diff --git "a/llava-v1.5-7b-task-lora/trainer_state.json" "b/llava-v1.5-7b-task-lora/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/llava-v1.5-7b-task-lora/trainer_state.json"
@@ -0,0 +1,5670 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 940,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 2.1876,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3793103448275863e-05,
+      "loss": 2.2886,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0689655172413793e-05,
+      "loss": 2.1823,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7586206896551727e-05,
+      "loss": 2.3409,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.4482758620689657e-05,
+      "loss": 2.2662,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.1379310344827587e-05,
+      "loss": 1.9962,
+      "step": 6
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.827586206896552e-05,
+      "loss": 1.9879,
+      "step": 7
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.517241379310345e-05,
+      "loss": 1.7576,
+      "step": 8
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.206896551724138e-05,
+      "loss": 1.683,
+      "step": 9
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.896551724137931e-05,
+      "loss": 1.5743,
+      "step": 10
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.586206896551724e-05,
+      "loss": 1.4835,
+      "step": 11
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 8.275862068965517e-05,
+      "loss": 1.5786,
+      "step": 12
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.96551724137931e-05,
+      "loss": 1.4103,
+      "step": 13
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.655172413793105e-05,
+      "loss": 1.3709,
+      "step": 14
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010344827586206898,
+      "loss": 1.4257,
+      "step": 15
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001103448275862069,
+      "loss": 1.3711,
+      "step": 16
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011724137931034482,
+      "loss": 1.2682,
+      "step": 17
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012413793103448277,
+      "loss": 1.3635,
+      "step": 18
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013103448275862068,
+      "loss": 1.2354,
+      "step": 19
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00013793103448275863,
+      "loss": 1.25,
+      "step": 20
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014482758620689657,
+      "loss": 1.1641,
+      "step": 21
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015172413793103449,
+      "loss": 1.2721,
+      "step": 22
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015862068965517243,
+      "loss": 1.1605,
+      "step": 23
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00016551724137931035,
+      "loss": 1.228,
+      "step": 24
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00017241379310344826,
+      "loss": 1.2069,
+      "step": 25
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001793103448275862,
+      "loss": 1.2013,
+      "step": 26
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00018620689655172415,
+      "loss": 1.2213,
+      "step": 27
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001931034482758621,
+      "loss": 1.1444,
+      "step": 28
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 1.263,
+      "step": 29
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999940538951178,
+      "loss": 1.1873,
+      "step": 30
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001999976215651183,
+      "loss": 1.1814,
+      "step": 31
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001999946485480332,
+      "loss": 1.1387,
+      "step": 32
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019999048637361222,
+      "loss": 1.1123,
+      "step": 33
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001999851350913528,
+      "loss": 1.0635,
+      "step": 34
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019997859476489355,
+      "loss": 1.0912,
+      "step": 35
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019997086547201336,
+      "loss": 1.0614,
+      "step": 36
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019996194730463062,
+      "loss": 1.1857,
+      "step": 37
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019995184036880205,
+      "loss": 1.1534,
+      "step": 38
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019994054478472144,
+      "loss": 1.1395,
+      "step": 39
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019992806068671822,
+      "loss": 1.0477,
+      "step": 40
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019991438822325596,
+      "loss": 1.1387,
+      "step": 41
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019989952755693048,
+      "loss": 1.0391,
+      "step": 42
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019988347886446784,
+      "loss": 1.0089,
+      "step": 43
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019986624233672253,
+      "loss": 1.1454,
+      "step": 44
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019984781817867492,
+      "loss": 1.109,
+      "step": 45
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000199828206609429,
+      "loss": 1.1386,
+      "step": 46
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019980740786220967,
+      "loss": 1.0947,
+      "step": 47
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019978542218435997,
+      "loss": 1.0971,
+      "step": 48
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019976224983733816,
+      "loss": 1.0616,
+      "step": 49
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001997378910967147,
+      "loss": 1.0902,
+      "step": 50
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019971234625216884,
+      "loss": 1.0984,
+      "step": 51
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001996856156074852,
+      "loss": 1.0847,
+      "step": 52
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019965769948055025,
+      "loss": 1.0508,
+      "step": 53
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019962859820334838,
+      "loss": 1.0532,
+      "step": 54
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019959831212195813,
+      "loss": 1.0539,
+      "step": 55
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019956684159654794,
+      "loss": 1.1491,
+      "step": 56
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019953418700137184,
+      "loss": 1.0944,
+      "step": 57
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019950034872476518,
+      "loss": 1.1604,
+      "step": 58
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001994653271691398,
+      "loss": 1.043,
+      "step": 59
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019942912275097943,
+      "loss": 1.1162,
+      "step": 60
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019939173590083456,
+      "loss": 0.9745,
+      "step": 61
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001993531670633175,
+      "loss": 0.984,
+      "step": 62
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019931341669709695,
+      "loss": 1.0841,
+      "step": 63
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019927248527489258,
+      "loss": 1.037,
+      "step": 64
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019923037328346946,
+      "loss": 1.1158,
+      "step": 65
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019918708122363218,
+      "loss": 0.9966,
+      "step": 66
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001991426096102191,
+      "loss": 1.0151,
+      "step": 67
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019909695897209588,
+      "loss": 1.0946,
+      "step": 68
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019905012985214953,
+      "loss": 1.1756,
+      "step": 69
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019900212280728178,
+      "loss": 1.0429,
+      "step": 70
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001989529384084024,
+      "loss": 1.1326,
+      "step": 71
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019890257724042268,
+      "loss": 1.1368,
+      "step": 72
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019885103990224812,
+      "loss": 0.9641,
+      "step": 73
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001987983270067716,
+      "loss": 0.9642,
+      "step": 74
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001987444391808659,
+      "loss": 1.0429,
+      "step": 75
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001986893770653764,
+      "loss": 1.1404,
+      "step": 76
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019863314131511324,
+      "loss": 1.0498,
+      "step": 77
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001985757325988438,
+      "loss": 1.0293,
+      "step": 78
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019851715159928466,
+      "loss": 1.0084,
+      "step": 79
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001984573990130932,
+      "loss": 1.0712,
+      "step": 80
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019839647555085983,
+      "loss": 1.1065,
+      "step": 81
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019833438193709912,
+      "loss": 1.0316,
+      "step": 82
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001982711189102413,
+      "loss": 1.0076,
+      "step": 83
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001982066872226236,
+      "loss": 1.0379,
+      "step": 84
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019814108764048115,
+      "loss": 1.0316,
+      "step": 85
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019807432094393792,
+      "loss": 1.0108,
+      "step": 86
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001980063879269975,
+      "loss": 0.9731,
+      "step": 87
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019793728939753363,
+      "loss": 1.0154,
+      "step": 88
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001978670261772804,
+      "loss": 1.0085,
+      "step": 89
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019779559910182282,
+      "loss": 0.9388,
+      "step": 90
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019772300902058668,
+      "loss": 1.0841,
+      "step": 91
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001976492567968284,
+      "loss": 1.0179,
+      "step": 92
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019757434330762495,
+      "loss": 0.9289,
+      "step": 93
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019749826944386322,
+      "loss": 1.0816,
+      "step": 94
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019742103611022959,
+      "loss": 0.9418,
+      "step": 95
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019734264422519902,
+      "loss": 1.0085,
+      "step": 96
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019726309472102428,
+      "loss": 1.0727,
+      "step": 97
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019718238854372477,
+      "loss": 1.0057,
+      "step": 98
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019710052665307523,
+      "loss": 1.0178,
+      "step": 99
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019701751002259448,
+      "loss": 1.0333,
+      "step": 100
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019693333963953372,
+      "loss": 1.0743,
+      "step": 101
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019684801650486473,
+      "loss": 1.0868,
+      "step": 102
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001967615416332682,
+      "loss": 0.9358,
+      "step": 103
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001966739160531214,
+      "loss": 0.9894,
+      "step": 104
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019658514080648614,
+      "loss": 0.9653,
+      "step": 105
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001964952169490962,
+      "loss": 0.9682,
+      "step": 106
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019640414555034506,
+      "loss": 1.012,
+      "step": 107
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019631192769327285,
+      "loss": 1.0601,
+      "step": 108
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00019621856447455364,
+      "loss": 1.0882,
+      "step": 109
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019612405700448245,
+      "loss": 1.0279,
+      "step": 110
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019602840640696193,
+      "loss": 0.9825,
+      "step": 111
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00019593161381948907,
+      "loss": 1.004,
+      "step": 112
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001958336803931416,
+      "loss": 1.0466,
+      "step": 113
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019573460729256433,
+      "loss": 0.9698,
+      "step": 114
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019563439569595542,
+      "loss": 1.0157,
+      "step": 115
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00019553304679505217,
+      "loss": 1.0221,
+      "step": 116
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00019543056179511695,
+      "loss": 1.051,
+      "step": 117
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019532694191492294,
+      "loss": 1.093,
+      "step": 118
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019522218838673944,
+      "loss": 1.0751,
+      "step": 119
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001951163024563174,
+      "loss": 1.0015,
