diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5430 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5679,
+      "step": 1
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.9276,
+      "step": 2
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6e-06,
+      "loss": 0.8136,
+      "step": 3
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6904,
+      "step": 4
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.8459,
+      "step": 5
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6818,
+      "step": 6
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.4e-05,
+      "loss": 0.765,
+      "step": 7
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.7109,
+      "step": 8
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.8e-05,
+      "loss": 0.7278,
+      "step": 9
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 0.8511,
+      "step": 10
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999993769982229e-05,
+      "loss": 0.7941,
+      "step": 11
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9999750800065415e-05,
+      "loss": 0.7982,
+      "step": 12
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.999943930305816e-05,
+      "loss": 0.7972,
+      "step": 13
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.999900321268178e-05,
+      "loss": 0.8692,
+      "step": 14
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9998442534369986e-05,
+      "loss": 0.6429,
+      "step": 15
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9997757275108847e-05,
+      "loss": 0.6293,
+      "step": 16
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.999694744343671e-05,
+      "loss": 0.7597,
+      "step": 17
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9996013049444117e-05,
+      "loss": 0.5985,
+      "step": 18
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9994954104773647e-05,
+      "loss": 0.6738,
+      "step": 19
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9993770622619784e-05,
+      "loss": 0.871,
+      "step": 20
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9992462617728763e-05,
+      "loss": 0.6064,
+      "step": 21
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9991030106398367e-05,
+      "loss": 0.6053,
+      "step": 22
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9989473106477742e-05,
+      "loss": 0.5962,
+      "step": 23
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9987791637367157e-05,
+      "loss": 0.6548,
+      "step": 24
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9985985720017786e-05,
+      "loss": 0.7324,
+      "step": 25
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9984055376931414e-05,
+      "loss": 0.6753,
+      "step": 26
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.998200063216019e-05,
+      "loss": 0.5786,
+      "step": 27
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9979821511306308e-05,
+      "loss": 0.5941,
+      "step": 28
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9977518041521683e-05,
+      "loss": 0.5032,
+      "step": 29
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9975090251507637e-05,
+      "loss": 0.5617,
+      "step": 30
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9972538171514525e-05,
+      "loss": 0.6562,
+      "step": 31
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.996986183334134e-05,
+      "loss": 0.5879,
+      "step": 32
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9967061270335363e-05,
+      "loss": 0.538,
+      "step": 33
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9964136517391708e-05,
+      "loss": 0.5936,
+      "step": 34
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9961087610952893e-05,
+      "loss": 0.459,
+      "step": 35
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9957914589008405e-05,
+      "loss": 0.5873,
+      "step": 36
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.995461749109421e-05,
+      "loss": 0.5291,
+      "step": 37
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9951196358292266e-05,
+      "loss": 0.6035,
+      "step": 38
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.994765123323001e-05,
+      "loss": 0.6045,
+      "step": 39
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9943982160079823e-05,
+      "loss": 0.5251,
+      "step": 40
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9940189184558487e-05,
+      "loss": 0.5626,
+      "step": 41
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9936272353926616e-05,
+      "loss": 0.54,
+      "step": 42
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9932231716988057e-05,
+      "loss": 0.5188,
+      "step": 43
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9928067324089286e-05,
+      "loss": 0.5038,
+      "step": 44
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9923779227118793e-05,
+      "loss": 0.4336,
+      "step": 45
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9919367479506413e-05,
+      "loss": 0.5293,
+      "step": 46
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.991483213622268e-05,
+      "loss": 0.5262,
+      "step": 47
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9910173253778136e-05,
+      "loss": 0.474,
+      "step": 48
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.990539089022262e-05,
+      "loss": 0.5727,
+      "step": 49
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9900485105144544e-05,
+      "loss": 0.4906,
+      "step": 50
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.989545595967018e-05,
+      "loss": 0.52,
+      "step": 51
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9890303516462842e-05,
+      "loss": 0.545,
+      "step": 52
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9885027839722166e-05,
+      "loss": 0.4373,
+      "step": 53
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.9879628995183274e-05,
+      "loss": 0.455,
+      "step": 54
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9874107050115954e-05,
+      "loss": 0.4469,
+      "step": 55
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.986846207332384e-05,
+      "loss": 0.5483,
+      "step": 56
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9862694135143547e-05,
+      "loss": 0.4492,
+      "step": 57
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9856803307443782e-05,
+      "loss": 0.553,
+      "step": 58
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9850789663624473e-05,
+      "loss": 0.4403,
+      "step": 59
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.9844653278615836e-05,
+      "loss": 0.2905,
+      "step": 60
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9838394228877443e-05,
+      "loss": 0.4364,
+      "step": 61
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.983201259239728e-05,
+      "loss": 0.4628,
+      "step": 62
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9825508448690764e-05,
+      "loss": 0.5168,
+      "step": 63
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.9818881878799755e-05,
+      "loss": 0.5607,
+      "step": 64
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.9812132965291548e-05,
+      "loss": 0.5577,
+      "step": 65
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.980526179225785e-05,
+      "loss": 0.3578,
+      "step": 66
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9798268445313713e-05,
+      "loss": 0.3537,
+      "step": 67
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.9791153011596497e-05,
+      "loss": 0.286,
+      "step": 68
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.9783915579764755e-05,
+      "loss": 0.431,
+      "step": 69
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.9776556239997146e-05,
+      "loss": 0.4343,
+      "step": 70
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.97690750839913e-05,
+      "loss": 0.5138,
+      "step": 71
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.97614722049627e-05,
+      "loss": 0.4607,
+      "step": 72
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.9753747697643477e-05,
+      "loss": 0.4031,
+      "step": 73
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.9745901658281267e-05,
+      "loss": 0.5382,
+      "step": 74
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.9737934184638006e-05,
+      "loss": 0.5144,
+      "step": 75
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9729845375988694e-05,
+      "loss": 0.5064,
+      "step": 76
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9721635333120176e-05,
+      "loss": 0.4038,
+      "step": 77
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.9713304158329873e-05,
+      "loss": 0.4594,
+      "step": 78
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.9704851955424526e-05,
+      "loss": 0.393,
+      "step": 79
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.9696278829718882e-05,
+      "loss": 0.4057,
+      "step": 80
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.968758488803439e-05,
+      "loss": 0.3919,
+      "step": 81
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.9678770238697876e-05,
+      "loss": 0.433,
+      "step": 82
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.966983499154018e-05,
+      "loss": 0.377,
+      "step": 83
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.96607792578948e-05,
+      "loss": 0.473,
+      "step": 84
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9651603150596497e-05,
+      "loss": 0.426,
+      "step": 85
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9642306783979902e-05,
+      "loss": 0.2673,
+      "step": 86
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.9632890273878063e-05,
+      "loss": 0.3363,
+      "step": 87
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9623353737621035e-05,
+      "loss": 0.2908,
+      "step": 88
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.9613697294034403e-05,
+      "loss": 0.3527,
+      "step": 89
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.9603921063437795e-05,
+      "loss": 0.3323,
+      "step": 90
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.959402516764339e-05,
+      "loss": 0.3735,
+      "step": 91
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.9584009729954395e-05,
+      "loss": 0.3092,
+      "step": 92
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.9573874875163533e-05,
+      "loss": 0.5028,
+      "step": 93
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.9563620729551448e-05,
+      "loss": 0.3488,
+      "step": 94
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.955324742088516e-05,
+      "loss": 0.2567,
+      "step": 95
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.954275507841646e-05,
+      "loss": 0.3016,
+      "step": 96
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.9532143832880307e-05,
+      "loss": 0.3654,
+      "step": 97
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.9521413816493206e-05,
+      "loss": 0.3559,
+      "step": 98
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.4253,
+      "step": 99
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.949959800742991e-05,
+      "loss": 0.2719,
+      "step": 100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.948851248657947e-05,
+      "loss": 0.3276,
+      "step": 101
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.9477308738526207e-05,
+      "loss": 0.3586,
+      "step": 102
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.9465986902869215e-05,
+      "loss": 0.3834,
+      "step": 103
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.9454547120678966e-05,
+      "loss": 0.324,
+      "step": 104
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.9442989534495557e-05,
+      "loss": 0.3047,
+      "step": 105
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9431314288326925e-05,
+      "loss": 0.3083,
+      "step": 106
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.941952152764705e-05,
+      "loss": 0.3903,
+      "step": 107
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9407611399394145e-05,
+      "loss": 0.2626,
+      "step": 108
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9395584051968833e-05,
+      "loss": 0.341,
+      "step": 109
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9383439635232296e-05,
+      "loss": 0.2571,
+      "step": 110
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9371178300504394e-05,
+      "loss": 0.251,
+      "step": 111
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.935880020056179e-05,
+      "loss": 0.2741,
+      "step": 112
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.934630548963606e-05,
+      "loss": 0.3759,
+      "step": 113
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.9333694323411732e-05,
+      "loss": 0.2496,
+      "step": 114
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.9320966859024397e-05,
+      "loss": 0.2505,
+      "step": 115
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.930812325505871e-05,
+      "loss": 0.3331,
+      "step": 116
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.929516367154643e-05,
