diff --git "a/checkpoint-1098/trainer_state.json" "b/checkpoint-1098/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1098/trainer_state.json"
@@ -0,0 +1,6613 @@
+{
+  "best_metric": 0.046589694917201996,
+  "best_model_checkpoint": "./phishing-email-detection/checkpoint-1098",
+  "epoch": 1.0,
+  "global_step": 1098,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.7022,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6837,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 0.6956,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6936,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 0.687,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-06,
+      "loss": 0.6838,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.6982,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6948,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9e-06,
+      "loss": 0.6805,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.6995,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7028,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6965,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6835,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.6879,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-05,
+      "loss": 0.6769,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7155,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.6811,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6727,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6834,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 0.6674,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6672,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.6812,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6874,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6365,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5e-05,
+      "loss": 0.6393,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.6681,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.6282,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.6342,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9e-05,
+      "loss": 0.6132,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3e-05,
+      "loss": 0.5506,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.1e-05,
+      "loss": 0.5207,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.4944,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3e-05,
+      "loss": 0.5201,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.3777,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.5e-05,
+      "loss": 0.5256,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4247,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.7e-05,
+      "loss": 0.4869,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.8e-05,
+      "loss": 0.3235,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.3241,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4e-05,
+      "loss": 0.4529,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.1e-05,
+      "loss": 0.1684,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3919,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.3e-05,
+      "loss": 0.2534,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.3468,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1609,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.2447,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7e-05,
+      "loss": 0.3794,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.8e-05,
+      "loss": 0.2932,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9e-05,
+      "loss": 0.2164,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 0.3335,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9984586929716405e-05,
+      "loss": 0.5085,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99691738594328e-05,
+      "loss": 0.1875,
+      "step": 52
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99537607891492e-05,
+      "loss": 0.1964,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9938347718865605e-05,
+      "loss": 0.2442,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9922934648582e-05,
+      "loss": 0.2489,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99075215782984e-05,
+      "loss": 0.3017,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.98921085080148e-05,
+      "loss": 0.3468,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9876695437731194e-05,
+      "loss": 0.6424,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9861282367447596e-05,
+      "loss": 0.0898,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9845869297164e-05,
+      "loss": 0.1585,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9830456226880394e-05,
+      "loss": 0.185,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9815043156596796e-05,
+      "loss": 0.2936,
+      "step": 62
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.97996300863132e-05,
+      "loss": 0.6241,
+      "step": 63
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9784217016029594e-05,
+      "loss": 0.2769,
+      "step": 64
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9768803945745996e-05,
+      "loss": 0.1915,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.97533908754624e-05,
+      "loss": 0.1106,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9737977805178794e-05,
+      "loss": 0.4,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9722564734895196e-05,
+      "loss": 0.2603,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.97071516646116e-05,
+      "loss": 0.2849,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9691738594327994e-05,
+      "loss": 0.0758,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.967632552404439e-05,
+      "loss": 0.0657,
+      "step": 71
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.966091245376079e-05,
+      "loss": 0.1914,
+      "step": 72
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.964549938347719e-05,
+      "loss": 0.3203,
+      "step": 73
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.963008631319359e-05,
+      "loss": 0.0725,
+      "step": 74
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.961467324290999e-05,
+      "loss": 0.4044,
+      "step": 75
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.959926017262639e-05,
+      "loss": 0.1029,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.958384710234279e-05,
+      "loss": 0.2754,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.956843403205919e-05,
+      "loss": 0.0372,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.955302096177559e-05,
+      "loss": 0.1509,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.953760789149199e-05,
+      "loss": 0.0387,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.952219482120839e-05,
+      "loss": 0.0669,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.950678175092479e-05,
+      "loss": 0.0175,
+      "step": 82
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.949136868064119e-05,
+      "loss": 0.1144,
+      "step": 83
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9475955610357585e-05,
+      "loss": 0.0177,
+      "step": 84
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.946054254007399e-05,
+      "loss": 0.065,
+      "step": 85
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.944512946979038e-05,
+      "loss": 0.1989,
+      "step": 86
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.942971639950678e-05,
+      "loss": 0.2918,
+      "step": 87
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.941430332922318e-05,
+      "loss": 0.4348,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.939889025893958e-05,
+      "loss": 0.4145,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.938347718865598e-05,
+      "loss": 0.0472,
+      "step": 90
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.936806411837238e-05,
+      "loss": 0.2278,
+      "step": 91
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.935265104808878e-05,
+      "loss": 0.1183,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.933723797780518e-05,
+      "loss": 0.0903,
+      "step": 93
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.932182490752158e-05,
+      "loss": 0.0395,
+      "step": 94
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.930641183723798e-05,
+      "loss": 0.089,
+      "step": 95
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.929099876695438e-05,
+      "loss": 0.041,
+      "step": 96
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.927558569667078e-05,
+      "loss": 0.0672,
+      "step": 97
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.926017262638718e-05,
+      "loss": 0.2325,
+      "step": 98
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.924475955610358e-05,
+      "loss": 0.1052,
+      "step": 99
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9229346485819974e-05,
+      "loss": 0.087,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9213933415536376e-05,
+      "loss": 0.0363,
+      "step": 101
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.919852034525277e-05,
+      "loss": 0.0185,
+      "step": 102
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9183107274969174e-05,
+      "loss": 0.0082,
+      "step": 103
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9167694204685577e-05,
+      "loss": 0.056,
+      "step": 104
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.915228113440197e-05,
+      "loss": 0.0162,
+      "step": 105
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9136868064118374e-05,
+      "loss": 0.0162,
+      "step": 106
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.912145499383478e-05,
+      "loss": 0.0134,
+      "step": 107
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.910604192355117e-05,
+      "loss": 0.2928,
+      "step": 108
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9090628853267574e-05,
+      "loss": 0.2841,
+      "step": 109
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.907521578298398e-05,
+      "loss": 0.4202,
+      "step": 110
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.905980271270037e-05,
+      "loss": 0.0033,
+      "step": 111
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9044389642416774e-05,
+      "loss": 0.5551,
+      "step": 112
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.902897657213317e-05,
+      "loss": 0.1007,
+      "step": 113
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9013563501849565e-05,
+      "loss": 0.0045,
+      "step": 114
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.899815043156597e-05,
+      "loss": 0.0088,
+      "step": 115
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.898273736128237e-05,
+      "loss": 0.1765,
+      "step": 116
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8967324290998766e-05,
+      "loss": 0.0855,
+      "step": 117
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.895191122071517e-05,
+      "loss": 0.0027,
+      "step": 118
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.893649815043157e-05,
+      "loss": 0.315,
+      "step": 119
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8921085080147966e-05,
+      "loss": 0.3212,
+      "step": 120
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.890567200986437e-05,
+      "loss": 0.2402,
+      "step": 121
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.889025893958077e-05,
+      "loss": 0.0099,
+      "step": 122
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8874845869297166e-05,
+      "loss": 0.9157,
+      "step": 123
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.885943279901357e-05,
+      "loss": 0.1004,
+      "step": 124
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.884401972872997e-05,
+      "loss": 0.0088,
+      "step": 125
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8828606658446366e-05,
+      "loss": 0.0038,
+      "step": 126
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.881319358816276e-05,
+      "loss": 0.4306,
+      "step": 127
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8797780517879164e-05,
+      "loss": 0.0142,
+      "step": 128
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.878236744759556e-05,
+      "loss": 0.4888,
+      "step": 129
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.876695437731196e-05,
+      "loss": 0.0365,
+      "step": 130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8751541307028364e-05,
+      "loss": 0.0066,
+      "step": 131
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.873612823674476e-05,
+      "loss": 0.5105,
+      "step": 132
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.872071516646116e-05,
+      "loss": 0.0654,
+      "step": 133
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8705302096177564e-05,
+      "loss": 0.0038,
+      "step": 134
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.868988902589396e-05,
+      "loss": 0.0984,
+      "step": 135
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.867447595561036e-05,
+      "loss": 0.0202,
+      "step": 136
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8659062885326764e-05,
+      "loss": 0.1943,
+      "step": 137
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.864364981504316e-05,
+      "loss": 0.7161,
+      "step": 138
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.862823674475956e-05,
+      "loss": 0.2464,
+      "step": 139
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.861282367447596e-05,
+      "loss": 0.2055,
+      "step": 140
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.859741060419236e-05,
+      "loss": 0.0071,
+      "step": 141
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8581997533908755e-05,
+      "loss": 0.0043,
+      "step": 142
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.856658446362516e-05,
+      "loss": 0.231,
+      "step": 143
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.855117139334155e-05,
+      "loss": 0.0978,
+      "step": 144
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8535758323057955e-05,
+      "loss": 0.0053,
+      "step": 145
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.852034525277436e-05,
+      "loss": 0.0053,
+      "step": 146
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.850493218249075e-05,
+      "loss": 0.0063,
+      "step": 147
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.8489519112207155e-05,
+      "loss": 0.0138,
+      "step": 148
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.847410604192356e-05,
+      "loss": 0.2408,
+      "step": 149
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.845869297163995e-05,
+      "loss": 0.1458,
+      "step": 150
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8443279901356355e-05,
+      "loss": 0.0037,
+      "step": 151
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.842786683107276e-05,
+      "loss": 0.0182,
+      "step": 152
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.841245376078915e-05,
+      "loss": 0.0178,
+      "step": 153
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8397040690505555e-05,
+      "loss": 0.3435,
+      "step": 154
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.838162762022195e-05,
+      "loss": 0.3426,
+      "step": 155
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8366214549938346e-05,
+      "loss": 0.1803,
+      "step": 156
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.835080147965475e-05,
+      "loss": 0.1841,
+      "step": 157
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.833538840937115e-05,
+      "loss": 0.0268,
+      "step": 158