+      "step": 120
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00019500928538287448,
+      "loss": 1.0284,
+      "step": 121
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001949011384390802,
+      "loss": 0.9958,
+      "step": 122
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001947918629110407,
+      "loss": 1.1206,
+      "step": 123
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00019468146009828346,
+      "loss": 0.9973,
+      "step": 124
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001945699313137419,
+      "loss": 0.9493,
+      "step": 125
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019445727788373974,
+      "loss": 1.1093,
+      "step": 126
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001943435011479752,
+      "loss": 1.0352,
+      "step": 127
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00019422860245950504,
+      "loss": 0.9579,
+      "step": 128
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00019411258318472862,
+      "loss": 1.0053,
+      "step": 129
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00019399544470337146,
+      "loss": 1.0048,
+      "step": 130
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00019387718840846896,
+      "loss": 1.0891,
+      "step": 131
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00019375781570634976,
+      "loss": 1.1069,
+      "step": 132
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00019363732801661913,
+      "loss": 1.0058,
+      "step": 133
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001935157267721419,
+      "loss": 1.0393,
+      "step": 134
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00019339301341902564,
+      "loss": 1.0531,
+      "step": 135
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001932691894166032,
+      "loss": 0.9608,
+      "step": 136
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00019314425623741566,
+      "loss": 1.0752,
+      "step": 137
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00019301821536719452,
+      "loss": 0.9667,
+      "step": 138
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00019289106830484436,
+      "loss": 1.0902,
+      "step": 139
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00019276281656242463,
+      "loss": 0.9993,
+      "step": 140
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00019263346166513196,
+      "loss": 0.9524,
+      "step": 141
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00019250300515128194,
+      "loss": 0.9226,
+      "step": 142
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00019237144857229082,
+      "loss": 1.0367,
+      "step": 143
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00019223879349265695,
+      "loss": 0.9414,
+      "step": 144
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00019210504148994248,
+      "loss": 0.9305,
+      "step": 145
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00019197019415475422,
+      "loss": 0.8975,
+      "step": 146
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00019183425309072494,
+      "loss": 0.9615,
+      "step": 147
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001916972199144943,
+      "loss": 1.0114,
+      "step": 148
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00019155909625568967,
+      "loss": 0.9631,
+      "step": 149
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00019141988375690646,
+      "loss": 0.9684,
+      "step": 150
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00019127958407368895,
+      "loss": 1.0255,
+      "step": 151
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001911381988745104,
+      "loss": 1.1141,
+      "step": 152
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00019099572984075325,
+      "loss": 1.078,
+      "step": 153
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00019085217866668917,
+      "loss": 0.9364,
+      "step": 154
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00019070754705945877,
+      "loss": 0.9256,
+      "step": 155
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001905618367390515,
+      "loss": 0.8849,
+      "step": 156
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00019041504943828508,
+      "loss": 1.1235,
+      "step": 157
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00019026718690278485,
+      "loss": 1.1084,
+      "step": 158
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001901182508909631,
+      "loss": 0.9836,
+      "step": 159
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001899682431739981,
+      "loss": 1.0317,
+      "step": 160
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00018981716553581314,
+      "loss": 0.9848,
+      "step": 161
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00018966501977305512,
+      "loss": 0.9545,
+      "step": 162
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00018951180769507343,
+      "loss": 0.9357,
+      "step": 163
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00018935753112389825,
+      "loss": 1.0412,
+      "step": 164
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00018920219189421883,
+      "loss": 0.957,
+      "step": 165
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00018904579185336195,
+      "loss": 1.0153,
+      "step": 166
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001888883328612697,
+      "loss": 1.0068,
+      "step": 167
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001887298167904774,
+      "loss": 0.9777,
+      "step": 168
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00018857024552609145,
+      "loss": 0.9894,
+      "step": 169
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00018840962096576682,
+      "loss": 1.0518,
+      "step": 170
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00018824794501968444,
+      "loss": 1.0353,
+      "step": 171
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00018808521961052857,
+      "loss": 1.0303,
+      "step": 172
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00018792144667346391,
+      "loss": 0.889,
+      "step": 173
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00018775662815611262,
+      "loss": 0.9982,
+      "step": 174
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00018759076601853106,
+      "loss": 0.9959,
+      "step": 175
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00018742386223318653,
+      "loss": 0.8984,
+      "step": 176
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00018725591878493387,
+      "loss": 0.9193,
+      "step": 177
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00018708693767099184,
+      "loss": 0.9917,
+      "step": 178
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00018691692090091927,
+      "loss": 1.046,
+      "step": 179
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001867458704965912,
+      "loss": 0.9424,
+      "step": 180
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.000186573788492175,
+      "loss": 0.9681,
+      "step": 181
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001864006769341059,
+      "loss": 0.9762,
+      "step": 182
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001862265378810629,
+      "loss": 0.9811,
+      "step": 183
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00018605137340394414,
+      "loss": 1.0102,
+      "step": 184
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00018587518558584233,
+      "loss": 0.9727,
+      "step": 185
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00018569797652201992,
+      "loss": 1.018,
+      "step": 186
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001855197483198843,
+      "loss": 0.9156,
+      "step": 187
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00018534050309896265,
+      "loss": 0.9272,
+      "step": 188
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0001851602429908767,
+      "loss": 0.6938,
+      "step": 189
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00018497897013931754,
+      "loss": 0.7238,
+      "step": 190
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00018479668670001983,
+      "loss": 0.7324,
+      "step": 191
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00018461339484073658,
+      "loss": 0.7076,
+      "step": 192
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00018442909674121295,
+      "loss": 0.6871,
+      "step": 193
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00018424379459316067,
+      "loss": 0.7156,
+      "step": 194
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00018405749060023167,
+      "loss": 0.7082,
+      "step": 195
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00018387018697799218,
+      "loss": 0.7174,
+      "step": 196
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00018368188595389615,
+      "loss": 0.655,
+      "step": 197
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00018349258976725884,
+      "loss": 0.6597,
+      "step": 198
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00018330230066923022,
+      "loss": 0.6928,
+      "step": 199
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00018311102092276815,
+      "loss": 0.6567,
+      "step": 200
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001829187528026115,
+      "loss": 0.7052,
+      "step": 201
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00018272549859525312,
+      "loss": 0.7075,
+      "step": 202
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00018253126059891256,
+      "loss": 0.7101,
+      "step": 203
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00018233604112350877,
+      "loss": 0.6877,
+      "step": 204
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00018213984249063278,
+      "loss": 0.6454,
+      "step": 205
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00018194266703351982,
+      "loss": 0.6476,
+      "step": 206
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00018174451709702187,
+      "loss": 0.7597,
+      "step": 207
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00018154539503757944,
+      "loss": 0.7288,
+      "step": 208
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001813453032231939,
+      "loss": 0.6766,
+      "step": 209
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00018114424403339904,
+      "loss": 0.6553,
+      "step": 210
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00018094221985923294,
+      "loss": 0.6661,
+      "step": 211
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00018073923310320947,
+      "loss": 0.8169,
+      "step": 212
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00018053528617928966,
+      "loss": 0.7249,
+      "step": 213
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00018033038151285319,
+      "loss": 0.6402,
+      "step": 214
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00018012452154066928,
+      "loss": 0.7054,
+      "step": 215
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001799177087108679,
+      "loss": 0.6768,
+      "step": 216
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00017970994548291062,
+      "loss": 0.6639,
+      "step": 217
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00017950123432756133,
+      "loss": 0.7245,
+      "step": 218
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00017929157772685685,
+      "loss": 0.6367,
+      "step": 219
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00017908097817407746,
+      "loss": 0.6771,
+      "step": 220
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00017886943817371726,
+      "loss": 0.7223,
+      "step": 221
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00017865696024145424,
+      "loss": 0.6609,
+      "step": 222
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00017844354690412062,
+      "loss": 0.6434,
+      "step": 223
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00017822920069967248,
+      "loss": 0.8253,
+      "step": 224
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001780139241771599,
+      "loss": 0.7738,
+      "step": 225
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00017779771989669647,
+      "loss": 0.6607,
+      "step": 226
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00017758059042942878,
+      "loss": 0.7007,
+      "step": 227
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00017736253835750604,
+      "loss": 0.7111,
+      "step": 228
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00017714356627404924,
+      "loss": 0.6154,
+      "step": 229
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001769236767831203,
+      "loss": 0.7109,
+      "step": 230
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00017670287249969116,
+      "loss": 0.6744,
+      "step": 231
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00017648115604961272,
+      "loss": 0.7274,
+      "step": 232
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00017625853006958351,
+      "loss": 0.7171,
+      "step": 233
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00017603499720711836,
+      "loss": 0.6993,
+      "step": 234
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00017581056012051696,
+      "loss": 0.6791,
+      "step": 235
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001755852214788322,
+      "loss": 0.7034,
+      "step": 236
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00017535898396183852,
+      "loss": 0.6822,
+      "step": 237
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00017513185025999989,
+      "loss": 0.7047,
+      "step": 238