+      "loss": 0.2619,
+      "step": 117
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.928208826996443e-05,
+      "loss": 0.3748,
+      "step": 118
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.926889721323268e-05,
+      "loss": 0.2056,
+      "step": 119
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.9255590665712214e-05,
+      "loss": 0.1981,
+      "step": 120
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.9242168793203088e-05,
+      "loss": 0.326,
+      "step": 121
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.9228631762942307e-05,
+      "loss": 0.1955,
+      "step": 122
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.9214979743601755e-05,
+      "loss": 0.2742,
+      "step": 123
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.9201212905286074e-05,
+      "loss": 0.3218,
+      "step": 124
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.918733141953056e-05,
+      "loss": 0.2482,
+      "step": 125
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.9173335459299022e-05,
+      "loss": 0.1862,
+      "step": 126
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.915922519898162e-05,
+      "loss": 0.3037,
+      "step": 127
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.9145000814392696e-05,
+      "loss": 0.3153,
+      "step": 128
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.9130662482768586e-05,
+      "loss": 0.2396,
+      "step": 129
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.911621038276542e-05,
+      "loss": 0.2857,
+      "step": 130
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.9101644694456873e-05,
+      "loss": 0.2374,
+      "step": 131
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.9086965599331938e-05,
+      "loss": 0.2653,
+      "step": 132
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.907217328029266e-05,
+      "loss": 0.1779,
+      "step": 133
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.9057267921651865e-05,
+      "loss": 0.2881,
+      "step": 134
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.904224970913085e-05,
+      "loss": 0.2646,
+      "step": 135
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.902711882985708e-05,
+      "loss": 0.2802,
+      "step": 136
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.901187547236184e-05,
+      "loss": 0.1165,
+      "step": 137
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8996519826577907e-05,
+      "loss": 0.1581,
+      "step": 138
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8981052083837186e-05,
+      "loss": 0.2183,
+      "step": 139
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8965472436868288e-05,
+      "loss": 0.2791,
+      "step": 140
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.894978107979417e-05,
+      "loss": 0.1321,
+      "step": 141
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8933978208129705e-05,
+      "loss": 0.239,
+      "step": 142
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.891806401877923e-05,
+      "loss": 0.2046,
+      "step": 143
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.8902038710034113e-05,
+      "loss": 0.2405,
+      "step": 144
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.888590248157027e-05,
+      "loss": 0.3314,
+      "step": 145
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.886965553444568e-05,
+      "loss": 0.2327,
+      "step": 146
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.8853298071097885e-05,
+      "loss": 0.1801,
+      "step": 147
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.883683029534145e-05,
+      "loss": 0.2249,
+      "step": 148
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.882025241236546e-05,
+      "loss": 0.2297,
+      "step": 149
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8803564628730916e-05,
+      "loss": 0.2726,
+      "step": 150
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.8786767152368197e-05,
+      "loss": 0.1512,
+      "step": 151
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.876986019257446e-05,
+      "loss": 0.2375,
+      "step": 152
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.875284396001102e-05,
+      "loss": 0.1998,
+      "step": 153
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.873571866670074e-05,
+      "loss": 0.1675,
+      "step": 154
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.8718484526025387e-05,
+      "loss": 0.1944,
+      "step": 155
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.8701141752722966e-05,
+      "loss": 0.2193,
+      "step": 156
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.8683690562885043e-05,
+      "loss": 0.1624,
+      "step": 157
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.866613117395407e-05,
+      "loss": 0.1033,
+      "step": 158
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.8648463804720655e-05,
+      "loss": 0.1854,
+      "step": 159
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.8630688675320844e-05,
+      "loss": 0.1867,
+      "step": 160
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.8612806007233382e-05,
+      "loss": 0.1184,
+      "step": 161
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.8594816023276954e-05,
+      "loss": 0.2329,
+      "step": 162
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.8576718947607394e-05,
+      "loss": 0.1073,
+      "step": 163
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.855851500571491e-05,
+      "loss": 0.2052,
+      "step": 164
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.8540204424421264e-05,
+      "loss": 0.2606,
+      "step": 165
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.8521787431876954e-05,
+      "loss": 0.131,
+      "step": 166
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.8503264257558356e-05,
+      "loss": 0.1578,
+      "step": 167
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.848463513226488e-05,
+      "loss": 0.1752,
+      "step": 168
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8465900288116098e-05,
+      "loss": 0.2435,
+      "step": 169
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.8447059958548822e-05,
+      "loss": 0.2357,
+      "step": 170
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.8428114378314237e-05,
+      "loss": 0.2525,
+      "step": 171
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.840906378347494e-05,
+      "loss": 0.1547,
+      "step": 172
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8389908411402026e-05,
+      "loss": 0.1657,
+      "step": 173
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8370648500772107e-05,
+      "loss": 0.2618,
+      "step": 174
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8351284291564358e-05,
+      "loss": 0.1614,
+      "step": 175
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8331816025057508e-05,
+      "loss": 0.2087,
+      "step": 176
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.8312243943826856e-05,
+      "loss": 0.1646,
+      "step": 177
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.8292568291741228e-05,
+      "loss": 0.2844,
+      "step": 178
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.8272789313959944e-05,
+      "loss": 0.2034,
+      "step": 179
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8252907256929777e-05,
+      "loss": 0.1028,
+      "step": 180
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.8232922368381855e-05,
+      "loss": 0.1161,
+      "step": 181
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.8212834897328614e-05,
+      "loss": 0.1904,
+      "step": 182
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.8192645094060643e-05,
+      "loss": 0.1951,
+      "step": 183
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.8172353210143613e-05,
+      "loss": 0.2307,
+      "step": 184
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.815195949841512e-05,
+      "loss": 0.1616,
+      "step": 185
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.813146421298154e-05,
+      "loss": 0.4231,
+      "step": 186
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.8110867609214854e-05,
+      "loss": 0.1671,
+      "step": 187
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.23,
+      "step": 188
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.8069371474479055e-05,
+      "loss": 0.145,
+      "step": 189
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.804847246055326e-05,
+      "loss": 0.1641,
+      "step": 190
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.802747316237454e-05,
+      "loss": 0.1772,
+      "step": 191
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.8006373841594905e-05,
+      "loss": 0.1625,
+      "step": 192
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.798517476111264e-05,
+      "loss": 0.1477,
+      "step": 193
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.7963876185069032e-05,
+      "loss": 0.0559,
+      "step": 194
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.794247837884511e-05,
+      "loss": 0.1295,
+      "step": 195
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.792098160905829e-05,
+      "loss": 0.1457,
+      "step": 196
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.7899386143559093e-05,
+      "loss": 0.159,
+      "step": 197
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.7877692251427783e-05,
+      "loss": 0.1064,
+      "step": 198
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.785590020297103e-05,
+      "loss": 0.2972,
+      "step": 199
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.7834010269718526e-05,
+      "loss": 0.1485,
+      "step": 200
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.7812022724419624e-05,
+      "loss": 0.2076,
+      "step": 201
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.778993784103992e-05,
+      "loss": 0.2338,
+      "step": 202
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.776775589475784e-05,
+      "loss": 0.3067,
+      "step": 203
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.774547716196123e-05,
+      "loss": 0.1041,
+      "step": 204
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.772310192024389e-05,
+      "loss": 0.1699,
+      "step": 205
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.7700630448402125e-05,
+      "loss": 0.2843,
+      "step": 206
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.7678063026431275e-05,
+      "loss": 0.1703,
+      "step": 207
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.7655399935522216e-05,
+      "loss": 0.1649,
+      "step": 208
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.7632641458057874e-05,
+      "loss": 0.2231,
+      "step": 209
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.7609787877609678e-05,
+      "loss": 0.1005,
+      "step": 210
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.758683947893406e-05,
+      "loss": 0.2017,
+      "step": 211
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.756379654796888e-05,
+      "loss": 0.1489,
+      "step": 212
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.7540659371829877e-05,
+      "loss": 0.137,
+      "step": 213
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.7517428238807085e-05,
+      "loss": 0.0878,
+      "step": 214
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.7494103438361252e-05,
+      "loss": 0.3383,
+      "step": 215
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.747068526112022e-05,
+      "loss": 0.2366,
+      "step": 216
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.744717399887531e-05,
+      "loss": 0.0469,
+      "step": 217
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.7423569944577677e-05,
+      "loss": 0.1412,
+      "step": 218
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7399873392334692e-05,
+      "loss": 0.1721,
+      "step": 219
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7376084637406222e-05,
+      "loss": 0.1308,
+      "step": 220
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.735220397620101e-05,
+      "loss": 0.2229,
+      "step": 221
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.732823170627294e-05,
+      "loss": 0.094,
+      "step": 222
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.7304168126317348e-05,
+      "loss": 0.0742,
+      "step": 223
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.728001353616729e-05,
+      "loss": 0.1997,
+      "step": 224
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.7255768236789826e-05,
+      "loss": 0.1241,
+      "step": 225
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.7231432530282246e-05,
+      "loss": 0.2281,
+      "step": 226
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.7207006719868314e-05,
+      "loss": 0.2825,
+      "step": 227
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.71824911098945e-05,
+      "loss": 0.1323,
+      "step": 228
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7157886005826173e-05,
+      "loss": 0.1151,
+      "step": 229
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7133191714243805e-05,
+      "loss": 0.1231,
+      "step": 230