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8319975339087546e-05,
+      "loss": 0.0522,
+      "step": 159
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.830456226880395e-05,
+      "loss": 0.9797,
+      "step": 160
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8289149198520344e-05,
+      "loss": 0.3329,
+      "step": 161
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8273736128236746e-05,
+      "loss": 0.4755,
+      "step": 162
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.825832305795315e-05,
+      "loss": 0.4914,
+      "step": 163
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8242909987669544e-05,
+      "loss": 0.7619,
+      "step": 164
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8227496917385946e-05,
+      "loss": 0.1818,
+      "step": 165
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.821208384710235e-05,
+      "loss": 0.0092,
+      "step": 166
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8196670776818744e-05,
+      "loss": 0.0152,
+      "step": 167
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8181257706535146e-05,
+      "loss": 0.2107,
+      "step": 168
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.816584463625154e-05,
+      "loss": 0.1817,
+      "step": 169
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.815043156596794e-05,
+      "loss": 0.0845,
+      "step": 170
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.813501849568434e-05,
+      "loss": 0.0233,
+      "step": 171
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.811960542540074e-05,
+      "loss": 0.0471,
+      "step": 172
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.810419235511714e-05,
+      "loss": 0.0261,
+      "step": 173
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.808877928483354e-05,
+      "loss": 0.0201,
+      "step": 174
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.807336621454994e-05,
+      "loss": 0.3822,
+      "step": 175
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.805795314426634e-05,
+      "loss": 0.0799,
+      "step": 176
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.804254007398274e-05,
+      "loss": 0.0055,
+      "step": 177
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.802712700369914e-05,
+      "loss": 0.0126,
+      "step": 178
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.801171393341554e-05,
+      "loss": 0.3471,
+      "step": 179
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.799630086313194e-05,
+      "loss": 0.0118,
+      "step": 180
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.798088779284834e-05,
+      "loss": 0.0196,
+      "step": 181
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.796547472256474e-05,
+      "loss": 0.1366,
+      "step": 182
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.795006165228113e-05,
+      "loss": 0.3279,
+      "step": 183
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7934648581997536e-05,
+      "loss": 0.014,
+      "step": 184
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.791923551171393e-05,
+      "loss": 0.1363,
+      "step": 185
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.790382244143033e-05,
+      "loss": 0.4293,
+      "step": 186
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7888409371146736e-05,
+      "loss": 0.3484,
+      "step": 187
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.787299630086313e-05,
+      "loss": 0.1952,
+      "step": 188
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.785758323057953e-05,
+      "loss": 0.0179,
+      "step": 189
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7842170160295936e-05,
+      "loss": 0.2721,
+      "step": 190
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.782675709001233e-05,
+      "loss": 0.0802,
+      "step": 191
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7811344019728733e-05,
+      "loss": 0.1416,
+      "step": 192
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7795930949445136e-05,
+      "loss": 0.8448,
+      "step": 193
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.778051787916153e-05,
+      "loss": 0.3513,
+      "step": 194
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7765104808877934e-05,
+      "loss": 0.0042,
+      "step": 195
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.774969173859433e-05,
+      "loss": 0.0436,
+      "step": 196
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.773427866831073e-05,
+      "loss": 0.6608,
+      "step": 197
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.771886559802713e-05,
+      "loss": 0.2199,
+      "step": 198
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.770345252774353e-05,
+      "loss": 0.2829,
+      "step": 199
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7688039457459925e-05,
+      "loss": 0.498,
+      "step": 200
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.767262638717633e-05,
+      "loss": 0.0246,
+      "step": 201
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.765721331689273e-05,
+      "loss": 0.1832,
+      "step": 202
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7641800246609125e-05,
+      "loss": 0.2133,
+      "step": 203
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.762638717632553e-05,
+      "loss": 0.3909,
+      "step": 204
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.761097410604193e-05,
+      "loss": 0.0236,
+      "step": 205
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7595561035758325e-05,
+      "loss": 0.1417,
+      "step": 206
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.758014796547473e-05,
+      "loss": 0.1494,
+      "step": 207
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.756473489519113e-05,
+      "loss": 0.3789,
+      "step": 208
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.7549321824907525e-05,
+      "loss": 0.1206,
+      "step": 209
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.753390875462392e-05,
+      "loss": 0.0822,
+      "step": 210
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.751849568434032e-05,
+      "loss": 0.2448,
+      "step": 211
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.750308261405672e-05,
+      "loss": 0.1909,
+      "step": 212
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.748766954377312e-05,
+      "loss": 0.0174,
+      "step": 213
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.747225647348952e-05,
+      "loss": 0.1579,
+      "step": 214
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.745684340320592e-05,
+      "loss": 0.1925,
+      "step": 215
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.744143033292232e-05,
+      "loss": 0.1041,
+      "step": 216
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.742601726263872e-05,
+      "loss": 0.0214,
+      "step": 217
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.741060419235512e-05,
+      "loss": 0.0632,
+      "step": 218
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.739519112207152e-05,
+      "loss": 0.1283,
+      "step": 219
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.737977805178792e-05,
+      "loss": 0.0242,
+      "step": 220
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.736436498150432e-05,
+      "loss": 0.1727,
+      "step": 221
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.734895191122072e-05,
+      "loss": 0.1659,
+      "step": 222
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.7333538840937116e-05,
+      "loss": 0.0115,
+      "step": 223
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.731812577065352e-05,
+      "loss": 0.0066,
+      "step": 224
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.7302712700369914e-05,
+      "loss": 0.0089,
+      "step": 225
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7287299630086316e-05,
+      "loss": 0.0138,
+      "step": 226
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.727188655980271e-05,
+      "loss": 0.0949,
+      "step": 227
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7256473489519114e-05,
+      "loss": 0.3039,
+      "step": 228
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7241060419235516e-05,
+      "loss": 0.0098,
+      "step": 229
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.722564734895191e-05,
+      "loss": 0.0063,
+      "step": 230
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7210234278668314e-05,
+      "loss": 0.1198,
+      "step": 231
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7194821208384716e-05,
+      "loss": 0.0718,
+      "step": 232
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.717940813810111e-05,
+      "loss": 0.0059,
+      "step": 233
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7163995067817514e-05,
+      "loss": 0.2145,
+      "step": 234
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.714858199753391e-05,
+      "loss": 0.0372,
+      "step": 235
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.713316892725031e-05,
+      "loss": 0.1689,
+      "step": 236
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.7117755856966714e-05,
+      "loss": 0.198,
+      "step": 237
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.710234278668311e-05,
+      "loss": 0.3523,
+      "step": 238
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.7086929716399505e-05,
+      "loss": 0.3983,
+      "step": 239
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.707151664611591e-05,
+      "loss": 0.1888,
+      "step": 240
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.70561035758323e-05,
+      "loss": 0.0172,
+      "step": 241
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.7040690505548705e-05,
+      "loss": 0.0229,
+      "step": 242
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.702527743526511e-05,
+      "loss": 0.0715,
+      "step": 243
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.70098643649815e-05,
+      "loss": 0.1341,
+      "step": 244
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6994451294697905e-05,
+      "loss": 0.0272,
+      "step": 245
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.697903822441431e-05,
+      "loss": 0.6443,
+      "step": 246
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.69636251541307e-05,
+      "loss": 0.1703,
+      "step": 247
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6948212083847105e-05,
+      "loss": 0.2898,
+      "step": 248
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.693279901356351e-05,
+      "loss": 0.0112,
+      "step": 249
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.69173859432799e-05,
+      "loss": 0.032,
+      "step": 250
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6901972872996305e-05,
+      "loss": 0.0832,
+      "step": 251
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.68865598027127e-05,
+      "loss": 0.0179,
+      "step": 252
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6871146732429096e-05,
+      "loss": 0.0313,
+      "step": 253
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.68557336621455e-05,
+      "loss": 0.0056,
+      "step": 254
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.68403205918619e-05,
+      "loss": 0.0614,
+      "step": 255
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6824907521578297e-05,
+      "loss": 0.03,
+      "step": 256
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.68094944512947e-05,
+      "loss": 0.0227,
+      "step": 257
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.67940813810111e-05,
+      "loss": 0.0078,
+      "step": 258
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6778668310727497e-05,
+      "loss": 0.0037,
+      "step": 259
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.67632552404439e-05,
+      "loss": 0.1599,
+      "step": 260
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.67478421701603e-05,
+      "loss": 0.0341,
+      "step": 261
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.67324290998767e-05,
+      "loss": 0.0599,
+      "step": 262
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.67170160295931e-05,
+      "loss": 0.0171,
+      "step": 263
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.67016029593095e-05,
+      "loss": 0.0897,
+      "step": 264
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.66861898890259e-05,
+      "loss": 0.3561,
+      "step": 265
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.667077681874229e-05,
+      "loss": 0.0027,
+      "step": 266
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6655363748458695e-05,
+      "loss": 0.1664,
+      "step": 267
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.663995067817509e-05,
+      "loss": 0.0046,
+      "step": 268
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.662453760789149e-05,
+      "loss": 0.3711,
+      "step": 269
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6609124537607895e-05,
+      "loss": 0.0132,
+      "step": 270
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.659371146732429e-05,
+      "loss": 0.0898,
+      "step": 271
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.657829839704069e-05,
+      "loss": 0.0032,
+      "step": 272
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6562885326757095e-05,
+      "loss": 0.2465,
+      "step": 273
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.654747225647349e-05,
+      "loss": 0.0031,
+      "step": 274
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.653205918618989e-05,
+      "loss": 0.0055,
+      "step": 275
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.6516646115906295e-05,
+      "loss": 0.0187,
+      "step": 276
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.650123304562269e-05,
+      "loss": 0.1055,
+      "step": 277
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.648581997533909e-05,
+      "loss": 0.3103,
+      "step": 278
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.647040690505549e-05,
+      "loss": 0.2236,
+      "step": 279
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.645499383477189e-05,
+      "loss": 0.0025,
+      "step": 280
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6439580764488286e-05,
+      "loss": 0.0195,
+      "step": 281
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.642416769420469e-05,
+      "loss": 0.0365,
+      "step": 282
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6408754623921084e-05,
+      "loss": 0.2471,
+      "step": 283
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6393341553637486e-05,
+      "loss": 0.339,
+      "step": 284
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.637792848335389e-05,
+      "loss": 0.1557,
+      "step": 285
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6362515413070284e-05,
+      "loss": 0.0088,
+      "step": 286
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6347102342786686e-05,
+      "loss": 0.0099,
+      "step": 287
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.633168927250309e-05,
+      "loss": 0.0029,
+      "step": 288
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6316276202219484e-05,
+      "loss": 0.0047,
+      "step": 289
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.6300863131935886e-05,
+      "loss": 0.2115,
+      "step": 290
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.628545006165229e-05,
+      "loss": 0.004,
+      "step": 291
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.6270036991368684e-05,
+      "loss": 0.0033,
+      "step": 292
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.6254623921085086e-05,
+      "loss": 0.0025,
