diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.12195121951219512,
+  "epoch": 20.0,
   "eval_steps": 500,
-  "global_step": 20,
+  "global_step": 2180,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,138 +11,13278 @@
     {
       "epoch": 0.01,
       "learning_rate": 0.0,
-      "loss": 1.1529,
+      "loss": 4.4528,
       "step": 1
     },
     {
-      "epoch": 0.01,
+      "epoch": 0.02,
       "learning_rate": 0.0,
-      "loss": 1.7234,
+      "loss": 3.4935,
       "step": 2
     },
     {
-      "epoch": 0.02,
+      "epoch": 0.03,
       "learning_rate": 0.0,
-      "loss": 1.1025,
+      "loss": 4.2635,
       "step": 3
     },
     {
-      "epoch": 0.02,
+      "epoch": 0.04,
       "learning_rate": 0.0,
-      "loss": 0.8796,
+      "loss": 4.7088,
       "step": 4
     },
     {
-      "epoch": 0.03,
+      "epoch": 0.05,
       "learning_rate": 0.0,
-      "loss": 0.9621,
+      "loss": 4.8945,
       "step": 5
     },
     {
-      "epoch": 0.04,
+      "epoch": 0.06,
       "learning_rate": 0.0001,
-      "loss": 1.2671,
+      "loss": 2.5235,
       "step": 6
     },
     {
-      "epoch": 0.04,
+      "epoch": 0.06,
       "learning_rate": 0.0002,
-      "loss": 0.6072,
+      "loss": 3.547,
       "step": 7
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 0.00018888888888888888,
-      "loss": 0.0546,
+      "epoch": 0.07,
+      "learning_rate": 0.00019990817263544538,
+      "loss": 2.391,
       "step": 8
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 0.00017777777777777779,
-      "loss": 0.0,
+      "epoch": 0.08,
+      "learning_rate": 0.00019981634527089072,
+      "loss": 6.2259,
       "step": 9
     },
     {
-      "epoch": 0.06,
-      "learning_rate": 0.0001666666666666667,
-      "loss": 0.0,
+      "epoch": 0.09,
+      "learning_rate": 0.00019972451790633612,
+      "loss": 1.4949,
       "step": 10
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 0.00015555555555555556,
-      "loss": 0.0,
+      "epoch": 0.1,
+      "learning_rate": 0.00019963269054178146,
+      "loss": 3.1107,
       "step": 11
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 0.00014444444444444444,
-      "loss": 0.0,
+      "epoch": 0.11,
+      "learning_rate": 0.00019954086317722683,
+      "loss": 3.1171,
       "step": 12
     },
     {
-      "epoch": 0.08,
-      "learning_rate": 0.00013333333333333334,
-      "loss": 0.0,
+      "epoch": 0.12,
+      "learning_rate": 0.00019944903581267217,
+      "loss": 5.1984,
       "step": 13
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 0.00012222222222222224,
-      "loss": 0.0,
+      "epoch": 0.13,
+      "learning_rate": 0.00019935720844811754,
+      "loss": 1.6979,
       "step": 14
     },
     {
-      "epoch": 0.09,
-      "learning_rate": 0.00011111111111111112,
-      "loss": 0.0,
+      "epoch": 0.14,
+      "learning_rate": 0.0001992653810835629,
+      "loss": 2.6234,
       "step": 15
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 0.0001,
-      "loss": 0.0,
+      "epoch": 0.15,
+      "learning_rate": 0.00019917355371900828,
+      "loss": 1.5695,
       "step": 16
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 8.888888888888889e-05,
-      "loss": 0.0,
+      "epoch": 0.16,
+      "learning_rate": 0.00019908172635445362,
+      "loss": 3.6877,
       "step": 17
     },
     {
-      "epoch": 0.11,
-      "learning_rate": 7.777777777777778e-05,
-      "loss": 0.0,
+      "epoch": 0.17,
+      "learning_rate": 0.000198989898989899,
+      "loss": 1.4101,
       "step": 18
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 0.0,
+      "epoch": 0.17,
+      "learning_rate": 0.00019889807162534436,
+      "loss": 4.173,
       "step": 19
     },
     {
-      "epoch": 0.12,
-      "learning_rate": 5.555555555555556e-05,
-      "loss": 0.0,
+      "epoch": 0.18,
+      "learning_rate": 0.00019880624426078973,
+      "loss": 1.2081,
       "step": 20
     },
     {
-      "epoch": 0.12,
-      "step": 20,
-      "total_flos": 825752514723840.0,
-      "train_loss": 0.38747100830078124,
-      "train_runtime": 30.2879,
-      "train_samples_per_second": 2.641,
-      "train_steps_per_second": 0.66
+      "epoch": 0.19,
+      "learning_rate": 0.0001987144168962351,
+      "loss": 3.698,
+      "step": 21
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019862258953168044,
+      "loss": 1.6489,
+      "step": 22
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0001985307621671258,
+      "loss": 2.89,
+      "step": 23
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019843893480257118,
+      "loss": 2.1044,
+      "step": 24
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019834710743801655,
+      "loss": 3.8467,
+      "step": 25
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001982552800734619,
+      "loss": 2.2487,
+      "step": 26
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019816345270890726,
+      "loss": 1.8136,
+      "step": 27
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019807162534435263,
+      "loss": 2.7894,
+      "step": 28
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000197979797979798,
+      "loss": 2.1635,
+      "step": 29
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019788797061524337,
+      "loss": 3.6555,
+      "step": 30
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001977961432506887,
+      "loss": 2.7806,
+      "step": 31
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019770431588613408,
+      "loss": 2.9857,
+      "step": 32
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019761248852157942,
+      "loss": 1.8047,
+      "step": 33
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019752066115702482,
+      "loss": 1.8327,
+      "step": 34
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019742883379247016,
+      "loss": 2.1286,
+      "step": 35
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019733700642791553,
