288102261553592e-05,
+      "loss": 0.288,
+      "step": 1183
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1263520157325466e-05,
+      "loss": 0.3574,
+      "step": 1184
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1238938053097346e-05,
+      "loss": 0.2549,
+      "step": 1185
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1214355948869223e-05,
+      "loss": 0.274,
+      "step": 1186
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1189773844641104e-05,
+      "loss": 0.2858,
+      "step": 1187
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.116519174041298e-05,
+      "loss": 0.3356,
+      "step": 1188
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1140609636184858e-05,
+      "loss": 0.2602,
+      "step": 1189
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1116027531956735e-05,
+      "loss": 0.2931,
+      "step": 1190
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1091445427728615e-05,
+      "loss": 0.2633,
+      "step": 1191
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1066863323500492e-05,
+      "loss": 0.2761,
+      "step": 1192
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.104228121927237e-05,
+      "loss": 0.2625,
+      "step": 1193
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.101769911504425e-05,
+      "loss": 0.3214,
+      "step": 1194
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0993117010816127e-05,
+      "loss": 0.266,
+      "step": 1195
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0968534906588004e-05,
+      "loss": 0.3566,
+      "step": 1196
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.0943952802359884e-05,
+      "loss": 0.2443,
+      "step": 1197
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.091937069813176e-05,
+      "loss": 0.3123,
+      "step": 1198
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.089478859390364e-05,
+      "loss": 0.2671,
+      "step": 1199
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.087020648967552e-05,
+      "loss": 0.2614,
+      "step": 1200
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.0845624385447396e-05,
+      "loss": 0.2834,
+      "step": 1201
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.0821042281219276e-05,
+      "loss": 0.278,
+      "step": 1202
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.079646017699115e-05,
+      "loss": 0.2856,
+      "step": 1203
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.077187807276303e-05,
+      "loss": 0.2899,
+      "step": 1204
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.0747295968534907e-05,
+      "loss": 0.2661,
+      "step": 1205
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0722713864306788e-05,
+      "loss": 0.2839,
+      "step": 1206
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0698131760078665e-05,
+      "loss": 0.272,
+      "step": 1207
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0673549655850542e-05,
+      "loss": 0.273,
+      "step": 1208
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.064896755162242e-05,
+      "loss": 0.2698,
+      "step": 1209
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.06243854473943e-05,
+      "loss": 0.2751,
+      "step": 1210
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0599803343166176e-05,
+      "loss": 0.3199,
+      "step": 1211
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.0575221238938054e-05,
+      "loss": 0.2864,
+      "step": 1212
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.055063913470993e-05,
+      "loss": 0.2853,
+      "step": 1213
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.052605703048181e-05,
+      "loss": 0.352,
+      "step": 1214
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.0501474926253688e-05,
+      "loss": 0.3372,
+      "step": 1215
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.047689282202557e-05,
+      "loss": 0.2809,
+      "step": 1216
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.0452310717797442e-05,
+      "loss": 0.2728,
+      "step": 1217
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.0427728613569323e-05,
+      "loss": 0.289,
+      "step": 1218
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.04031465093412e-05,
+      "loss": 0.3203,
+      "step": 1219
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.037856440511308e-05,
+      "loss": 0.2703,
+      "step": 1220
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.0353982300884957e-05,
+      "loss": 0.2818,
+      "step": 1221
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.0329400196656834e-05,
+      "loss": 0.2666,
+      "step": 1222
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.030481809242871e-05,
+      "loss": 0.2848,
+      "step": 1223
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.0280235988200592e-05,
+      "loss": 0.3284,
+      "step": 1224
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.025565388397247e-05,
+      "loss": 0.2866,
+      "step": 1225
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.023107177974435e-05,
+      "loss": 0.3175,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0206489675516223e-05,
+      "loss": 0.2748,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0181907571288103e-05,
+      "loss": 0.2746,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.015732546705998e-05,
+      "loss": 0.2968,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.013274336283186e-05,
+      "loss": 0.2447,
+      "step": 1230
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0108161258603738e-05,
+      "loss": 0.2926,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0083579154375615e-05,
+      "loss": 0.2806,
+      "step": 1232
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0058997050147492e-05,
+      "loss": 0.3005,
+      "step": 1233
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0034414945919372e-05,
+      "loss": 0.2865,
+      "step": 1234
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.000983284169125e-05,
+      "loss": 0.2751,
+      "step": 1235
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9985250737463127e-05,
+      "loss": 0.3042,
+      "step": 1236
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9960668633235004e-05,
+      "loss": 0.2725,
+      "step": 1237
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.9936086529006884e-05,
+      "loss": 0.2947,
+      "step": 1238
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.991150442477876e-05,
+      "loss": 0.285,
+      "step": 1239
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.988692232055064e-05,
+      "loss": 0.2764,
+      "step": 1240
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.986234021632252e-05,
+      "loss": 0.2738,
+      "step": 1241
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9837758112094396e-05,
+      "loss": 0.2824,
+      "step": 1242
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9813176007866273e-05,
+      "loss": 0.2972,
+      "step": 1243
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9788593903638153e-05,
+      "loss": 0.263,
+      "step": 1244
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.976401179941003e-05,
+      "loss": 0.2752,
+      "step": 1245
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.9739429695181907e-05,
+      "loss": 0.2451,
+      "step": 1246
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9714847590953788e-05,
+      "loss": 0.2723,
+      "step": 1247
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9690265486725665e-05,
+      "loss": 0.2607,
+      "step": 1248
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9665683382497542e-05,
+      "loss": 0.2822,
+      "step": 1249
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9641101278269422e-05,
+      "loss": 0.279,
+      "step": 1250
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.96165191740413e-05,
+      "loss": 0.2663,
+      "step": 1251
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9591937069813176e-05,
+      "loss": 0.3455,
+      "step": 1252
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.9567354965585057e-05,
+      "loss": 0.2566,
+      "step": 1253
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9542772861356934e-05,
+      "loss": 0.2683,
+      "step": 1254
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.951819075712881e-05,
+      "loss": 0.2743,
+      "step": 1255
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9493608652900688e-05,
+      "loss": 0.257,
+      "step": 1256
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.946902654867257e-05,
+      "loss": 0.2721,
+      "step": 1257
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 0.2434,
+      "step": 1258
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.9419862340216326e-05,
+      "loss": 0.31,
+      "step": 1259
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.93952802359882e-05,
+      "loss": 0.3205,
+      "step": 1260
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.937069813176008e-05,
+      "loss": 0.2594,
+      "step": 1261
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9346116027531957e-05,
+      "loss": 0.2673,
+      "step": 1262
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9321533923303837e-05,
+      "loss": 0.2698,
+      "step": 1263
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9296951819075714e-05,
+      "loss": 0.3474,
+      "step": 1264
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.927236971484759e-05,
+      "loss": 0.2559,
+      "step": 1265
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.924778761061947e-05,
+      "loss": 0.2546,
+      "step": 1266
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.922320550639135e-05,
+      "loss": 0.2809,
+      "step": 1267
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9198623402163226e-05,
+      "loss": 0.2644,
+      "step": 1268
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9174041297935107e-05,
+      "loss": 0.273,
+      "step": 1269
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.914945919370698e-05,
+      "loss": 0.2653,
+      "step": 1270
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.912487708947886e-05,
+      "loss": 0.3041,
+      "step": 1271
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9100294985250738e-05,
+      "loss": 0.2783,
+      "step": 1272
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9075712881022618e-05,
+      "loss": 0.2692,
+      "step": 1273
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9051130776794495e-05,
+      "loss": 0.2894,
+      "step": 1274
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9026548672566372e-05,
+      "loss": 0.2682,
+      "step": 1275
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.900196656833825e-05,
+      "loss": 0.269,
+      "step": 1276
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.897738446411013e-05,
+      "loss": 0.2622,
+      "step": 1277
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.8952802359882007e-05,
+      "loss": 0.2878,
+      "step": 1278
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.8928220255653887e-05,
+      "loss": 0.2714,
+      "step": 1279
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.890363815142576e-05,
+      "loss": 0.2718,
+      "step": 1280
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.887905604719764e-05,
+      "loss": 0.252,
+      "step": 1281
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.885447394296952e-05,
+      "loss": 0.2861,
+      "step": 1282
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.88298918387414e-05,
+      "loss": 0.2693,
+      "step": 1283
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8805309734513272e-05,
+      "loss": 0.3149,
+      "step": 1284
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8780727630285153e-05,
+      "loss": 0.2707,
+      "step": 1285
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.875614552605703e-05,
+      "loss": 0.2522,
+      "step": 1286
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.873156342182891e-05,
+      "loss": 0.2657,
+      "step": 1287
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8706981317600787e-05,
+      "loss": 0.2686,
+      "step": 1288
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8682399213372664e-05,
+      "loss": 0.2678,
+      "step": 1289
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.865781710914454e-05,
+      "loss": 0.2709,
+      "step": 1290
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8633235004916422e-05,
+      "loss": 0.3105,
+      "step": 1291
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.86086529006883e-05,
+      "loss": 0.2571,