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.000174903823074438,
+      "loss": 0.6997,
+      "step": 239
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00017467490511689993,
+      "loss": 0.7198,
+      "step": 240
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00017444509910972603,
+      "loss": 0.7137,
+      "step": 241
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00017421440778581755,
+      "loss": 0.7461,
+      "step": 242
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00017398283388860413,
+      "loss": 0.7134,
+      "step": 243
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001737503801720111,
+      "loss": 0.697,
+      "step": 244
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00017351704940042688,
+      "loss": 0.7388,
+      "step": 245
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001732828443486699,
+      "loss": 0.713,
+      "step": 246
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017304776780195576,
+      "loss": 0.7775,
+      "step": 247
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017281182255586406,
+      "loss": 0.7171,
+      "step": 248
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017257501141630516,
+      "loss": 0.6963,
+      "step": 249
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001723373371994869,
+      "loss": 0.7245,
+      "step": 250
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017209880273188075,
+      "loss": 0.6883,
+      "step": 251
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017185941085018879,
+      "loss": 0.6677,
+      "step": 252
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001716191644013094,
+      "loss": 0.687,
+      "step": 253
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001713780662423038,
+      "loss": 0.6464,
+      "step": 254
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017113611924036182,
+      "loss": 0.6727,
+      "step": 255
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.000170893326272768,
+      "loss": 0.7104,
+      "step": 256
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017064969022686724,
+      "loss": 0.6962,
+      "step": 257
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001704052140000305,
+      "loss": 0.7072,
+      "step": 258
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001701599004996203,
+      "loss": 0.7634,
+      "step": 259
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016991375264295634,
+      "loss": 0.7518,
+      "step": 260
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016966677335728046,
+      "loss": 0.6743,
+      "step": 261
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001694189655797222,
+      "loss": 0.7077,
+      "step": 262
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016917033225726368,
+      "loss": 0.7159,
+      "step": 263
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016892087634670438,
+      "loss": 0.7228,
+      "step": 264
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016867060081462643,
+      "loss": 0.7195,
+      "step": 265
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001684195086373589,
+      "loss": 0.7129,
+      "step": 266
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016816760280094267,
+      "loss": 0.7258,
+      "step": 267
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001679148863010948,
+      "loss": 0.672,
+      "step": 268
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016766136214317286,
+      "loss": 0.7452,
+      "step": 269
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016740703334213936,
+      "loss": 0.7328,
+      "step": 270
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016715190292252577,
+      "loss": 0.7458,
+      "step": 271
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001668959739183965,
+      "loss": 0.7344,
+      "step": 272
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00016663924937331296,
+      "loss": 0.6737,
+      "step": 273
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00016638173234029732,
+      "loss": 0.686,
+      "step": 274
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00016612342588179617,
+      "loss": 0.6873,
+      "step": 275
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00016586433306964402,
+      "loss": 0.6912,
+      "step": 276
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00016560445698502702,
+      "loss": 0.6801,
+      "step": 277
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00016534380071844606,
+      "loss": 0.7085,
+      "step": 278
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00016508236736968016,
+      "loss": 0.6999,
+      "step": 279
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00016482016004774948,
+      "loss": 0.6793,
+      "step": 280
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001645571818708785,
+      "loss": 0.6892,
+      "step": 281
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00016429343596645896,
+      "loss": 0.7138,
+      "step": 282
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0001640289254710124,
+      "loss": 0.65,
+      "step": 283
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00016376365353015313,
+      "loss": 0.7034,
+      "step": 284
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00016349762329855068,
+      "loss": 0.7561,
+      "step": 285
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00016323083793989243,
+      "loss": 0.6998,
+      "step": 286
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00016296330062684578,
+      "loss": 0.7439,
+      "step": 287
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00016269501454102065,
+      "loss": 0.6734,
+      "step": 288
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00016242598287293135,
+      "loss": 0.7231,
+      "step": 289
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00016215620882195894,
+      "loss": 0.6827,
+      "step": 290
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0001618856955963131,
+      "loss": 0.7203,
+      "step": 291
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00016161444641299376,
+      "loss": 0.6595,
+      "step": 292
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00016134246449775314,
+      "loss": 0.6964,
+      "step": 293
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00016106975308505722,
+      "loss": 0.6844,
+      "step": 294
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00016079631541804736,
+      "loss": 0.6621,
+      "step": 295
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00016052215474850158,
+      "loss": 0.6574,
+      "step": 296
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00016024727433679617,
+      "loss": 0.7115,
+      "step": 297
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00015997167745186657,
+      "loss": 0.689,
+      "step": 298
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00015969536737116879,
+      "loss": 0.6422,
+      "step": 299
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00015941834738064024,
+      "loss": 0.7619,
+      "step": 300
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00015914062077466076,
+      "loss": 0.7776,
+      "step": 301
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001588621908560134,
+      "loss": 0.7157,
+      "step": 302
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00015858306093584522,
+      "loss": 0.6986,
+      "step": 303
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00015830323433362772,
+      "loss": 0.6644,
+      "step": 304
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00015802271437711753,
+      "loss": 0.6756,
+      "step": 305
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00015774150440231685,
+      "loss": 0.6343,
+      "step": 306
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00015745960775343372,
+      "loss": 0.6833,
+      "step": 307
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00015717702778284218,
+      "loss": 0.686,
+      "step": 308
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0001568937678510425,
+      "loss": 0.7058,
+      "step": 309
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00015660983132662125,
+      "loss": 0.6421,
+      "step": 310
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00015632522158621112,
+      "loss": 0.7206,
+      "step": 311
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00015603994201445083,
+      "loss": 0.6923,
+      "step": 312
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0001557539960039449,
+      "loss": 0.7092,
+      "step": 313
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00015546738695522332,
+      "loss": 0.7465,
+      "step": 314
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00015518011827670095,
+      "loss": 0.701,
+      "step": 315
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00015489219338463714,
+      "loss": 0.7157,
+      "step": 316
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00015460361570309518,
+      "loss": 0.6585,
+      "step": 317
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00015431438866390137,
+      "loss": 0.6455,
+      "step": 318
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00015402451570660437,
+      "loss": 0.7563,
+      "step": 319
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00015373400027843408,
+      "loss": 0.6688,
+      "step": 320
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00015344284583426103,
+      "loss": 0.6508,
+      "step": 321
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0001531510558365549,
+      "loss": 0.6647,
+      "step": 322
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00015285863375534353,
+      "loss": 0.6401,
+      "step": 323
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00015256558306817164,
+      "loss": 0.6478,
+      "step": 324
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00015227190726005953,
+      "loss": 0.7038,
+      "step": 325
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001519776098234615,
+      "loss": 0.6994,
+      "step": 326
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00015168269425822433,
+      "loss": 0.7038,
+      "step": 327
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00015138716407154588,
+      "loss": 0.6898,
+      "step": 328
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00015109102277793306,
+      "loss": 0.6907,
+      "step": 329
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00015079427389916026,
+      "loss": 0.6597,
+      "step": 330
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001504969209642274,
+      "loss": 0.8138,
+      "step": 331
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.000150198967509318,
+      "loss": 0.6845,
+      "step": 332
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.000149900417077757,
+      "loss": 0.7397,
+      "step": 333
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00014960127321996875,
+      "loss": 0.6613,
+      "step": 334
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00014930153949343477,
+      "loss": 0.6948,
+      "step": 335
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00014900121946265143,
+      "loss": 0.7035,
+      "step": 336
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00014870031669908748,
+      "loss": 0.6901,
+      "step": 337
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00014839883478114172,
+      "loss": 0.6756,
+      "step": 338
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00014809677729410046,
+      "loss": 0.6576,
+      "step": 339
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00014779414783009455,
+      "loss": 0.6971,
+      "step": 340
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00014749094998805714,
+      "loss": 0.6282,
+      "step": 341
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00014718718737368055,
+      "loss": 0.7039,
+      "step": 342
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0001468828635993735,
+      "loss": 0.6784,
+      "step": 343
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00014657798228421816,
+      "loss": 0.7283,
+      "step": 344
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00014627254705392706,
+      "loss": 0.7035,
+      "step": 345
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00014596656154080003,
+      "loss": 0.6997,
+      "step": 346
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00014566002938368105,
+      "loss": 0.7096,
+      "step": 347
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00014535295422791474,
+      "loss": 0.7273,
+      "step": 348
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0001450453397253033,
+      "loss": 0.7222,
+      "step": 349
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00014473718953406296,
+      "loss": 0.6742,
+      "step": 350
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00014442850731878038,
+      "loss": 0.6749,
+      "step": 351
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0001441192967503692,
+      "loss": 0.7543,
+      "step": 352
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00014380956150602644,
+      "loss": 0.6966,
+      "step": 353
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00014349930526918854,
+      "loss": 0.6952,
+      "step": 354
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00014318853172948774,
+      "loss": 0.6639,
+      "step": 355
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00014287724458270822,
+      "loss": 0.6729,
+      "step": 356
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00014256544753074194,
+      "loss": 0.647,
+      "step": 357
