diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.02880658436214,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 0.7979,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.781035935540111e-06,
+      "loss": 0.837,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8228751702678988e-06,
+      "loss": 0.8691,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.562071871080222e-06,
+      "loss": 0.8361,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.135437376734581e-06,
+      "loss": 0.7944,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.603911105808009e-06,
+      "loss": 0.8049,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 0.8178,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.343107806620333e-06,
+      "loss": 0.8495,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.6457503405357975e-06,
+      "loss": 0.8288,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.916473312274693e-06,
+      "loss": 0.8166,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.161372029336719e-06,
+      "loss": 0.8196,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.384947041348121e-06,
+      "loss": 0.8146,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.590616115309205e-06,
+      "loss": 0.7934,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.781035935540111e-06,
+      "loss": 0.7958,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.958312547002481e-06,
+      "loss": 0.7876,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.124143742160444e-06,
+      "loss": 0.7363,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.279918205451739e-06,
+      "loss": 0.7644,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.426786276075908e-06,
+      "loss": 0.7935,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.5657115530125735e-06,
+      "loss": 0.7577,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.697509247814803e-06,
+      "loss": 0.7831,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.8228751702679e-06,
+      "loss": 0.746,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.94240796487683e-06,
+      "loss": 0.7343,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.056626400379656e-06,
+      "loss": 0.7205,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.165982976888233e-06,
+      "loss": 0.7139,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.270874753469162e-06,
+      "loss": 0.7135,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.371652050849318e-06,
+      "loss": 0.6998,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.468625510803695e-06,
+      "loss": 0.6967,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.562071871080222e-06,
+      "loss": 0.7194,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.652238726493115e-06,
+      "loss": 0.724,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.73934848254259e-06,
+      "loss": 0.6683,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.823601660519256e-06,
+      "loss": 0.6822,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.905179677700556e-06,
+      "loss": 0.6887,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.984247199604618e-06,
+      "loss": 0.6723,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.06095414099185e-06,
+      "loss": 0.7092,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.135437376734582e-06,
+      "loss": 0.6577,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.207822211616019e-06,
+      "loss": 0.6642,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.278223648705538e-06,
+      "loss": 0.6476,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.346747488552685e-06,
+      "loss": 0.647,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.413491285577103e-06,
+      "loss": 0.6712,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.478545183354915e-06,
+      "loss": 0.6348,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.541992646749767e-06,
+      "loss": 0.6164,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.60391110580801e-06,
+      "loss": 0.6321,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.664372523879183e-06,
+      "loss": 0.6181,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.723443900416943e-06,
+      "loss": 0.6219,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.781187717270378e-06,
+      "loss": 0.624,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.837662335919769e-06,
+      "loss": 0.5858,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.892922351988353e-06,
+      "loss": 0.5858,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947018912428344e-06,
+      "loss": 0.5807,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.5795,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.5921,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.6104,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.5226,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.5683,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5351,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5291,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5551,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5426,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5426,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5179,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5203,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5135,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5002,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.5012,
+      "step": 63
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4854,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.5138,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4956,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4943,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4952,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4955,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4676,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4624,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.4548,
+      "step": 72
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4619,
+      "step": 73
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4641,
+      "step": 74
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4311,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.445,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4585,
+      "step": 77
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4624,
+      "step": 78
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4096,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4414,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4439,
+      "step": 81
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.4254,
+      "step": 82
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4059,
+      "step": 83
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4201,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4256,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4119,
+      "step": 86
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.41,
+      "step": 87
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.393,
+      "step": 88
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4011,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.4015,
+      "step": 90
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.3899,
+      "step": 91
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.3961,
+      "step": 92
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4019,
+      "step": 93
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3906,
+      "step": 94
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3788,
+      "step": 95
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3959,
+      "step": 96
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3911,
+      "step": 97
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3991,
+      "step": 98
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.4025,
+      "step": 99
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3915,
+      "step": 100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3597,
+      "step": 101
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.3807,
+      "step": 102
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3746,
+      "step": 103
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3731,
+      "step": 104
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3653,
+      "step": 105
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3973,
+      "step": 106
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3616,
+      "step": 107
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3783,
+      "step": 108
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.392,
+      "step": 109
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3904,
+      "step": 110
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3874,
+      "step": 111
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3473,
+      "step": 112
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3705,
+      "step": 113
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3567,
+      "step": 114
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3526,
+      "step": 115
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3582,
+      "step": 116
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.364,
+      "step": 117
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3869,
+      "step": 118
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3456,
+      "step": 119
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3663,
+      "step": 120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.3556,
+      "step": 121