+      "step": 293
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.623921085080148e-05,
+      "loss": 0.0025,
+      "step": 294
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.622379778051788e-05,
+      "loss": 0.0028,
+      "step": 295
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.620838471023428e-05,
+      "loss": 0.0059,
+      "step": 296
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.619297163995068e-05,
+      "loss": 0.0029,
+      "step": 297
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.617755856966708e-05,
+      "loss": 0.0445,
+      "step": 298
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.616214549938348e-05,
+      "loss": 0.0432,
+      "step": 299
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.614673242909988e-05,
+      "loss": 0.2325,
+      "step": 300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.613131935881628e-05,
+      "loss": 0.0195,
+      "step": 301
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.611590628853268e-05,
+      "loss": 0.1756,
+      "step": 302
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.610049321824908e-05,
+      "loss": 0.0044,
+      "step": 303
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.608508014796548e-05,
+      "loss": 0.0018,
+      "step": 304
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.606966707768188e-05,
+      "loss": 0.3524,
+      "step": 305
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.605425400739828e-05,
+      "loss": 0.0033,
+      "step": 306
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.603884093711468e-05,
+      "loss": 0.3018,
+      "step": 307
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.602342786683107e-05,
+      "loss": 0.0471,
+      "step": 308
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.600801479654747e-05,
+      "loss": 0.0902,
+      "step": 309
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.599260172626387e-05,
+      "loss": 0.3098,
+      "step": 310
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.597718865598027e-05,
+      "loss": 0.0041,
+      "step": 311
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.596177558569667e-05,
+      "loss": 0.5342,
+      "step": 312
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.594636251541307e-05,
+      "loss": 0.1162,
+      "step": 313
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.593094944512947e-05,
+      "loss": 0.0098,
+      "step": 314
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.591553637484587e-05,
+      "loss": 0.1333,
+      "step": 315
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.590012330456227e-05,
+      "loss": 0.3174,
+      "step": 316
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.588471023427867e-05,
+      "loss": 0.0164,
+      "step": 317
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.586929716399507e-05,
+      "loss": 0.0033,
+      "step": 318
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.585388409371147e-05,
+      "loss": 0.0046,
+      "step": 319
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.583847102342787e-05,
+      "loss": 0.2797,
+      "step": 320
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.582305795314427e-05,
+      "loss": 0.0084,
+      "step": 321
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.5807644882860664e-05,
+      "loss": 0.0113,
+      "step": 322
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.5792231812577066e-05,
+      "loss": 0.0334,
+      "step": 323
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.577681874229346e-05,
+      "loss": 0.0412,
+      "step": 324
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5761405672009864e-05,
+      "loss": 0.0242,
+      "step": 325
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5745992601726267e-05,
+      "loss": 0.031,
+      "step": 326
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.573057953144266e-05,
+      "loss": 0.0121,
+      "step": 327
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5715166461159064e-05,
+      "loss": 0.0037,
+      "step": 328
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.569975339087547e-05,
+      "loss": 0.0361,
+      "step": 329
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.568434032059186e-05,
+      "loss": 0.0137,
+      "step": 330
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5668927250308264e-05,
+      "loss": 0.0029,
+      "step": 331
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.565351418002467e-05,
+      "loss": 0.5282,
+      "step": 332
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.563810110974106e-05,
+      "loss": 0.0581,
+      "step": 333
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.5622688039457464e-05,
+      "loss": 0.0033,
+      "step": 334
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.560727496917386e-05,
+      "loss": 0.0397,
+      "step": 335
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.559186189889026e-05,
+      "loss": 0.0731,
+      "step": 336
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.557644882860666e-05,
+      "loss": 0.3565,
+      "step": 337
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.556103575832306e-05,
+      "loss": 0.2693,
+      "step": 338
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5545622688039456e-05,
+      "loss": 0.0085,
+      "step": 339
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.553020961775586e-05,
+      "loss": 0.0079,
+      "step": 340
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.551479654747226e-05,
+      "loss": 0.1846,
+      "step": 341
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5499383477188656e-05,
+      "loss": 0.0116,
+      "step": 342
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.548397040690506e-05,
+      "loss": 0.2786,
+      "step": 343
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.546855733662146e-05,
+      "loss": 0.1938,
+      "step": 344
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.5453144266337856e-05,
+      "loss": 0.0055,
+      "step": 345
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.543773119605426e-05,
+      "loss": 0.0034,
+      "step": 346
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.542231812577066e-05,
+      "loss": 0.0059,
+      "step": 347
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.5406905055487056e-05,
+      "loss": 0.005,
+      "step": 348
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.539149198520346e-05,
+      "loss": 0.0043,
+      "step": 349
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.5376078914919854e-05,
+      "loss": 0.2746,
+      "step": 350
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.536066584463625e-05,
+      "loss": 0.0065,
+      "step": 351
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.534525277435265e-05,
+      "loss": 0.0865,
+      "step": 352
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.5329839704069054e-05,
+      "loss": 0.0968,
+      "step": 353
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.531442663378545e-05,
+      "loss": 0.1466,
+      "step": 354
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.529901356350185e-05,
+      "loss": 0.0039,
+      "step": 355
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.5283600493218254e-05,
+      "loss": 0.0368,
+      "step": 356
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.526818742293465e-05,
+      "loss": 0.0161,
+      "step": 357
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.525277435265105e-05,
+      "loss": 0.0031,
+      "step": 358
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5237361282367454e-05,
+      "loss": 0.0017,
+      "step": 359
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.522194821208385e-05,
+      "loss": 0.0015,
+      "step": 360
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.520653514180025e-05,
+      "loss": 0.368,
+      "step": 361
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5191122071516654e-05,
+      "loss": 0.5389,
+      "step": 362
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.517570900123305e-05,
+      "loss": 0.0112,
+      "step": 363
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5160295930949445e-05,
+      "loss": 0.0014,
+      "step": 364
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.514488286066585e-05,
+      "loss": 0.0042,
+      "step": 365
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.512946979038224e-05,
+      "loss": 0.1527,
+      "step": 366
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.5114056720098645e-05,
+      "loss": 0.0016,
+      "step": 367
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.509864364981505e-05,
+      "loss": 0.0027,
+      "step": 368
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.508323057953144e-05,
+      "loss": 0.002,
+      "step": 369
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.5067817509247845e-05,
+      "loss": 0.2218,
+      "step": 370
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.505240443896425e-05,
+      "loss": 0.3571,
+      "step": 371
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.503699136868064e-05,
+      "loss": 0.24,
+      "step": 372
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.5021578298397045e-05,
+      "loss": 0.0017,
+      "step": 373
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.500616522811345e-05,
+      "loss": 0.0031,
+      "step": 374
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.499075215782984e-05,
+      "loss": 0.0017,
+      "step": 375
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.4975339087546245e-05,
+      "loss": 0.6421,
+      "step": 376
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.495992601726264e-05,
+      "loss": 0.0031,
+      "step": 377
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.4944512946979036e-05,
+      "loss": 0.0145,
+      "step": 378
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.492909987669544e-05,
+      "loss": 0.1832,
+      "step": 379
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.491368680641184e-05,
+      "loss": 0.0018,
+      "step": 380
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4898273736128236e-05,
+      "loss": 0.0018,
+      "step": 381
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.488286066584464e-05,
+      "loss": 0.3842,
+      "step": 382
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4867447595561034e-05,
+      "loss": 0.0555,
+      "step": 383
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4852034525277436e-05,
+      "loss": 0.0066,
+      "step": 384
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.483662145499384e-05,
+      "loss": 0.0361,
+      "step": 385
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4821208384710234e-05,
+      "loss": 0.0094,
+      "step": 386
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4805795314426636e-05,
+      "loss": 0.007,
+      "step": 387
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.479038224414304e-05,
+      "loss": 0.0205,
+      "step": 388
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.4774969173859434e-05,
+      "loss": 0.0018,
+      "step": 389
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4759556103575836e-05,
+      "loss": 0.1393,
+      "step": 390
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.474414303329223e-05,
+      "loss": 0.3189,
+      "step": 391
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.4728729963008634e-05,
+      "loss": 0.0034,
+      "step": 392
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.471331689272503e-05,
+      "loss": 0.0074,
+      "step": 393
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.469790382244143e-05,
+      "loss": 0.0236,
+      "step": 394
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.468249075215783e-05,
+      "loss": 0.01,
+      "step": 395
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.466707768187423e-05,
+      "loss": 0.015,
+      "step": 396
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.465166461159063e-05,
+      "loss": 0.0033,
+      "step": 397
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.463625154130703e-05,
+      "loss": 0.285,
+      "step": 398
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.462083847102343e-05,
+      "loss": 0.0074,
+      "step": 399
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.460542540073983e-05,
+      "loss": 0.001,
+      "step": 400
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.459001233045623e-05,
+      "loss": 0.3007,
+      "step": 401
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.457459926017263e-05,
+      "loss": 0.0019,
+      "step": 402
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.455918618988903e-05,
+      "loss": 0.0029,
+      "step": 403
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.454377311960543e-05,
+      "loss": 0.0125,
+      "step": 404
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.452836004932183e-05,
+      "loss": 0.0415,
+      "step": 405
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.4512946979038226e-05,
+      "loss": 0.0031,
+      "step": 406
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.449753390875462e-05,
+      "loss": 0.0025,
+      "step": 407
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.448212083847102e-05,
+      "loss": 0.0048,
+      "step": 408
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.4466707768187426e-05,
+      "loss": 0.0755,
+      "step": 409
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.445129469790382e-05,
+      "loss": 0.1365,
+      "step": 410
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.443588162762022e-05,
+      "loss": 0.0036,
+      "step": 411
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4420468557336626e-05,
+      "loss": 0.0262,
+      "step": 412
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.440505548705302e-05,
+      "loss": 0.0023,
+      "step": 413
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4389642416769423e-05,
+      "loss": 0.1611,
+      "step": 414
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4374229346485826e-05,
+      "loss": 0.1367,
+      "step": 415
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.435881627620222e-05,
+      "loss": 0.0034,
+      "step": 416
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4343403205918624e-05,
+      "loss": 0.1921,
+      "step": 417
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4327990135635026e-05,
+      "loss": 0.0025,
+      "step": 418
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.431257706535142e-05,
+      "loss": 0.0452,
+      "step": 419
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.429716399506782e-05,
+      "loss": 0.0165,
+      "step": 420
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.428175092478422e-05,
+      "loss": 0.0011,
+      "step": 421
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.4266337854500615e-05,
+      "loss": 0.001,
+      "step": 422
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.425092478421702e-05,
+      "loss": 0.0041,
+      "step": 423
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.423551171393342e-05,
+      "loss": 0.0158,
+      "step": 424
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.4220098643649815e-05,
+      "loss": 0.4164,
+      "step": 425
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.420468557336622e-05,
+      "loss": 0.0099,
+      "step": 426
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.418927250308262e-05,
+      "loss": 0.0015,
+      "step": 427