+      "step": 1292
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.858407079646018e-05,
+      "loss": 0.2647,
+      "step": 1293
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8559488692232057e-05,
+      "loss": 0.3274,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8534906588003934e-05,
+      "loss": 0.2669,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.851032448377581e-05,
+      "loss": 0.2867,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.848574237954769e-05,
+      "loss": 0.2525,
+      "step": 1297
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8461160275319568e-05,
+      "loss": 0.3913,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8436578171091445e-05,
+      "loss": 0.2571,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8411996066863326e-05,
+      "loss": 0.268,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8387413962635203e-05,
+      "loss": 0.2549,
+      "step": 1301
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.836283185840708e-05,
+      "loss": 0.281,
+      "step": 1302
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.833824975417896e-05,
+      "loss": 0.2556,
+      "step": 1303
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8313667649950837e-05,
+      "loss": 0.2828,
+      "step": 1304
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8289085545722714e-05,
+      "loss": 0.2544,
+      "step": 1305
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8264503441494595e-05,
+      "loss": 0.2567,
+      "step": 1306
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8239921337266472e-05,
+      "loss": 0.2787,
+      "step": 1307
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.821533923303835e-05,
+      "loss": 0.2454,
+      "step": 1308
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8190757128810226e-05,
+      "loss": 0.2563,
+      "step": 1309
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8166175024582106e-05,
+      "loss": 0.2817,
+      "step": 1310
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8141592920353983e-05,
+      "loss": 0.2456,
+      "step": 1311
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8117010816125864e-05,
+      "loss": 0.2586,
+      "step": 1312
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8092428711897737e-05,
+      "loss": 0.3284,
+      "step": 1313
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8067846607669618e-05,
+      "loss": 0.272,
+      "step": 1314
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8043264503441495e-05,
+      "loss": 0.2808,
+      "step": 1315
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8018682399213375e-05,
+      "loss": 0.2619,
+      "step": 1316
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7994100294985252e-05,
+      "loss": 0.2771,
+      "step": 1317
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.796951819075713e-05,
+      "loss": 0.2778,
+      "step": 1318
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7944936086529007e-05,
+      "loss": 0.2658,
+      "step": 1319
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7920353982300887e-05,
+      "loss": 0.2798,
+      "step": 1320
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7895771878072764e-05,
+      "loss": 0.2592,
+      "step": 1321
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7871189773844644e-05,
+      "loss": 0.2793,
+      "step": 1322
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7846607669616518e-05,
+      "loss": 0.2751,
+      "step": 1323
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.78220255653884e-05,
+      "loss": 0.274,
+      "step": 1324
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7797443461160276e-05,
+      "loss": 0.2599,
+      "step": 1325
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7772861356932156e-05,
+      "loss": 0.2745,
+      "step": 1326
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7748279252704033e-05,
+      "loss": 0.2754,
+      "step": 1327
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.772369714847591e-05,
+      "loss": 0.2602,
+      "step": 1328
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7699115044247787e-05,
+      "loss": 0.2723,
+      "step": 1329
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7674532940019668e-05,
+      "loss": 0.2988,
+      "step": 1330
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7649950835791545e-05,
+      "loss": 0.2492,
+      "step": 1331
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7625368731563422e-05,
+      "loss": 0.2747,
+      "step": 1332
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.76007866273353e-05,
+      "loss": 0.2781,
+      "step": 1333
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.757620452310718e-05,
+      "loss": 0.2625,
+      "step": 1334
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.7551622418879056e-05,
+      "loss": 0.2629,
+      "step": 1335
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7527040314650937e-05,
+      "loss": 0.2644,
+      "step": 1336
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.750245821042281e-05,
+      "loss": 0.2898,
+      "step": 1337
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.747787610619469e-05,
+      "loss": 0.2639,
+      "step": 1338
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7453294001966568e-05,
+      "loss": 0.2783,
+      "step": 1339
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.742871189773845e-05,
+      "loss": 0.2861,
+      "step": 1340
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7404129793510325e-05,
+      "loss": 0.269,
+      "step": 1341
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.7379547689282202e-05,
+      "loss": 0.3152,
+      "step": 1342
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.735496558505408e-05,
+      "loss": 0.2786,
+      "step": 1343
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.733038348082596e-05,
+      "loss": 0.2679,
+      "step": 1344
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7305801376597837e-05,
+      "loss": 0.2582,
+      "step": 1345
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7281219272369717e-05,
+      "loss": 0.2562,
+      "step": 1346
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7256637168141594e-05,
+      "loss": 0.2684,
+      "step": 1347
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.723205506391347e-05,
+      "loss": 0.2649,
+      "step": 1348
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.720747295968535e-05,
+      "loss": 0.2831,
+      "step": 1349
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.718289085545723e-05,
+      "loss": 0.2716,
+      "step": 1350
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7158308751229106e-05,
+      "loss": 0.2605,
+      "step": 1351
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7133726647000983e-05,
+      "loss": 0.2601,
+      "step": 1352
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7109144542772864e-05,
+      "loss": 0.2726,
+      "step": 1353
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.708456243854474e-05,
+      "loss": 0.274,
+      "step": 1354
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7059980334316618e-05,
+      "loss": 0.2612,
+      "step": 1355
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7035398230088495e-05,
+      "loss": 0.265,
+      "step": 1356
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7010816125860375e-05,
+      "loss": 0.2618,
+      "step": 1357
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6986234021632252e-05,
+      "loss": 0.2475,
+      "step": 1358
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6961651917404133e-05,
+      "loss": 0.2566,
+      "step": 1359
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.693706981317601e-05,
+      "loss": 0.2587,
+      "step": 1360
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6912487708947887e-05,
+      "loss": 0.2764,
+      "step": 1361
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.6887905604719764e-05,
+      "loss": 0.2668,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6863323500491644e-05,
+      "loss": 0.2594,
+      "step": 1363
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.683874139626352e-05,
+      "loss": 0.261,
+      "step": 1364
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6814159292035402e-05,
+      "loss": 0.2802,
+      "step": 1365
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6789577187807275e-05,
+      "loss": 0.3199,
+      "step": 1366
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6764995083579156e-05,
+      "loss": 0.3255,
+      "step": 1367
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6740412979351033e-05,
+      "loss": 0.2591,
+      "step": 1368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.6715830875122913e-05,
+      "loss": 0.2289,
+      "step": 1369
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.669124877089479e-05,
+      "loss": 0.2442,
+      "step": 1370
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.2172,
+      "step": 1371
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6642084562438545e-05,
+      "loss": 0.2834,
+      "step": 1372
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6617502458210425e-05,
+      "loss": 0.2358,
+      "step": 1373
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6592920353982302e-05,
+      "loss": 0.2431,
+      "step": 1374
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.656833824975418e-05,
+      "loss": 0.248,
+      "step": 1375
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.6543756145526056e-05,
+      "loss": 0.2431,
+      "step": 1376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6519174041297937e-05,
+      "loss": 0.2574,
+      "step": 1377
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6494591937069814e-05,
+      "loss": 0.2253,
+      "step": 1378
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6470009832841694e-05,
+      "loss": 0.2329,
+      "step": 1379
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6445427728613568e-05,
+      "loss": 0.2515,
+      "step": 1380
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6420845624385448e-05,
+      "loss": 0.2351,
+      "step": 1381
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6396263520157325e-05,
+      "loss": 0.2479,
+      "step": 1382
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.6371681415929206e-05,
+      "loss": 0.2494,
+      "step": 1383
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6347099311701083e-05,
+      "loss": 0.25,
+      "step": 1384
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.632251720747296e-05,
+      "loss": 0.2176,
+      "step": 1385
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6297935103244837e-05,
+      "loss": 0.2629,
+      "step": 1386
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6273352999016717e-05,
+      "loss": 0.2515,
+      "step": 1387
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6248770894788594e-05,
+      "loss": 0.2474,
+      "step": 1388
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6224188790560475e-05,
+      "loss": 0.2321,
+      "step": 1389
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.619960668633235e-05,
+      "loss": 0.2453,
+      "step": 1390
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.617502458210423e-05,
+      "loss": 0.2368,
+      "step": 1391
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6150442477876106e-05,
+      "loss": 0.2469,
+      "step": 1392
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6125860373647986e-05,
+      "loss": 0.2589,
+      "step": 1393
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6101278269419863e-05,
+      "loss": 0.255,
+      "step": 1394
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.607669616519174e-05,
+      "loss": 0.293,
+      "step": 1395
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6052114060963617e-05,
+      "loss": 0.2542,
+      "step": 1396
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6027531956735498e-05,
+      "loss": 0.2259,
+      "step": 1397
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6002949852507375e-05,
+      "loss": 0.2619,
+      "step": 1398
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5978367748279252e-05,
+      "loss": 0.2531,
+      "step": 1399
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.595378564405113e-05,
+      "loss": 0.2532,
+      "step": 1400
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.592920353982301e-05,
+      "loss": 0.2613,
+      "step": 1401
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5904621435594887e-05,