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3641,
+      "step": 122
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3185,
+      "step": 123
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3776,
+      "step": 124
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.342,
+      "step": 125
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3457,
+      "step": 126
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3211,
+      "step": 127
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3432,
+      "step": 128
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3755,
+      "step": 129
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3689,
+      "step": 130
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.3486,
+      "step": 131
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3658,
+      "step": 132
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3382,
+      "step": 133
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3281,
+      "step": 134
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3648,
+      "step": 135
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3584,
+      "step": 136
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3401,
+      "step": 137
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3261,
+      "step": 138
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3512,
+      "step": 139
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.3422,
+      "step": 140
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3526,
+      "step": 141
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3176,
+      "step": 142
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3644,
+      "step": 143
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3321,
+      "step": 144
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3514,
+      "step": 145
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3383,
+      "step": 146
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3413,
+      "step": 147
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3489,
+      "step": 148
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3138,
+      "step": 149
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.345,
+      "step": 150
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3354,
+      "step": 151
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3315,
+      "step": 152
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3347,
+      "step": 153
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3266,
+      "step": 154
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3396,
+      "step": 155
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3415,
+      "step": 156
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3145,
+      "step": 157
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3248,
+      "step": 158
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3414,
+      "step": 159
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.3236,
+      "step": 160
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3415,
+      "step": 161
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3201,
+      "step": 162
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3237,
+      "step": 163
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3401,
+      "step": 164
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3169,
+      "step": 165
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.343,
+      "step": 166
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.341,
+      "step": 167
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3089,
+      "step": 168
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3455,
+      "step": 169
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.3592,
+      "step": 170
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3434,
+      "step": 171
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3209,
+      "step": 172
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3249,
+      "step": 173
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3387,
+      "step": 174
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3381,
+      "step": 175
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3187,
+      "step": 176
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3288,
+      "step": 177
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3013,
+      "step": 178
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.3254,
+      "step": 179
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3152,
+      "step": 180
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3201,
+      "step": 181
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3142,
+      "step": 182
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3196,
+      "step": 183
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3082,
+      "step": 184
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.2923,
+      "step": 185
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.2982,
+      "step": 186
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3376,
+      "step": 187
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.3418,
+      "step": 188
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.2858,
+      "step": 189
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3023,
+      "step": 190
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3127,
+      "step": 191
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3141,
+      "step": 192
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.2992,
+      "step": 193
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3259,
+      "step": 194
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3212,
+      "step": 195
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3287,
+      "step": 196
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.2984,
+      "step": 197
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.3114,
+      "step": 198
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.307,
+      "step": 199
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3002,
+      "step": 200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3127,
+      "step": 201
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3135,
+      "step": 202
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3173,
+      "step": 203
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3102,
+      "step": 204
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3267,
+      "step": 205
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3202,
+      "step": 206
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3341,
+      "step": 207
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.3112,
+      "step": 208
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3116,
+      "step": 209
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3113,
+      "step": 210
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3174,
+      "step": 211
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3199,
+      "step": 212
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3333,
+      "step": 213
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.2887,
+      "step": 214
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3124,
+      "step": 215
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3125,
+      "step": 216
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3122,
+      "step": 217
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.3086,
+      "step": 218
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.3058,
+      "step": 219
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.2943,
+      "step": 220
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.2954,
+      "step": 221
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.3137,
+      "step": 222
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.2931,
+      "step": 223
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.2925,
+      "step": 224
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.3081,
+      "step": 225
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.323,
+      "step": 226
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.307,
+      "step": 227
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.3096,
+      "step": 228
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.3089,
+      "step": 229
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.2965,
+      "step": 230
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.3042,
+      "step": 231
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.2873,
+      "step": 232
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.28,
+      "step": 233
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.2956,
+      "step": 234
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.2875,
+      "step": 235
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.3204,
+      "step": 236
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.289,
+      "step": 237
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.3079,
+      "step": 238
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.2955,
+      "step": 239
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.2922,
+      "step": 240
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.3248,
+      "step": 241
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.2899,
+      "step": 242
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.3018,
+      "step": 243
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.2893,
+      "step": 244
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.3034,
+      "step": 245
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.27,
+      "step": 246