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00014225314428154494,
+      "loss": 0.711,
+      "step": 358
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0001419403385490929,
+      "loss": 0.6997,
+      "step": 359
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00014162703405333726,
+      "loss": 0.7596,
+      "step": 360
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00014131323452016073,
+      "loss": 0.6618,
+      "step": 361
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00014099894368133325,
+      "loss": 0.7145,
+      "step": 362
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0001406841652744674,
+      "loss": 0.7416,
+      "step": 363
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00014036890304297396,
+      "loss": 0.6452,
+      "step": 364
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0001400531607360176,
+      "loss": 0.6843,
+      "step": 365
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.000139736942108472,
+      "loss": 0.7051,
+      "step": 366
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00013942025092087548,
+      "loss": 0.7183,
+      "step": 367
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00013910309093938597,
+      "loss": 0.6799,
+      "step": 368
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00013878546593573663,
+      "loss": 0.6735,
+      "step": 369
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00013846737968719052,
+      "loss": 0.632,
+      "step": 370
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00013814883597649606,
+      "loss": 0.6892,
+      "step": 371
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00013782983859184192,
+      "loss": 0.6802,
+      "step": 372
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00013751039132681186,
+      "loss": 0.6761,
+      "step": 373
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00013719049798033978,
+      "loss": 0.6824,
+      "step": 374
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00013687016235666447,
+      "loss": 0.72,
+      "step": 375
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0001365493882652844,
+      "loss": 0.6919,
+      "step": 376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00013622817952091225,
+      "loss": 0.3804,
+      "step": 377
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00013590653994342987,
+      "loss": 0.4047,
+      "step": 378
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00013558447335784254,
+      "loss": 0.4268,
+      "step": 379
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0001352619835942337,
+      "loss": 0.389,
+      "step": 380
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0001349390744877192,
+      "loss": 0.4203,
+      "step": 381
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00013461574987840192,
+      "loss": 0.4585,
+      "step": 382
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00013429201361132595,
+      "loss": 0.4049,
+      "step": 383
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00013396786953643084,
+      "loss": 0.4205,
+      "step": 384
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00013364332150850596,
+      "loss": 0.3821,
+      "step": 385
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0001333183733871445,
+      "loss": 0.3748,
+      "step": 386
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0001329930290366977,
+      "loss": 0.3675,
+      "step": 387
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00013266729232622884,
+      "loss": 0.3359,
+      "step": 388
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00013234116712946715,
+      "loss": 0.3514,
+      "step": 389
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00013201465732476194,
+      "loss": 0.3909,
+      "step": 390
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0001316877667950363,
+      "loss": 0.3805,
+      "step": 391
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00013136049942774096,
+      "loss": 0.4049,
+      "step": 392
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0001310328591148081,
+      "loss": 0.4273,
+      "step": 393
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00013070484975260502,
+      "loss": 0.356,
+      "step": 394
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00013037647524188792,
+      "loss": 0.4221,
+      "step": 395
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00013004773948775532,
+      "loss": 0.4015,
+      "step": 396
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00012971864639960175,
+      "loss": 0.3346,
+      "step": 397
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0001293891998910713,
+      "loss": 0.3899,
+      "step": 398
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00012905940388001092,
+      "loss": 0.3619,
+      "step": 399
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00012872926228842398,
+      "loss": 0.3837,
+      "step": 400
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00012839877904242348,
+      "loss": 0.3598,
+      "step": 401
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0001280679580721856,
+      "loss": 0.4221,
+      "step": 402
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0001277368033119026,
+      "loss": 0.4154,
+      "step": 403
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00012740531869973647,
+      "loss": 0.4163,
+      "step": 404
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00012707350817777163,
+      "loss": 0.3702,
+      "step": 405
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00012674137569196857,
+      "loss": 0.4351,
+      "step": 406
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00012640892519211628,
+      "loss": 0.3819,
+      "step": 407
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.000126076160631786,
+      "loss": 0.3664,
+      "step": 408
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00012574308596828364,
+      "loss": 0.3746,
+      "step": 409
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00012540970516260292,
+      "loss": 0.4059,
+      "step": 410
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00012507602217937835,
+      "loss": 0.4267,
+      "step": 411
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00012474204098683795,
+      "loss": 0.4026,
+      "step": 412
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00012440776555675616,
+      "loss": 0.386,
+      "step": 413
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00012407319986440648,
+      "loss": 0.4094,
+      "step": 414
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00012373834788851427,
+      "loss": 0.401,
+      "step": 415
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00012340321361120954,
+      "loss": 0.3938,
+      "step": 416
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00012306780101797936,
+      "loss": 0.4015,
+      "step": 417
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00012273211409762066,
+      "loss": 0.4115,
+      "step": 418
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.00012239615684219273,
+      "loss": 0.4172,
+      "step": 419
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0001220599332469697,
+      "loss": 0.3897,
+      "step": 420
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00012172344731039311,
+      "loss": 0.4019,
+      "step": 421
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00012138670303402429,
+      "loss": 0.3838,
+      "step": 422
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00012104970442249679,
+      "loss": 0.3961,
+      "step": 423
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00012071245548346883,
+      "loss": 0.3596,
+      "step": 424
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00012037496022757552,
+      "loss": 0.4054,
+      "step": 425
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00012003722266838124,
+      "loss": 0.3964,
+      "step": 426
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00011969924682233187,
+      "loss": 0.3895,
+      "step": 427
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0001193610367087071,
+      "loss": 0.3695,
+      "step": 428
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0001190225963495725,
+      "loss": 0.3914,
+      "step": 429
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00011868392976973188,
+      "loss": 0.3781,
+      "step": 430
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0001183450409966792,
+      "loss": 0.412,
+      "step": 431
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00011800593406055082,
+      "loss": 0.4239,
+      "step": 432
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00011766661299407763,
+      "loss": 0.4357,
+      "step": 433
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00011732708183253685,
+      "loss": 0.3945,
+      "step": 434
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00011698734461370428,
+      "loss": 0.4635,
+      "step": 435
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00011664740537780626,
+      "loss": 0.4222,
+      "step": 436
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00011630726816747143,
+      "loss": 0.4205,
+      "step": 437
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00011596693702768288,
+      "loss": 0.3976,
+      "step": 438
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00011562641600572993,
+      "loss": 0.4068,
+      "step": 439
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00011528570915115988,
+      "loss": 0.3564,
+      "step": 440
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00011494482051573027,
+      "loss": 0.4185,
+      "step": 441
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00011460375415336015,
+      "loss": 0.3978,
+      "step": 442
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00011426251412008233,
+      "loss": 0.3957,
+      "step": 443
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0001139211044739948,
+      "loss": 0.4026,
+      "step": 444
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00011357952927521269,
+      "loss": 0.381,
+      "step": 445
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00011323779258581997,
+      "loss": 0.3935,
+      "step": 446
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00011289589846982096,
+      "loss": 0.4046,
+      "step": 447
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00011255385099309228,
+      "loss": 0.363,
+      "step": 448
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0001122116542233342,
+      "loss": 0.3909,
+      "step": 449
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00011186931223002255,
+      "loss": 0.4179,
+      "step": 450
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00011152682908436007,
+      "loss": 0.4005,
+      "step": 451
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00011118420885922822,
+      "loss": 0.431,
+      "step": 452
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00011084145562913856,
+      "loss": 0.414,
+      "step": 453
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0001104985734701844,
+      "loss": 0.4168,
+      "step": 454
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0001101555664599923,
+      "loss": 0.374,
+      "step": 455
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00010981243867767357,
+      "loss": 0.3913,
+      "step": 456
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00010946919420377581,
+      "loss": 0.418,
+      "step": 457
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00010912583712023425,
+      "loss": 0.4128,
+      "step": 458
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0001087823715103234,
+      "loss": 0.4182,
+      "step": 459
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0001084388014586083,
+      "loss": 0.3988,
+      "step": 460
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00010809513105089607,
+      "loss": 0.4154,
+      "step": 461
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00010775136437418736,
+      "loss": 0.412,
+      "step": 462
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00010740750551662752,
+      "loss": 0.3769,
+      "step": 463
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00010706355856745824,
+      "loss": 0.3595,
+      "step": 464
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0001067195276169688,
+      "loss": 0.425,
+      "step": 465
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00010637541675644742,
+      "loss": 0.3763,
+      "step": 466
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00010603123007813263,
+      "loss": 0.4234,
+      "step": 467
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00010568697167516461,
+      "loss": 0.4064,
+      "step": 468
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00010534264564153651,
+      "loss": 0.3525,
+      "step": 469
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00010499825607204574,
+      "loss": 0.4296,
+      "step": 470
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0001046538070622453,
+      "loss": 0.387,
+      "step": 471
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00010430930270839507,
+      "loss": 0.4323,
+      "step": 472
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00010396474710741311,
+      "loss": 0.4123,
+      "step": 473
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00010362014435682688,
+      "loss": 0.39,
+      "step": 474
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00010327549855472455,
+      "loss": 0.4337,
+      "step": 475
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00010293081379970637,