+      "loss": 2.2817,
+      "step": 36
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019724517906336087,
+      "loss": 2.1979,
+      "step": 37
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019715335169880624,
+      "loss": 2.4925,
+      "step": 38
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001970615243342516,
+      "loss": 3.034,
+      "step": 39
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019696969696969698,
+      "loss": 3.0346,
+      "step": 40
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019687786960514235,
+      "loss": 2.3958,
+      "step": 41
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001967860422405877,
+      "loss": 3.2426,
+      "step": 42
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0001966942148760331,
+      "loss": 2.0532,
+      "step": 43
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019660238751147843,
+      "loss": 2.2306,
+      "step": 44
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001965105601469238,
+      "loss": 1.9101,
+      "step": 45
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019641873278236914,
+      "loss": 1.9577,
+      "step": 46
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001963269054178145,
+      "loss": 2.4992,
+      "step": 47
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019623507805325988,
+      "loss": 2.2513,
+      "step": 48
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019614325068870525,
+      "loss": 2.5588,
+      "step": 49
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019605142332415062,
+      "loss": 1.8592,
+      "step": 50
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019595959595959596,
+      "loss": 2.0871,
+      "step": 51
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019586776859504133,
+      "loss": 1.6297,
+      "step": 52
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001957759412304867,
+      "loss": 1.3555,
+      "step": 53
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019568411386593207,
+      "loss": 2.0615,
+      "step": 54
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001955922865013774,
+      "loss": 1.6861,
+      "step": 55
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019550045913682278,
+      "loss": 1.8923,
+      "step": 56
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019540863177226812,
+      "loss": 2.597,
+      "step": 57
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019531680440771352,
+      "loss": 2.5299,
+      "step": 58
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019522497704315886,
+      "loss": 2.1354,
+      "step": 59
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019513314967860423,
+      "loss": 1.9797,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001950413223140496,
+      "loss": 1.3097,
+      "step": 61
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019494949494949494,
+      "loss": 1.5033,
+      "step": 62
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00019485766758494034,
+      "loss": 2.4062,
+      "step": 63
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019476584022038568,
+      "loss": 2.8677,
+      "step": 64
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00019467401285583105,
+      "loss": 1.1687,
+      "step": 65
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001945821854912764,
+      "loss": 2.4832,
+      "step": 66
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001944903581267218,
+      "loss": 1.083,
+      "step": 67
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00019439853076216713,
+      "loss": 1.669,
+      "step": 68
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001943067033976125,
+      "loss": 1.8857,
+      "step": 69
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00019421487603305787,
+      "loss": 2.561,
+      "step": 70
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001941230486685032,
+      "loss": 1.9139,
+      "step": 71
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001940312213039486,
+      "loss": 2.133,
+      "step": 72
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019393939393939395,
+      "loss": 2.5081,
+      "step": 73
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00019384756657483932,
+      "loss": 2.9829,
+      "step": 74
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00019375573921028466,
+      "loss": 2.1406,
+      "step": 75
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00019366391184573003,
+      "loss": 1.6052,
+      "step": 76
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001935720844811754,
+      "loss": 1.586,
+      "step": 77
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00019348025711662077,
+      "loss": 1.7505,
+      "step": 78
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001933884297520661,
+      "loss": 2.547,
+      "step": 79
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00019329660238751148,
+      "loss": 4.5113,
+      "step": 80
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00019320477502295685,
+      "loss": 1.6663,
+      "step": 81
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00019311294765840222,
+      "loss": 3.168,
+      "step": 82
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001930211202938476,
+      "loss": 1.5986,
+      "step": 83
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00019292929292929293,
+      "loss": 2.3072,
+      "step": 84
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001928374655647383,
+      "loss": 1.5781,
+      "step": 85
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00019274563820018364,
+      "loss": 1.846,
+      "step": 86
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00019265381083562904,
+      "loss": 1.6553,
+      "step": 87
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00019256198347107438,