+      "loss": 0.2471,
+      "step": 1402
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.5880039331366767e-05,
+      "loss": 0.2392,
+      "step": 1403
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5855457227138644e-05,
+      "loss": 0.2364,
+      "step": 1404
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.583087512291052e-05,
+      "loss": 0.2576,
+      "step": 1405
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.58062930186824e-05,
+      "loss": 0.2604,
+      "step": 1406
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.578171091445428e-05,
+      "loss": 0.2988,
+      "step": 1407
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5757128810226156e-05,
+      "loss": 0.2859,
+      "step": 1408
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5732546705998033e-05,
+      "loss": 0.2589,
+      "step": 1409
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5707964601769913e-05,
+      "loss": 0.2524,
+      "step": 1410
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.568338249754179e-05,
+      "loss": 0.2342,
+      "step": 1411
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.565880039331367e-05,
+      "loss": 0.2429,
+      "step": 1412
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.5634218289085548e-05,
+      "loss": 0.2339,
+      "step": 1413
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.5609636184857425e-05,
+      "loss": 0.2732,
+      "step": 1414
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.5585054080629302e-05,
+      "loss": 0.2423,
+      "step": 1415
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.5560471976401182e-05,
+      "loss": 0.2329,
+      "step": 1416
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.553588987217306e-05,
+      "loss": 0.2411,
+      "step": 1417
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.551130776794494e-05,
+      "loss": 0.2633,
+      "step": 1418
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5486725663716813e-05,
+      "loss": 0.2404,
+      "step": 1419
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5462143559488694e-05,
+      "loss": 0.2422,
+      "step": 1420
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.543756145526057e-05,
+      "loss": 0.2343,
+      "step": 1421
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.541297935103245e-05,
+      "loss": 0.3232,
+      "step": 1422
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5388397246804325e-05,
+      "loss": 0.2546,
+      "step": 1423
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5363815142576205e-05,
+      "loss": 0.251,
+      "step": 1424
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5339233038348082e-05,
+      "loss": 0.233,
+      "step": 1425
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5314650934119963e-05,
+      "loss": 0.237,
+      "step": 1426
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.529006882989184e-05,
+      "loss": 0.2499,
+      "step": 1427
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5265486725663717e-05,
+      "loss": 0.2478,
+      "step": 1428
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5240904621435596e-05,
+      "loss": 0.227,
+      "step": 1429
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.5216322517207473e-05,
+      "loss": 0.2435,
+      "step": 1430
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5191740412979352e-05,
+      "loss": 0.2444,
+      "step": 1431
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.516715830875123e-05,
+      "loss": 0.2362,
+      "step": 1432
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5142576204523107e-05,
+      "loss": 0.2788,
+      "step": 1433
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5117994100294986e-05,
+      "loss": 0.2394,
+      "step": 1434
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5093411996066865e-05,
+      "loss": 0.2284,
+      "step": 1435
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.5068829891838742e-05,
+      "loss": 0.2979,
+      "step": 1436
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.504424778761062e-05,
+      "loss": 0.2268,
+      "step": 1437
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.5019665683382498e-05,
+      "loss": 0.2527,
+      "step": 1438
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4995083579154376e-05,
+      "loss": 0.2303,
+      "step": 1439
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4970501474926255e-05,
+      "loss": 0.2401,
+      "step": 1440
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4945919370698134e-05,
+      "loss": 0.2366,
+      "step": 1441
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4921337266470011e-05,
+      "loss": 0.2871,
+      "step": 1442
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4896755162241888e-05,
+      "loss": 0.2312,
+      "step": 1443
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4872173058013767e-05,
+      "loss": 0.2492,
+      "step": 1444
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4847590953785646e-05,
+      "loss": 0.2593,
+      "step": 1445
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4823008849557524e-05,
+      "loss": 0.2445,
+      "step": 1446
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.47984267453294e-05,
+      "loss": 0.2508,
+      "step": 1447
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4773844641101278e-05,
+      "loss": 0.2467,
+      "step": 1448
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4749262536873157e-05,
+      "loss": 0.2457,
+      "step": 1449
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4724680432645036e-05,
+      "loss": 0.228,
+      "step": 1450
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.4700098328416915e-05,
+      "loss": 0.2828,
+      "step": 1451
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.467551622418879e-05,
+      "loss": 0.2436,
+      "step": 1452
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.4650934119960669e-05,
+      "loss": 0.3162,
+      "step": 1453
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.4626352015732547e-05,
+      "loss": 0.2466,
+      "step": 1454
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.4601769911504426e-05,
+      "loss": 0.2338,
+      "step": 1455
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.4577187807276305e-05,
+      "loss": 0.2571,
+      "step": 1456
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.455260570304818e-05,
+      "loss": 0.2257,
+      "step": 1457
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.4528023598820059e-05,
+      "loss": 0.2377,
+      "step": 1458
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4503441494591938e-05,
+      "loss": 0.2056,
+      "step": 1459
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4478859390363817e-05,
+      "loss": 0.2503,
+      "step": 1460
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4454277286135695e-05,
+      "loss": 0.2921,
+      "step": 1461
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.442969518190757e-05,
+      "loss": 0.2971,
+      "step": 1462
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.440511307767945e-05,
+      "loss": 0.2333,
+      "step": 1463
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4380530973451328e-05,
+      "loss": 0.2496,
+      "step": 1464
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.4355948869223207e-05,
+      "loss": 0.2638,
+      "step": 1465
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4331366764995086e-05,
+      "loss": 0.2404,
+      "step": 1466
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4306784660766961e-05,
+      "loss": 0.273,
+      "step": 1467
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.428220255653884e-05,
+      "loss": 0.2548,
+      "step": 1468
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4257620452310719e-05,
+      "loss": 0.2269,
+      "step": 1469
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4233038348082597e-05,
+      "loss": 0.2287,
+      "step": 1470
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.4208456243854473e-05,
+      "loss": 0.2444,
+      "step": 1471
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4183874139626351e-05,
+      "loss": 0.2351,
+      "step": 1472
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.415929203539823e-05,
+      "loss": 0.224,
+      "step": 1473
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4134709931170109e-05,
+      "loss": 0.2852,
+      "step": 1474
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4110127826941988e-05,
+      "loss": 0.292,
+      "step": 1475
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4085545722713863e-05,
+      "loss": 0.2478,
+      "step": 1476
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.4060963618485742e-05,
+      "loss": 0.2793,
+      "step": 1477
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.403638151425762e-05,
+      "loss": 0.2224,
+      "step": 1478
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.40117994100295e-05,
+      "loss": 0.2693,
+      "step": 1479
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3987217305801378e-05,
+      "loss": 0.2383,
+      "step": 1480
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3962635201573255e-05,
+      "loss": 0.2246,
+      "step": 1481
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3938053097345132e-05,
+      "loss": 0.2188,
+      "step": 1482
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.391347099311701e-05,
+      "loss": 0.2526,
+      "step": 1483
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 0.2511,
+      "step": 1484
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.3864306784660768e-05,
+      "loss": 0.2363,
+      "step": 1485
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3839724680432645e-05,
+      "loss": 0.2077,
+      "step": 1486
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3815142576204524e-05,
+      "loss": 0.2373,
+      "step": 1487
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3790560471976401e-05,
+      "loss": 0.2369,
+      "step": 1488
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.376597836774828e-05,
+      "loss": 0.2263,
+      "step": 1489
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3741396263520159e-05,
+      "loss": 0.2565,
+      "step": 1490
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3716814159292036e-05,
+      "loss": 0.2871,
+      "step": 1491
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.3692232055063914e-05,
+      "loss": 0.2274,
+      "step": 1492
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3667649950835793e-05,
+      "loss": 0.2384,
+      "step": 1493
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3643067846607672e-05,
+      "loss": 0.2855,
+      "step": 1494
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3618485742379547e-05,
+      "loss": 0.2479,
+      "step": 1495
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3593903638151426e-05,
+      "loss": 0.2744,
+      "step": 1496
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3569321533923305e-05,
+      "loss": 0.2184,
+      "step": 1497
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3544739429695184e-05,
+      "loss": 0.2459,
+      "step": 1498
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.3520157325467062e-05,
+      "loss": 0.2134,
+      "step": 1499
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3495575221238938e-05,
+      "loss": 0.2775,
+      "step": 1500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3470993117010816e-05,
+      "loss": 0.2219,
+      "step": 1501
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3446411012782695e-05,
+      "loss": 0.2339,
+      "step": 1502
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3421828908554574e-05,
+      "loss": 0.2821,
+      "step": 1503
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3397246804326453e-05,
+      "loss": 0.2249,
+      "step": 1504
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3372664700098328e-05,
+      "loss": 0.2736,
+      "step": 1505
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.3348082595870207e-05,
+      "loss": 0.2475,
+      "step": 1506
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3323500491642085e-05,
+      "loss": 0.2333,
+      "step": 1507
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3298918387413964e-05,
+      "loss": 0.2511,
+      "step": 1508
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3274336283185843e-05,
+      "loss": 0.2291,
+      "step": 1509
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3249754178957718e-05,
+      "loss": 0.2269,
+      "step": 1510