+      "loss": 0.3811,
+      "step": 476
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00010258609419083567,
+      "loss": 0.3828,
+      "step": 477
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00010224134382759041,
+      "loss": 0.4279,
+      "step": 478
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00010189656680981415,
+      "loss": 0.4081,
+      "step": 479
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00010155176723766757,
+      "loss": 0.3947,
+      "step": 480
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00010120694921157952,
+      "loss": 0.3972,
+      "step": 481
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00010086211683219824,
+      "loss": 0.4335,
+      "step": 482
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0001005172742003428,
+      "loss": 0.4024,
+      "step": 483
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.000100172425416954,
+      "loss": 0.4127,
+      "step": 484
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.9827574583046e-05,
+      "loss": 0.3778,
+      "step": 485
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.948272579965723e-05,
+      "loss": 0.3849,
+      "step": 486
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.913788316780177e-05,
+      "loss": 0.3924,
+      "step": 487
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.879305078842052e-05,
+      "loss": 0.3981,
+      "step": 488
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.844823276233243e-05,
+      "loss": 0.4112,
+      "step": 489
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.810343319018588e-05,
+      "loss": 0.3772,
+      "step": 490
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.775865617240964e-05,
+      "loss": 0.4012,
+      "step": 491
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.741390580916434e-05,
+      "loss": 0.4257,
+      "step": 492
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.706918620029364e-05,
+      "loss": 0.4266,
+      "step": 493
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.672450144527546e-05,
+      "loss": 0.4075,
+      "step": 494
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.637985564317316e-05,
+      "loss": 0.3916,
+      "step": 495
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.603525289258689e-05,
+      "loss": 0.4133,
+      "step": 496
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 9.569069729160494e-05,
+      "loss": 0.415,
+      "step": 497
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.534619293775471e-05,
+      "loss": 0.4424,
+      "step": 498
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.500174392795431e-05,
+      "loss": 0.3833,
+      "step": 499
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 9.465735435846348e-05,
+      "loss": 0.4125,
+      "step": 500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 9.43130283248354e-05,
+      "loss": 0.415,
+      "step": 501
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.396876992186738e-05,
+      "loss": 0.4123,
+      "step": 502
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 9.362458324355257e-05,
+      "loss": 0.4073,
+      "step": 503
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 9.328047238303121e-05,
+      "loss": 0.4492,
+      "step": 504
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.293644143254177e-05,
+      "loss": 0.3842,
+      "step": 505
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.259249448337253e-05,
+      "loss": 0.3957,
+      "step": 506
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.224863562581265e-05,
+      "loss": 0.4234,
+      "step": 507
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.190486894910394e-05,
+      "loss": 0.394,
+      "step": 508
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.156119854139173e-05,
+      "loss": 0.438,
+      "step": 509
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.121762848967666e-05,
+      "loss": 0.4308,
+      "step": 510
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.087416287976577e-05,
+      "loss": 0.3848,
+      "step": 511
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.053080579622422e-05,
+      "loss": 0.4004,
+      "step": 512
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.018756132232647e-05,
+      "loss": 0.4204,
+      "step": 513
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.984443354000774e-05,
+      "loss": 0.4189,
+      "step": 514
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.950142652981563e-05,
+      "loss": 0.3693,
+      "step": 515
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 8.915854437086146e-05,
+      "loss": 0.393,
+      "step": 516
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.881579114077181e-05,
+      "loss": 0.3968,
+      "step": 517
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.847317091563994e-05,
+      "loss": 0.3969,
+      "step": 518
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 8.813068776997747e-05,
+      "loss": 0.3808,
+      "step": 519
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.778834577666584e-05,
+      "loss": 0.376,
+      "step": 520
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.744614900690777e-05,
+      "loss": 0.4335,
+      "step": 521
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.710410153017905e-05,
+      "loss": 0.4409,
+      "step": 522
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.676220741418005e-05,
+      "loss": 0.3661,
+      "step": 523
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.642047072478734e-05,
+      "loss": 0.4364,
+      "step": 524
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 8.607889552600524e-05,
+      "loss": 0.4184,
+      "step": 525
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.573748587991769e-05,
+      "loss": 0.4287,
+      "step": 526
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.539624584663986e-05,
+      "loss": 0.3787,
+      "step": 527
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.505517948426976e-05,
+      "loss": 0.3736,
+      "step": 528
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 8.471429084884012e-05,
+      "loss": 0.3949,
+      "step": 529
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.437358399427011e-05,
+      "loss": 0.3967,
+      "step": 530
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.403306297231714e-05,
+      "loss": 0.3834,
+      "step": 531
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.369273183252858e-05,
+      "loss": 0.3839,
+      "step": 532
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.335259462219374e-05,
+      "loss": 0.3951,
+      "step": 533
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 8.301265538629573e-05,
+      "loss": 0.3814,
+      "step": 534
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.267291816746319e-05,
+      "loss": 0.4207,
+      "step": 535
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.233338700592244e-05,
+      "loss": 0.4081,
+      "step": 536
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.199406593944916e-05,
+      "loss": 0.4028,
+      "step": 537
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 8.165495900332083e-05,
+      "loss": 0.3982,
+      "step": 538
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.131607023026814e-05,
+      "loss": 0.3836,
+      "step": 539
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.097740365042748e-05,
+      "loss": 0.4145,
+      "step": 540
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.063896329129292e-05,
+      "loss": 0.3982,
+      "step": 541
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.030075317766814e-05,
+      "loss": 0.3979,
+      "step": 542
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.996277733161881e-05,
+      "loss": 0.414,
+      "step": 543
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.962503977242449e-05,
+      "loss": 0.3386,
+      "step": 544
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 7.928754451653119e-05,
+      "loss": 0.379,
+      "step": 545
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 7.895029557750324e-05,
+      "loss": 0.4198,
+      "step": 546
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 7.861329696597576e-05,
+      "loss": 0.3939,
+      "step": 547
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 7.827655268960692e-05,
+      "loss": 0.3767,
+      "step": 548
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 7.794006675303031e-05,
+      "loss": 0.3972,
+      "step": 549
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 7.760384315780731e-05,
+      "loss": 0.3748,
+      "step": 550
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 7.726788590237933e-05,
+      "loss": 0.3914,
+      "step": 551
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 7.693219898202067e-05,
+      "loss": 0.3768,
+      "step": 552
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 7.659678638879048e-05,
+      "loss": 0.3819,
+      "step": 553
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 7.626165211148575e-05,
+      "loss": 0.4186,
+      "step": 554
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 7.592680013559354e-05,
+      "loss": 0.3752,
+      "step": 555
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.559223444324385e-05,
+      "loss": 0.4123,
+      "step": 556
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.525795901316207e-05,
+      "loss": 0.408,
+      "step": 557
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 7.492397782062169e-05,
+      "loss": 0.3921,
+      "step": 558
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 7.459029483739711e-05,
+      "loss": 0.3909,
+      "step": 559
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 7.425691403171639e-05,
+      "loss": 0.3834,
+      "step": 560
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 7.392383936821403e-05,
+      "loss": 0.3637,
+      "step": 561
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.359107480788372e-05,
+      "loss": 0.4098,
+      "step": 562
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 7.325862430803148e-05,
+      "loss": 0.3947,
+      "step": 563
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 7.292649182222837e-05,
+      "loss": 0.4422,
+      "step": 564
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 7.259468130026356e-05,
+      "loss": 0.2145,
+      "step": 565
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 7.22631966880974e-05,
+      "loss": 0.226,
+      "step": 566
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 7.193204192781443e-05,
+      "loss": 0.2171,
+      "step": 567
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 7.160122095757653e-05,
+      "loss": 0.2361,
+      "step": 568
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.127073771157606e-05,
+      "loss": 0.2122,
+      "step": 569
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.094059611998909e-05,
+      "loss": 0.2158,
+      "step": 570
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.061080010892873e-05,
+      "loss": 0.2161,
+      "step": 571
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 7.028135360039827e-05,
+      "loss": 0.2232,
+      "step": 572
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 6.995226051224475e-05,
+      "loss": 0.2088,
+      "step": 573
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 6.96235247581121e-05,
+      "loss": 0.2123,
+      "step": 574
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 6.9295150247395e-05,
+      "loss": 0.2266,
+      "step": 575
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 6.896714088519193e-05,
+      "loss": 0.1801,
+      "step": 576
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 6.863950057225904e-05,
+      "loss": 0.2187,
+      "step": 577
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 6.831223320496371e-05,
+      "loss": 0.21,
+      "step": 578
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 6.798534267523806e-05,
+      "loss": 0.2233,
+      "step": 579
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 6.765883287053287e-05,
+      "loss": 0.2025,
+      "step": 580
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 6.733270767377118e-05,
+      "loss": 0.219,
+      "step": 581
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 6.700697096330233e-05,
+      "loss": 0.2202,
+      "step": 582
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 6.668162661285552e-05,
+      "loss": 0.1953,
+      "step": 583
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 6.63566784914941e-05,
+      "loss": 0.2078,
+      "step": 584
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 6.603213046356918e-05,
+      "loss": 0.1914,
+      "step": 585
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 6.570798638867407e-05,
+      "loss": 0.2202,
+      "step": 586
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 6.53842501215981e-05,
+      "loss": 0.2239,
+      "step": 587
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 6.506092551228079e-05,
+      "loss": 0.2178,
+      "step": 588
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 6.473801640576633e-05,
+      "loss": 0.2101,
+      "step": 589
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 6.441552664215748e-05,
+      "loss": 0.2105,
+      "step": 590
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 6.409346005657018e-05,
+      "loss": 0.201,
+      "step": 591
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 6.377182047908776e-05,
+      "loss": 0.2115,
+      "step": 592
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 6.345061173471564e-05,