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.4173859432799015e-05,
+      "loss": 0.0012,
+      "step": 428
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.415844636251542e-05,
+      "loss": 0.377,
+      "step": 429
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.414303329223182e-05,
+      "loss": 0.0014,
+      "step": 430
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.4127620221948215e-05,
+      "loss": 0.0296,
+      "step": 431
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.411220715166462e-05,
+      "loss": 0.0012,
+      "step": 432
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.409679408138101e-05,
+      "loss": 0.0028,
+      "step": 433
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.408138101109741e-05,
+      "loss": 0.003,
+      "step": 434
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.406596794081381e-05,
+      "loss": 0.0013,
+      "step": 435
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.405055487053021e-05,
+      "loss": 0.0015,
+      "step": 436
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.403514180024661e-05,
+      "loss": 0.003,
+      "step": 437
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.401972872996301e-05,
+      "loss": 0.0028,
+      "step": 438
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.400431565967941e-05,
+      "loss": 0.0009,
+      "step": 439
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.398890258939581e-05,
+      "loss": 0.3179,
+      "step": 440
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.397348951911221e-05,
+      "loss": 0.0871,
+      "step": 441
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.395807644882861e-05,
+      "loss": 0.0103,
+      "step": 442
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.394266337854501e-05,
+      "loss": 0.0018,
+      "step": 443
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.392725030826141e-05,
+      "loss": 0.5116,
+      "step": 444
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.391183723797781e-05,
+      "loss": 0.0027,
+      "step": 445
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.389642416769421e-05,
+      "loss": 0.0624,
+      "step": 446
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3881011097410604e-05,
+      "loss": 0.0013,
+      "step": 447
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3865598027127006e-05,
+      "loss": 0.2554,
+      "step": 448
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.38501849568434e-05,
+      "loss": 0.3956,
+      "step": 449
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3834771886559804e-05,
+      "loss": 0.0018,
+      "step": 450
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3819358816276206e-05,
+      "loss": 0.0011,
+      "step": 451
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.38039457459926e-05,
+      "loss": 0.0025,
+      "step": 452
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3788532675709004e-05,
+      "loss": 0.0023,
+      "step": 453
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.3773119605425406e-05,
+      "loss": 0.0072,
+      "step": 454
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.37577065351418e-05,
+      "loss": 0.0182,
+      "step": 455
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3742293464858204e-05,
+      "loss": 0.0015,
+      "step": 456
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.37268803945746e-05,
+      "loss": 0.0016,
+      "step": 457
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3711467324291e-05,
+      "loss": 0.0012,
+      "step": 458
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3696054254007404e-05,
+      "loss": 0.0419,
+      "step": 459
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.36806411837238e-05,
+      "loss": 0.0132,
+      "step": 460
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.36652281134402e-05,
+      "loss": 0.6488,
+      "step": 461
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.36498150431566e-05,
+      "loss": 0.0024,
+      "step": 462
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.363440197287299e-05,
+      "loss": 0.0036,
+      "step": 463
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.3618988902589395e-05,
+      "loss": 0.0013,
+      "step": 464
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.36035758323058e-05,
+      "loss": 0.0018,
+      "step": 465
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.358816276202219e-05,
+      "loss": 0.0015,
+      "step": 466
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.3572749691738595e-05,
+      "loss": 0.3114,
+      "step": 467
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.3557336621455e-05,
+      "loss": 0.0012,
+      "step": 468
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.354192355117139e-05,
+      "loss": 0.0012,
+      "step": 469
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.3526510480887795e-05,
+      "loss": 0.0016,
+      "step": 470
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.35110974106042e-05,
+      "loss": 0.0063,
+      "step": 471
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.349568434032059e-05,
+      "loss": 0.0113,
+      "step": 472
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.3480271270036995e-05,
+      "loss": 0.0013,
+      "step": 473
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.34648581997534e-05,
+      "loss": 0.0202,
+      "step": 474
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.344944512946979e-05,
+      "loss": 0.8024,
+      "step": 475
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.343403205918619e-05,
+      "loss": 0.1431,
+      "step": 476
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.341861898890259e-05,
+      "loss": 0.0395,
+      "step": 477
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.3403205918618987e-05,
+      "loss": 0.0492,
+      "step": 478
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.338779284833539e-05,
+      "loss": 0.0033,
+      "step": 479
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.337237977805179e-05,
+      "loss": 0.3942,
+      "step": 480
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.3356966707768187e-05,
+      "loss": 0.0011,
+      "step": 481
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.334155363748459e-05,
+      "loss": 0.0992,
+      "step": 482
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.332614056720099e-05,
+      "loss": 0.0046,
+      "step": 483
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.331072749691739e-05,
+      "loss": 0.0047,
+      "step": 484
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.329531442663379e-05,
+      "loss": 0.388,
+      "step": 485
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.327990135635019e-05,
+      "loss": 0.4356,
+      "step": 486
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.326448828606659e-05,
+      "loss": 0.1234,
+      "step": 487
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.324907521578299e-05,
+      "loss": 0.0022,
+      "step": 488
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.3233662145499385e-05,
+      "loss": 0.2838,
+      "step": 489
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.321824907521578e-05,
+      "loss": 0.0012,
+      "step": 490
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.320283600493218e-05,
+      "loss": 0.3868,
+      "step": 491
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.3187422934648585e-05,
+      "loss": 0.0014,
+      "step": 492
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.317200986436498e-05,
+      "loss": 0.0931,
+      "step": 493
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.315659679408138e-05,
+      "loss": 0.0479,
+      "step": 494
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.3141183723797785e-05,
+      "loss": 0.3443,
+      "step": 495
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.312577065351418e-05,
+      "loss": 0.0019,
+      "step": 496
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.311035758323058e-05,
+      "loss": 0.0217,
+      "step": 497
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.3094944512946985e-05,
+      "loss": 0.0034,
+      "step": 498
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.307953144266338e-05,
+      "loss": 0.0029,
+      "step": 499
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.306411837237978e-05,
+      "loss": 0.0033,
+      "step": 500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3048705302096185e-05,
+      "loss": 0.0286,
+      "step": 501
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.303329223181258e-05,
+      "loss": 0.3277,
+      "step": 502
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3017879161528976e-05,
+      "loss": 0.0323,
+      "step": 503
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.300246609124538e-05,
+      "loss": 0.0044,
+      "step": 504
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2987053020961774e-05,
+      "loss": 0.0076,
+      "step": 505
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2971639950678176e-05,
+      "loss": 0.0179,
+      "step": 506
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.295622688039458e-05,
+      "loss": 0.0271,
+      "step": 507
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2940813810110974e-05,
+      "loss": 0.0072,
+      "step": 508
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2925400739827376e-05,
+      "loss": 0.0058,
+      "step": 509
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.290998766954378e-05,
+      "loss": 0.0026,
+      "step": 510
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2894574599260174e-05,
+      "loss": 0.0565,
+      "step": 511
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2879161528976576e-05,
+      "loss": 0.0288,
+      "step": 512
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.286374845869298e-05,
+      "loss": 0.0157,
+      "step": 513
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2848335388409374e-05,
+      "loss": 0.0716,
+      "step": 514
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.2832922318125776e-05,
+      "loss": 0.0282,
+      "step": 515
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.281750924784217e-05,
+      "loss": 0.1884,
+      "step": 516
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.280209617755857e-05,
+      "loss": 0.0066,
+      "step": 517
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.278668310727497e-05,
+      "loss": 0.004,
+      "step": 518
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.277127003699137e-05,
+      "loss": 0.1473,
+      "step": 519
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.275585696670777e-05,
+      "loss": 0.0018,
+      "step": 520
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.274044389642417e-05,
+      "loss": 0.0013,
+      "step": 521
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.272503082614057e-05,
+      "loss": 0.0166,
+      "step": 522
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.270961775585697e-05,
+      "loss": 0.3395,
+      "step": 523
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.269420468557337e-05,
+      "loss": 0.3455,
+      "step": 524
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.267879161528977e-05,
+      "loss": 0.1434,
+      "step": 525
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.266337854500617e-05,
+      "loss": 0.011,
+      "step": 526
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.264796547472257e-05,
+      "loss": 0.2195,
+      "step": 527
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.263255240443897e-05,
+      "loss": 0.0471,
+      "step": 528
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.261713933415537e-05,
+      "loss": 0.0145,
+      "step": 529
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.260172626387176e-05,
+      "loss": 0.0846,
+      "step": 530
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.2586313193588165e-05,
+      "loss": 0.0244,
+      "step": 531
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.257090012330456e-05,
+      "loss": 0.0035,
+      "step": 532
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.255548705302096e-05,
+      "loss": 0.3924,
+      "step": 533
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.254007398273736e-05,
+      "loss": 0.0021,
+      "step": 534
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.252466091245376e-05,
+      "loss": 0.4137,
+      "step": 535
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.250924784217016e-05,
+      "loss": 0.2125,
+      "step": 536
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.249383477188656e-05,
+      "loss": 0.2046,
+      "step": 537
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.247842170160296e-05,
+      "loss": 0.2159,
+      "step": 538
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.246300863131936e-05,
+      "loss": 0.0051,
+      "step": 539
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.244759556103576e-05,
+      "loss": 0.0145,
+      "step": 540
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.243218249075216e-05,
+      "loss": 0.0304,
+      "step": 541
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.241676942046856e-05,
+      "loss": 0.0055,
+      "step": 542
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.240135635018496e-05,
+      "loss": 0.0047,
+      "step": 543
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.238594327990136e-05,
+      "loss": 0.0092,
+      "step": 544
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2370530209617756e-05,
+      "loss": 0.1466,
+      "step": 545
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.235511713933415e-05,
+      "loss": 0.43,
+      "step": 546
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2339704069050554e-05,
+      "loss": 0.0028,
+      "step": 547
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2324290998766957e-05,
+      "loss": 0.0354,
+      "step": 548
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.230887792848335e-05,
+      "loss": 0.1144,
+      "step": 549
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2293464858199754e-05,
+      "loss": 0.0106,
+      "step": 550
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.227805178791616e-05,
+      "loss": 0.0047,
+      "step": 551
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.226263871763255e-05,
+      "loss": 0.0167,
+      "step": 552
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.2247225647348954e-05,
+      "loss": 0.2297,
+      "step": 553
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.223181257706536e-05,
+      "loss": 0.0049,
+      "step": 554
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.221639950678175e-05,
+      "loss": 0.009,
+      "step": 555
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.2200986436498154e-05,
+      "loss": 0.0541,
+      "step": 556
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.218557336621456e-05,
+      "loss": 0.0051,
+      "step": 557
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.217016029593095e-05,
+      "loss": 0.0477,
+      "step": 558
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.215474722564735e-05,
+      "loss": 0.0592,
+      "step": 559
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.213933415536375e-05,
+      "loss": 0.0048,
+      "step": 560
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.2123921085080146e-05,
+      "loss": 0.008,
+      "step": 561
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.210850801479655e-05,
+      "loss": 0.0052,