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.3036,
+      "step": 247
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.2835,
+      "step": 248
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3442,
+      "step": 249
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.291,
+      "step": 250
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3169,
+      "step": 251
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3033,
+      "step": 252
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.3032,
+      "step": 253
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.2779,
+      "step": 254
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.2794,
+      "step": 255
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.2985,
+      "step": 256
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.2956,
+      "step": 257
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2927,
+      "step": 258
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2736,
+      "step": 259
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2861,
+      "step": 260
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3033,
+      "step": 261
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2963,
+      "step": 262
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2852,
+      "step": 263
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3027,
+      "step": 264
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.2889,
+      "step": 265
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.31,
+      "step": 266
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.3064,
+      "step": 267
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2735,
+      "step": 268
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2826,
+      "step": 269
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.29,
+      "step": 270
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.289,
+      "step": 271
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2854,
+      "step": 272
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3029,
+      "step": 273
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2879,
+      "step": 274
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2894,
+      "step": 275
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.3017,
+      "step": 276
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.2922,
+      "step": 277
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.3098,
+      "step": 278
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2926,
+      "step": 279
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2661,
+      "step": 280
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.282,
+      "step": 281
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2752,
+      "step": 282
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2928,
+      "step": 283
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2696,
+      "step": 284
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2817,
+      "step": 285
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.2967,
+      "step": 286
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2946,
+      "step": 287
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2953,
+      "step": 288
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2899,
+      "step": 289
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2888,
+      "step": 290
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2641,
+      "step": 291
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2852,
+      "step": 292
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.292,
+      "step": 293
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2837,
+      "step": 294
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.275,
+      "step": 295
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.2859,
+      "step": 296
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2806,
+      "step": 297
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3106,
+      "step": 298
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2569,
+      "step": 299
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2908,
+      "step": 300
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.3007,
+      "step": 301
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2834,
+      "step": 302
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2811,
+      "step": 303
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2772,
+      "step": 304
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2986,
+      "step": 305
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.2654,
+      "step": 306
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2892,
+      "step": 307
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2783,
+      "step": 308
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2843,
+      "step": 309
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2847,
+      "step": 310
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2639,
+      "step": 311
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2781,
+      "step": 312
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2669,
+      "step": 313
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.2705,
+      "step": 314
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.3079,
+      "step": 315
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2478,
+      "step": 316
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.294,
+      "step": 317
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2735,
+      "step": 318
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2851,
+      "step": 319
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2798,
+      "step": 320
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2771,
+      "step": 321
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2792,
+      "step": 322
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2843,
+      "step": 323
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2917,
+      "step": 324
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2984,
+      "step": 325
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2783,
+      "step": 326
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2911,
+      "step": 327
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2912,
+      "step": 328
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2701,
+      "step": 329
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2984,
+      "step": 330
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2734,
+      "step": 331
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2636,
+      "step": 332
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2815,
+      "step": 333
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2729,
+      "step": 334
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2849,
+      "step": 335
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2789,
+      "step": 336
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2737,
+      "step": 337
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2668,
+      "step": 338
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.273,
+      "step": 339
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.285,
+      "step": 340
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2847,
+      "step": 341
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2941,
+      "step": 342
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2846,
+      "step": 343
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2742,
+      "step": 344
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2712,
+      "step": 345
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2847,
+      "step": 346
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.27,
+      "step": 347
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2875,
+      "step": 348
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2752,
+      "step": 349
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2642,
+      "step": 350
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2853,
+      "step": 351
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2753,
+      "step": 352
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2775,
+      "step": 353
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.2859,
+      "step": 354
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2832,
+      "step": 355
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2862,
+      "step": 356
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2691,
+      "step": 357
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2751,
+      "step": 358
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2777,
+      "step": 359
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2772,
+      "step": 360
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2731,
+      "step": 361
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2727,
+      "step": 362
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2712,
+      "step": 363
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.2709,
+      "step": 364
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2769,
+      "step": 365
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2747,
+      "step": 366
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.3117,
+      "step": 367
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2668,
+      "step": 368
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2824,
+      "step": 369
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2697,
+      "step": 370
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2739,
+      "step": 371
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.271,