+      "loss": 0.2137,
+      "step": 593
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 6.312983764333555e-05,
+      "loss": 0.2025,
+      "step": 594
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 6.280950201966025e-05,
+      "loss": 0.2072,
+      "step": 595
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 6.248960867318816e-05,
+      "loss": 0.2125,
+      "step": 596
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 6.21701614081581e-05,
+      "loss": 0.2134,
+      "step": 597
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 6.185116402350397e-05,
+      "loss": 0.1979,
+      "step": 598
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.153262031280951e-05,
+      "loss": 0.2166,
+      "step": 599
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 6.121453406426341e-05,
+      "loss": 0.1843,
+      "step": 600
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.089690906061405e-05,
+      "loss": 0.2279,
+      "step": 601
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 6.057974907912457e-05,
+      "loss": 0.2293,
+      "step": 602
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 6.0263057891528e-05,
+      "loss": 0.1968,
+      "step": 603
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 5.994683926398241e-05,
+      "loss": 0.2056,
+      "step": 604
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5.9631096957026065e-05,
+      "loss": 0.2134,
+      "step": 605
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 5.931583472553265e-05,
+      "loss": 0.2124,
+      "step": 606
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.900105631866676e-05,
+      "loss": 0.1952,
+      "step": 607
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 5.86867654798393e-05,
+      "loss": 0.2294,
+      "step": 608
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5.8372965946662815e-05,
+      "loss": 0.2078,
+      "step": 609
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 5.805966145090709e-05,
+      "loss": 0.2053,
+      "step": 610
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 5.7746855718455085e-05,
+      "loss": 0.1753,
+      "step": 611
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5.7434552469258085e-05,
+      "loss": 0.1954,
+      "step": 612
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 5.712275541729184e-05,
+      "loss": 0.245,
+      "step": 613
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.6811468270512266e-05,
+      "loss": 0.2103,
+      "step": 614
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 5.650069473081148e-05,
+      "loss": 0.2288,
+      "step": 615
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5.619043849397358e-05,
+      "loss": 0.1956,
+      "step": 616
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 5.588070324963082e-05,
+      "loss": 0.2034,
+      "step": 617
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.557149268121965e-05,
+      "loss": 0.2061,
+      "step": 618
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 5.5262810465937045e-05,
+      "loss": 0.2272,
+      "step": 619
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5.49546602746967e-05,
+      "loss": 0.2099,
+      "step": 620
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 5.464704577208526e-05,
+      "loss": 0.2174,
+      "step": 621
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.433997061631897e-05,
+      "loss": 0.1958,
+      "step": 622
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 5.4033438459199994e-05,
+      "loss": 0.2127,
+      "step": 623
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.3727452946072965e-05,
+      "loss": 0.1877,
+      "step": 624
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 5.3422017715781856e-05,
+      "loss": 0.2127,
+      "step": 625
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 5.311713640062653e-05,
+      "loss": 0.2122,
+      "step": 626
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.2812812626319495e-05,
+      "loss": 0.1962,
+      "step": 627
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 5.2509050011942885e-05,
+      "loss": 0.1989,
+      "step": 628
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.2205852169905434e-05,
+      "loss": 0.2066,
+      "step": 629
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 5.1903222705899553e-05,
+      "loss": 0.216,
+      "step": 630
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.160116521885827e-05,
+      "loss": 0.1945,
+      "step": 631
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 5.129968330091253e-05,
+      "loss": 0.22,
+      "step": 632
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.099878053734859e-05,
+      "loss": 0.2085,
+      "step": 633
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 5.069846050656525e-05,
+      "loss": 0.2149,
+      "step": 634
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.039872678003129e-05,
+      "loss": 0.2147,
+      "step": 635
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 5.0099582922243017e-05,
+      "loss": 0.2025,
+      "step": 636
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 4.980103249068203e-05,
+      "loss": 0.209,
+      "step": 637
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 4.950307903577259e-05,
+      "loss": 0.2066,
+      "step": 638
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 4.920572610083977e-05,
+      "loss": 0.1999,
+      "step": 639
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 4.8908977222066966e-05,
+      "loss": 0.1888,
+      "step": 640
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 4.861283592845416e-05,
+      "loss": 0.203,
+      "step": 641
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 4.831730574177568e-05,
+      "loss": 0.1905,
+      "step": 642
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 4.8022390176538555e-05,
+      "loss": 0.1772,
+      "step": 643
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 4.772809273994048e-05,
+      "loss": 0.2002,
+      "step": 644
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 4.7434416931828384e-05,
+      "loss": 0.2037,
+      "step": 645
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 4.714136624465654e-05,
+      "loss": 0.2121,
+      "step": 646
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 4.684894416344511e-05,
+      "loss": 0.2193,
+      "step": 647
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 4.655715416573898e-05,
+      "loss": 0.2119,
+      "step": 648
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 4.626599972156593e-05,
+      "loss": 0.2147,
+      "step": 649
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.5975484293395696e-05,
+      "loss": 0.1987,
+      "step": 650
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.568561133609865e-05,
+      "loss": 0.23,
+      "step": 651
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.5396384296904815e-05,
+      "loss": 0.2032,
+      "step": 652
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.510780661536288e-05,
+      "loss": 0.1954,
+      "step": 653
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.481988172329913e-05,
+      "loss": 0.2175,
+      "step": 654
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.453261304477672e-05,
+      "loss": 0.1893,
+      "step": 655
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 4.4246003996055083e-05,
+      "loss": 0.2012,
+      "step": 656
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 4.396005798554919e-05,
+      "loss": 0.1918,
+      "step": 657
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.367477841378889e-05,
+      "loss": 0.1682,
+      "step": 658
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.3390168673378775e-05,
+      "loss": 0.2308,
+      "step": 659
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.310623214895754e-05,
+      "loss": 0.1955,
+      "step": 660
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 4.282297221715785e-05,
+      "loss": 0.1907,
+      "step": 661
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 4.2540392246566285e-05,
+      "loss": 0.216,
+      "step": 662
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 4.225849559768316e-05,
+      "loss": 0.214,
+      "step": 663
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 4.197728562288251e-05,
+      "loss": 0.1959,
+      "step": 664
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.169676566637232e-05,
+      "loss": 0.221,
+      "step": 665
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.1416939064154783e-05,
+      "loss": 0.2136,
+      "step": 666
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.113780914398659e-05,
+      "loss": 0.1989,
+      "step": 667
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.0859379225339265e-05,
+      "loss": 0.2084,
+      "step": 668
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.0581652619359775e-05,
+      "loss": 0.2125,
+      "step": 669
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.030463262883121e-05,
+      "loss": 0.2134,
+      "step": 670
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.0028322548133436e-05,
+      "loss": 0.1939,
+      "step": 671
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 3.975272566320386e-05,
+      "loss": 0.1979,
+      "step": 672
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 3.9477845251498414e-05,
+      "loss": 0.2245,
+      "step": 673
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.920368458195268e-05,
+      "loss": 0.2091,
+      "step": 674
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 3.893024691494279e-05,
+      "loss": 0.2384,
+      "step": 675
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.86575355022469e-05,
+      "loss": 0.1949,
+      "step": 676
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 3.8385553587006266e-05,
+      "loss": 0.1908,
+      "step": 677
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.811430440368694e-05,
+      "loss": 0.1898,
+      "step": 678
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.7843791178041054e-05,
+      "loss": 0.21,
+      "step": 679
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.7574017127068653e-05,
+      "loss": 0.2023,
+      "step": 680
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 3.730498545897938e-05,
+      "loss": 0.1823,
+      "step": 681
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.703669937315423e-05,
+      "loss": 0.2009,
+      "step": 682
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 3.676916206010761e-05,
+      "loss": 0.2006,
+      "step": 683
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.65023767014493e-05,
+      "loss": 0.2102,
+      "step": 684
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 3.6236346469846895e-05,
+      "loss": 0.2274,
+      "step": 685
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.597107452898763e-05,
+      "loss": 0.2016,
+      "step": 686
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 3.570656403354107e-05,
+      "loss": 0.1788,
+      "step": 687
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.5442818129121494e-05,
+      "loss": 0.203,
+      "step": 688
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.5179839952250524e-05,
+      "loss": 0.203,
+      "step": 689
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.491763263031987e-05,
+      "loss": 0.2054,
+      "step": 690
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.465619928155397e-05,
+      "loss": 0.2175,
+      "step": 691
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.4395543014972985e-05,
+      "loss": 0.1948,
+      "step": 692
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.413566693035598e-05,
+      "loss": 0.2305,
+      "step": 693
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.387657411820388e-05,
+      "loss": 0.208,
+      "step": 694
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.3618267659702686e-05,
+      "loss": 0.1987,
+      "step": 695
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.336075062668707e-05,
+      "loss": 0.2087,
+      "step": 696
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.310402608160356e-05,
+      "loss": 0.1867,
+      "step": 697
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 3.284809707747426e-05,
+      "loss": 0.193,
+      "step": 698
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.2592966657860625e-05,
+      "loss": 0.2181,
+      "step": 699
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 3.2338637856827145e-05,
+      "loss": 0.2251,
+      "step": 700
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.208511369890523e-05,
+      "loss": 0.2046,
+      "step": 701
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 3.1832397199057326e-05,
+      "loss": 0.1922,
+      "step": 702
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.1580491362641085e-05,
+      "loss": 0.2165,
+      "step": 703
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 3.13293991853736e-05,
+      "loss": 0.1802,
+      "step": 704
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.107912365329566e-05,
+      "loss": 0.1901,
+      "step": 705
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.082966774273637e-05,
+      "loss": 0.213,
+      "step": 706
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.0581034420277776e-05,
+      "loss": 0.2196,
+      "step": 707
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.0333226642719548e-05,
+      "loss": 0.2127,
+      "step": 708
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 3.0086247357043705e-05,
+      "loss": 0.1955,
+      "step": 709
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.9840099500379702e-05,