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3225172074729597e-05,
+      "loss": 0.2412,
+      "step": 1511
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.3200589970501476e-05,
+      "loss": 0.2732,
+      "step": 1512
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3176007866273355e-05,
+      "loss": 0.2554,
+      "step": 1513
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3151425762045233e-05,
+      "loss": 0.232,
+      "step": 1514
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3126843657817109e-05,
+      "loss": 0.2443,
+      "step": 1515
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3102261553588987e-05,
+      "loss": 0.2367,
+      "step": 1516
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3077679449360866e-05,
+      "loss": 0.2041,
+      "step": 1517
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.3053097345132745e-05,
+      "loss": 0.2463,
+      "step": 1518
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.302851524090462e-05,
+      "loss": 0.2255,
+      "step": 1519
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.3003933136676499e-05,
+      "loss": 0.2951,
+      "step": 1520
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2979351032448378e-05,
+      "loss": 0.2327,
+      "step": 1521
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2954768928220256e-05,
+      "loss": 0.3448,
+      "step": 1522
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2930186823992135e-05,
+      "loss": 0.2324,
+      "step": 1523
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.290560471976401e-05,
+      "loss": 0.2398,
+      "step": 1524
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.288102261553589e-05,
+      "loss": 0.2346,
+      "step": 1525
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2856440511307768e-05,
+      "loss": 0.2422,
+      "step": 1526
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2831858407079647e-05,
+      "loss": 0.3009,
+      "step": 1527
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2807276302851526e-05,
+      "loss": 0.2258,
+      "step": 1528
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2782694198623401e-05,
+      "loss": 0.2553,
+      "step": 1529
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.275811209439528e-05,
+      "loss": 0.2812,
+      "step": 1530
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2733529990167158e-05,
+      "loss": 0.2225,
+      "step": 1531
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2708947885939037e-05,
+      "loss": 0.2454,
+      "step": 1532
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2684365781710916e-05,
+      "loss": 0.2885,
+      "step": 1533
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2659783677482793e-05,
+      "loss": 0.2335,
+      "step": 1534
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.263520157325467e-05,
+      "loss": 0.2566,
+      "step": 1535
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2610619469026549e-05,
+      "loss": 0.3034,
+      "step": 1536
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2586037364798428e-05,
+      "loss": 0.2459,
+      "step": 1537
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2561455260570306e-05,
+      "loss": 0.2231,
+      "step": 1538
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2536873156342183e-05,
+      "loss": 0.2317,
+      "step": 1539
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2512291052114062e-05,
+      "loss": 0.2193,
+      "step": 1540
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2487708947885939e-05,
+      "loss": 0.2316,
+      "step": 1541
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2463126843657818e-05,
+      "loss": 0.23,
+      "step": 1542
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2438544739429697e-05,
+      "loss": 0.2746,
+      "step": 1543
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2413962635201574e-05,
+      "loss": 0.233,
+      "step": 1544
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2389380530973452e-05,
+      "loss": 0.2776,
+      "step": 1545
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2364798426745331e-05,
+      "loss": 0.2307,
+      "step": 1546
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2340216322517208e-05,
+      "loss": 0.2391,
+      "step": 1547
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2315634218289087e-05,
+      "loss": 0.2391,
+      "step": 1548
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2291052114060964e-05,
+      "loss": 0.2394,
+      "step": 1549
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2266470009832843e-05,
+      "loss": 0.227,
+      "step": 1550
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2241887905604722e-05,
+      "loss": 0.28,
+      "step": 1551
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2217305801376599e-05,
+      "loss": 0.2608,
+      "step": 1552
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2192723697148477e-05,
+      "loss": 0.2454,
+      "step": 1553
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2168141592920354e-05,
+      "loss": 0.2366,
+      "step": 1554
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2143559488692233e-05,
+      "loss": 0.2367,
+      "step": 1555
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2118977384464112e-05,
+      "loss": 0.229,
+      "step": 1556
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2094395280235989e-05,
+      "loss": 0.248,
+      "step": 1557
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2069813176007868e-05,
+      "loss": 0.2544,
+      "step": 1558
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2045231071779745e-05,
+      "loss": 0.2486,
+      "step": 1559
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2020648967551623e-05,
+      "loss": 0.2335,
+      "step": 1560
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.19960668633235e-05,
+      "loss": 0.2575,
+      "step": 1561
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.197148475909538e-05,
+      "loss": 0.2311,
+      "step": 1562
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.1946902654867258e-05,
+      "loss": 0.2759,
+      "step": 1563
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.1922320550639135e-05,
+      "loss": 0.2308,
+      "step": 1564
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.1897738446411014e-05,
+      "loss": 0.2806,
+      "step": 1565
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.187315634218289e-05,
+      "loss": 0.2535,
+      "step": 1566
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.184857423795477e-05,
+      "loss": 0.2315,
+      "step": 1567
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1823992133726648e-05,
+      "loss": 0.2523,
+      "step": 1568
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1799410029498525e-05,
+      "loss": 0.2524,
+      "step": 1569
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1774827925270404e-05,
+      "loss": 0.2305,
+      "step": 1570
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1750245821042281e-05,
+      "loss": 0.2412,
+      "step": 1571
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.172566371681416e-05,
+      "loss": 0.2115,
+      "step": 1572
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1701081612586037e-05,
+      "loss": 0.3154,
+      "step": 1573
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.1676499508357916e-05,
+      "loss": 0.2475,
+      "step": 1574
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1651917404129794e-05,
+      "loss": 0.2668,
+      "step": 1575
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1627335299901672e-05,
+      "loss": 0.2544,
+      "step": 1576
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.160275319567355e-05,
+      "loss": 0.2225,
+      "step": 1577
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1578171091445427e-05,
+      "loss": 0.2399,
+      "step": 1578
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1553588987217306e-05,
+      "loss": 0.2606,
+      "step": 1579
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1529006882989185e-05,
+      "loss": 0.2352,
+      "step": 1580
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1504424778761062e-05,
+      "loss": 0.2303,
+      "step": 1581
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.147984267453294e-05,
+      "loss": 0.2813,
+      "step": 1582
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1455260570304818e-05,
+      "loss": 0.2538,
+      "step": 1583
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1430678466076696e-05,
+      "loss": 0.236,
+      "step": 1584
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1406096361848573e-05,
+      "loss": 0.3073,
+      "step": 1585
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1381514257620452e-05,
+      "loss": 0.2282,
+      "step": 1586
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1356932153392331e-05,
+      "loss": 0.2466,
+      "step": 1587
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1332350049164208e-05,
+      "loss": 0.2701,
+      "step": 1588
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1307767944936087e-05,
+      "loss": 0.2324,
+      "step": 1589
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1283185840707964e-05,
+      "loss": 0.2466,
+      "step": 1590
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1258603736479843e-05,
+      "loss": 0.2259,
+      "step": 1591
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1234021632251721e-05,
+      "loss": 0.2247,
+      "step": 1592
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.12094395280236e-05,
+      "loss": 0.244,
+      "step": 1593
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1184857423795477e-05,
+      "loss": 0.3768,
+      "step": 1594
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1160275319567356e-05,
+      "loss": 0.2419,
+      "step": 1595
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1135693215339235e-05,
+      "loss": 0.2286,
+      "step": 1596
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.2389,
+      "step": 1597
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.108652900688299e-05,
+      "loss": 0.2296,
+      "step": 1598
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1061946902654869e-05,
+      "loss": 0.243,
+      "step": 1599
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1037364798426746e-05,
+      "loss": 0.232,
+      "step": 1600
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1012782694198625e-05,
+      "loss": 0.2423,
+      "step": 1601
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0988200589970502e-05,
+      "loss": 0.2447,
+      "step": 1602
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.096361848574238e-05,
+      "loss": 0.2764,
+      "step": 1603
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.093903638151426e-05,
+      "loss": 0.2579,
+      "step": 1604
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0914454277286137e-05,
+      "loss": 0.2359,
+      "step": 1605
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0889872173058015e-05,
+      "loss": 0.2488,
+      "step": 1606
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0865290068829892e-05,
+      "loss": 0.2167,
+      "step": 1607
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0840707964601771e-05,
+      "loss": 0.2732,
+      "step": 1608
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0816125860373648e-05,
+      "loss": 0.254,
+      "step": 1609
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0791543756145527e-05,
+      "loss": 0.2348,
+      "step": 1610
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0766961651917406e-05,
+      "loss": 0.245,
+      "step": 1611
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0742379547689283e-05,
+      "loss": 0.2345,
+      "step": 1612
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0717797443461161e-05,
+      "loss": 0.2273,
+      "step": 1613
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0693215339233038e-05,
+      "loss": 0.2059,
+      "step": 1614
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.0668633235004917e-05,
+      "loss": 0.2274,
+      "step": 1615
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0644051130776796e-05,
+      "loss": 0.261,
+      "step": 1616
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0619469026548673e-05,
+      "loss": 0.315,
+      "step": 1617
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0594886922320552e-05,
+      "loss": 0.2269,
+      "step": 1618
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0570304818092429e-05,