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7108408542839146e-05,
+      "loss": 0.1965,
+      "step": 231
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.7083536800411392e-05,
+      "loss": 0.0313,
+      "step": 232
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.7058576796863342e-05,
+      "loss": 0.0978,
+      "step": 233
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.7033528843197523e-05,
+      "loss": 0.2916,
+      "step": 234
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.700839325151233e-05,
+      "loss": 0.1508,
+      "step": 235
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.698317033499813e-05,
+      "loss": 0.1523,
+      "step": 236
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.6957860407933354e-05,
+      "loss": 0.1806,
+      "step": 237
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.69324637856806e-05,
+      "loss": 0.1856,
+      "step": 238
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.6906980784682677e-05,
+      "loss": 0.1587,
+      "step": 239
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.6881411722458688e-05,
+      "loss": 0.1596,
+      "step": 240
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.6855756917600057e-05,
+      "loss": 0.1812,
+      "step": 241
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.683001668976656e-05,
+      "loss": 0.1902,
+      "step": 242
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.6804191359682356e-05,
+      "loss": 0.2641,
+      "step": 243
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.6778281249131973e-05,
+      "loss": 0.1197,
+      "step": 244
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6752286680956306e-05,
+      "loss": 0.1497,
+      "step": 245
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.6726207979048604e-05,
+      "loss": 0.1216,
+      "step": 246
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.6700045468350418e-05,
+      "loss": 0.2281,
+      "step": 247
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.667379947484756e-05,
+      "loss": 0.0862,
+      "step": 248
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.6647470325566045e-05,
+      "loss": 0.1606,
+      "step": 249
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.6621058348568008e-05,
+      "loss": 0.1685,
+      "step": 250
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.6594563872947613e-05,
+      "loss": 0.0948,
+      "step": 251
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.656798722882698e-05,
+      "loss": 0.1327,
+      "step": 252
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.6541328747352033e-05,
+      "loss": 0.1896,
+      "step": 253
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.6514588760688397e-05,
+      "loss": 0.1581,
+      "step": 254
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6487767602017264e-05,
+      "loss": 0.0961,
+      "step": 255
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.6460865605531214e-05,
+      "loss": 0.167,
+      "step": 256
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.6433883106430085e-05,
+      "loss": 0.1535,
+      "step": 257
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6406820440916778e-05,
+      "loss": 0.1329,
+      "step": 258
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.6379677946193064e-05,
+      "loss": 0.0954,
+      "step": 259
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.6352455960455385e-05,
+      "loss": 0.1772,
+      "step": 260
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.6325154822890663e-05,
+      "loss": 0.2473,
+      "step": 261
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.6297774873672036e-05,
+      "loss": 0.218,
+      "step": 262
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.6270316453954643e-05,
+      "loss": 0.1639,
+      "step": 263
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.6242779905871375e-05,
+      "loss": 0.0807,
+      "step": 264
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.6215165572528595e-05,
+      "loss": 0.1949,
+      "step": 265
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.618747379800188e-05,
+      "loss": 0.047,
+      "step": 266
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.6159704927331723e-05,
+      "loss": 0.0766,
+      "step": 267
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6131859306519243e-05,
+      "loss": 0.1252,
+      "step": 268
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.610393728252186e-05,
+      "loss": 0.0595,
+      "step": 269
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.607593920324899e-05,
+      "loss": 0.1346,
+      "step": 270
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.6047865417557695e-05,
+      "loss": 0.1994,
+      "step": 271
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.6019716275248342e-05,
+      "loss": 0.1699,
+      "step": 272
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.5991492127060244e-05,
+      "loss": 0.0973,
+      "step": 273
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.596319332466729e-05,
+      "loss": 0.1836,
+      "step": 274
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.5934820220673564e-05,
+      "loss": 0.2331,
+      "step": 275
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.5906373168608952e-05,
+      "loss": 0.2077,
+      "step": 276
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.0968,
+      "step": 277
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.5849258638989166e-05,
+      "loss": 0.0766,
+      "step": 278
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.5820591873083057e-05,
+      "loss": 0.1773,
+      "step": 279
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.5791852582395334e-05,
+      "loss": 0.1721,
+      "step": 280
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.576304112501858e-05,
+      "loss": 0.1446,
+      "step": 281
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.5734157859944574e-05,
+      "loss": 0.0778,
+      "step": 282
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.570520314705983e-05,
+      "loss": 0.1777,
+      "step": 283
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.5676177347141096e-05,
+      "loss": 0.1835,
+      "step": 284
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.5647080821850868e-05,
+      "loss": 0.0859,
+      "step": 285
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.5617913933732892e-05,
+      "loss": 0.078,
+      "step": 286
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.558867704620763e-05,
+      "loss": 0.1068,
+      "step": 287
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.5559370523567734e-05,
+      "loss": 0.0324,
+      "step": 288
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.5529994730973522e-05,
+      "loss": 0.1911,
+      "step": 289
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.5500550034448415e-05,
+      "loss": 0.1982,
+      "step": 290
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.5471036800874375e-05,
+      "loss": 0.182,
+      "step": 291
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.5441455397987342e-05,
+      "loss": 0.0543,
+      "step": 292
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.5411806194372645e-05,
+      "loss": 0.1248,
+      "step": 293
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.5382089559460423e-05,
+      "loss": 0.1653,
+      "step": 294
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.535230586352099e-05,
+      "loss": 0.1096,
+      "step": 295
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.5322455477660274e-05,
+      "loss": 0.0903,
+      "step": 296
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.5292538773815128e-05,
+      "loss": 0.1182,
+      "step": 297
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.5262556124748754e-05,
+      "loss": 0.0673,
+      "step": 298
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.523250790404602e-05,
+      "loss": 0.093,
+      "step": 299
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.5202394486108823e-05,
+      "loss": 0.1897,
+      "step": 300
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.5172216246151426e-05,
+      "loss": 0.0933,
+      "step": 301
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.5141973560195768e-05,
+      "loss": 0.1609,
+      "step": 302
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.5111666805066796e-05,
+      "loss": 0.1047,
+      "step": 303
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.508129635838775e-05,
+      "loss": 0.1929,
+      "step": 304
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.5050862598575474e-05,
+      "loss": 0.1296,
+      "step": 305
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.50203659048357e-05,
+      "loss": 0.1553,
+      "step": 306
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.4989806657158321e-05,
+      "loss": 0.1143,
+      "step": 307
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.4959185236312642e-05,
+      "loss": 0.1663,
+      "step": 308
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.492850202384266e-05,
+      "loss": 0.0577,
+      "step": 309
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.4897757402062285e-05,
+      "loss": 0.2528,
+      "step": 310
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.4866951754050605e-05,
+      "loss": 0.0547,
+      "step": 311
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.4836085463647088e-05,
+      "loss": 0.1592,
+      "step": 312
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.4805158915446804e-05,
+      "loss": 0.1176,
+      "step": 313
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.4774172494795651e-05,
+      "loss": 0.0831,
+      "step": 314
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.4743126587785523e-05,
+      "loss": 0.1111,
+      "step": 315
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.4712021581249534e-05,
+      "loss": 0.1827,
+      "step": 316
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.4680857862757163e-05,
+      "loss": 0.2238,
+      "step": 317
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.4649635820609457e-05,
+      "loss": 0.1041,
+      "step": 318
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.4618355843834162e-05,
+      "loss": 0.0725,
+      "step": 319
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4587018322180906e-05,
+      "loss": 0.2012,
+      "step": 320
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4555623646116325e-05,
+      "loss": 0.2067,
+      "step": 321
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4524172206819195e-05,
+      "loss": 0.1836,
+      "step": 322
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4492664396175569e-05,
+      "loss": 0.0738,
+      "step": 323
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.4461100606773884e-05,
+      "loss": 0.0907,
+      "step": 324
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.4429481231900082e-05,
+      "loss": 0.1252,
+      "step": 325
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.4397806665532693e-05,
+      "loss": 0.0857,
+      "step": 326
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.4366077302337948e-05,
+      "loss": 0.0647,
+      "step": 327
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.4334293537664836e-05,
+      "loss": 0.1304,
+      "step": 328
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.4302455767540189e-05,
+      "loss": 0.1349,
+      "step": 329
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.4270564388663761e-05,
+      "loss": 0.1823,
+      "step": 330
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.4238619798403266e-05,
+      "loss": 0.1129,
+      "step": 331
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.4206622394789432e-05,
+      "loss": 0.1305,
+      "step": 332
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.417457257651104e-05,
+      "loss": 0.1353,
+      "step": 333
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.4142470742909976e-05,
+      "loss": 0.1181,
+      "step": 334
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.411031729397622e-05,
+      "loss": 0.179,
+      "step": 335
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.4078112630342891e-05,
+      "loss": 0.1187,
+      "step": 336
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.4045857153281243e-05,
+      "loss": 0.1068,
+      "step": 337
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.4013551264695663e-05,
+      "loss": 0.051,
+      "step": 338
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.3981195367118676e-05,
+      "loss": 0.2048,
+      "step": 339
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.3948789863705914e-05,
+      "loss": 0.1283,
+      "step": 340
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.3916335158231101e-05,
+      "loss": 0.218,
+      "step": 341
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.388383165508102e-05,
+      "loss": 0.1566,
+      "step": 342
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.3851279759250473e-05,
+      "loss": 0.1169,
+      "step": 343