+      "step": 372
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.275,
+      "step": 373
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.2672,
+      "step": 374
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2621,
+      "step": 375
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2721,
+      "step": 376
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2515,
+      "step": 377
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2911,
+      "step": 378
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2613,
+      "step": 379
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2893,
+      "step": 380
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2548,
+      "step": 381
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2687,
+      "step": 382
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2777,
+      "step": 383
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2702,
+      "step": 384
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2696,
+      "step": 385
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2621,
+      "step": 386
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2805,
+      "step": 387
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2788,
+      "step": 388
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.251,
+      "step": 389
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2638,
+      "step": 390
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2694,
+      "step": 391
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2742,
+      "step": 392
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.2861,
+      "step": 393
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2588,
+      "step": 394
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2796,
+      "step": 395
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2749,
+      "step": 396
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2605,
+      "step": 397
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.272,
+      "step": 398
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2897,
+      "step": 399
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2907,
+      "step": 400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2586,
+      "step": 401
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2716,
+      "step": 402
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2676,
+      "step": 403
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2899,
+      "step": 404
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.267,
+      "step": 405
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2746,
+      "step": 406
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2452,
+      "step": 407
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2591,
+      "step": 408
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2573,
+      "step": 409
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2729,
+      "step": 410
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2821,
+      "step": 411
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2504,
+      "step": 412
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2886,
+      "step": 413
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2853,
+      "step": 414
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2527,
+      "step": 415
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2672,
+      "step": 416
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2577,
+      "step": 417
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2646,
+      "step": 418
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.27,
+      "step": 419
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2586,
+      "step": 420
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2624,
+      "step": 421
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.2563,
+      "step": 422
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2535,
+      "step": 423
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2651,
+      "step": 424
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2438,
+      "step": 425
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2606,
+      "step": 426
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2694,
+      "step": 427
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2884,
+      "step": 428
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2693,
+      "step": 429
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.266,
+      "step": 430
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2922,
+      "step": 431
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.2787,
+      "step": 432
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2703,
+      "step": 433
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2754,
+      "step": 434
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.25,
+      "step": 435
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2686,
+      "step": 436
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2823,
+      "step": 437
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2745,
+      "step": 438
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2535,
+      "step": 439
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2647,
+      "step": 440
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2745,
+      "step": 441
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.2644,
+      "step": 442
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2672,
+      "step": 443
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2497,
+      "step": 444
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2597,
+      "step": 445
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2913,
+      "step": 446
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2741,
+      "step": 447
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2591,
+      "step": 448
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2562,
+      "step": 449
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.2663,
+      "step": 450
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.244,
+      "step": 451
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.281,
+      "step": 452
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2614,
+      "step": 453
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2644,
+      "step": 454
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2553,
+      "step": 455
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2788,
+      "step": 456
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2663,
+      "step": 457
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.261,
+      "step": 458
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2952,
+      "step": 459
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2701,
+      "step": 460
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.2554,
+      "step": 461
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2461,
+      "step": 462
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2652,
+      "step": 463
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2601,
+      "step": 464
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2641,
+      "step": 465
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2622,
+      "step": 466
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.263,
+      "step": 467
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2815,
+      "step": 468
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2839,
+      "step": 469
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2567,
+      "step": 470
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.2742,
+      "step": 471
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2642,
+      "step": 472
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2677,
+      "step": 473
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2507,
+      "step": 474
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2686,
+      "step": 475
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2504,
+      "step": 476
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2661,
+      "step": 477
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2839,
+      "step": 478
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2635,
+      "step": 479
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.291,
+      "step": 480
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2627,
+      "step": 481
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2552,
+      "step": 482
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2432,
+      "step": 483
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2667,
+      "step": 484
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2501,
+      "step": 485
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.236,
+      "step": 486
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2419,
+      "step": 487
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2592,
+      "step": 488
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.253,
+      "step": 489
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.2625,
+      "step": 490
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.256,
+      "step": 491
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2427,
+      "step": 492
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2615,
+      "step": 493
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2683,
+      "step": 494
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2799,
+      "step": 495
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2552,
+      "step": 496
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2462,