+      "step": 562
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.209309494451295e-05,
+      "loss": 0.0039,
+      "step": 563
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.2077681874229346e-05,
+      "loss": 0.006,
+      "step": 564
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.206226880394575e-05,
+      "loss": 0.0022,
+      "step": 565
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.204685573366215e-05,
+      "loss": 0.007,
+      "step": 566
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2031442663378546e-05,
+      "loss": 0.0026,
+      "step": 567
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.201602959309495e-05,
+      "loss": 0.0023,
+      "step": 568
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.200061652281135e-05,
+      "loss": 0.0022,
+      "step": 569
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1985203452527746e-05,
+      "loss": 0.219,
+      "step": 570
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.196979038224415e-05,
+      "loss": 0.0018,
+      "step": 571
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1954377311960544e-05,
+      "loss": 0.004,
+      "step": 572
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.193896424167694e-05,
+      "loss": 0.3496,
+      "step": 573
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.192355117139334e-05,
+      "loss": 0.0058,
+      "step": 574
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.1908138101109744e-05,
+      "loss": 0.0141,
+      "step": 575
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.189272503082614e-05,
+      "loss": 0.3665,
+      "step": 576
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.187731196054254e-05,
+      "loss": 0.0655,
+      "step": 577
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1861898890258944e-05,
+      "loss": 0.0012,
+      "step": 578
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.184648581997534e-05,
+      "loss": 0.0067,
+      "step": 579
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.183107274969174e-05,
+      "loss": 0.0015,
+      "step": 580
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1815659679408144e-05,
+      "loss": 0.0015,
+      "step": 581
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.180024660912454e-05,
+      "loss": 0.0019,
+      "step": 582
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.178483353884094e-05,
+      "loss": 0.0356,
+      "step": 583
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1769420468557344e-05,
+      "loss": 0.0012,
+      "step": 584
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.175400739827374e-05,
+      "loss": 0.0147,
+      "step": 585
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1738594327990135e-05,
+      "loss": 0.0632,
+      "step": 586
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.172318125770654e-05,
+      "loss": 0.0017,
+      "step": 587
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.170776818742293e-05,
+      "loss": 0.0017,
+      "step": 588
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1692355117139335e-05,
+      "loss": 0.0049,
+      "step": 589
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.167694204685574e-05,
+      "loss": 0.0012,
+      "step": 590
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.166152897657213e-05,
+      "loss": 0.0012,
+      "step": 591
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1646115906288535e-05,
+      "loss": 0.3058,
+      "step": 592
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.163070283600494e-05,
+      "loss": 0.0026,
+      "step": 593
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.161528976572133e-05,
+      "loss": 0.3416,
+      "step": 594
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1599876695437735e-05,
+      "loss": 0.0092,
+      "step": 595
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.158446362515414e-05,
+      "loss": 0.0016,
+      "step": 596
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.156905055487053e-05,
+      "loss": 0.002,
+      "step": 597
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.1553637484586935e-05,
+      "loss": 0.0014,
+      "step": 598
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.153822441430333e-05,
+      "loss": 0.6758,
+      "step": 599
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.152281134401973e-05,
+      "loss": 0.0049,
+      "step": 600
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.150739827373613e-05,
+      "loss": 0.0047,
+      "step": 601
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.149198520345253e-05,
+      "loss": 0.003,
+      "step": 602
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1476572133168926e-05,
+      "loss": 0.0064,
+      "step": 603
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.146115906288533e-05,
+      "loss": 0.0634,
+      "step": 604
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1445745992601724e-05,
+      "loss": 0.0036,
+      "step": 605
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1430332922318126e-05,
+      "loss": 0.0051,
+      "step": 606
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.141491985203453e-05,
+      "loss": 0.238,
+      "step": 607
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1399506781750924e-05,
+      "loss": 0.003,
+      "step": 608
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.1384093711467326e-05,
+      "loss": 0.3641,
+      "step": 609
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.136868064118373e-05,
+      "loss": 0.005,
+      "step": 610
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1353267570900124e-05,
+      "loss": 0.0023,
+      "step": 611
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1337854500616526e-05,
+      "loss": 0.0011,
+      "step": 612
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.132244143033293e-05,
+      "loss": 0.3954,
+      "step": 613
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.1307028360049324e-05,
+      "loss": 0.5307,
+      "step": 614
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.129161528976572e-05,
+      "loss": 0.0015,
+      "step": 615
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.127620221948212e-05,
+      "loss": 0.0011,
+      "step": 616
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.126078914919852e-05,
+      "loss": 0.0016,
+      "step": 617
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.124537607891492e-05,
+      "loss": 0.0022,
+      "step": 618
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.122996300863132e-05,
+      "loss": 0.0029,
+      "step": 619
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.121454993834772e-05,
+      "loss": 0.0032,
+      "step": 620
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.119913686806412e-05,
+      "loss": 0.0022,
+      "step": 621
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.118372379778052e-05,
+      "loss": 0.0015,
+      "step": 622
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.116831072749692e-05,
+      "loss": 0.0012,
+      "step": 623
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.115289765721332e-05,
+      "loss": 0.0029,
+      "step": 624
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.113748458692972e-05,
+      "loss": 0.0022,
+      "step": 625
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.112207151664612e-05,
+      "loss": 0.0015,
+      "step": 626
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.110665844636252e-05,
+      "loss": 0.4074,
+      "step": 627
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1091245376078915e-05,
+      "loss": 0.0013,
+      "step": 628
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.107583230579531e-05,
+      "loss": 0.0129,
+      "step": 629
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.106041923551171e-05,
+      "loss": 0.0019,
+      "step": 630
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1045006165228116e-05,
+      "loss": 0.0011,
+      "step": 631
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.102959309494451e-05,
+      "loss": 0.0033,
+      "step": 632
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.101418002466091e-05,
+      "loss": 0.0014,
+      "step": 633
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0998766954377316e-05,
+      "loss": 0.0027,
+      "step": 634
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.098335388409371e-05,
+      "loss": 0.0016,
+      "step": 635
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0967940813810113e-05,
+      "loss": 0.0015,
+      "step": 636
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0952527743526516e-05,
+      "loss": 0.0067,
+      "step": 637
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.093711467324291e-05,
+      "loss": 0.3461,
+      "step": 638
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0921701602959313e-05,
+      "loss": 0.001,
+      "step": 639
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0906288532675716e-05,
+      "loss": 0.0378,
+      "step": 640
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.089087546239211e-05,
+      "loss": 0.0035,
+      "step": 641
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.087546239210851e-05,
+      "loss": 0.0021,
+      "step": 642
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.086004932182491e-05,
+      "loss": 0.0138,
+      "step": 643
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0844636251541305e-05,
+      "loss": 0.2242,
+      "step": 644
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.082922318125771e-05,
+      "loss": 0.1115,
+      "step": 645
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.081381011097411e-05,
+      "loss": 0.0015,
+      "step": 646
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0798397040690505e-05,
+      "loss": 0.0011,
+      "step": 647
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.078298397040691e-05,
+      "loss": 0.0009,
+      "step": 648
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.076757090012331e-05,
+      "loss": 0.2452,
+      "step": 649
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0752157829839705e-05,
+      "loss": 0.0534,
+      "step": 650
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.073674475955611e-05,
+      "loss": 0.2856,
+      "step": 651
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.072133168927251e-05,
+      "loss": 0.1185,
+      "step": 652
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.0705918618988905e-05,
+      "loss": 0.0034,
+      "step": 653
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.069050554870531e-05,
+      "loss": 0.0039,
+      "step": 654
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.06750924784217e-05,
+      "loss": 0.0022,
+      "step": 655
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.0659679408138105e-05,
+      "loss": 0.001,
+      "step": 656
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.06442663378545e-05,
+      "loss": 0.0013,
+      "step": 657
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.06288532675709e-05,
+      "loss": 0.0306,
+      "step": 658
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.06134401972873e-05,
+      "loss": 0.0477,
+      "step": 659
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05980271270037e-05,
+      "loss": 0.2355,
+      "step": 660
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05826140567201e-05,
+      "loss": 0.0856,
+      "step": 661
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05672009864365e-05,
+      "loss": 0.2559,
+      "step": 662
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05517879161529e-05,
+      "loss": 0.1194,
+      "step": 663
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.05363748458693e-05,
+      "loss": 0.0016,
+      "step": 664
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.05209617755857e-05,
+      "loss": 0.0014,
+      "step": 665
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.05055487053021e-05,
+      "loss": 0.2945,
+      "step": 666
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.04901356350185e-05,
+      "loss": 0.2178,
+      "step": 667
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.04747225647349e-05,
+      "loss": 0.3807,
+      "step": 668
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.04593094944513e-05,
+      "loss": 0.0016,
+      "step": 669
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0443896424167696e-05,
+      "loss": 0.2671,
+      "step": 670
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.042848335388409e-05,
+      "loss": 0.0023,
+      "step": 671
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0413070283600494e-05,
+      "loss": 0.0036,
+      "step": 672
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0397657213316896e-05,
+      "loss": 0.0031,
+      "step": 673
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.038224414303329e-05,
+      "loss": 0.0132,
+      "step": 674
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.0366831072749694e-05,
+      "loss": 0.0129,
+      "step": 675
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0351418002466096e-05,
+      "loss": 0.0027,
+      "step": 676
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.033600493218249e-05,
+      "loss": 0.0015,
+      "step": 677
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0320591861898894e-05,
+      "loss": 0.0826,
+      "step": 678
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.030517879161529e-05,
+      "loss": 0.0075,
+      "step": 679
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.028976572133169e-05,
+      "loss": 0.0381,
+      "step": 680
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0274352651048094e-05,
+      "loss": 0.0065,
+      "step": 681
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.025893958076449e-05,
+      "loss": 0.0099,
+      "step": 682
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.024352651048089e-05,
+      "loss": 0.0035,
+      "step": 683
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.022811344019729e-05,
+      "loss": 0.0036,
+      "step": 684
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.021270036991368e-05,
+      "loss": 0.0048,
+      "step": 685
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.0197287299630085e-05,
+      "loss": 0.1165,
+      "step": 686
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.018187422934649e-05,
+      "loss": 0.0055,
+      "step": 687
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.016646115906288e-05,
+      "loss": 0.0151,
+      "step": 688
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0151048088779285e-05,
+      "loss": 0.0049,
+      "step": 689
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.013563501849569e-05,
+      "loss": 0.0022,
+      "step": 690
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.012022194821208e-05,
+      "loss": 0.0017,
+      "step": 691
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0104808877928485e-05,
+      "loss": 0.0088,
+      "step": 692
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.008939580764489e-05,
+      "loss": 0.0021,
+      "step": 693
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.007398273736128e-05,
+      "loss": 0.0025,
+      "step": 694
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0058569667077685e-05,
+      "loss": 0.1682,
+      "step": 695
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.004315659679409e-05,
+      "loss": 0.001,
+      "step": 696
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.002774352651048e-05,