+      "loss": 0.2434,
+      "step": 1619
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0545722713864308e-05,
+      "loss": 0.2414,
+      "step": 1620
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0521140609636185e-05,
+      "loss": 0.2781,
+      "step": 1621
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0496558505408063e-05,
+      "loss": 0.2311,
+      "step": 1622
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0471976401179942e-05,
+      "loss": 0.2429,
+      "step": 1623
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.044739429695182e-05,
+      "loss": 0.213,
+      "step": 1624
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0422812192723698e-05,
+      "loss": 0.2236,
+      "step": 1625
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0398230088495575e-05,
+      "loss": 0.2259,
+      "step": 1626
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0373647984267454e-05,
+      "loss": 0.2397,
+      "step": 1627
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0349065880039332e-05,
+      "loss": 0.2423,
+      "step": 1628
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.032448377581121e-05,
+      "loss": 0.224,
+      "step": 1629
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0299901671583088e-05,
+      "loss": 0.239,
+      "step": 1630
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0275319567354965e-05,
+      "loss": 0.2402,
+      "step": 1631
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0250737463126844e-05,
+      "loss": 0.2547,
+      "step": 1632
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0226155358898721e-05,
+      "loss": 0.2488,
+      "step": 1633
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.02015732546706e-05,
+      "loss": 0.2361,
+      "step": 1634
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0176991150442479e-05,
+      "loss": 0.2533,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0152409046214356e-05,
+      "loss": 0.2249,
+      "step": 1636
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0127826941986234e-05,
+      "loss": 0.2284,
+      "step": 1637
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0103244837758111e-05,
+      "loss": 0.2326,
+      "step": 1638
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.007866273352999e-05,
+      "loss": 0.2362,
+      "step": 1639
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0054080629301869e-05,
+      "loss": 0.2313,
+      "step": 1640
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0029498525073746e-05,
+      "loss": 0.2481,
+      "step": 1641
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.0004916420845625e-05,
+      "loss": 0.2269,
+      "step": 1642
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.980334316617502e-06,
+      "loss": 0.2496,
+      "step": 1643
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.95575221238938e-06,
+      "loss": 0.2297,
+      "step": 1644
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.93117010816126e-06,
+      "loss": 0.273,
+      "step": 1645
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.906588003933136e-06,
+      "loss": 0.3092,
+      "step": 1646
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.882005899705015e-06,
+      "loss": 0.273,
+      "step": 1647
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.857423795476894e-06,
+      "loss": 0.2349,
+      "step": 1648
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.832841691248771e-06,
+      "loss": 0.2578,
+      "step": 1649
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.80825958702065e-06,
+      "loss": 0.2458,
+      "step": 1650
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.783677482792528e-06,
+      "loss": 0.2422,
+      "step": 1651
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.759095378564405e-06,
+      "loss": 0.2253,
+      "step": 1652
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.734513274336284e-06,
+      "loss": 0.2327,
+      "step": 1653
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.709931170108163e-06,
+      "loss": 0.2404,
+      "step": 1654
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.68534906588004e-06,
+      "loss": 0.23,
+      "step": 1655
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.660766961651919e-06,
+      "loss": 0.2463,
+      "step": 1656
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.636184857423796e-06,
+      "loss": 0.2551,
+      "step": 1657
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.611602753195675e-06,
+      "loss": 0.2404,
+      "step": 1658
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.587020648967553e-06,
+      "loss": 0.2439,
+      "step": 1659
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.56243854473943e-06,
+      "loss": 0.2484,
+      "step": 1660
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.537856440511309e-06,
+      "loss": 0.3494,
+      "step": 1661
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.513274336283186e-06,
+      "loss": 0.2872,
+      "step": 1662
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.488692232055065e-06,
+      "loss": 0.2489,
+      "step": 1663
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.464110127826944e-06,
+      "loss": 0.232,
+      "step": 1664
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.43952802359882e-06,
+      "loss": 0.2303,
+      "step": 1665
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.4149459193707e-06,
+      "loss": 0.2365,
+      "step": 1666
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.390363815142576e-06,
+      "loss": 0.2319,
+      "step": 1667
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.365781710914455e-06,
+      "loss": 0.2213,
+      "step": 1668
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.341199606686332e-06,
+      "loss": 0.2286,
+      "step": 1669
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.316617502458211e-06,
+      "loss": 0.2331,
+      "step": 1670
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.29203539823009e-06,
+      "loss": 0.2202,
+      "step": 1671
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.267453294001967e-06,
+      "loss": 0.2743,
+      "step": 1672
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.242871189773846e-06,
+      "loss": 0.2138,
+      "step": 1673
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.218289085545723e-06,
+      "loss": 0.2197,
+      "step": 1674
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.193706981317601e-06,
+      "loss": 0.2305,
+      "step": 1675
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 9.16912487708948e-06,
+      "loss": 0.2409,
+      "step": 1676
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.144542772861357e-06,
+      "loss": 0.2248,
+      "step": 1677
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.119960668633236e-06,
+      "loss": 0.2418,
+      "step": 1678
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.095378564405113e-06,
+      "loss": 0.2546,
+      "step": 1679
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.070796460176992e-06,
+      "loss": 0.243,
+      "step": 1680
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.046214355948869e-06,
+      "loss": 0.2307,
+      "step": 1681
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.021632251720747e-06,
+      "loss": 0.2327,
+      "step": 1682
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.997050147492626e-06,
+      "loss": 0.2354,
+      "step": 1683
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.972468043264503e-06,
+      "loss": 0.2135,
+      "step": 1684
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.947885939036382e-06,
+      "loss": 0.2238,
+      "step": 1685
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.923303834808259e-06,
+      "loss": 0.2576,
+      "step": 1686
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.898721730580138e-06,
+      "loss": 0.239,
+      "step": 1687
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.874139626352017e-06,
+      "loss": 0.2627,
+      "step": 1688
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.849557522123894e-06,
+      "loss": 0.2331,
+      "step": 1689
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.824975417895772e-06,
+      "loss": 0.2354,
+      "step": 1690
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.80039331366765e-06,
+      "loss": 0.2057,
+      "step": 1691
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.775811209439528e-06,
+      "loss": 0.2384,
+      "step": 1692
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.751229105211405e-06,
+      "loss": 0.2173,
+      "step": 1693
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.726647000983284e-06,
+      "loss": 0.22,
+      "step": 1694
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.702064896755163e-06,
+      "loss": 0.2278,
+      "step": 1695
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.67748279252704e-06,
+      "loss": 0.2548,
+      "step": 1696
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.652900688298919e-06,
+      "loss": 0.233,
+      "step": 1697
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.628318584070797e-06,
+      "loss": 0.2323,
+      "step": 1698
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.603736479842674e-06,
+      "loss": 0.254,
+      "step": 1699
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.579154375614553e-06,
+      "loss": 0.3021,
+      "step": 1700
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.554572271386432e-06,
+      "loss": 0.226,
+      "step": 1701
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.529990167158309e-06,
+      "loss": 0.2717,
+      "step": 1702
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.505408062930188e-06,
+      "loss": 0.2074,
+      "step": 1703
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 8.480825958702066e-06,
+      "loss": 0.2557,
+      "step": 1704
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.456243854473943e-06,
+      "loss": 0.2506,
+      "step": 1705
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.431661750245822e-06,
+      "loss": 0.2157,
+      "step": 1706
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.407079646017701e-06,
+      "loss": 0.2203,
+      "step": 1707
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.382497541789578e-06,
+      "loss": 0.2342,
+      "step": 1708
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.357915437561457e-06,
+      "loss": 0.2214,
+      "step": 1709
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.2509,
+      "step": 1710
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.308751229105212e-06,
+      "loss": 0.2195,
+      "step": 1711
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.28416912487709e-06,
+      "loss": 0.228,
+      "step": 1712
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.259587020648968e-06,
+      "loss": 0.2412,
+      "step": 1713
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.235004916420847e-06,
+      "loss": 0.2292,
+      "step": 1714
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.210422812192724e-06,
+      "loss": 0.2084,
+      "step": 1715
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.185840707964603e-06,
+      "loss": 0.2386,
+      "step": 1716
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.16125860373648e-06,
+      "loss": 0.2333,
+      "step": 1717
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.136676499508359e-06,
+      "loss": 0.2361,
+      "step": 1718
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.112094395280237e-06,
+      "loss": 0.2264,
+      "step": 1719
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.087512291052114e-06,
+      "loss": 0.2312,
+      "step": 1720
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.062930186823993e-06,
+      "loss": 0.256,
+      "step": 1721
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.03834808259587e-06,
+      "loss": 0.3104,
+      "step": 1722
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.013765978367749e-06,
+      "loss": 0.2353,
+      "step": 1723
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 7.989183874139626e-06,
+      "loss": 0.2332,
+      "step": 1724
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.964601769911505e-06,
+      "loss": 0.2211,
+      "step": 1725
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.940019665683384e-06,
+      "loss": 0.2455,
+      "step": 1726
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.91543756145526e-06,
+      "loss": 0.2272,
+      "step": 1727
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.89085545722714e-06,