+      "step": 497
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2428,
+      "step": 498
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2395,
+      "step": 499
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2384,
+      "step": 501
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2739,
+      "step": 502
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2792,
+      "step": 503
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2562,
+      "step": 504
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2451,
+      "step": 505
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2518,
+      "step": 506
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2658,
+      "step": 507
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2574,
+      "step": 508
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2606,
+      "step": 509
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2448,
+      "step": 510
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2752,
+      "step": 511
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2628,
+      "step": 512
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2467,
+      "step": 513
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2618,
+      "step": 514
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.249,
+      "step": 515
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2582,
+      "step": 516
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.288,
+      "step": 517
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2567,
+      "step": 518
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 519
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.2596,
+      "step": 520
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2515,
+      "step": 521
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2604,
+      "step": 522
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2632,
+      "step": 523
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2699,
+      "step": 524
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2748,
+      "step": 525
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2639,
+      "step": 526
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2679,
+      "step": 527
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2457,
+      "step": 528
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2639,
+      "step": 529
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.26,
+      "step": 530
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2579,
+      "step": 531
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2609,
+      "step": 532
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2641,
+      "step": 533
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2789,
+      "step": 534
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2628,
+      "step": 535
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2589,
+      "step": 536
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2463,
+      "step": 537
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2539,
+      "step": 538
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.2498,
+      "step": 539
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2594,
+      "step": 540
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2491,
+      "step": 541
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2292,
+      "step": 542
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2732,
+      "step": 543
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2656,
+      "step": 544
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2585,
+      "step": 545
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2303,
+      "step": 546
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2513,
+      "step": 547
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2487,
+      "step": 548
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 549
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2556,
+      "step": 550
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2648,
+      "step": 551
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2568,
+      "step": 552
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2476,
+      "step": 553
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2552,
+      "step": 554
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.244,
+      "step": 555
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2474,
+      "step": 556
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2675,
+      "step": 557
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.2606,
+      "step": 558
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2618,
+      "step": 559
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2447,
+      "step": 560
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2384,
+      "step": 561
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.265,
+      "step": 562
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2576,
+      "step": 563
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2465,
+      "step": 564
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2527,
+      "step": 565
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2321,
+      "step": 566
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2619,
+      "step": 567
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.2614,
+      "step": 568
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.256,
+      "step": 569
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 570
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2497,
+      "step": 571
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2436,
+      "step": 572
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2407,
+      "step": 573
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2472,
+      "step": 574
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2424,
+      "step": 575
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2501,
+      "step": 576
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2558,
+      "step": 577
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.2547,
+      "step": 578
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2778,
+      "step": 579
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2585,
+      "step": 580
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.25,
+      "step": 581
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2537,
+      "step": 582
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2685,
+      "step": 583
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2577,
+      "step": 584
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.239,
+      "step": 585
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2481,
+      "step": 586
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2491,
+      "step": 587
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.264,
+      "step": 588
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2611,
+      "step": 589
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2626,
+      "step": 590
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 591
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2378,
+      "step": 592
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2594,
+      "step": 593
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2414,
+      "step": 594
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2707,
+      "step": 595
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2395,
+      "step": 596
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.2456,
+      "step": 597
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2589,
+      "step": 598
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2422,
+      "step": 599
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.259,
+      "step": 600
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2552,
+      "step": 601
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2673,
+      "step": 602
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2367,
+      "step": 603
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 604
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2451,
+      "step": 605
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2633,
+      "step": 606
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.2671,
+      "step": 607
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2515,
+      "step": 608
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.234,
+      "step": 609
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2525,
+      "step": 610
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.261,
+      "step": 611
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2483,
+      "step": 612
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2364,
+      "step": 613
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2588,
+      "step": 614
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2723,
+      "step": 615
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2564,
+      "step": 616
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.2556,
+      "step": 617
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.26,
+      "step": 618
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.264,
+      "step": 619
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2565,
+      "step": 620
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2285,
+      "step": 621
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.264,
+      "step": 622
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2429,
+      "step": 623
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2426,
+      "step": 624