+      "loss": 0.0016,
+      "step": 697
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.001233045622688e-05,
+      "loss": 0.0027,
+      "step": 698
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.999691738594328e-05,
+      "loss": 0.3733,
+      "step": 699
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9981504315659677e-05,
+      "loss": 0.001,
+      "step": 700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.996609124537608e-05,
+      "loss": 0.0011,
+      "step": 701
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.995067817509248e-05,
+      "loss": 0.0626,
+      "step": 702
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.9935265104808877e-05,
+      "loss": 0.002,
+      "step": 703
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.991985203452528e-05,
+      "loss": 0.1558,
+      "step": 704
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.990443896424168e-05,
+      "loss": 0.0077,
+      "step": 705
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.988902589395808e-05,
+      "loss": 0.0016,
+      "step": 706
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.987361282367448e-05,
+      "loss": 0.0012,
+      "step": 707
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 3.985819975339088e-05,
+      "loss": 0.001,
+      "step": 708
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.984278668310728e-05,
+      "loss": 0.0009,
+      "step": 709
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.982737361282368e-05,
+      "loss": 0.0011,
+      "step": 710
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9811960542540075e-05,
+      "loss": 0.0892,
+      "step": 711
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.979654747225648e-05,
+      "loss": 0.1312,
+      "step": 712
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.978113440197287e-05,
+      "loss": 0.0013,
+      "step": 713
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9765721331689275e-05,
+      "loss": 0.0219,
+      "step": 714
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.975030826140567e-05,
+      "loss": 0.0009,
+      "step": 715
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.973489519112207e-05,
+      "loss": 0.0377,
+      "step": 716
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.9719482120838475e-05,
+      "loss": 0.0623,
+      "step": 717
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.970406905055487e-05,
+      "loss": 0.001,
+      "step": 718
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 3.968865598027127e-05,
+      "loss": 0.2698,
+      "step": 719
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9673242909987675e-05,
+      "loss": 0.1464,
+      "step": 720
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.965782983970407e-05,
+      "loss": 0.0027,
+      "step": 721
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.964241676942047e-05,
+      "loss": 0.0016,
+      "step": 722
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9627003699136875e-05,
+      "loss": 0.001,
+      "step": 723
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.961159062885327e-05,
+      "loss": 0.0009,
+      "step": 724
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.959617755856967e-05,
+      "loss": 0.0256,
+      "step": 725
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.958076448828607e-05,
+      "loss": 0.0073,
+      "step": 726
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9565351418002464e-05,
+      "loss": 0.0014,
+      "step": 727
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9549938347718866e-05,
+      "loss": 0.001,
+      "step": 728
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.953452527743527e-05,
+      "loss": 0.1047,
+      "step": 729
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 3.9519112207151664e-05,
+      "loss": 0.714,
+      "step": 730
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9503699136868066e-05,
+      "loss": 0.0011,
+      "step": 731
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.948828606658447e-05,
+      "loss": 0.0008,
+      "step": 732
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9472872996300864e-05,
+      "loss": 0.2677,
+      "step": 733
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9457459926017266e-05,
+      "loss": 0.0008,
+      "step": 734
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.944204685573367e-05,
+      "loss": 0.4921,
+      "step": 735
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9426633785450064e-05,
+      "loss": 0.0467,
+      "step": 736
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9411220715166466e-05,
+      "loss": 0.0624,
+      "step": 737
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.939580764488287e-05,
+      "loss": 0.0007,
+      "step": 738
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.9380394574599264e-05,
+      "loss": 0.0022,
+      "step": 739
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.936498150431566e-05,
+      "loss": 0.0178,
+      "step": 740
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.934956843403206e-05,
+      "loss": 0.0025,
+      "step": 741
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.933415536374846e-05,
+      "loss": 0.1864,
+      "step": 742
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.931874229346486e-05,
+      "loss": 0.0007,
+      "step": 743
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.930332922318126e-05,
+      "loss": 0.0027,
+      "step": 744
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.928791615289766e-05,
+      "loss": 0.0123,
+      "step": 745
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.927250308261406e-05,
+      "loss": 0.0016,
+      "step": 746
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.925709001233046e-05,
+      "loss": 0.0013,
+      "step": 747
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.924167694204686e-05,
+      "loss": 0.0012,
+      "step": 748
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.922626387176326e-05,
+      "loss": 0.0061,
+      "step": 749
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.921085080147966e-05,
+      "loss": 0.0011,
+      "step": 750
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.919543773119606e-05,
+      "loss": 0.0061,
+      "step": 751
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 3.918002466091246e-05,
+      "loss": 0.001,
+      "step": 752
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.9164611590628855e-05,
+      "loss": 0.0018,
+      "step": 753
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.914919852034525e-05,
+      "loss": 0.0008,
+      "step": 754
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.913378545006165e-05,
+      "loss": 0.0007,
+      "step": 755
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.911837237977805e-05,
+      "loss": 0.0008,
+      "step": 756
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.910295930949445e-05,
+      "loss": 0.0008,
+      "step": 757
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.908754623921085e-05,
+      "loss": 0.1326,
+      "step": 758
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.907213316892725e-05,
+      "loss": 0.0006,
+      "step": 759
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.905672009864365e-05,
+      "loss": 0.0012,
+      "step": 760
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.904130702836005e-05,
+      "loss": 0.1199,
+      "step": 761
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.902589395807645e-05,
+      "loss": 0.0006,
+      "step": 762
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.901048088779285e-05,
+      "loss": 0.0006,
+      "step": 763
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.899506781750925e-05,
+      "loss": 0.2864,
+      "step": 764
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.897965474722565e-05,
+      "loss": 0.0006,
+      "step": 765
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.896424167694205e-05,
+      "loss": 0.0014,
+      "step": 766
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8948828606658446e-05,
+      "loss": 0.0564,
+      "step": 767
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.893341553637485e-05,
+      "loss": 0.0008,
+      "step": 768
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8918002466091244e-05,
+      "loss": 0.0014,
+      "step": 769
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8902589395807647e-05,
+      "loss": 0.0022,
+      "step": 770
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.888717632552404e-05,
+      "loss": 0.0093,
+      "step": 771
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8871763255240444e-05,
+      "loss": 0.0008,
+      "step": 772
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.8856350184956847e-05,
+      "loss": 0.4581,
+      "step": 773
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.884093711467324e-05,
+      "loss": 0.007,
+      "step": 774
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.8825524044389644e-05,
+      "loss": 0.2129,
+      "step": 775
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.881011097410605e-05,
+      "loss": 0.0026,
+      "step": 776
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.879469790382244e-05,
+      "loss": 0.0014,
+      "step": 777
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.8779284833538844e-05,
+      "loss": 0.0021,
+      "step": 778
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.876387176325525e-05,
+      "loss": 0.0013,
+      "step": 779
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.874845869297164e-05,
+      "loss": 0.0018,
+      "step": 780
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.8733045622688045e-05,
+      "loss": 0.0096,
+      "step": 781
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.871763255240444e-05,
+      "loss": 0.0028,
+      "step": 782
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.8702219482120836e-05,
+      "loss": 0.0742,
+      "step": 783
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.868680641183724e-05,
+      "loss": 0.3255,
+      "step": 784
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.867139334155364e-05,
+      "loss": 0.4515,
+      "step": 785
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8655980271270036e-05,
+      "loss": 0.0019,
+      "step": 786
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.864056720098644e-05,
+      "loss": 0.0023,
+      "step": 787
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.862515413070284e-05,
+      "loss": 0.0096,
+      "step": 788
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8609741060419236e-05,
+      "loss": 0.1568,
+      "step": 789
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.859432799013564e-05,
+      "loss": 0.1501,
+      "step": 790
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.857891491985204e-05,
+      "loss": 0.278,
+      "step": 791
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8563501849568436e-05,
+      "loss": 0.2169,
+      "step": 792
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.854808877928484e-05,
+      "loss": 0.2421,
+      "step": 793
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8532675709001234e-05,
+      "loss": 0.1288,
+      "step": 794
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8517262638717636e-05,
+      "loss": 0.0946,
+      "step": 795
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.850184956843403e-05,
+      "loss": 0.0706,
+      "step": 796
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8486436498150434e-05,
+      "loss": 0.0273,
+      "step": 797
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.847102342786683e-05,
+      "loss": 0.0427,
+      "step": 798
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.845561035758323e-05,
+      "loss": 0.0752,
+      "step": 799
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8440197287299634e-05,
+      "loss": 0.6192,
+      "step": 800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.842478421701603e-05,
+      "loss": 0.0923,
+      "step": 801
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.840937114673243e-05,
+      "loss": 0.0593,
+      "step": 802
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8393958076448834e-05,
+      "loss": 0.0135,
+      "step": 803
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.837854500616523e-05,
+      "loss": 0.0103,
+      "step": 804
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.836313193588163e-05,
+      "loss": 0.0453,
+      "step": 805
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.8347718865598034e-05,
+      "loss": 0.0046,
+      "step": 806
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.833230579531443e-05,
+      "loss": 0.0033,
+      "step": 807
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.831689272503083e-05,
+      "loss": 0.0636,
+      "step": 808
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.830147965474723e-05,
+      "loss": 0.0041,
+      "step": 809
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.828606658446362e-05,
+      "loss": 0.0028,
+      "step": 810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.8270653514180025e-05,
+      "loss": 0.0011,
+      "step": 811
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.825524044389643e-05,
+      "loss": 0.0013,
+      "step": 812
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.823982737361282e-05,
+      "loss": 0.0022,
+      "step": 813
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.8224414303329225e-05,
+      "loss": 0.0692,
+      "step": 814
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.820900123304563e-05,
+      "loss": 0.0011,
+      "step": 815
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.819358816276202e-05,
+      "loss": 0.0015,
+      "step": 816
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.8178175092478425e-05,
+      "loss": 0.0013,
+      "step": 817
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.816276202219483e-05,
+      "loss": 0.0008,
+      "step": 818
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.814734895191122e-05,
+      "loss": 0.0598,
+      "step": 819
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8131935881627625e-05,
+      "loss": 0.0122,
+      "step": 820
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.811652281134403e-05,
+      "loss": 0.0009,
+      "step": 821
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.810110974106042e-05,
+      "loss": 0.0014,
+      "step": 822
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.808569667077682e-05,
+      "loss": 0.0012,
+      "step": 823
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.807028360049322e-05,
+      "loss": 0.0015,
+      "step": 824
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8054870530209616e-05,
+      "loss": 0.2001,
+      "step": 825
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.803945745992602e-05,
+      "loss": 0.0011,
+      "step": 826
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8024044389642414e-05,
+      "loss": 0.4526,
+      "step": 827
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8008631319358816e-05,
+      "loss": 0.0006,
+      "step": 828
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.799321824907522e-05,
+      "loss": 0.2262,
+      "step": 829
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7977805178791614e-05,
+      "loss": 0.0106,
+      "step": 830
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7962392108508016e-05,
+      "loss": 0.0349,
+      "step": 831