+      "loss": 0.2532,
+      "step": 1728
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.866273352999016e-06,
+      "loss": 0.2287,
+      "step": 1729
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.841691248770895e-06,
+      "loss": 0.2192,
+      "step": 1730
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 7.817109144542774e-06,
+      "loss": 0.2246,
+      "step": 1731
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.792527040314651e-06,
+      "loss": 0.2451,
+      "step": 1732
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.76794493608653e-06,
+      "loss": 0.2366,
+      "step": 1733
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.743362831858407e-06,
+      "loss": 0.2692,
+      "step": 1734
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.718780727630285e-06,
+      "loss": 0.2585,
+      "step": 1735
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.694198623402162e-06,
+      "loss": 0.2146,
+      "step": 1736
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.669616519174041e-06,
+      "loss": 0.2338,
+      "step": 1737
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 7.64503441494592e-06,
+      "loss": 0.2406,
+      "step": 1738
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.620452310717798e-06,
+      "loss": 0.2177,
+      "step": 1739
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.595870206489676e-06,
+      "loss": 0.2189,
+      "step": 1740
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.571288102261554e-06,
+      "loss": 0.2427,
+      "step": 1741
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.5467059980334324e-06,
+      "loss": 0.232,
+      "step": 1742
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.52212389380531e-06,
+      "loss": 0.2319,
+      "step": 1743
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.497541789577188e-06,
+      "loss": 0.2537,
+      "step": 1744
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 7.472959685349067e-06,
+      "loss": 0.2491,
+      "step": 1745
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.448377581120944e-06,
+      "loss": 0.2411,
+      "step": 1746
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.423795476892823e-06,
+      "loss": 0.2119,
+      "step": 1747
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.3992133726647e-06,
+      "loss": 0.2526,
+      "step": 1748
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.374631268436579e-06,
+      "loss": 0.2389,
+      "step": 1749
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.350049164208457e-06,
+      "loss": 0.2337,
+      "step": 1750
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.325467059980334e-06,
+      "loss": 0.2381,
+      "step": 1751
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 7.300884955752213e-06,
+      "loss": 0.2509,
+      "step": 1752
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.27630285152409e-06,
+      "loss": 0.233,
+      "step": 1753
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.251720747295969e-06,
+      "loss": 0.2393,
+      "step": 1754
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.227138643067848e-06,
+      "loss": 0.2301,
+      "step": 1755
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.202556538839725e-06,
+      "loss": 0.2435,
+      "step": 1756
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.1779744346116035e-06,
+      "loss": 0.2123,
+      "step": 1757
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 7.1533923303834805e-06,
+      "loss": 0.2299,
+      "step": 1758
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.128810226155359e-06,
+      "loss": 0.2379,
+      "step": 1759
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.104228121927236e-06,
+      "loss": 0.2319,
+      "step": 1760
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.079646017699115e-06,
+      "loss": 0.2412,
+      "step": 1761
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.055063913470994e-06,
+      "loss": 0.2267,
+      "step": 1762
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.030481809242871e-06,
+      "loss": 0.2221,
+      "step": 1763
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 7.00589970501475e-06,
+      "loss": 0.2599,
+      "step": 1764
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 6.9813176007866275e-06,
+      "loss": 0.2241,
+      "step": 1765
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.956735496558505e-06,
+      "loss": 0.2158,
+      "step": 1766
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.932153392330384e-06,
+      "loss": 0.2427,
+      "step": 1767
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.907571288102262e-06,
+      "loss": 0.196,
+      "step": 1768
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.88298918387414e-06,
+      "loss": 0.2197,
+      "step": 1769
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.858407079646018e-06,
+      "loss": 0.229,
+      "step": 1770
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.833824975417897e-06,
+      "loss": 0.2774,
+      "step": 1771
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 6.809242871189774e-06,
+      "loss": 0.2403,
+      "step": 1772
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.784660766961652e-06,
+      "loss": 0.2864,
+      "step": 1773
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.760078662733531e-06,
+      "loss": 0.2167,
+      "step": 1774
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.735496558505408e-06,
+      "loss": 0.2396,
+      "step": 1775
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.710914454277287e-06,
+      "loss": 0.2201,
+      "step": 1776
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.686332350049164e-06,
+      "loss": 0.2116,
+      "step": 1777
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.661750245821043e-06,
+      "loss": 0.2397,
+      "step": 1778
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 6.6371681415929215e-06,
+      "loss": 0.242,
+      "step": 1779
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.6125860373647985e-06,
+      "loss": 0.284,
+      "step": 1780
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.588003933136677e-06,
+      "loss": 0.2365,
+      "step": 1781
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.563421828908554e-06,
+      "loss": 0.2239,
+      "step": 1782
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.538839724680433e-06,
+      "loss": 0.2338,
+      "step": 1783
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.51425762045231e-06,
+      "loss": 0.2846,
+      "step": 1784
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.489675516224189e-06,
+      "loss": 0.2511,
+      "step": 1785
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 6.465093411996068e-06,
+      "loss": 0.2169,
+      "step": 1786
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.440511307767945e-06,
+      "loss": 0.232,
+      "step": 1787
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.415929203539823e-06,
+      "loss": 0.2791,
+      "step": 1788
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.3913470993117005e-06,
+      "loss": 0.2368,
+      "step": 1789
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.366764995083579e-06,
+      "loss": 0.2493,
+      "step": 1790
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.342182890855458e-06,
+      "loss": 0.2149,
+      "step": 1791
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 6.317600786627335e-06,
+      "loss": 0.2408,
+      "step": 1792
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.293018682399214e-06,
+      "loss": 0.2369,
+      "step": 1793
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.268436578171092e-06,
+      "loss": 0.2175,
+      "step": 1794
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.2438544739429696e-06,
+      "loss": 0.2304,
+      "step": 1795
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.219272369714848e-06,
+      "loss": 0.2279,
+      "step": 1796
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.194690265486726e-06,
+      "loss": 0.2253,
+      "step": 1797
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.170108161258604e-06,
+      "loss": 0.2706,
+      "step": 1798
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 6.145526057030482e-06,
+      "loss": 0.2045,
+      "step": 1799
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.120943952802361e-06,
+      "loss": 0.2486,
+      "step": 1800
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.096361848574239e-06,
+      "loss": 0.2241,
+      "step": 1801
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.0717797443461166e-06,
+      "loss": 0.2309,
+      "step": 1802
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.0471976401179945e-06,
+      "loss": 0.2465,
+      "step": 1803
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.022615535889872e-06,
+      "loss": 0.2081,
+      "step": 1804
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.99803343166175e-06,
+      "loss": 0.2198,
+      "step": 1805
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.973451327433629e-06,
+      "loss": 0.2295,
+      "step": 1806
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.948869223205507e-06,
+      "loss": 0.2327,
+      "step": 1807
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.924287118977385e-06,
+      "loss": 0.2367,
+      "step": 1808
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.899705014749263e-06,
+      "loss": 0.2325,
+      "step": 1809
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.875122910521141e-06,
+      "loss": 0.244,
+      "step": 1810
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.8505408062930185e-06,
+      "loss": 0.226,
+      "step": 1811
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.825958702064897e-06,
+      "loss": 0.2207,
+      "step": 1812
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.801376597836775e-06,
+      "loss": 0.239,
+      "step": 1813
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.776794493608653e-06,
+      "loss": 0.252,
+      "step": 1814
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.752212389380531e-06,
+      "loss": 0.2331,
+      "step": 1815
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.727630285152409e-06,
+      "loss": 0.2566,
+      "step": 1816
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.703048180924287e-06,
+      "loss": 0.251,
+      "step": 1817
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.6784660766961655e-06,
+      "loss": 0.2128,
+      "step": 1818
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.653883972468043e-06,
+      "loss": 0.2496,
+      "step": 1819
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.629301868239921e-06,
+      "loss": 0.232,
+      "step": 1820
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.6047197640118e-06,
+      "loss": 0.294,
+      "step": 1821
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.580137659783678e-06,
+      "loss": 0.2293,
+      "step": 1822
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.2318,
+      "step": 1823
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.5309734513274346e-06,
+      "loss": 0.2286,
+      "step": 1824
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.5063913470993125e-06,
+      "loss": 0.2293,
+      "step": 1825
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.48180924287119e-06,
+      "loss": 0.2376,
+      "step": 1826
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.457227138643068e-06,
+      "loss": 0.25,
+      "step": 1827
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.432645034414946e-06,
+      "loss": 0.2078,
+      "step": 1828
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.408062930186824e-06,
+      "loss": 0.237,
+      "step": 1829
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.383480825958703e-06,
+      "loss": 0.2129,
+      "step": 1830
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.358898721730581e-06,
+      "loss": 0.2363,
+      "step": 1831
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.334316617502459e-06,
+      "loss": 0.2851,
+      "step": 1832
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.3097345132743365e-06,
+      "loss": 0.2211,
+      "step": 1833
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.285152409046214e-06,
+      "loss": 0.2273,
+      "step": 1834
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.260570304818092e-06,
+      "loss": 0.2448,
+      "step": 1835
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.235988200589971e-06,
+      "loss": 0.3382,