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2514,
+      "step": 625
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.2558,
+      "step": 626
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2393,
+      "step": 627
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2578,
+      "step": 628
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2276,
+      "step": 629
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2563,
+      "step": 630
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2557,
+      "step": 631
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2603,
+      "step": 632
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.239,
+      "step": 633
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2439,
+      "step": 634
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.243,
+      "step": 635
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.2482,
+      "step": 636
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2433,
+      "step": 637
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2457,
+      "step": 638
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.249,
+      "step": 639
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 640
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.262,
+      "step": 641
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2524,
+      "step": 642
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2438,
+      "step": 643
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2332,
+      "step": 644
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2561,
+      "step": 645
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.2319,
+      "step": 646
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2443,
+      "step": 647
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2392,
+      "step": 648
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.249,
+      "step": 649
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.267,
+      "step": 650
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.244,
+      "step": 651
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2371,
+      "step": 652
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2575,
+      "step": 653
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2467,
+      "step": 654
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2489,
+      "step": 655
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.2431,
+      "step": 656
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2582,
+      "step": 657
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2327,
+      "step": 658
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2486,
+      "step": 659
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2532,
+      "step": 660
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2579,
+      "step": 661
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2388,
+      "step": 662
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2533,
+      "step": 663
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2402,
+      "step": 664
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.2542,
+      "step": 665
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2543,
+      "step": 666
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2518,
+      "step": 667
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2565,
+      "step": 668
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2459,
+      "step": 669
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2297,
+      "step": 670
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2515,
+      "step": 671
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.258,
+      "step": 672
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.264,
+      "step": 673
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2469,
+      "step": 674
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.2576,
+      "step": 675
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2275,
+      "step": 676
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2542,
+      "step": 677
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2402,
+      "step": 678
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2333,
+      "step": 679
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2414,
+      "step": 680
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2338,
+      "step": 681
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2657,
+      "step": 682
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2445,
+      "step": 683
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2414,
+      "step": 684
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2461,
+      "step": 685
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2431,
+      "step": 686
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2476,
+      "step": 687
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2518,
+      "step": 688
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2362,
+      "step": 689
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.253,
+      "step": 690
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2392,
+      "step": 691
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2379,
+      "step": 692
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2357,
+      "step": 693
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.2369,
+      "step": 694
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2445,
+      "step": 695
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2524,
+      "step": 696
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2284,
+      "step": 697
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2436,
+      "step": 698
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2535,
+      "step": 699
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2612,
+      "step": 700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2511,
+      "step": 701
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2445,
+      "step": 702
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2545,
+      "step": 703
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.2401,
+      "step": 704
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2389,
+      "step": 705
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2509,
+      "step": 706
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2374,
+      "step": 707
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2391,
+      "step": 708
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2457,
+      "step": 709
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2395,
+      "step": 710
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2506,
+      "step": 711
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2376,
+      "step": 712
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2315,
+      "step": 713
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.2429,
+      "step": 714
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2484,
+      "step": 715
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2401,
+      "step": 716
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2541,
+      "step": 717
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2316,
+      "step": 718
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2656,
+      "step": 719
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.248,
+      "step": 720
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2569,
+      "step": 721
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2442,
+      "step": 722
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2473,
+      "step": 723
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2476,
+      "step": 724
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2353,
+      "step": 725
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2505,
+      "step": 726
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2395,
+      "step": 727
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2392,
+      "step": 728
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2393,
+      "step": 729
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2286,
+      "step": 730
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.2515,
+      "step": 731
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.233,
+      "step": 732
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.234,
+      "step": 733
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2607,
+      "step": 734
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2402,
+      "step": 735
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2498,
+      "step": 736
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2657,
+      "step": 737
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.256,
+      "step": 738
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2292,
+      "step": 739
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2541,
+      "step": 740
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2346,
+      "step": 741
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2469,
+      "step": 742
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.2685,
+      "step": 743
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2691,
+      "step": 744
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2289,
+      "step": 745
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2348,
+      "step": 746
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2204,
+      "step": 747
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2519,
+      "step": 748
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2394,
+      "step": 749
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.24,