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.794697903822442e-05,
+      "loss": 0.0037,
+      "step": 832
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7931565967940814e-05,
+      "loss": 0.2165,
+      "step": 833
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7916152897657216e-05,
+      "loss": 0.0013,
+      "step": 834
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.790073982737362e-05,
+      "loss": 0.1627,
+      "step": 835
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.7885326757090014e-05,
+      "loss": 0.3021,
+      "step": 836
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.786991368680641e-05,
+      "loss": 0.0008,
+      "step": 837
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.785450061652281e-05,
+      "loss": 0.0006,
+      "step": 838
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.783908754623921e-05,
+      "loss": 0.1849,
+      "step": 839
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.782367447595561e-05,
+      "loss": 0.0009,
+      "step": 840
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.780826140567201e-05,
+      "loss": 0.0008,
+      "step": 841
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.779284833538841e-05,
+      "loss": 0.0006,
+      "step": 842
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.777743526510481e-05,
+      "loss": 0.0008,
+      "step": 843
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.776202219482121e-05,
+      "loss": 0.0011,
+      "step": 844
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.774660912453761e-05,
+      "loss": 0.0068,
+      "step": 845
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.773119605425401e-05,
+      "loss": 0.001,
+      "step": 846
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.771578298397041e-05,
+      "loss": 0.0007,
+      "step": 847
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.770036991368681e-05,
+      "loss": 0.0128,
+      "step": 848
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.768495684340321e-05,
+      "loss": 0.0011,
+      "step": 849
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7669543773119605e-05,
+      "loss": 0.0006,
+      "step": 850
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.765413070283601e-05,
+      "loss": 0.0006,
+      "step": 851
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.76387176325524e-05,
+      "loss": 0.0108,
+      "step": 852
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7623304562268806e-05,
+      "loss": 0.0005,
+      "step": 853
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.76078914919852e-05,
+      "loss": 0.0023,
+      "step": 854
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.75924784217016e-05,
+      "loss": 0.0008,
+      "step": 855
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7577065351418006e-05,
+      "loss": 0.001,
+      "step": 856
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.75616522811344e-05,
+      "loss": 0.0043,
+      "step": 857
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7546239210850803e-05,
+      "loss": 0.0004,
+      "step": 858
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7530826140567206e-05,
+      "loss": 0.003,
+      "step": 859
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.75154130702836e-05,
+      "loss": 0.0089,
+      "step": 860
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.0262,
+      "step": 861
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7484586929716406e-05,
+      "loss": 0.0048,
+      "step": 862
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.74691738594328e-05,
+      "loss": 0.0007,
+      "step": 863
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7453760789149204e-05,
+      "loss": 0.0008,
+      "step": 864
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.74383477188656e-05,
+      "loss": 0.0005,
+      "step": 865
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7422934648581995e-05,
+      "loss": 0.0005,
+      "step": 866
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.74075215782984e-05,
+      "loss": 0.0008,
+      "step": 867
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.73921085080148e-05,
+      "loss": 0.1344,
+      "step": 868
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7376695437731195e-05,
+      "loss": 0.001,
+      "step": 869
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.73612823674476e-05,
+      "loss": 0.0005,
+      "step": 870
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7345869297164e-05,
+      "loss": 0.001,
+      "step": 871
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.7330456226880395e-05,
+      "loss": 0.0254,
+      "step": 872
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.73150431565968e-05,
+      "loss": 0.0006,
+      "step": 873
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.72996300863132e-05,
+      "loss": 0.0004,
+      "step": 874
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.7284217016029595e-05,
+      "loss": 0.0172,
+      "step": 875
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.7268803945746e-05,
+      "loss": 0.0008,
+      "step": 876
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.72533908754624e-05,
+      "loss": 0.0004,
+      "step": 877
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.7237977805178795e-05,
+      "loss": 0.0031,
+      "step": 878
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.722256473489519e-05,
+      "loss": 0.0017,
+      "step": 879
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.720715166461159e-05,
+      "loss": 0.0003,
+      "step": 880
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.719173859432799e-05,
+      "loss": 0.0004,
+      "step": 881
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.717632552404439e-05,
+      "loss": 0.0004,
+      "step": 882
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.716091245376079e-05,
+      "loss": 0.0003,
+      "step": 883
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.714549938347719e-05,
+      "loss": 0.0003,
+      "step": 884
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.713008631319359e-05,
+      "loss": 0.0006,
+      "step": 885
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.711467324290999e-05,
+      "loss": 0.0004,
+      "step": 886
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.709926017262639e-05,
+      "loss": 0.002,
+      "step": 887
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.708384710234279e-05,
+      "loss": 0.0004,
+      "step": 888
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.706843403205919e-05,
+      "loss": 0.0002,
+      "step": 889
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.705302096177559e-05,
+      "loss": 0.0005,
+      "step": 890
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.703760789149199e-05,
+      "loss": 0.0003,
+      "step": 891
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7022194821208386e-05,
+      "loss": 0.0006,
+      "step": 892
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.700678175092478e-05,
+      "loss": 0.0005,
+      "step": 893
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.6991368680641184e-05,
+      "loss": 0.0006,
+      "step": 894
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6975955610357586e-05,
+      "loss": 0.3184,
+      "step": 895
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.696054254007398e-05,
+      "loss": 0.0006,
+      "step": 896
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6945129469790384e-05,
+      "loss": 0.0004,
+      "step": 897
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.692971639950678e-05,
+      "loss": 0.963,
+      "step": 898
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.691430332922318e-05,
+      "loss": 0.366,
+      "step": 899
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6898890258939584e-05,
+      "loss": 0.0003,
+      "step": 900
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.688347718865598e-05,
+      "loss": 0.0002,
+      "step": 901
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.686806411837238e-05,
+      "loss": 0.0004,
+      "step": 902
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.6852651048088784e-05,
+      "loss": 0.0003,
+      "step": 903
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.683723797780518e-05,
+      "loss": 0.0004,
+      "step": 904
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.682182490752158e-05,
+      "loss": 0.0003,
+      "step": 905
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.680641183723798e-05,
+      "loss": 0.0004,
+      "step": 906
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.679099876695438e-05,
+      "loss": 0.0009,
+      "step": 907
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6775585696670775e-05,
+      "loss": 0.4215,
+      "step": 908
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.676017262638718e-05,
+      "loss": 0.0005,
+      "step": 909
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.674475955610357e-05,
+      "loss": 0.0004,
+      "step": 910
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6729346485819975e-05,
+      "loss": 0.0003,
+      "step": 911
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.671393341553638e-05,
+      "loss": 0.0004,
+      "step": 912
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.669852034525277e-05,
+      "loss": 0.3228,
+      "step": 913
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.6683107274969175e-05,
+      "loss": 0.0496,
+      "step": 914
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.666769420468558e-05,
+      "loss": 0.0004,
+      "step": 915
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.665228113440197e-05,
+      "loss": 0.0003,
+      "step": 916
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6636868064118375e-05,
+      "loss": 0.0004,
+      "step": 917
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.662145499383478e-05,
+      "loss": 0.0004,
+      "step": 918
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.660604192355117e-05,
+      "loss": 0.0004,
+      "step": 919
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6590628853267576e-05,
+      "loss": 0.0005,
+      "step": 920
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.657521578298397e-05,
+      "loss": 0.0004,
+      "step": 921
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6559802712700367e-05,
+      "loss": 0.0005,
+      "step": 922
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.654438964241677e-05,
+      "loss": 0.0003,
+      "step": 923
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.652897657213317e-05,
+      "loss": 0.1785,
+      "step": 924
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.6513563501849567e-05,
+      "loss": 0.0004,
+      "step": 925
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.649815043156597e-05,
+      "loss": 0.0004,
+      "step": 926
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.648273736128237e-05,
+      "loss": 0.0005,
+      "step": 927
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.646732429099877e-05,
+      "loss": 0.0049,
+      "step": 928
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.645191122071517e-05,
+      "loss": 0.0004,
+      "step": 929
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.643649815043157e-05,
+      "loss": 0.0008,
+      "step": 930
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.642108508014797e-05,
+      "loss": 0.0003,
+      "step": 931
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.640567200986437e-05,
+      "loss": 0.0004,
+      "step": 932
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.639025893958077e-05,
+      "loss": 0.0005,
+      "step": 933
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.637484586929717e-05,
+      "loss": 0.0005,
+      "step": 934
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.635943279901356e-05,
+      "loss": 0.0006,
+      "step": 935
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.6344019728729965e-05,
+      "loss": 0.0007,
+      "step": 936
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.632860665844636e-05,
+      "loss": 0.0794,
+      "step": 937
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.631319358816276e-05,
+      "loss": 0.004,
+      "step": 938
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6297780517879165e-05,
+      "loss": 0.0006,
+      "step": 939
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.628236744759556e-05,
+      "loss": 0.0003,
+      "step": 940
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.626695437731196e-05,
+      "loss": 0.0005,
+      "step": 941
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6251541307028365e-05,
+      "loss": 0.0003,
+      "step": 942
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.623612823674476e-05,
+      "loss": 0.0003,
+      "step": 943
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.622071516646116e-05,
+      "loss": 0.5048,
+      "step": 944
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6205302096177565e-05,
+      "loss": 0.4882,
+      "step": 945
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.618988902589396e-05,
+      "loss": 0.0007,
+      "step": 946
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.617447595561036e-05,
+      "loss": 0.0006,
+      "step": 947
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.615906288532676e-05,
+      "loss": 0.0005,
+      "step": 948
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6143649815043154e-05,
+      "loss": 0.001,
+      "step": 949
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6128236744759556e-05,
+      "loss": 0.0004,
+      "step": 950
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.611282367447596e-05,
+      "loss": 0.0008,
+      "step": 951
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6097410604192354e-05,
+      "loss": 0.0147,
+      "step": 952
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6081997533908756e-05,
+      "loss": 0.0006,
+      "step": 953
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.606658446362516e-05,
+      "loss": 0.0008,
+      "step": 954
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6051171393341554e-05,
+      "loss": 0.0009,
+      "step": 955
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6035758323057956e-05,
+      "loss": 0.2269,
+      "step": 956
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.602034525277436e-05,
+      "loss": 0.001,
+      "step": 957
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6004932182490754e-05,
+      "loss": 0.012,
+      "step": 958
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.5989519112207156e-05,
+      "loss": 0.0018,
+      "step": 959
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.597410604192356e-05,
+      "loss": 0.0012,
+      "step": 960
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.5958692971639954e-05,
+      "loss": 0.001,
+      "step": 961
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.594327990135635e-05,
+      "loss": 0.0021,
+      "step": 962
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.592786683107275e-05,
+      "loss": 0.0015,
+      "step": 963
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.591245376078915e-05,
+      "loss": 0.0012,
+      "step": 964
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.589704069050555e-05,
+      "loss": 0.0017,
+      "step": 965
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.588162762022195e-05,
+      "loss": 0.0372,
+      "step": 966