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.381867987633725e-05,
+      "loss": 0.1577,
+      "step": 344
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.3786032412537037e-05,
+      "loss": 0.1951,
+      "step": 345
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.3753337774638397e-05,
+      "loss": 0.2035,
+      "step": 346
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.3720596370017687e-05,
+      "loss": 0.1015,
+      "step": 347
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.3687808606633965e-05,
+      "loss": 0.135,
+      "step": 348
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.3654974893023934e-05,
+      "loss": 0.0781,
+      "step": 349
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.3622095638296827e-05,
+      "loss": 0.1211,
+      "step": 350
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.3589171252129329e-05,
+      "loss": 0.0978,
+      "step": 351
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.3556202144760461e-05,
+      "loss": 0.0827,
+      "step": 352
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.3523188726986476e-05,
+      "loss": 0.0728,
+      "step": 353
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.349013141015573e-05,
+      "loss": 0.1944,
+      "step": 354
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.3457030606163564e-05,
+      "loss": 0.0596,
+      "step": 355
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.3423886727447176e-05,
+      "loss": 0.0672,
+      "step": 356
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.339070018698047e-05,
+      "loss": 0.1356,
+      "step": 357
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.335747139826892e-05,
+      "loss": 0.1934,
+      "step": 358
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.3324200775344416e-05,
+      "loss": 0.1326,
+      "step": 359
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.32908887327601e-05,
+      "loss": 0.1379,
+      "step": 360
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.325753568558521e-05,
+      "loss": 0.1601,
+      "step": 361
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.3224142049399896e-05,
+      "loss": 0.0933,
+      "step": 362
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.3190708240290051e-05,
+      "loss": 0.1535,
+      "step": 363
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.315723467484213e-05,
+      "loss": 0.1245,
+      "step": 364
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.3123721770137945e-05,
+      "loss": 0.048,
+      "step": 365
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.1028,
+      "step": 366
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.3056579613733675e-05,
+      "loss": 0.1159,
+      "step": 367
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.3022951198627254e-05,
+      "loss": 0.0539,
+      "step": 368
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.2989285117441452e-05,
+      "loss": 0.1,
+      "step": 369
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.2955581789656844e-05,
+      "loss": 0.1866,
+      "step": 370
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.2921841635218082e-05,
+      "loss": 0.1799,
+      "step": 371
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.28880650745287e-05,
+      "loss": 0.1304,
+      "step": 372
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.2854252528445841e-05,
+      "loss": 0.0713,
+      "step": 373
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.282040441827503e-05,
+      "loss": 0.0738,
+      "step": 374
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.2786521165764923e-05,
+      "loss": 0.1433,
+      "step": 375
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.275260319310205e-05,
+      "loss": 0.1236,
+      "step": 376
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.271865092290556e-05,
+      "loss": 0.1821,
+      "step": 377
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.2684664778221943e-05,
+      "loss": 0.0852,
+      "step": 378
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.2650645182519768e-05,
+      "loss": 0.1182,
+      "step": 379
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.2616592559684408e-05,
+      "loss": 0.1491,
+      "step": 380
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.2582507334012755e-05,
+      "loss": 0.0218,
+      "step": 381
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.2548389930207932e-05,
+      "loss": 0.1093,
+      "step": 382
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.2514240773374004e-05,
+      "loss": 0.0267,
+      "step": 383
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.2480060289010677e-05,
+      "loss": 0.1571,
+      "step": 384
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.2445848903008001e-05,
+      "loss": 0.088,
+      "step": 385
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.2411607041641062e-05,
+      "loss": 0.1832,
+      "step": 386
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.2377335131564669e-05,
+      "loss": 0.1053,
+      "step": 387
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.2343033599808044e-05,
+      "loss": 0.1378,
+      "step": 388
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.2308702873769486e-05,
+      "loss": 0.0896,
+      "step": 389
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.2274343381211067e-05,
+      "loss": 0.0887,
+      "step": 390
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.2239955550253277e-05,
+      "loss": 0.1197,
+      "step": 391
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.2205539809369719e-05,
+      "loss": 0.0696,
+      "step": 392
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.217109658738175e-05,
+      "loss": 0.097,
+      "step": 393
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.2136626313453136e-05,
+      "loss": 0.1334,
+      "step": 394
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.2102129417084713e-05,
+      "loss": 0.1402,
+      "step": 395
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.2067606328109038e-05,
+      "loss": 0.1815,
+      "step": 396
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.2033057476685029e-05,
+      "loss": 0.1832,
+      "step": 397
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.1998483293292602e-05,
+      "loss": 0.0747,
+      "step": 398
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.1963884208727306e-05,
+      "loss": 0.1618,
+      "step": 399
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.192926065409497e-05,
+      "loss": 0.1851,
+      "step": 400
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.1894613060806312e-05,
+      "loss": 0.1275,
+      "step": 401
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.185994186057158e-05,
+      "loss": 0.1287,
+      "step": 402
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.1825247485395158e-05,
+      "loss": 0.1105,
+      "step": 403
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.1790530367570194e-05,
+      "loss": 0.2055,
+      "step": 404
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.1755790939673208e-05,
+      "loss": 0.1011,
+      "step": 405
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1.172102963455871e-05,
+      "loss": 0.1992,
+      "step": 406
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.1686246885353798e-05,
+      "loss": 0.0954,
+      "step": 407
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.165144312545276e-05,
+      "loss": 0.0403,
+      "step": 408
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.1616618788511684e-05,
+      "loss": 0.0246,
+      "step": 409
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.1581774308443042e-05,
+      "loss": 0.11,
+      "step": 410
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.1546910119410297e-05,
+      "loss": 0.0918,
+      "step": 411
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.1512026655822483e-05,
+      "loss": 0.0756,
+      "step": 412
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.1477124352328799e-05,
+      "loss": 0.0948,
+      "step": 413
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.1442203643813184e-05,
+      "loss": 0.0924,
+      "step": 414
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.1407264965388907e-05,
+      "loss": 0.0627,
+      "step": 415
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.1372308752393144e-05,
+      "loss": 0.1171,
+      "step": 416
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.1337335440381552e-05,
+      "loss": 0.119,
+      "step": 417
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.1302345465122839e-05,
+      "loss": 0.0812,
+      "step": 418
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.1267339262593342e-05,
+      "loss": 0.131,
+      "step": 419
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.1232317268971586e-05,
+      "loss": 0.0628,
+      "step": 420
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.1197279920632864e-05,
+      "loss": 0.3083,
+      "step": 421
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.1162227654143777e-05,
+      "loss": 0.0955,
+      "step": 422
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.1127160906256811e-05,
+      "loss": 0.1338,
+      "step": 423
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.1092080113904886e-05,
+      "loss": 0.0703,
+      "step": 424
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.1056985714195931e-05,
+      "loss": 0.0571,
+      "step": 425
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.1021878144407408e-05,
+      "loss": 0.1294,
+      "step": 426
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.0986757841980886e-05,
+      "loss": 0.1442,
+      "step": 427
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.0951625244516584e-05,
+      "loss": 0.0917,
+      "step": 428
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.0916480789767907e-05,
+      "loss": 0.0897,
+      "step": 429
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.088132491563602e-05,
+      "loss": 0.1631,
+      "step": 430
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.0846158060164359e-05,
+      "loss": 0.0956,
+      "step": 431
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.081098066153319e-05,
+      "loss": 0.092,
+      "step": 432
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.0775793158054158e-05,
+      "loss": 0.1717,
+      "step": 433
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.07405959881648e-05,
+      "loss": 0.0835,
+      "step": 434
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.070538959042311e-05,
+      "loss": 0.0962,
+      "step": 435
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0670174403502051e-05,
+      "loss": 0.0923,
+      "step": 436
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.0634950866184109e-05,
+      "loss": 0.044,
+      "step": 437
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.0599719417355801e-05,
+      "loss": 0.06,
+      "step": 438
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.0564480496002237e-05,
+      "loss": 0.0681,
+      "step": 439
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.0529234541201631e-05,
+      "loss": 0.1336,
+      "step": 440
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.0493981992119833e-05,
+      "loss": 0.059,
+      "step": 441
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.0458723288004858e-05,
+      "loss": 0.1313,
+      "step": 442
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.0423458868181404e-05,
+      "loss": 0.1512,
+      "step": 443
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.0388189172045407e-05,
+      "loss": 0.0324,
+      "step": 444
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.0352914639058527e-05,
+      "loss": 0.1013,
+      "step": 445
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.03176357087427e-05,
+      "loss": 0.1223,
+      "step": 446
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.0282352820674655e-05,
+      "loss": 0.0727,
+      "step": 447
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.0247066414480424e-05,
+      "loss": 0.1489,
+      "step": 448
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.0211776929829893e-05,
+      "loss": 0.0941,
+      "step": 449
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.0176484806431288e-05,
+      "loss": 0.0076,
+      "step": 450
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.0141190484025722e-05,
+      "loss": 0.062,
+      "step": 451
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.0105894402381703e-05,
+      "loss": 0.0602,
+      "step": 452
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.007059700128967e-05,
+      "loss": 0.2056,
+      "step": 453
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.0035298720556493e-05,
+      "loss": 0.0693,
+      "step": 454
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0723,
+      "step": 455
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 9.964701279443509e-06,
+      "loss": 0.1812,
+      "step": 456
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 9.929402998710333e-06,