+      "loss": 0.193,
+      "step": 710
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.9594785999969542e-05,
+      "loss": 0.1934,
+      "step": 711
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.9350309773132765e-05,
+      "loss": 0.2202,
+      "step": 712
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.9106673727232015e-05,
+      "loss": 0.1906,
+      "step": 713
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.8863880759638184e-05,
+      "loss": 0.2021,
+      "step": 714
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.8621933757696228e-05,
+      "loss": 0.1927,
+      "step": 715
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.8380835598690602e-05,
+      "loss": 0.1975,
+      "step": 716
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.8140589149811214e-05,
+      "loss": 0.2174,
+      "step": 717
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.790119726811925e-05,
+      "loss": 0.1967,
+      "step": 718
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.766266280051316e-05,
+      "loss": 0.1885,
+      "step": 719
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 2.7424988583694832e-05,
+      "loss": 0.2141,
+      "step": 720
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.718817744413594e-05,
+      "loss": 0.2245,
+      "step": 721
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.695223219804427e-05,
+      "loss": 0.2199,
+      "step": 722
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 2.671715565133015e-05,
+      "loss": 0.2163,
+      "step": 723
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 2.6482950599573163e-05,
+      "loss": 0.1899,
+      "step": 724
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.6249619827988915e-05,
+      "loss": 0.1866,
+      "step": 725
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.6017166111395884e-05,
+      "loss": 0.1954,
+      "step": 726
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.5785592214182476e-05,
+      "loss": 0.2127,
+      "step": 727
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.555490089027399e-05,
+      "loss": 0.2049,
+      "step": 728
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.5325094883100108e-05,
+      "loss": 0.2079,
+      "step": 729
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.509617692556201e-05,
+      "loss": 0.2061,
+      "step": 730
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.4868149740000113e-05,
+      "loss": 0.2302,
+      "step": 731
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.464101603816148e-05,
+      "loss": 0.1819,
+      "step": 732
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.4414778521167814e-05,
+      "loss": 0.189,
+      "step": 733
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.418943987948309e-05,
+      "loss": 0.2262,
+      "step": 734
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.3965002792881664e-05,
+      "loss": 0.2094,
+      "step": 735
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.3741469930416495e-05,
+      "loss": 0.1811,
+      "step": 736
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.351884395038727e-05,
+      "loss": 0.1906,
+      "step": 737
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.3297127500308845e-05,
+      "loss": 0.196,
+      "step": 738
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.3076323216879714e-05,
+      "loss": 0.1987,
+      "step": 739
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.2856433725950767e-05,
+      "loss": 0.1889,
+      "step": 740
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.2637461642493963e-05,
+      "loss": 0.2074,
+      "step": 741
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.2419409570571248e-05,
+      "loss": 0.2005,
+      "step": 742
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 2.2202280103303553e-05,
+      "loss": 0.1966,
+      "step": 743
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.198607582284009e-05,
+      "loss": 0.2121,
+      "step": 744
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 2.1770799300327538e-05,
+      "loss": 0.2034,
+      "step": 745
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.1556453095879424e-05,
+      "loss": 0.2022,
+      "step": 746
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.134303975854576e-05,
+      "loss": 0.2034,
+      "step": 747
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.113056182628277e-05,
+      "loss": 0.1977,
+      "step": 748
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 2.0919021825922536e-05,
+      "loss": 0.1987,
+      "step": 749
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.070842227314317e-05,
+      "loss": 0.1859,
+      "step": 750
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 2.049876567243869e-05,
+      "loss": 0.1722,
+      "step": 751
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 2.0290054517089408e-05,
+      "loss": 0.2052,
+      "step": 752
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 2.0082291289132106e-05,
+      "loss": 0.1345,
+      "step": 753
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.987547845933072e-05,
+      "loss": 0.1436,
+      "step": 754
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.9669618487146813e-05,
+      "loss": 0.1271,
+      "step": 755
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.9464713820710334e-05,
+      "loss": 0.1301,
+      "step": 756
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.926076689679054e-05,
+      "loss": 0.1289,
+      "step": 757
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.9057780140767044e-05,
+      "loss": 0.1329,
+      "step": 758
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.8855755966600962e-05,
+      "loss": 0.1352,
+      "step": 759
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.8654696776806125e-05,
+      "loss": 0.1274,
+      "step": 760
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.8454604962420595e-05,
+      "loss": 0.1184,
+      "step": 761
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.825548290297817e-05,
+      "loss": 0.1267,
+      "step": 762
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.805733296648018e-05,
+      "loss": 0.1272,
+      "step": 763
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.786015750936725e-05,
+      "loss": 0.1241,
+      "step": 764
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.766395887649124e-05,
+      "loss": 0.124,
+      "step": 765
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.7468739401087486e-05,
+      "loss": 0.1202,
+      "step": 766
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.727450140474689e-05,
+      "loss": 0.1194,
+      "step": 767
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.7081247197388516e-05,
+      "loss": 0.1261,
+      "step": 768
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.688897907723186e-05,
+      "loss": 0.1311,
+      "step": 769
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.669769933076981e-05,
+      "loss": 0.1217,
+      "step": 770
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.6507410232741206e-05,
+      "loss": 0.1228,
+      "step": 771
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.631811404610387e-05,
+      "loss": 0.1211,
+      "step": 772
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.612981302200782e-05,
+      "loss": 0.1238,
+      "step": 773
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.5942509399768336e-05,
+      "loss": 0.1278,
+      "step": 774
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.5756205406839364e-05,
+      "loss": 0.1278,
+      "step": 775
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.557090325878705e-05,
+      "loss": 0.1291,
+      "step": 776
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.5386605159263424e-05,
+      "loss": 0.1329,
+      "step": 777
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.520331329998017e-05,
+      "loss": 0.128,
+      "step": 778
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.50210298606825e-05,
+      "loss": 0.129,
+      "step": 779
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.4839757009123301e-05,
+      "loss": 0.1235,
+      "step": 780
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.465949690103735e-05,
+      "loss": 0.1157,
+      "step": 781
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.448025168011572e-05,
+      "loss": 0.1201,
+      "step": 782
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.430202347798012e-05,
+      "loss": 0.1132,
+      "step": 783
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.4124814414157706e-05,
+      "loss": 0.1335,
+      "step": 784
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.394862659605589e-05,
+      "loss": 0.113,
+      "step": 785
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.377346211893712e-05,
+      "loss": 0.1235,
+      "step": 786
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.3599323065894098e-05,
+      "loss": 0.1154,
+      "step": 787
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.3426211507825026e-05,
+      "loss": 0.1252,
+      "step": 788
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.325412950340882e-05,
+      "loss": 0.1206,
+      "step": 789
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.3083079099080764e-05,
+      "loss": 0.1281,
+      "step": 790
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.2913062329008164e-05,
+      "loss": 0.1279,
+      "step": 791
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.2744081215066128e-05,
+      "loss": 0.1254,
+      "step": 792
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.25761377668135e-05,
+      "loss": 0.1273,
+      "step": 793
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.240923398146897e-05,
+      "loss": 0.1165,
+      "step": 794
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.2243371843887375e-05,
+      "loss": 0.1161,
+      "step": 795
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.207855332653609e-05,
+      "loss": 0.1232,
+      "step": 796
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.1914780389471446e-05,
+      "loss": 0.1292,
+      "step": 797
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.1752054980315597e-05,
+      "loss": 0.1286,
+      "step": 798
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.159037903423319e-05,
+      "loss": 0.1284,
+      "step": 799
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.1429754473908538e-05,
+      "loss": 0.1233,
+      "step": 800
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.1270183209522611e-05,
+      "loss": 0.1252,
+      "step": 801
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.111166713873032e-05,
+      "loss": 0.1158,
+      "step": 802
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.0954208146638067e-05,
+      "loss": 0.1246,
+      "step": 803
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.0797808105781181e-05,
+      "loss": 0.1287,
+      "step": 804
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.0642468876101786e-05,
+      "loss": 0.1275,
+      "step": 805
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.0488192304926559e-05,
+      "loss": 0.119,
+      "step": 806
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.0334980226944879e-05,
+      "loss": 0.1237,
+      "step": 807
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.0182834464186874e-05,
+      "loss": 0.1294,
+      "step": 808
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.0031756826001914e-05,
+      "loss": 0.1249,
+      "step": 809
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 9.881749109036931e-06,
+      "loss": 0.1317,
+      "step": 810
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 9.732813097215176e-06,
+      "loss": 0.1169,
+      "step": 811
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 9.584950561714944e-06,
+      "loss": 0.1273,
+      "step": 812
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 9.438163260948495e-06,
+      "loss": 0.1264,
+      "step": 813
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 9.292452940541229e-06,
+      "loss": 0.1166,
+      "step": 814
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 9.147821333310858e-06,
+      "loss": 0.1342,
+      "step": 815
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 9.004270159246764e-06,
+      "loss": 0.1256,
+      "step": 816
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.86180112548961e-06,
+      "loss": 0.125,
+      "step": 817
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.720415926311053e-06,
+      "loss": 0.1257,
+      "step": 818
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.580116243093562e-06,
+      "loss": 0.1273,
+      "step": 819
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.440903744310357e-06,
+      "loss": 0.1264,
+      "step": 820
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.302780085505679e-06,
+      "loss": 0.1157,
+      "step": 821
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.165746909275084e-06,
+      "loss": 0.1276,
+      "step": 822
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.029805845245819e-06,
+      "loss": 0.1327,
+      "step": 823
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 7.894958510057527e-06,
+      "loss": 0.1287,
+      "step": 824
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.761206507343044e-06,
+      "loss": 0.123,
+      "step": 825
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.628551427709218e-06,
+      "loss": 0.1272,
+      "step": 826
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 7.496994848718075e-06,