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.586621454993835e-05,
+      "loss": 0.0024,
+      "step": 967
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.585080147965475e-05,
+      "loss": 0.0018,
+      "step": 968
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.583538840937115e-05,
+      "loss": 0.0013,
+      "step": 969
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.581997533908755e-05,
+      "loss": 0.3861,
+      "step": 970
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.580456226880395e-05,
+      "loss": 0.281,
+      "step": 971
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5789149198520345e-05,
+      "loss": 0.0025,
+      "step": 972
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.577373612823675e-05,
+      "loss": 0.0006,
+      "step": 973
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.575832305795315e-05,
+      "loss": 0.001,
+      "step": 974
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.5742909987669545e-05,
+      "loss": 0.3288,
+      "step": 975
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.572749691738595e-05,
+      "loss": 0.0015,
+      "step": 976
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.571208384710234e-05,
+      "loss": 0.0148,
+      "step": 977
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.569667077681874e-05,
+      "loss": 0.0049,
+      "step": 978
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.568125770653514e-05,
+      "loss": 0.001,
+      "step": 979
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.566584463625154e-05,
+      "loss": 0.2691,
+      "step": 980
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.565043156596794e-05,
+      "loss": 0.0017,
+      "step": 981
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.563501849568434e-05,
+      "loss": 0.004,
+      "step": 982
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.561960542540074e-05,
+      "loss": 0.0037,
+      "step": 983
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.560419235511714e-05,
+      "loss": 0.1971,
+      "step": 984
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.558877928483354e-05,
+      "loss": 0.002,
+      "step": 985
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.557336621454994e-05,
+      "loss": 0.1229,
+      "step": 986
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.555795314426634e-05,
+      "loss": 0.0672,
+      "step": 987
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.554254007398274e-05,
+      "loss": 0.0057,
+      "step": 988
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.552712700369914e-05,
+      "loss": 0.0055,
+      "step": 989
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.551171393341554e-05,
+      "loss": 0.0078,
+      "step": 990
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5496300863131934e-05,
+      "loss": 0.0752,
+      "step": 991
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.5480887792848337e-05,
+      "loss": 0.0034,
+      "step": 992
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.546547472256473e-05,
+      "loss": 0.0117,
+      "step": 993
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5450061652281134e-05,
+      "loss": 0.0052,
+      "step": 994
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5434648581997537e-05,
+      "loss": 0.0074,
+      "step": 995
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.541923551171393e-05,
+      "loss": 0.0018,
+      "step": 996
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5403822441430334e-05,
+      "loss": 0.001,
+      "step": 997
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.538840937114674e-05,
+      "loss": 0.075,
+      "step": 998
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.537299630086313e-05,
+      "loss": 0.2192,
+      "step": 999
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5357583230579534e-05,
+      "loss": 0.0014,
+      "step": 1000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.534217016029594e-05,
+      "loss": 0.0011,
+      "step": 1001
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.532675709001233e-05,
+      "loss": 0.0063,
+      "step": 1002
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.5311344019728735e-05,
+      "loss": 0.0039,
+      "step": 1003
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.529593094944513e-05,
+      "loss": 0.0007,
+      "step": 1004
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5280517879161526e-05,
+      "loss": 0.0015,
+      "step": 1005
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.526510480887793e-05,
+      "loss": 0.0015,
+      "step": 1006
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.524969173859433e-05,
+      "loss": 0.004,
+      "step": 1007
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5234278668310726e-05,
+      "loss": 0.2604,
+      "step": 1008
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.521886559802713e-05,
+      "loss": 0.0016,
+      "step": 1009
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.520345252774353e-05,
+      "loss": 0.1464,
+      "step": 1010
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5188039457459926e-05,
+      "loss": 0.0013,
+      "step": 1011
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.517262638717633e-05,
+      "loss": 0.0009,
+      "step": 1012
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.515721331689273e-05,
+      "loss": 0.0005,
+      "step": 1013
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5141800246609126e-05,
+      "loss": 0.0008,
+      "step": 1014
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.512638717632553e-05,
+      "loss": 0.0011,
+      "step": 1015
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.511097410604193e-05,
+      "loss": 0.0007,
+      "step": 1016
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5095561035758326e-05,
+      "loss": 0.0006,
+      "step": 1017
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.508014796547472e-05,
+      "loss": 0.0007,
+      "step": 1018
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5064734895191124e-05,
+      "loss": 0.005,
+      "step": 1019
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.504932182490752e-05,
+      "loss": 0.0126,
+      "step": 1020
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.503390875462392e-05,
+      "loss": 0.0005,
+      "step": 1021
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5018495684340324e-05,
+      "loss": 0.001,
+      "step": 1022
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.500308261405672e-05,
+      "loss": 0.0008,
+      "step": 1023
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.498766954377312e-05,
+      "loss": 0.0005,
+      "step": 1024
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.4972256473489524e-05,
+      "loss": 0.0005,
+      "step": 1025
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.495684340320592e-05,
+      "loss": 0.0061,
+      "step": 1026
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.494143033292232e-05,
+      "loss": 0.001,
+      "step": 1027
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4926017262638724e-05,
+      "loss": 0.0004,
+      "step": 1028
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.491060419235512e-05,
+      "loss": 0.0006,
+      "step": 1029
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.489519112207152e-05,
+      "loss": 0.0008,
+      "step": 1030
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.487977805178792e-05,
+      "loss": 0.4525,
+      "step": 1031
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.486436498150432e-05,
+      "loss": 0.4309,
+      "step": 1032
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4848951911220715e-05,
+      "loss": 0.0904,
+      "step": 1033
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.483353884093712e-05,
+      "loss": 0.3958,
+      "step": 1034
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.481812577065351e-05,
+      "loss": 0.0362,
+      "step": 1035
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4802712700369915e-05,
+      "loss": 0.001,
+      "step": 1036
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.478729963008632e-05,
+      "loss": 0.253,
+      "step": 1037
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.477188655980271e-05,
+      "loss": 0.0017,
+      "step": 1038
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4756473489519115e-05,
+      "loss": 0.0467,
+      "step": 1039
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.474106041923552e-05,
+      "loss": 0.2376,
+      "step": 1040
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.472564734895191e-05,
+      "loss": 0.0014,
+      "step": 1041
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4710234278668315e-05,
+      "loss": 0.0014,
+      "step": 1042
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.469482120838472e-05,
+      "loss": 0.0013,
+      "step": 1043
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.467940813810111e-05,
+      "loss": 0.3713,
+      "step": 1044
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4663995067817515e-05,
+      "loss": 0.0062,
+      "step": 1045
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.464858199753391e-05,
+      "loss": 0.0093,
+      "step": 1046
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.4633168927250306e-05,
+      "loss": 0.0013,
+      "step": 1047
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.461775585696671e-05,
+      "loss": 0.0628,
+      "step": 1048
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4602342786683104e-05,
+      "loss": 0.2886,
+      "step": 1049
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4586929716399506e-05,
+      "loss": 0.0012,
+      "step": 1050
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.457151664611591e-05,
+      "loss": 0.0017,
+      "step": 1051
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4556103575832304e-05,
+      "loss": 0.2664,
+      "step": 1052
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4540690505548706e-05,
+      "loss": 0.002,
+      "step": 1053
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.452527743526511e-05,
+      "loss": 0.3553,
+      "step": 1054
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4509864364981504e-05,
+      "loss": 0.1123,
+      "step": 1055
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4494451294697906e-05,
+      "loss": 0.3231,
+      "step": 1056
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.447903822441431e-05,
+      "loss": 0.0034,
+      "step": 1057
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4463625154130704e-05,
+      "loss": 0.1834,
+      "step": 1058
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.4448212083847106e-05,
+      "loss": 0.0121,
+      "step": 1059
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.44327990135635e-05,
+      "loss": 0.1922,
+      "step": 1060
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.44173859432799e-05,
+      "loss": 0.0315,
+      "step": 1061
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.44019728729963e-05,
+      "loss": 0.0335,
+      "step": 1062
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43865598027127e-05,
+      "loss": 0.0226,
+      "step": 1063
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43711467324291e-05,
+      "loss": 0.0236,
+      "step": 1064
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43557336621455e-05,
+      "loss": 0.005,
+      "step": 1065
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43403205918619e-05,
+      "loss": 0.0091,
+      "step": 1066
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43249075215783e-05,
+      "loss": 0.0053,
+      "step": 1067
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.43094944512947e-05,
+      "loss": 0.0068,
+      "step": 1068
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.42940813810111e-05,
+      "loss": 0.0026,
+      "step": 1069
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.42786683107275e-05,
+      "loss": 0.0027,
+      "step": 1070
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.42632552404439e-05,
+      "loss": 0.0541,
+      "step": 1071
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.42478421701603e-05,
+      "loss": 0.0021,
+      "step": 1072
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.42324290998767e-05,
+      "loss": 0.0019,
+      "step": 1073
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.421701602959309e-05,
+      "loss": 0.0014,
+      "step": 1074
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4201602959309496e-05,
+      "loss": 0.0218,
+      "step": 1075
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.418618988902589e-05,
+      "loss": 0.0016,
+      "step": 1076
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.417077681874229e-05,
+      "loss": 0.001,
+      "step": 1077
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4155363748458696e-05,
+      "loss": 0.0012,
+      "step": 1078
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.413995067817509e-05,
+      "loss": 0.2197,
+      "step": 1079
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4124537607891493e-05,
+      "loss": 0.0009,
+      "step": 1080
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4109124537607896e-05,
+      "loss": 0.4255,
+      "step": 1081
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.409371146732429e-05,
+      "loss": 0.0007,
+      "step": 1082
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4078298397040693e-05,
+      "loss": 0.0018,
+      "step": 1083
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4062885326757096e-05,
+      "loss": 0.0013,
+      "step": 1084
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.404747225647349e-05,
+      "loss": 0.0015,
+      "step": 1085
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4032059186189894e-05,
+      "loss": 0.0027,
+      "step": 1086
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.401664611590629e-05,
+      "loss": 0.0017,
+      "step": 1087
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.400123304562269e-05,
+      "loss": 0.3513,
+      "step": 1088
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.398581997533909e-05,
+      "loss": 0.0033,
+      "step": 1089
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.397040690505549e-05,
+      "loss": 0.0328,
+      "step": 1090
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.3954993834771885e-05,
+      "loss": 0.0497,
+      "step": 1091
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.393958076448829e-05,
+      "loss": 0.0048,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.392416769420469e-05,
+      "loss": 0.003,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3908754623921085e-05,
+      "loss": 0.0018,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.389334155363749e-05,
+      "loss": 0.0925,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.387792848335389e-05,
+      "loss": 0.0055,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3862515413070285e-05,
+      "loss": 0.0048,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.384710234278669e-05,
+      "loss": 0.002,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9888433515482696,
+      "eval_loss": 0.046589694917201996,
+      "eval_runtime": 41.2835,
+      "eval_samples_per_second": 106.386,
+      "eval_steps_per_second": 6.661,
+      "step": 1098
+    }
+  ],
+  "max_steps": 3294,
+  "num_train_epochs": 3,
+  "total_flos": 2325713559778560.0,
+  "trial_name": null,
+  "trial_params": null
+}