+      "loss": 0.1627,
+      "step": 457
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 9.894105597618297e-06,
+      "loss": 0.1258,
+      "step": 458
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 9.858809515974283e-06,
+      "loss": 0.1079,
+      "step": 459
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 9.823515193568715e-06,
+      "loss": 0.0355,
+      "step": 460
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 9.788223070170112e-06,
+      "loss": 0.0295,
+      "step": 461
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 9.752933585519578e-06,
+      "loss": 0.1113,
+      "step": 462
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 9.717647179325347e-06,
+      "loss": 0.0883,
+      "step": 463
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 9.682364291257304e-06,
+      "loss": 0.1155,
+      "step": 464
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 9.647085360941476e-06,
+      "loss": 0.1176,
+      "step": 465
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 9.6118108279546e-06,
+      "loss": 0.1611,
+      "step": 466
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 9.576541131818599e-06,
+      "loss": 0.0669,
+      "step": 467
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.541276711995149e-06,
+      "loss": 0.1038,
+      "step": 468
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.506018007880169e-06,
+      "loss": 0.1918,
+      "step": 469
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 9.470765458798369e-06,
+      "loss": 0.0618,
+      "step": 470
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 9.435519503997767e-06,
+      "loss": 0.094,
+      "step": 471
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 9.400280582644204e-06,
+      "loss": 0.0436,
+      "step": 472
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 9.365049133815898e-06,
+      "loss": 0.0931,
+      "step": 473
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 9.32982559649795e-06,
+      "loss": 0.1703,
+      "step": 474
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 9.29461040957689e-06,
+      "loss": 0.1413,
+      "step": 475
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 9.259404011835203e-06,
+      "loss": 0.0133,
+      "step": 476
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 9.224206841945844e-06,
+      "loss": 0.06,
+      "step": 477
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 9.189019338466812e-06,
+      "loss": 0.0607,
+      "step": 478
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 9.153841939835645e-06,
+      "loss": 0.0456,
+      "step": 479
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 9.118675084363986e-06,
+      "loss": 0.1682,
+      "step": 480
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 9.083519210232096e-06,
+      "loss": 0.0619,
+      "step": 481
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 9.04837475548342e-06,
+      "loss": 0.0894,
+      "step": 482
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 9.013242158019116e-06,
+      "loss": 0.0853,
+      "step": 483
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 8.978121855592593e-06,
+      "loss": 0.1291,
+      "step": 484
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 8.943014285804072e-06,
+      "loss": 0.0908,
+      "step": 485
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 8.907919886095115e-06,
+      "loss": 0.0506,
+      "step": 486
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 8.872839093743195e-06,
+      "loss": 0.2595,
+      "step": 487
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 8.837772345856226e-06,
+      "loss": 0.1209,
+      "step": 488
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 8.802720079367136e-06,
+      "loss": 0.1269,
+      "step": 489
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 8.767682731028415e-06,
+      "loss": 0.0954,
+      "step": 490
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 8.732660737406663e-06,
+      "loss": 0.0518,
+      "step": 491
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 8.697654534877166e-06,
+      "loss": 0.0313,
+      "step": 492
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 8.662664559618452e-06,
+      "loss": 0.1055,
+      "step": 493
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 8.627691247606862e-06,
+      "loss": 0.0845,
+      "step": 494
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 8.592735034611097e-06,
+      "loss": 0.0545,
+      "step": 495
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 8.557796356186818e-06,
+      "loss": 0.1807,
+      "step": 496
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 8.522875647671205e-06,
+      "loss": 0.058,
+      "step": 497
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 8.487973344177517e-06,
+      "loss": 0.1292,
+      "step": 498
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 8.453089880589706e-06,
+      "loss": 0.0256,
+      "step": 499
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 8.418225691556962e-06,
+      "loss": 0.1098,
+      "step": 500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 8.383381211488321e-06,
+      "loss": 0.0777,
+      "step": 501
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 8.348556874547242e-06,
+      "loss": 0.1937,
+      "step": 502
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 8.313753114646204e-06,
+      "loss": 0.1641,
+      "step": 503
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 8.278970365441292e-06,
+      "loss": 0.0931,
+      "step": 504
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 8.244209060326794e-06,
+      "loss": 0.0384,
+      "step": 505
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 8.209469632429811e-06,
+      "loss": 0.0222,
+      "step": 506
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 8.174752514604844e-06,
+      "loss": 0.0296,
+      "step": 507
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 8.140058139428425e-06,
+      "loss": 0.1177,
+      "step": 508
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 8.10538693919369e-06,
+      "loss": 0.1979,
+      "step": 509
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 8.070739345905032e-06,
+      "loss": 0.0402,
+      "step": 510
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 8.036115791272697e-06,
+      "loss": 0.1115,
+      "step": 511
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 8.001516706707401e-06,
+      "loss": 0.0556,
+      "step": 512
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 7.966942523314976e-06,
+      "loss": 0.1767,
+      "step": 513
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 7.932393671890965e-06,
+      "loss": 0.1191,
+      "step": 514
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 7.89787058291529e-06,
+      "loss": 0.1683,
+      "step": 515
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 7.863373686546868e-06,
+      "loss": 0.0887,
+      "step": 516
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 7.82890341261825e-06,
+      "loss": 0.1171,
+      "step": 517
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 7.794460190630283e-06,
+      "loss": 0.0898,
+      "step": 518
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 7.760044449746725e-06,
+      "loss": 0.0259,
+      "step": 519
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 7.725656618788938e-06,
+      "loss": 0.059,
+      "step": 520
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 7.691297126230515e-06,
+      "loss": 0.137,
+      "step": 521
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 7.656966400191956e-06,
+      "loss": 0.0523,
+      "step": 522
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 7.622664868435333e-06,
+      "loss": 0.1678,
+      "step": 523
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 7.58839295835894e-06,
+      "loss": 0.0061,
+      "step": 524
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 7.554151096992002e-06,
+      "loss": 0.1454,
+      "step": 525
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 7.519939710989326e-06,
+      "loss": 0.0851,
+      "step": 526
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 7.485759226626002e-06,
+      "loss": 0.0491,
+      "step": 527
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 7.45161006979207e-06,
+      "loss": 0.1054,
+      "step": 528
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 7.417492665987247e-06,
+      "loss": 0.0537,
+      "step": 529
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 7.383407440315595e-06,
+      "loss": 0.0878,
+      "step": 530
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 7.349354817480234e-06,
+      "loss": 0.0771,
+      "step": 531
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 7.315335221778064e-06,
+      "loss": 0.1762,
+      "step": 532
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 7.281349077094444e-06,
+      "loss": 0.1396,
+      "step": 533
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 7.247396806897953e-06,
+      "loss": 0.0754,
+      "step": 534
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 7.213478834235079e-06,
+      "loss": 0.1278,
+      "step": 535
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 7.179595581724972e-06,
+      "loss": 0.1281,
+      "step": 536
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 7.145747471554163e-06,
+      "loss": 0.1049,
+      "step": 537
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 7.111934925471302e-06,
+      "loss": 0.1971,
+      "step": 538
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 7.078158364781919e-06,
+      "loss": 0.0472,
+      "step": 539
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 7.044418210343161e-06,
+      "loss": 0.2505,
+      "step": 540
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.010714882558551e-06,
+      "loss": 0.0784,
+      "step": 541
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 6.97704880137275e-06,
+      "loss": 0.0451,
+      "step": 542
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 6.9434203862663244e-06,
+      "loss": 0.0579,
+      "step": 543
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.0581,
+      "step": 544
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 6.876278229862059e-06,
+      "loss": 0.0672,
+      "step": 545
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 6.842765325157874e-06,
+      "loss": 0.0549,
+      "step": 546
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 6.8092917597099504e-06,
+      "loss": 0.168,
+      "step": 547
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 6.775857950600107e-06,
+      "loss": 0.0587,
+      "step": 548
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 6.742464314414791e-06,
+      "loss": 0.0502,
+      "step": 549
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 6.7091112672399e-06,
+      "loss": 0.235,
+      "step": 550
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 6.675799224655588e-06,
+      "loss": 0.161,
+      "step": 551
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 6.642528601731082e-06,
+      "loss": 0.1872,
+      "step": 552
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 6.609299813019534e-06,
+      "loss": 0.1231,
+      "step": 553
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 6.5761132725528265e-06,
+      "loss": 0.142,
+      "step": 554
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 6.542969393836436e-06,
+      "loss": 0.123,
+      "step": 555
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 6.509868589844274e-06,
+      "loss": 0.0388,
+      "step": 556
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 6.4768112730135255e-06,
+      "loss": 0.1143,
+      "step": 557
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 6.44379785523954e-06,
+      "loss": 0.0874,
+      "step": 558
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 6.410828747870672e-06,
+      "loss": 0.1333,
+      "step": 559
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 6.3779043617031775e-06,
+      "loss": 0.095,
+      "step": 560
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 6.34502510697607e-06,
+      "loss": 0.0825,
+      "step": 561
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 6.312191393366036e-06,
+      "loss": 0.0773,
+      "step": 562
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 6.279403629982316e-06,
+      "loss": 0.2261,
+      "step": 563
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 6.246662225361603e-06,
+      "loss": 0.1245,
+      "step": 564
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 6.213967587462969e-06,
+      "loss": 0.0852,
+      "step": 565
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 6.181320123662755e-06,
+      "loss": 0.1585,
+      "step": 566
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 6.148720240749528e-06,
+      "loss": 0.0888,
+      "step": 567
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 6.116168344918982e-06,
+      "loss": 0.0683,
+      "step": 568
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 6.083664841768901e-06,