+      "step": 750
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 751
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2712,
+      "step": 752
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 753
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2508,
+      "step": 754
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2511,
+      "step": 755
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2335,
+      "step": 756
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2509,
+      "step": 757
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 758
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2306,
+      "step": 759
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2514,
+      "step": 760
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2524,
+      "step": 761
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2354,
+      "step": 762
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2476,
+      "step": 763
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2418,
+      "step": 764
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2556,
+      "step": 765
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2326,
+      "step": 766
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2651,
+      "step": 767
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.252,
+      "step": 768
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2414,
+      "step": 769
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2558,
+      "step": 770
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2399,
+      "step": 771
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.2419,
+      "step": 772
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2479,
+      "step": 773
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2445,
+      "step": 774
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2383,
+      "step": 775
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 776
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2561,
+      "step": 777
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2278,
+      "step": 778
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2535,
+      "step": 779
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2516,
+      "step": 780
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2377,
+      "step": 781
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2363,
+      "step": 782
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2464,
+      "step": 783
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2385,
+      "step": 784
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2433,
+      "step": 785
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2419,
+      "step": 786
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2462,
+      "step": 787
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2531,
+      "step": 788
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2455,
+      "step": 789
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2477,
+      "step": 790
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2393,
+      "step": 791
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.2511,
+      "step": 792
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2497,
+      "step": 793
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2325,
+      "step": 794
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2496,
+      "step": 795
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2376,
+      "step": 796
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2383,
+      "step": 797
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2343,
+      "step": 798
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2423,
+      "step": 799
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2392,
+      "step": 800
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.2311,
+      "step": 801
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2467,
+      "step": 802
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2365,
+      "step": 803
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.24,
+      "step": 804
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2396,
+      "step": 805
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2457,
+      "step": 806
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2482,
+      "step": 807
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 808
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 809
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2372,
+      "step": 810
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2466,
+      "step": 811
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2434,
+      "step": 812
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2569,
+      "step": 813
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.246,
+      "step": 814
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2448,
+      "step": 815
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2338,
+      "step": 816
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2381,
+      "step": 817
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2389,
+      "step": 818
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2433,
+      "step": 819
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2422,
+      "step": 820
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.2343,
+      "step": 821
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2231,
+      "step": 822
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2393,
+      "step": 823
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2482,
+      "step": 824
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2386,
+      "step": 825
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2369,
+      "step": 826
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2346,
+      "step": 827
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.286,
+      "step": 828
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2402,
+      "step": 829
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2403,
+      "step": 830
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.2493,
+      "step": 831
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2413,
+      "step": 832
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2253,
+      "step": 833
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 834
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2306,
+      "step": 835
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2309,
+      "step": 836
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2492,
+      "step": 837
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2567,
+      "step": 838
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2447,
+      "step": 839
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.2384,
+      "step": 840
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2531,
+      "step": 841
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2521,
+      "step": 842
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2359,
+      "step": 843
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2439,
+      "step": 844
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2331,
+      "step": 845
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2386,
+      "step": 846
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2901,
+      "step": 847
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.238,
+      "step": 848
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.25,
+      "step": 849
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.2409,
+      "step": 850
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2486,
+      "step": 851
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2353,
+      "step": 852
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2548,
+      "step": 853
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2278,
+      "step": 854
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2336,
+      "step": 855
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2443,
+      "step": 856
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2485,
+      "step": 857
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2444,
+      "step": 858
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2407,
+      "step": 859
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.2535,
+      "step": 860
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2464,
+      "step": 861
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2464,
+      "step": 862
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.242,
+      "step": 863
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2507,
+      "step": 864
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.219,
+      "step": 865
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2552,
+      "step": 866
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2335,
+      "step": 867
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.2448,
+      "step": 868
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.27,
+      "step": 869
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2208,
+      "step": 870
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2309,
+      "step": 871
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2278,
+      "step": 872
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2531,
+      "step": 873
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2406,
+      "step": 874
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.236,
+      "step": 875
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2437,
+      "step": 876
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2455,