+      "step": 1836
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.211406096361849e-06,
+      "loss": 0.2834,
+      "step": 1837
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.186823992133727e-06,
+      "loss": 0.2161,
+      "step": 1838
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.162241887905605e-06,
+      "loss": 0.2276,
+      "step": 1839
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.137659783677483e-06,
+      "loss": 0.2089,
+      "step": 1840
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.1130776794493605e-06,
+      "loss": 0.2775,
+      "step": 1841
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.088495575221239e-06,
+      "loss": 0.2244,
+      "step": 1842
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.063913470993117e-06,
+      "loss": 0.2638,
+      "step": 1843
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.039331366764995e-06,
+      "loss": 0.2267,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.014749262536873e-06,
+      "loss": 0.2219,
+      "step": 1845
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.990167158308751e-06,
+      "loss": 0.2513,
+      "step": 1846
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.96558505408063e-06,
+      "loss": 0.2524,
+      "step": 1847
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.9410029498525075e-06,
+      "loss": 0.2384,
+      "step": 1848
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.9164208456243854e-06,
+      "loss": 0.2322,
+      "step": 1849
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.891838741396264e-06,
+      "loss": 0.2269,
+      "step": 1850
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.867256637168142e-06,
+      "loss": 0.2817,
+      "step": 1851
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.84267453294002e-06,
+      "loss": 0.2247,
+      "step": 1852
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.818092428711898e-06,
+      "loss": 0.2388,
+      "step": 1853
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.793510324483777e-06,
+      "loss": 0.2255,
+      "step": 1854
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.7689282202556545e-06,
+      "loss": 0.2309,
+      "step": 1855
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.7443461160275324e-06,
+      "loss": 0.214,
+      "step": 1856
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.71976401179941e-06,
+      "loss": 0.2798,
+      "step": 1857
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.695181907571288e-06,
+      "loss": 0.221,
+      "step": 1858
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.670599803343166e-06,
+      "loss": 0.2283,
+      "step": 1859
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.646017699115045e-06,
+      "loss": 0.2076,
+      "step": 1860
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.621435594886923e-06,
+      "loss": 0.2326,
+      "step": 1861
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.596853490658801e-06,
+      "loss": 0.2133,
+      "step": 1862
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.5722713864306786e-06,
+      "loss": 0.2345,
+      "step": 1863
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.5476892822025565e-06,
+      "loss": 0.2266,
+      "step": 1864
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.523107177974434e-06,
+      "loss": 0.2065,
+      "step": 1865
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.498525073746313e-06,
+      "loss": 0.2135,
+      "step": 1866
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.473942969518191e-06,
+      "loss": 0.2331,
+      "step": 1867
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.449360865290069e-06,
+      "loss": 0.2575,
+      "step": 1868
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.424778761061947e-06,
+      "loss": 0.2695,
+      "step": 1869
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.400196656833825e-06,
+      "loss": 0.2176,
+      "step": 1870
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.375614552605703e-06,
+      "loss": 0.2366,
+      "step": 1871
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.351032448377581e-06,
+      "loss": 0.2344,
+      "step": 1872
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.326450344149459e-06,
+      "loss": 0.221,
+      "step": 1873
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.301868239921337e-06,
+      "loss": 0.2329,
+      "step": 1874
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.277286135693216e-06,
+      "loss": 0.2262,
+      "step": 1875
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.252704031465094e-06,
+      "loss": 0.2452,
+      "step": 1876
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.228121927236972e-06,
+      "loss": 0.2443,
+      "step": 1877
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.2035398230088504e-06,
+      "loss": 0.2299,
+      "step": 1878
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.178957718780728e-06,
+      "loss": 0.2076,
+      "step": 1879
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.154375614552606e-06,
+      "loss": 0.2328,
+      "step": 1880
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.129793510324484e-06,
+      "loss": 0.2278,
+      "step": 1881
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.105211406096362e-06,
+      "loss": 0.2549,
+      "step": 1882
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.08062930186824e-06,
+      "loss": 0.2473,
+      "step": 1883
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.056047197640119e-06,
+      "loss": 0.3125,
+      "step": 1884
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.031465093411997e-06,
+      "loss": 0.2198,
+      "step": 1885
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.0068829891838745e-06,
+      "loss": 0.2956,
+      "step": 1886
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.982300884955752e-06,
+      "loss": 0.2146,
+      "step": 1887
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.95771878072763e-06,
+      "loss": 0.2239,
+      "step": 1888
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.933136676499508e-06,
+      "loss": 0.2447,
+      "step": 1889
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.908554572271387e-06,
+      "loss": 0.2296,
+      "step": 1890
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.883972468043265e-06,
+      "loss": 0.2243,
+      "step": 1891
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.859390363815143e-06,
+      "loss": 0.2176,
+      "step": 1892
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.834808259587021e-06,
+      "loss": 0.239,
+      "step": 1893
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.810226155358899e-06,
+      "loss": 0.2568,
+      "step": 1894
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.785644051130777e-06,
+      "loss": 0.2674,
+      "step": 1895
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.761061946902655e-06,
+      "loss": 0.2171,
+      "step": 1896
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.7364798426745335e-06,
+      "loss": 0.2311,
+      "step": 1897
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.7118977384464114e-06,
+      "loss": 0.2594,
+      "step": 1898
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.6873156342182893e-06,
+      "loss": 0.2251,
+      "step": 1899
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.662733529990167e-06,
+      "loss": 0.2732,
+      "step": 1900
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.638151425762045e-06,
+      "loss": 0.266,
+      "step": 1901
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 3.613569321533924e-06,
+      "loss": 0.2274,
+      "step": 1902
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5889872173058017e-06,
+      "loss": 0.2111,
+      "step": 1903
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5644051130776796e-06,
+      "loss": 0.2085,
+      "step": 1904
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5398230088495575e-06,
+      "loss": 0.2466,
+      "step": 1905
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.5152409046214354e-06,
+      "loss": 0.2361,
+      "step": 1906
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.4906588003933138e-06,
+      "loss": 0.2322,
+      "step": 1907
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.466076696165192e-06,
+      "loss": 0.2415,
+      "step": 1908
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 3.44149459193707e-06,
+      "loss": 0.23,
+      "step": 1909
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.4169124877089483e-06,
+      "loss": 0.2263,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.392330383480826e-06,
+      "loss": 0.2302,
+      "step": 1911
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.367748279252704e-06,
+      "loss": 0.2055,
+      "step": 1912
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.343166175024582e-06,
+      "loss": 0.2392,
+      "step": 1913
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.3185840707964607e-06,
+      "loss": 0.2159,
+      "step": 1914
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 3.2940019665683386e-06,
+      "loss": 0.2159,
+      "step": 1915
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.2694198623402165e-06,
+      "loss": 0.2456,
+      "step": 1916
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.2448377581120944e-06,
+      "loss": 0.2232,
+      "step": 1917
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.2202556538839723e-06,
+      "loss": 0.2457,
+      "step": 1918
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.1956735496558502e-06,
+      "loss": 0.2348,
+      "step": 1919
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.171091445427729e-06,
+      "loss": 0.2308,
+      "step": 1920
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.146509341199607e-06,
+      "loss": 0.214,
+      "step": 1921
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 3.1219272369714848e-06,
+      "loss": 0.2171,
+      "step": 1922
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.097345132743363e-06,
+      "loss": 0.2493,
+      "step": 1923
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.072763028515241e-06,
+      "loss": 0.2275,
+      "step": 1924
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.0481809242871193e-06,
+      "loss": 0.2223,
+      "step": 1925
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 3.0235988200589972e-06,
+      "loss": 0.2339,
+      "step": 1926
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.999016715830875e-06,
+      "loss": 0.2178,
+      "step": 1927
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.9744346116027534e-06,
+      "loss": 0.2155,
+      "step": 1928
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.9498525073746313e-06,
+      "loss": 0.2166,
+      "step": 1929
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.9252704031465092e-06,
+      "loss": 0.2071,
+      "step": 1930
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.9006882989183876e-06,
+      "loss": 0.2183,
+      "step": 1931
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.8761061946902655e-06,
+      "loss": 0.218,
+      "step": 1932
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.8515240904621434e-06,
+      "loss": 0.2589,
+      "step": 1933
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.8269419862340217e-06,
+      "loss": 0.2276,
+      "step": 1934
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.8023598820059e-06,
+      "loss": 0.2397,
+      "step": 1935
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 0.2091,
+      "step": 1936
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.7531956735496562e-06,
+      "loss": 0.2278,
+      "step": 1937
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.728613569321534e-06,
+      "loss": 0.2428,
+      "step": 1938
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.704031465093412e-06,
+      "loss": 0.2623,
+      "step": 1939
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.6794493608652904e-06,
+      "loss": 0.2741,
+      "step": 1940
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.6548672566371683e-06,
+      "loss": 0.2374,
+      "step": 1941
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.630285152409046e-06,
+      "loss": 0.2217,
+      "step": 1942
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.6057030481809245e-06,
+      "loss": 0.2387,
+      "step": 1943
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.5811209439528024e-06,
+      "loss": 0.2776,
+      "step": 1944
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.5565388397246803e-06,