+      "loss": 0.0344,
+      "step": 569
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 6.051210136294089e-06,
+      "loss": 0.0684,
+      "step": 570
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 6.018804632881326e-06,
+      "loss": 0.085,
+      "step": 571
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 5.986448735304339e-06,
+      "loss": 0.0794,
+      "step": 572
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 5.95414284671876e-06,
+      "loss": 0.1032,
+      "step": 573
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 5.921887369657113e-06,
+      "loss": 0.0803,
+      "step": 574
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 5.8896827060237824e-06,
+      "loss": 0.1079,
+      "step": 575
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 5.857529257090027e-06,
+      "loss": 0.0858,
+      "step": 576
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 5.825427423488962e-06,
+      "loss": 0.088,
+      "step": 577
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 5.793377605210575e-06,
+      "loss": 0.1064,
+      "step": 578
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 5.76138020159674e-06,
+      "loss": 0.1505,
+      "step": 579
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 5.729435611336239e-06,
+      "loss": 0.0761,
+      "step": 580
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 5.6975442324598165e-06,
+      "loss": 0.1122,
+      "step": 581
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 5.665706462335167e-06,
+      "loss": 0.0576,
+      "step": 582
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 5.633922697662053e-06,
+      "loss": 0.1021,
+      "step": 583
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 5.602193334467307e-06,
+      "loss": 0.1051,
+      "step": 584
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 5.570518768099918e-06,
+      "loss": 0.1117,
+      "step": 585
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 5.538899393226122e-06,
+      "loss": 0.0574,
+      "step": 586
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 5.507335603824435e-06,
+      "loss": 0.0677,
+      "step": 587
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 5.475827793180808e-06,
+      "loss": 0.1251,
+      "step": 588
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 5.444376353883678e-06,
+      "loss": 0.1317,
+      "step": 589
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 5.412981677819094e-06,
+      "loss": 0.081,
+      "step": 590
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 5.381644156165844e-06,
+      "loss": 0.119,
+      "step": 591
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 5.350364179390547e-06,
+      "loss": 0.141,
+      "step": 592
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 5.319142137242839e-06,
+      "loss": 0.0467,
+      "step": 593
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 5.28797841875047e-06,
+      "loss": 0.0679,
+      "step": 594
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 5.256873412214476e-06,
+      "loss": 0.0373,
+      "step": 595
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 5.225827505204355e-06,
+      "loss": 0.0312,
+      "step": 596
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 5.194841084553196e-06,
+      "loss": 0.067,
+      "step": 597
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 5.163914536352919e-06,
+      "loss": 0.0799,
+      "step": 598
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 5.133048245949397e-06,
+      "loss": 0.1467,
+      "step": 599
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 5.1022425979377174e-06,
+      "loss": 0.0356,
+      "step": 600
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 5.071497976157346e-06,
+      "loss": 0.0865,
+      "step": 601
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 5.040814763687358e-06,
+      "loss": 0.055,
+      "step": 602
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 5.010193342841683e-06,
+      "loss": 0.1404,
+      "step": 603
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 4.979634095164298e-06,
+      "loss": 0.1654,
+      "step": 604
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 4.949137401424527e-06,
+      "loss": 0.1067,
+      "step": 605
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 4.918703641612255e-06,
+      "loss": 0.185,
+      "step": 606
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 4.8883331949332094e-06,
+      "loss": 0.1365,
+      "step": 607
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 4.858026439804235e-06,
+      "loss": 0.116,
+      "step": 608
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 4.827783753848575e-06,
+      "loss": 0.0145,
+      "step": 609
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 4.797605513891178e-06,
+      "loss": 0.1629,
+      "step": 610
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 4.767492095953985e-06,
+      "loss": 0.0282,
+      "step": 611
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 4.737443875251251e-06,
+      "loss": 0.1205,
+      "step": 612
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 4.707461226184875e-06,
+      "loss": 0.0808,
+      "step": 613
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 4.6775445223397306e-06,
+      "loss": 0.088,
+      "step": 614
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 4.647694136479008e-06,
+      "loss": 0.0382,
+      "step": 615
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 4.61791044053958e-06,
+      "loss": 0.1505,
+      "step": 616
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 4.5881938056273565e-06,
+      "loss": 0.012,
+      "step": 617
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 4.558544602012663e-06,
+      "loss": 0.0834,
+      "step": 618
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 4.52896319912563e-06,
+      "loss": 0.1898,
+      "step": 619
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 4.4994499655515865e-06,
+      "loss": 0.0717,
+      "step": 620
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 4.470005269026482e-06,
+      "loss": 0.0335,
+      "step": 621
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 4.440629476432268e-06,
+      "loss": 0.0744,
+      "step": 622
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 4.411322953792374e-06,
+      "loss": 0.033,
+      "step": 623
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 4.38208606626711e-06,
+      "loss": 0.0739,
+      "step": 624
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 4.3529191781491305e-06,
+      "loss": 0.0353,
+      "step": 625
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 4.323822652858911e-06,
+      "loss": 0.0445,
+      "step": 626
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 4.294796852940173e-06,
+      "loss": 0.1612,
+      "step": 627
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 4.265842140055428e-06,
+      "loss": 0.0408,
+      "step": 628
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 4.236958874981423e-06,
+      "loss": 0.0897,
+      "step": 629
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.208147417604665e-06,
+      "loss": 0.0647,
+      "step": 630
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 4.179408126916946e-06,
+      "loss": 0.1362,
+      "step": 631
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 4.150741361010837e-06,
+      "loss": 0.0979,
+      "step": 632
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.1128,
+      "step": 633
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 4.093626831391051e-06,
+      "loss": 0.2077,
+      "step": 634
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 4.0651797793264356e-06,
+      "loss": 0.0514,
+      "step": 635
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 4.036806675332715e-06,
+      "loss": 0.0951,
+      "step": 636
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 4.008507872939759e-06,
+      "loss": 0.0384,
+      "step": 637
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 3.98028372475166e-06,
+      "loss": 0.1407,
+      "step": 638
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 3.952134582442307e-06,
+      "loss": 0.1118,
+      "step": 639
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 3.924060796751012e-06,
+      "loss": 0.0251,
+      "step": 640
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 3.8960627174781425e-06,
+      "loss": 0.06,
+      "step": 641
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 3.8681406934807585e-06,
+      "loss": 0.1162,
+      "step": 642
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 3.840295072668279e-06,
+      "loss": 0.0737,
+      "step": 643
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 3.8125262019981224e-06,
+      "loss": 0.0381,
+      "step": 644
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 3.7848344274714075e-06,
+      "loss": 0.0772,
+      "step": 645
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 3.7572200941286284e-06,
+      "loss": 0.1061,
+      "step": 646
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 3.7296835460453593e-06,
+      "loss": 0.067,
+      "step": 647
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 3.702225126327965e-06,
+      "loss": 0.1586,
+      "step": 648
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 3.6748451771093374e-06,
+      "loss": 0.1414,
+      "step": 649
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 3.647544039544615e-06,
+      "loss": 0.0791,
+      "step": 650
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 3.6203220538069405e-06,
+      "loss": 0.1359,
+      "step": 651
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 3.593179559083225e-06,
+      "loss": 0.106,
+      "step": 652
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 3.5661168935699132e-06,
+      "loss": 0.068,
+      "step": 653
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 3.5391343944687906e-06,
+      "loss": 0.2397,
+      "step": 654
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 3.51223239798274e-06,
+      "loss": 0.1602,
+      "step": 655
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 3.4854112393116047e-06,
+      "loss": 0.0664,
+      "step": 656
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 3.4586712526479717e-06,
+      "loss": 0.0567,
+      "step": 657
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 3.432012771173021e-06,
+      "loss": 0.0942,
+      "step": 658
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 3.4054361270523906e-06,
+      "loss": 0.0762,
+      "step": 659
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 3.378941651431996e-06,
+      "loss": 0.1079,
+      "step": 660
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 3.3525296744339574e-06,
+      "loss": 0.1181,
+      "step": 661
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 3.326200525152441e-06,
+      "loss": 0.0886,
+      "step": 662
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 3.2999545316495816e-06,
+      "loss": 0.0869,
+      "step": 663
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 3.2737920209513994e-06,
+      "loss": 0.1142,
+      "step": 664
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 3.247713319043695e-06,
+      "loss": 0.1378,
+      "step": 665
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 3.2217187508680305e-06,
+      "loss": 0.0297,
+      "step": 666
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 3.195808640317648e-06,
+      "loss": 0.1139,
+      "step": 667
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 3.1699833102334397e-06,
+      "loss": 0.0585,
+      "step": 668
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 3.144243082399947e-06,
+      "loss": 0.0942,
+      "step": 669
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 3.1185882775413123e-06,
+      "loss": 0.0433,
+      "step": 670
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 3.0930192153173233e-06,
+      "loss": 0.0558,
+      "step": 671
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 3.067536214319402e-06,
+      "loss": 0.0668,
+      "step": 672
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 3.0421395920666464e-06,
+      "loss": 0.0936,
+      "step": 673
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 3.0168296650018736e-06,
+      "loss": 0.0995,
+      "step": 674
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 2.9916067484876697e-06,
+      "loss": 0.0822,
+      "step": 675
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 2.966471156802477e-06,
+      "loss": 0.074,
+      "step": 676
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 2.9414232031366598e-06,
+      "loss": 0.0729,
+      "step": 677
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 2.9164631995886095e-06,
+      "loss": 0.0376,
+      "step": 678
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 2.891591457160858e-06,
+      "loss": 0.0196,
+      "step": 679
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 2.8668082857562006e-06,
+      "loss": 0.1273,
+      "step": 680