+      "step": 877
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2435,
+      "step": 878
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.2389,
+      "step": 879
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2288,
+      "step": 880
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2544,
+      "step": 881
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2379,
+      "step": 882
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2217,
+      "step": 883
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2335,
+      "step": 884
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2373,
+      "step": 885
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 886
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2419,
+      "step": 887
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2408,
+      "step": 888
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2488,
+      "step": 889
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2583,
+      "step": 890
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2287,
+      "step": 891
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2431,
+      "step": 892
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2244,
+      "step": 893
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2462,
+      "step": 894
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2377,
+      "step": 895
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.234,
+      "step": 896
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2226,
+      "step": 897
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.23,
+      "step": 898
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.2339,
+      "step": 899
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2464,
+      "step": 900
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2486,
+      "step": 901
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2469,
+      "step": 902
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2505,
+      "step": 903
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2304,
+      "step": 904
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2385,
+      "step": 905
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2293,
+      "step": 906
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2374,
+      "step": 907
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2505,
+      "step": 908
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2465,
+      "step": 909
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2379,
+      "step": 910
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2183,
+      "step": 911
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2424,
+      "step": 912
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2437,
+      "step": 913
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2317,
+      "step": 914
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2494,
+      "step": 915
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2474,
+      "step": 916
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2277,
+      "step": 917
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.2473,
+      "step": 918
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2356,
+      "step": 919
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2334,
+      "step": 920
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2429,
+      "step": 921
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2351,
+      "step": 922
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2323,
+      "step": 923
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2593,
+      "step": 924
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2484,
+      "step": 925
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2447,
+      "step": 926
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2368,
+      "step": 927
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.2422,
+      "step": 928
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2438,
+      "step": 929
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2475,
+      "step": 930
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2476,
+      "step": 931
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2447,
+      "step": 932
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.258,
+      "step": 933
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2394,
+      "step": 934
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2336,
+      "step": 935
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2434,
+      "step": 936
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.2394,
+      "step": 937
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2359,
+      "step": 938
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2308,
+      "step": 939
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2241,
+      "step": 940
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2305,
+      "step": 941
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2391,
+      "step": 942
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2313,
+      "step": 943
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2414,
+      "step": 944
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2337,
+      "step": 945
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2356,
+      "step": 946
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2472,
+      "step": 947
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2266,
+      "step": 948
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.242,
+      "step": 949
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2323,
+      "step": 950
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2368,
+      "step": 951
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2279,
+      "step": 952
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2428,
+      "step": 953
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.227,
+      "step": 954
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2331,
+      "step": 955
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2462,
+      "step": 956
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.2447,
+      "step": 957
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2269,
+      "step": 958
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2617,
+      "step": 959
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2413,
+      "step": 960
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2305,
+      "step": 961
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2509,
+      "step": 962
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2416,
+      "step": 963
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2377,
+      "step": 964
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2413,
+      "step": 965
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2499,
+      "step": 966
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.2361,
+      "step": 967
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2317,
+      "step": 968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2303,
+      "step": 969
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2344,
+      "step": 970
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2247,
+      "step": 971
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2203,
+      "step": 972
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2418,
+      "step": 973
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2411,
+      "step": 974
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2376,
+      "step": 975
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.2275,
+      "step": 976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2245,
+      "step": 977
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2279,
+      "step": 978
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2324,
+      "step": 979
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2247,
+      "step": 980
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2589,
+      "step": 981
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2422,
+      "step": 982
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.226,
+      "step": 983
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2394,
+      "step": 984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.2439,
+      "step": 985
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.223,
+      "step": 986
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2262,
+      "step": 987
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.216,
+      "step": 988
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2195,
+      "step": 989
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2153,
+      "step": 990
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2328,
+      "step": 991
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2321,
+      "step": 992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2203,
+      "step": 993
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2395,
+      "step": 994
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2388,
+      "step": 995
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.2251,
+      "step": 996
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.2351,
+      "step": 997
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.2273,
+      "step": 998
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.2265,
+      "step": 999
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.2272,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4860,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 200,
+  "total_flos": 4.2001827364864e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}