+      "loss": 0.1268,
+      "step": 827
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 7.366538334868045e-06,
+      "loss": 0.1303,
+      "step": 828
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.23718343757539e-06,
+      "loss": 0.1186,
+      "step": 829
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.108931695155663e-06,
+      "loss": 0.1284,
+      "step": 830
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 6.981784632805466e-06,
+      "loss": 0.1328,
+      "step": 831
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 6.855743762584355e-06,
+      "loss": 0.1284,
+      "step": 832
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 6.730810583396807e-06,
+      "loss": 0.1213,
+      "step": 833
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 6.606986580974384e-06,
+      "loss": 0.1164,
+      "step": 834
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 6.4842732278580845e-06,
+      "loss": 0.1223,
+      "step": 835
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 6.362671983380886e-06,
+      "loss": 0.1207,
+      "step": 836
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 6.242184293650233e-06,
+      "loss": 0.1259,
+      "step": 837
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 6.122811591531064e-06,
+      "loss": 0.1221,
+      "step": 838
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 6.004555296628556e-06,
+      "loss": 0.1256,
+      "step": 839
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5.887416815271407e-06,
+      "loss": 0.1136,
+      "step": 840
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5.771397540494961e-06,
+      "loss": 0.1111,
+      "step": 841
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.656498852024828e-06,
+      "loss": 0.1175,
+      "step": 842
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.542722116260257e-06,
+      "loss": 0.1187,
+      "step": 843
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5.430068686258094e-06,
+      "loss": 0.111,
+      "step": 844
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5.3185399017165465e-06,
+      "loss": 0.1276,
+      "step": 845
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5.208137088959308e-06,
+      "loss": 0.1277,
+      "step": 846
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 5.09886156091981e-06,
+      "loss": 0.1189,
+      "step": 847
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 4.990714617125536e-06,
+      "loss": 0.1243,
+      "step": 848
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 4.88369754368263e-06,
+      "loss": 0.131,
+      "step": 849
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 4.777811613260574e-06,
+      "loss": 0.1235,
+      "step": 850
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 4.673058085077053e-06,
+      "loss": 0.1324,
+      "step": 851
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 4.569438204883036e-06,
+      "loss": 0.1274,
+      "step": 852
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 4.466953204947843e-06,
+      "loss": 0.1305,
+      "step": 853
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 4.36560430404459e-06,
+      "loss": 0.1341,
+      "step": 854
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.265392707435678e-06,
+      "loss": 0.1228,
+      "step": 855
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.166319606858438e-06,
+      "loss": 0.1261,
+      "step": 856
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.0683861805109524e-06,
+      "loss": 0.1248,
+      "step": 857
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 3.97159359303807e-06,
+      "loss": 0.1264,
+      "step": 858
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 3.8759429955175564e-06,
+      "loss": 0.1241,
+      "step": 859
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 3.781435525446364e-06,
+      "loss": 0.1189,
+      "step": 860
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 3.688072306727164e-06,
+      "loss": 0.1211,
+      "step": 861
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 3.595854449654945e-06,
+      "loss": 0.125,
+      "step": 862
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 3.5047830509038037e-06,
+      "loss": 0.1261,
+      "step": 863
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.414859193513886e-06,
+      "loss": 0.1318,
+      "step": 864
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.3260839468786e-06,
+      "loss": 0.1187,
+      "step": 865
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.2384583667318114e-06,
+      "loss": 0.1249,
+      "step": 866
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.15198349513528e-06,
+      "loss": 0.1298,
+      "step": 867
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.0666603604663e-06,
+      "loss": 0.13,
+      "step": 868
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.9824899774055113e-06,
+      "loss": 0.1189,
+      "step": 869
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.899473346924775e-06,
+      "loss": 0.1195,
+      "step": 870
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 2.817611456275249e-06,
+      "loss": 0.1191,
+      "step": 871
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.736905278975721e-06,
+      "loss": 0.1145,
+      "step": 872
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 2.657355774800985e-06,
+      "loss": 0.1215,
+      "step": 873
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.5789638897704247e-06,
+      "loss": 0.1224,
+      "step": 874
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.501730556136783e-06,
+      "loss": 0.1225,
+      "step": 875
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.4256566923750555e-06,
+      "loss": 0.1302,
+      "step": 876
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.3507432031716016e-06,
+      "loss": 0.1211,
+      "step": 877
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 2.27699097941334e-06,
+      "loss": 0.1404,
+      "step": 878
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.2044008981771947e-06,
+      "loss": 0.1164,
+      "step": 879
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.1329738227196196e-06,
+      "loss": 0.122,
+      "step": 880
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 2.062710602466411e-06,
+      "loss": 0.1225,
+      "step": 881
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.993612073002493e-06,
+      "loss": 0.1152,
+      "step": 882
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.9256790560620798e-06,
+      "loss": 0.116,
+      "step": 883
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.8589123595188718e-06,
+      "loss": 0.1332,
+      "step": 884
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.7933127773764214e-06,
+      "loss": 0.1254,
+      "step": 885
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.728881089758727e-06,
+      "loss": 0.1182,
+      "step": 886
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.6656180629009088e-06,
+      "loss": 0.1335,
+      "step": 887
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.6035244491401702e-06,
+      "loss": 0.119,
+      "step": 888
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.5426009869067948e-06,
+      "loss": 0.1227,
+      "step": 889
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.4828484007153753e-06,
+      "loss": 0.1295,
+      "step": 890
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.4242674011561875e-06,
+      "loss": 0.1073,
+      "step": 891
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.3668586848867737e-06,
+      "loss": 0.1244,
+      "step": 892
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.3106229346236398e-06,
+      "loss": 0.1261,
+      "step": 893
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.255560819134105e-06,
+      "loss": 0.1293,
+      "step": 894
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.2016729932284198e-06,
+      "loss": 0.122,
+      "step": 895
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.1489600977518833e-06,
+      "loss": 0.1316,
+      "step": 896
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.0974227595773378e-06,
+      "loss": 0.1292,
+      "step": 897
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.0470615915975978e-06,
+      "loss": 0.1167,
+      "step": 898
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 9.97877192718255e-07,
+      "loss": 0.1186,
+      "step": 899
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 9.498701478504845e-07,
+      "loss": 0.1239,
+      "step": 900
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 9.030410279041279e-07,
+      "loss": 0.1213,
+      "step": 901
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 8.573903897809099e-07,
+      "loss": 0.1263,
+      "step": 902
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 8.129187763678103e-07,
+      "loss": 0.1319,
+      "step": 903
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.696267165305693e-07,
+      "loss": 0.1263,
+      "step": 904
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.275147251074366e-07,
+      "loss": 0.12,
+      "step": 905
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 6.865833029030543e-07,
+      "loss": 0.1254,
+      "step": 906
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 6.468329366824955e-07,
+      "loss": 0.1478,
+      "step": 907
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.082640991654343e-07,
+      "loss": 0.1222,
+      "step": 908
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 5.708772490205849e-07,
+      "loss": 0.1241,
+      "step": 909
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 5.346728308602056e-07,
+      "loss": 0.1255,
+      "step": 910
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 4.99651275234847e-07,
+      "loss": 0.1216,
+      "step": 911
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 4.6581299862817894e-07,
+      "loss": 0.1223,
+      "step": 912
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 4.33158403452083e-07,
+      "loss": 0.1347,
+      "step": 913
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 4.016878780418787e-07,
+      "loss": 0.1196,
+      "step": 914
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.71401796651627e-07,
+      "loss": 0.1098,
+      "step": 915
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.423005194497786e-07,
+      "loss": 0.1406,
+      "step": 916
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.1438439251481046e-07,
+      "loss": 0.1219,
+      "step": 917
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.876537478311847e-07,
+      "loss": 0.1094,
+      "step": 918
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.621089032853075e-07,
+      "loss": 0.1261,
+      "step": 919
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 2.377501626618428e-07,
+      "loss": 0.1271,
+      "step": 920
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 2.1457781564004909e-07,
+      "loss": 0.1261,
+      "step": 921
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.9259213779033725e-07,
+      "loss": 0.1165,
+      "step": 922
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.7179339057098453e-07,
+      "loss": 0.1322,
+      "step": 923
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.521818213250703e-07,
+      "loss": 0.1276,
+      "step": 924
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.337576632774784e-07,
+      "loss": 0.1207,
+      "step": 925
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.1652113553217714e-07,
+      "loss": 0.1265,
+      "step": 926
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.0047244306955473e-07,
+      "loss": 0.1208,
+      "step": 927
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 8.561177674403232e-08,
+      "loss": 0.1207,
+      "step": 928
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.193931328177695e-08,
+      "loss": 0.1234,
+      "step": 929
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 5.9455215278581e-08,
+      "loss": 0.1291,
+      "step": 930
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 4.815963119796374e-08,
+      "loss": 0.125,
+      "step": 931
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.805269536939493e-08,
+      "loss": 0.1228,
+      "step": 932
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.9134527986651728e-08,
+      "loss": 0.144,
+      "step": 933
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.1405235106453093e-08,
+      "loss": 0.1341,
+      "step": 934
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.4864908647194143e-08,
+      "loss": 0.1113,
+      "step": 935
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 9.513626387780417e-09,
+      "loss": 0.124,
+      "step": 936
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 5.351451966806309e-09,
+      "loss": 0.1142,
+      "step": 937
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.3784348817113e-09,
+      "loss": 0.1143,
+      "step": 938
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 5.946104882359471e-10,
+      "loss": 0.1341,
+      "step": 939
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0,
+      "loss": 0.1388,
+      "step": 940
+    },
+    {
+      "epoch": 5.0,
+      "step": 940,
+      "total_flos": 8329575720960.0,
+      "train_loss": 0.5075884050432038,
+      "train_runtime": 2014.6169,
+      "train_samples_per_second": 7.443,
+      "train_steps_per_second": 0.467
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 940,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 50000,
+  "total_flos": 8329575720960.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}