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 2.84211399417383e-06,
+      "loss": 0.1264,
+      "step": 681
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 2.8175088901055026e-06,
+      "loss": 0.0461,
+      "step": 682
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 2.792993280131688e-06,
+      "loss": 0.2352,
+      "step": 683
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 2.7685674697177568e-06,
+      "loss": 0.1403,
+      "step": 684
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 2.7442317632101743e-06,
+      "loss": 0.0196,
+      "step": 685
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 2.719986463832708e-06,
+      "loss": 0.049,
+      "step": 686
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 2.695831873682656e-06,
+      "loss": 0.167,
+      "step": 687
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 2.6717682937270605e-06,
+      "loss": 0.0915,
+      "step": 688
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 2.647796023798991e-06,
+      "loss": 0.1057,
+      "step": 689
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 2.6239153625937786e-06,
+      "loss": 0.043,
+      "step": 690
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 2.6001266076653096e-06,
+      "loss": 0.1321,
+      "step": 691
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 2.576430055422324e-06,
+      "loss": 0.024,
+      "step": 692
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 2.552826001124693e-06,
+      "loss": 0.1472,
+      "step": 693
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 2.529314738879782e-06,
+      "loss": 0.048,
+      "step": 694
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 2.5058965616387485e-06,
+      "loss": 0.0357,
+      "step": 695
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 2.4825717611929144e-06,
+      "loss": 0.0977,
+      "step": 696
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 2.459340628170128e-06,
+      "loss": 0.0612,
+      "step": 697
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 2.4362034520311216e-06,
+      "loss": 0.1108,
+      "step": 698
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 2.4131605210659414e-06,
+      "loss": 0.0249,
+      "step": 699
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 2.390212122390323e-06,
+      "loss": 0.0722,
+      "step": 700
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 2.3673585419421275e-06,
+      "loss": 0.0467,
+      "step": 701
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 2.3446000644777856e-06,
+      "loss": 0.1096,
+      "step": 702
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 2.321936973568727e-06,
+      "loss": 0.1346,
+      "step": 703
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 2.2993695515978767e-06,
+      "loss": 0.0711,
+      "step": 704
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 2.2768980797561112e-06,
+      "loss": 0.0984,
+      "step": 705
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 2.2545228380387706e-06,
+      "loss": 0.147,
+      "step": 706
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 2.2322441052421606e-06,
+      "loss": 0.0853,
+      "step": 707
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 2.2100621589600813e-06,
+      "loss": 0.0207,
+      "step": 708
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 2.1879772755803763e-06,
+      "loss": 0.0978,
+      "step": 709
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 2.1659897302814757e-06,
+      "loss": 0.1167,
+      "step": 710
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 2.1440997970289744e-06,
+      "loss": 0.2338,
+      "step": 711
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 2.12230774857222e-06,
+      "loss": 0.1159,
+      "step": 712
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 2.100613856440911e-06,
+      "loss": 0.0416,
+      "step": 713
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 2.0790183909417096e-06,
+      "loss": 0.0619,
+      "step": 714
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 2.057521621154891e-06,
+      "loss": 0.0717,
+      "step": 715
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 2.036123814930967e-06,
+      "loss": 0.0874,
+      "step": 716
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 2.0148252388873645e-06,
+      "loss": 0.0301,
+      "step": 717
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 1.9936261584050974e-06,
+      "loss": 0.0501,
+      "step": 718
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 1.9725268376254593e-06,
+      "loss": 0.1024,
+      "step": 719
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.9515275394467446e-06,
+      "loss": 0.0519,
+      "step": 720
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 1.9306285255209467e-06,
+      "loss": 0.1614,
+      "step": 721
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.0687,
+      "step": 722
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 1.88913239078515e-06,
+      "loss": 0.0362,
+      "step": 723
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 1.8685357870184605e-06,
+      "loss": 0.209,
+      "step": 724
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 1.8480405015848824e-06,
+      "loss": 0.0325,
+      "step": 725
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 1.8276467898563887e-06,
+      "loss": 0.1137,
+      "step": 726
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 1.8073549059393602e-06,
+      "loss": 0.0332,
+      "step": 727
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 1.787165102671391e-06,
+      "loss": 0.0665,
+      "step": 728
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 1.7670776316181427e-06,
+      "loss": 0.0126,
+      "step": 729
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 1.7470927430702277e-06,
+      "loss": 0.1067,
+      "step": 730
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 1.7272106860400572e-06,
+      "loss": 0.1861,
+      "step": 731
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 1.7074317082587755e-06,
+      "loss": 0.1216,
+      "step": 732
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 1.6877560561731454e-06,
+      "loss": 0.1214,
+      "step": 733
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 1.668183974942491e-06,
+      "loss": 0.0446,
+      "step": 734
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 1.6487157084356454e-06,
+      "loss": 0.1755,
+      "step": 735
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 1.6293514992278935e-06,
+      "loss": 0.0672,
+      "step": 736
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 1.610091588597975e-06,
+      "loss": 0.0688,
+      "step": 737
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 1.5909362165250609e-06,
+      "loss": 0.0391,
+      "step": 738
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 1.5718856216857659e-06,
+      "loss": 0.204,
+      "step": 739
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 1.5529400414511809e-06,
+      "loss": 0.1083,
+      "step": 740
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 1.534099711883905e-06,
+      "loss": 0.1317,
+      "step": 741
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 1.5153648677351196e-06,
+      "loss": 0.0364,
+      "step": 742
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 1.4967357424416473e-06,
+      "loss": 0.1328,
+      "step": 743
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 1.4782125681230497e-06,
+      "loss": 0.1175,
+      "step": 744
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 1.4597955755787375e-06,
+      "loss": 0.0556,
+      "step": 745
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 1.4414849942850927e-06,
+      "loss": 0.12,
+      "step": 746
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 1.4232810523926066e-06,
+      "loss": 0.0377,
+      "step": 747
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 1.4051839767230479e-06,
+      "loss": 0.1299,
+      "step": 748
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 1.3871939927666189e-06,
+      "loss": 0.1263,
+      "step": 749
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 1.369311324679159e-06,
+      "loss": 0.1131,
+      "step": 750
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 1.3515361952793494e-06,
+      "loss": 0.0813,
+      "step": 751
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 1.333868826045932e-06,
+      "loss": 0.0639,
+      "step": 752
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 1.3163094371149598e-06,
+      "loss": 0.1011,
+      "step": 753
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 1.2988582472770372e-06,
+      "loss": 0.079,
+      "step": 754
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 1.281515473974614e-06,
+      "loss": 0.1537,
+      "step": 755
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 1.264281333299261e-06,
+      "loss": 0.2419,
+      "step": 756
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 1.2471560399889804e-06,
+      "loss": 0.1426,
+      "step": 757
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 1.2301398074255444e-06,
+      "loss": 0.061,
+      "step": 758
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 1.2132328476318033e-06,
+      "loss": 0.0603,
+      "step": 759
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 1.196435371269089e-06,
+      "loss": 0.1725,
+      "step": 760
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 1.1797475876345431e-06,
+      "loss": 0.0901,
+      "step": 761
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 1.1631697046585511e-06,
+      "loss": 0.1082,
+      "step": 762
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 1.1467019289021197e-06,
+      "loss": 0.1151,
+      "step": 763
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 1.1303444655543206e-06,
+      "loss": 0.1142,
+      "step": 764
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 1.1140975184297332e-06,
+      "loss": 0.0631,
+      "step": 765
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 1.0979612899658875e-06,
+      "loss": 0.0626,
+      "step": 766
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 1.0819359812207708e-06,
+      "loss": 0.1549,
+      "step": 767
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 1.0660217918702965e-06,
+      "loss": 0.0501,
+      "step": 768
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 1.05021892020583e-06,
+      "loss": 0.0599,
+      "step": 769
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 1.0345275631317165e-06,
+      "loss": 0.0376,
+      "step": 770
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 1.018947916162817e-06,
+      "loss": 0.1889,
+      "step": 771
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 1.0034801734220922e-06,
+      "loss": 0.0231,
+      "step": 772
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 9.881245276381646e-07,
+      "loss": 0.0564,
+      "step": 773
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 9.728811701429242e-07,
+      "loss": 0.0173,
+      "step": 774
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 9.577502908691516e-07,
+      "loss": 0.0863,
+      "step": 775
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 9.427320783481353e-07,
+      "loss": 0.0672,
+      "step": 776
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 9.27826719707341e-07,
+      "loss": 0.0871,
+      "step": 777
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 9.130344006680658e-07,
+      "loss": 0.0947,
+      "step": 778
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 8.983553055431305e-07,
+      "loss": 0.0186,
+      "step": 779
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 8.837896172345827e-07,
+      "loss": 0.0912,
+      "step": 780
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 8.693375172314133e-07,
+      "loss": 0.0403,
+      "step": 781
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 8.54999185607307e-07,
+      "loss": 0.1068,
+      "step": 782
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 8.407748010183824e-07,
+      "loss": 0.0304,
+      "step": 783
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 8.266645407009788e-07,
+      "loss": 0.039,
+      "step": 784
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 8.126685804694401e-07,
+      "loss": 0.0582,
+      "step": 785
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 7.987870947139276e-07,
+      "loss": 0.0706,
+      "step": 786
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 7.850202563982457e-07,
+      "loss": 0.1113,
+      "step": 787
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 7.713682370576947e-07,
+      "loss": 0.0539,
+      "step": 788
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 7.578312067969162e-07,
+      "loss": 0.0998,
+      "step": 789
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 7.4440933428779e-07,
+      "loss": 0.065,
+      "step": 790
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 7.311027867673237e-07,
+      "loss": 0.0216,
+      "step": 791
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 7.17911730035572e-07,
+      "loss": 0.0828,