+      "loss": 0.2324,
+      "step": 1945
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.5319567354965586e-06,
+      "loss": 0.2239,
+      "step": 1946
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.5073746312684365e-06,
+      "loss": 0.229,
+      "step": 1947
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.482792527040315e-06,
+      "loss": 0.225,
+      "step": 1948
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.4582104228121927e-06,
+      "loss": 0.2191,
+      "step": 1949
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.433628318584071e-06,
+      "loss": 0.2237,
+      "step": 1950
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.409046214355949e-06,
+      "loss": 0.2416,
+      "step": 1951
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.3844641101278273e-06,
+      "loss": 0.2449,
+      "step": 1952
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.359882005899705e-06,
+      "loss": 0.2182,
+      "step": 1953
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.335299901671583e-06,
+      "loss": 0.2323,
+      "step": 1954
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.3107177974434614e-06,
+      "loss": 0.2399,
+      "step": 1955
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.2861356932153393e-06,
+      "loss": 0.224,
+      "step": 1956
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.261553588987217e-06,
+      "loss": 0.2265,
+      "step": 1957
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.2369714847590955e-06,
+      "loss": 0.2487,
+      "step": 1958
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.2123893805309734e-06,
+      "loss": 0.2531,
+      "step": 1959
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.1878072763028513e-06,
+      "loss": 0.2251,
+      "step": 1960
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.1632251720747296e-06,
+      "loss": 0.2158,
+      "step": 1961
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.138643067846608e-06,
+      "loss": 0.2417,
+      "step": 1962
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.114060963618486e-06,
+      "loss": 0.2193,
+      "step": 1963
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.089478859390364e-06,
+      "loss": 0.252,
+      "step": 1964
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.064896755162242e-06,
+      "loss": 0.2248,
+      "step": 1965
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.04031465093412e-06,
+      "loss": 0.2242,
+      "step": 1966
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.0157325467059983e-06,
+      "loss": 0.2504,
+      "step": 1967
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.991150442477876e-06,
+      "loss": 0.2247,
+      "step": 1968
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.966568338249754e-06,
+      "loss": 0.2381,
+      "step": 1969
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.9419862340216324e-06,
+      "loss": 0.2211,
+      "step": 1970
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.9174041297935103e-06,
+      "loss": 0.2244,
+      "step": 1971
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8928220255653884e-06,
+      "loss": 0.2354,
+      "step": 1972
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8682399213372667e-06,
+      "loss": 0.238,
+      "step": 1973
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8436578171091446e-06,
+      "loss": 0.2114,
+      "step": 1974
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.8190757128810225e-06,
+      "loss": 0.2196,
+      "step": 1975
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.7944936086529009e-06,
+      "loss": 0.212,
+      "step": 1976
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.7699115044247788e-06,
+      "loss": 0.2233,
+      "step": 1977
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.7453294001966569e-06,
+      "loss": 0.221,
+      "step": 1978
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.720747295968535e-06,
+      "loss": 0.21,
+      "step": 1979
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.696165191740413e-06,
+      "loss": 0.2334,
+      "step": 1980
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.671583087512291e-06,
+      "loss": 0.2196,
+      "step": 1981
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.6470009832841693e-06,
+      "loss": 0.2331,
+      "step": 1982
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.6224188790560472e-06,
+      "loss": 0.2388,
+      "step": 1983
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.5978367748279251e-06,
+      "loss": 0.2486,
+      "step": 1984
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5732546705998034e-06,
+      "loss": 0.2205,
+      "step": 1985
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5486725663716816e-06,
+      "loss": 0.2609,
+      "step": 1986
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.5240904621435597e-06,
+      "loss": 0.2199,
+      "step": 1987
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4995083579154376e-06,
+      "loss": 0.2357,
+      "step": 1988
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4749262536873157e-06,
+      "loss": 0.2243,
+      "step": 1989
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4503441494591938e-06,
+      "loss": 0.2089,
+      "step": 1990
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.4257620452310717e-06,
+      "loss": 0.2733,
+      "step": 1991
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.40117994100295e-06,
+      "loss": 0.2318,
+      "step": 1992
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3765978367748281e-06,
+      "loss": 0.2249,
+      "step": 1993
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.352015732546706e-06,
+      "loss": 0.2202,
+      "step": 1994
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3274336283185841e-06,
+      "loss": 0.2319,
+      "step": 1995
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.3028515240904622e-06,
+      "loss": 0.299,
+      "step": 1996
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.2782694198623401e-06,
+      "loss": 0.2314,
+      "step": 1997
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2536873156342182e-06,
+      "loss": 0.2238,
+      "step": 1998
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2291052114060964e-06,
+      "loss": 0.2227,
+      "step": 1999
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.2045231071779745e-06,
+      "loss": 0.2241,
+      "step": 2000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.1799410029498526e-06,
+      "loss": 0.2228,
+      "step": 2001
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.1553588987217307e-06,
+      "loss": 0.2889,
+      "step": 2002
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.1307767944936086e-06,
+      "loss": 0.3582,
+      "step": 2003
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.1061946902654867e-06,
+      "loss": 0.3018,
+      "step": 2004
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0816125860373648e-06,
+      "loss": 0.2843,
+      "step": 2005
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.057030481809243e-06,
+      "loss": 0.2285,
+      "step": 2006
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.032448377581121e-06,
+      "loss": 0.2033,
+      "step": 2007
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0078662733529991e-06,
+      "loss": 0.2079,
+      "step": 2008
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.83284169124877e-07,
+      "loss": 0.2237,
+      "step": 2009
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.587020648967552e-07,
+      "loss": 0.2276,
+      "step": 2010
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.341199606686334e-07,
+      "loss": 0.2337,
+      "step": 2011
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.095378564405113e-07,
+      "loss": 0.236,
+      "step": 2012
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.849557522123894e-07,
+      "loss": 0.24,
+      "step": 2013
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.603736479842675e-07,
+      "loss": 0.2369,
+      "step": 2014
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.357915437561455e-07,
+      "loss": 0.2039,
+      "step": 2015
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.112094395280236e-07,
+      "loss": 0.247,
+      "step": 2016
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 7.866273352999017e-07,
+      "loss": 0.2372,
+      "step": 2017
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 7.620452310717798e-07,
+      "loss": 0.2288,
+      "step": 2018
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.374631268436578e-07,
+      "loss": 0.2341,
+      "step": 2019
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.128810226155358e-07,
+      "loss": 0.2359,
+      "step": 2020
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.882989183874141e-07,
+      "loss": 0.2194,
+      "step": 2021
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.637168141592921e-07,
+      "loss": 0.236,
+      "step": 2022
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.391347099311701e-07,
+      "loss": 0.216,
+      "step": 2023
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.145526057030482e-07,
+      "loss": 0.2284,
+      "step": 2024
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.899705014749263e-07,
+      "loss": 0.2152,
+      "step": 2025
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.653883972468043e-07,
+      "loss": 0.2257,
+      "step": 2026
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.408062930186824e-07,
+      "loss": 0.2362,
+      "step": 2027
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.162241887905605e-07,
+      "loss": 0.2279,
+      "step": 2028
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.916420845624385e-07,
+      "loss": 0.225,
+      "step": 2029
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.670599803343167e-07,
+      "loss": 0.239,
+      "step": 2030
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.424778761061947e-07,
+      "loss": 0.2448,
+      "step": 2031
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.1789577187807275e-07,
+      "loss": 0.2039,
+      "step": 2032
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.9331366764995086e-07,
+      "loss": 0.2316,
+      "step": 2033
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.687315634218289e-07,
+      "loss": 0.2652,
+      "step": 2034
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.4414945919370703e-07,
+      "loss": 0.2122,
+      "step": 2035
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.1956735496558503e-07,
+      "loss": 0.2185,
+      "step": 2036
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.9498525073746315e-07,
+      "loss": 0.2102,
+      "step": 2037
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.704031465093412e-07,
+      "loss": 0.2479,
+      "step": 2038
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.4582104228121926e-07,
+      "loss": 0.2113,
+      "step": 2039
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.2123893805309735e-07,
+      "loss": 0.2732,
+      "step": 2040
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.9665683382497543e-07,
+      "loss": 0.2202,
+      "step": 2041
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.7207472959685351e-07,
+      "loss": 0.2155,
+      "step": 2042
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.4749262536873157e-07,
+      "loss": 0.2358,
+      "step": 2043
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2291052114060963e-07,
+      "loss": 0.2521,
+      "step": 2044
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.832841691248772e-08,
+      "loss": 0.2373,
+      "step": 2045
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.374631268436579e-08,
+      "loss": 0.2807,
+      "step": 2046
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.916420845624386e-08,
+      "loss": 0.2181,
+      "step": 2047
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.458210422812193e-08,
+      "loss": 0.2248,
+      "step": 2048
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.2167,
+      "step": 2049
+    },
+    {
+      "epoch": 3.0,
+      "step": 2049,
+      "total_flos": 4.429251995854543e+19,
+      "train_loss": 0.30861409058682565,
+      "train_runtime": 65041.8761,
+      "train_samples_per_second": 8.065,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2049,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500000,
+  "total_flos": 4.429251995854543e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}