+      "step": 792
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 7.048363284535742e-07,
+      "loss": 0.1331,
+      "step": 793
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 6.918767449412933e-07,
+      "loss": 0.0332,
+      "step": 794
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 6.790331409756046e-07,
+      "loss": 0.0621,
+      "step": 795
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 6.663056765882692e-07,
+      "loss": 0.0588,
+      "step": 796
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 6.536945103639425e-07,
+      "loss": 0.0843,
+      "step": 797
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 6.411997994382102e-07,
+      "loss": 0.0688,
+      "step": 798
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 6.288216994956075e-07,
+      "loss": 0.142,
+      "step": 799
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 6.165603647677054e-07,
+      "loss": 0.0458,
+      "step": 800
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 6.044159480311673e-07,
+      "loss": 0.0396,
+      "step": 801
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.923886006058566e-07,
+      "loss": 0.089,
+      "step": 802
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 5.804784723529544e-07,
+      "loss": 0.0851,
+      "step": 803
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 5.68685711673076e-07,
+      "loss": 0.1189,
+      "step": 804
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.570104655044428e-07,
+      "loss": 0.1897,
+      "step": 805
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 5.454528793210356e-07,
+      "loss": 0.0368,
+      "step": 806
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.340130971307888e-07,
+      "loss": 0.0975,
+      "step": 807
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 5.226912614737967e-07,
+      "loss": 0.0529,
+      "step": 808
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 5.1148751342053e-07,
+      "loss": 0.0756,
+      "step": 809
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.004019925700921e-07,
+      "loss": 0.1461,
+      "step": 810
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.1067,
+      "step": 811
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.785861835067962e-07,
+      "loss": 0.1042,
+      "step": 812
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 4.678561671196946e-07,
+      "loss": 0.1052,
+      "step": 813
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 4.5724492158354397e-07,
+      "loss": 0.0779,
+      "step": 814
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 4.4675257911484526e-07,
+      "loss": 0.1499,
+      "step": 815
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 4.3637927044855476e-07,
+      "loss": 0.0695,
+      "step": 816
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 4.261251248364684e-07,
+      "loss": 0.0501,
+      "step": 817
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 4.1599027004560535e-07,
+      "loss": 0.0578,
+      "step": 818
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 4.0597483235661483e-07,
+      "loss": 0.1199,
+      "step": 819
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 3.960789365622075e-07,
+      "loss": 0.0526,
+      "step": 820
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 3.8630270596559774e-07,
+      "loss": 0.0517,
+      "step": 821
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 3.766462623789646e-07,
+      "loss": 0.164,
+      "step": 822
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 3.671097261219392e-07,
+      "loss": 0.061,
+      "step": 823
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 3.57693216020103e-07,
+      "loss": 0.0336,
+      "step": 824
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 3.483968494035039e-07,
+      "loss": 0.0872,
+      "step": 825
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 3.3922074210520407e-07,
+      "loss": 0.0199,
+      "step": 826
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 3.301650084598229e-07,
+      "loss": 0.135,
+      "step": 827
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 3.2122976130212644e-07,
+      "loss": 0.0802,
+      "step": 828
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 3.1241511196561045e-07,
+      "loss": 0.0254,
+      "step": 829
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.0372117028111825e-07,
+      "loss": 0.0889,
+      "step": 830
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 2.951480445754751e-07,
+      "loss": 0.0602,
+      "step": 831
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 2.866958416701271e-07,
+      "loss": 0.1432,
+      "step": 832
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 2.783646668798279e-07,
+      "loss": 0.0257,
+      "step": 833
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 2.7015462401130843e-07,
+      "loss": 0.0436,
+      "step": 834
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 2.620658153619959e-07,
+      "loss": 0.0553,
+      "step": 835
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 2.540983417187348e-07,
+      "loss": 0.0602,
+      "step": 836
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 2.4625230235652555e-07,
+      "loss": 0.0616,
+      "step": 837
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 2.3852779503730217e-07,
+      "loss": 0.0996,
+      "step": 838
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 2.3092491600869794e-07,
+      "loss": 0.0644,
+      "step": 839
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 2.2344376000285606e-07,
+      "loss": 0.0891,
+      "step": 840
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 2.1608442023524745e-07,
+      "loss": 0.1266,
+      "step": 841
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 2.0884698840350492e-07,
+      "loss": 0.0545,
+      "step": 842
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 2.0173155468628859e-07,
+      "loss": 0.126,
+      "step": 843
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 1.9473820774215557e-07,
+      "loss": 0.125,
+      "step": 844
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 1.8786703470845546e-07,
+      "loss": 0.1316,
+      "step": 845
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 1.8111812120024884e-07,
+      "loss": 0.0266,
+      "step": 846
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.7449155130923824e-07,
+      "loss": 0.0922,
+      "step": 847
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 1.6798740760272104e-07,
+      "loss": 0.035,
+      "step": 848
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 1.6160577112255827e-07,
+      "loss": 0.1335,
+      "step": 849
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 1.553467213841664e-07,
+      "loss": 0.1031,
+      "step": 850
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 1.4921033637552928e-07,
+      "loss": 0.0911,
+      "step": 851
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 1.4319669255622115e-07,
+      "loss": 0.1445,
+      "step": 852
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 1.3730586485645736e-07,
+      "loss": 0.0731,
+      "step": 853
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 1.3153792667616183e-07,
+      "loss": 0.0513,
+      "step": 854
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 1.2589294988404887e-07,
+      "loss": 0.0186,
+      "step": 855
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 1.2037100481672836e-07,
+      "loss": 0.1151,
+      "step": 856
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 1.149721602778342e-07,
+      "loss": 0.1106,
+      "step": 857
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 1.0969648353715945e-07,
+      "loss": 0.1174,
+      "step": 858
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 1.0454404032982591e-07,
+      "loss": 0.0454,
+      "step": 859
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 9.951489485545696e-08,
+      "loss": 0.0349,
+      "step": 860
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 9.460910977738492e-08,
+      "loss": 0.0404,
+      "step": 861
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 8.982674622186605e-08,
+      "loss": 0.0463,
+      "step": 862
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 8.516786377732012e-08,
+      "loss": 0.1219,
+      "step": 863
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 8.063252049358983e-08,
+      "loss": 0.0705,
+      "step": 864
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 7.622077288121033e-08,
+      "loss": 0.0455,
+      "step": 865
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 7.193267591071529e-08,
+      "loss": 0.2314,
+      "step": 866
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 6.776828301194638e-08,
+      "loss": 0.1068,
+      "step": 867
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 6.3727646073386e-08,
+      "loss": 0.0376,
+      "step": 868
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 5.981081544151446e-08,
+      "loss": 0.0626,
+      "step": 869
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 5.6017839920180506e-08,
+      "loss": 0.1064,
+      "step": 870
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 5.23487667699929e-08,
+      "loss": 0.1166,
+      "step": 871
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 4.880364170773533e-08,
+      "loss": 0.0656,
+      "step": 872
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 4.5382508905791324e-08,
+      "loss": 0.1329,
+      "step": 873
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 4.208541099159691e-08,
+      "loss": 0.1983,
+      "step": 874
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 3.891238904710881e-08,
+      "loss": 0.101,
+      "step": 875
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 3.586348260829486e-08,
+      "loss": 0.1629,
+      "step": 876
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 3.293872966463663e-08,
+      "loss": 0.0403,
+      "step": 877
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 3.013816665865976e-08,
+      "loss": 0.192,
+      "step": 878
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 2.746182848547885e-08,
+      "loss": 0.1605,
+      "step": 879
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 2.4909748492362162e-08,
+      "loss": 0.0636,
+      "step": 880
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 2.248195847831758e-08,
+      "loss": 0.1296,
+      "step": 881
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 2.0178488693695096e-08,
+      "loss": 0.0956,
+      "step": 882
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 1.7999367839810488e-08,
+      "loss": 0.1446,
+      "step": 883
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 1.5944623068586683e-08,
+      "loss": 0.0483,
+      "step": 884
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 1.4014279982216272e-08,
+      "loss": 0.1016,
+      "step": 885
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 1.2208362632842863e-08,
+      "loss": 0.0762,
+      "step": 886
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 1.0526893522260217e-08,
+      "loss": 0.063,
+      "step": 887
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 8.969893601634694e-09,
+      "loss": 0.0416,
+      "step": 888
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 7.5373822712399e-09,
+      "loss": 0.0938,
+      "step": 889
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 6.229377380218005e-09,
+      "loss": 0.042,
+      "step": 890
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.045895226355457e-09,
+      "loss": 0.1783,
+      "step": 891
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 3.986950555883162e-09,
+      "loss": 0.0617,
+      "step": 892
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 3.0525565632888445e-09,
+      "loss": 0.1284,
+      "step": 893
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 2.242724891156067e-09,
+      "loss": 0.0935,
+      "step": 894
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 1.5574656300143542e-09,
+      "loss": 0.1433,
+      "step": 895
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 9.96787318218173e-10,
+      "loss": 0.0548,
+      "step": 896
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5.606969418414654e-10,
+      "loss": 0.173,
+      "step": 897
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 2.4919993458549783e-10,
+      "loss": 0.1156,
+      "step": 898
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 6.230017771224895e-11,
+      "loss": 0.0657,
+      "step": 899
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.0174,
+      "step": 900
+    },
+    {
+      "epoch": 10.0,
+      "step": 900,
+      "total_flos": 8335010861383680.0,
+      "train_loss": 0.16723996378429648,
+      "train_runtime": 641.3051,
+      "train_samples_per_second": 11.165,
+      "train_steps_per_second": 1.403
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 900,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 100,
+  "total_flos": 8335010861383680.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}