+      "loss": 1.7912,
+      "step": 88
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00019247015610651975,
+      "loss": 2.2191,
+      "step": 89
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00019237832874196512,
+      "loss": 1.8041,
+      "step": 90
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001922865013774105,
+      "loss": 1.587,
+      "step": 91
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00019219467401285586,
+      "loss": 2.5109,
+      "step": 92
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001921028466483012,
+      "loss": 2.5659,
+      "step": 93
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00019201101928374657,
+      "loss": 2.0447,
+      "step": 94
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00019191919191919191,
+      "loss": 1.594,
+      "step": 95
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001918273645546373,
+      "loss": 2.3942,
+      "step": 96
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00019173553719008265,
+      "loss": 1.2796,
+      "step": 97
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00019164370982552802,
+      "loss": 1.7205,
+      "step": 98
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00019155188246097336,
+      "loss": 2.9089,
+      "step": 99
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00019146005509641873,
+      "loss": 2.9295,
+      "step": 100
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001913682277318641,
+      "loss": 1.7956,
+      "step": 101
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00019127640036730947,
+      "loss": 1.6725,
+      "step": 102
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00019118457300275484,
+      "loss": 1.364,
+      "step": 103
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00019109274563820018,
+      "loss": 1.9176,
+      "step": 104
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00019100091827364555,
+      "loss": 2.3724,
+      "step": 105
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00019090909090909092,
+      "loss": 1.894,
+      "step": 106
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001908172635445363,
+      "loss": 1.7133,
+      "step": 107
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00019072543617998163,
+      "loss": 2.6717,
+      "step": 108
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000190633608815427,
+      "loss": 1.6688,
+      "step": 109
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.29878048780487804,
+      "eval_loss": 2.229534149169922,
+      "eval_runtime": 13.9417,
+      "eval_samples_per_second": 11.763,
+      "eval_steps_per_second": 1.506,
+      "step": 109
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00019054178145087237,
+      "loss": 1.2822,
+      "step": 110
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00019044995408631774,
+      "loss": 1.7568,
+      "step": 111
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00019035812672176308,
+      "loss": 1.1837,
+      "step": 112
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00019026629935720845,
+      "loss": 1.6861,
+      "step": 113
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00019017447199265382,
+      "loss": 1.957,
+      "step": 114
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001900826446280992,
+      "loss": 1.6983,
+      "step": 115
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00018999081726354456,
+      "loss": 1.4546,
+      "step": 116
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001898989898989899,
+      "loss": 1.9029,
+      "step": 117
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00018980716253443527,
+      "loss": 2.7485,
+      "step": 118
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00018971533516988061,
+      "loss": 2.0669,
+      "step": 119
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.000189623507805326,
+      "loss": 1.8441,
+      "step": 120
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00018953168044077135,
+      "loss": 0.8765,
+      "step": 121
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00018943985307621672,
+      "loss": 2.8991,
+      "step": 122
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001893480257116621,
+      "loss": 1.4626,
+      "step": 123
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00018925619834710743,
+      "loss": 2.5054,
+      "step": 124
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00018916437098255283,
+      "loss": 1.0539,
+      "step": 125
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00018907254361799817,
+      "loss": 1.1429,
+      "step": 126
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00018898071625344354,
+      "loss": 1.7878,
+      "step": 127
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00018888888888888888,
+      "loss": 1.4984,
+      "step": 128
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00018879706152433425,
+      "loss": 1.3146,
+      "step": 129
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00018870523415977962,
+      "loss": 1.7478,
+      "step": 130
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.000188613406795225,
+      "loss": 1.6455,
+      "step": 131
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00018852157943067033,
+      "loss": 1.3298,
+      "step": 132
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001884297520661157,
+      "loss": 2.1197,
+      "step": 133
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00018833792470156107,
+      "loss": 1.551,
+      "step": 134
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00018824609733700644,
+      "loss": 1.4791,
+      "step": 135
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001881542699724518,
+      "loss": 0.6671,
+      "step": 136
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018806244260789715,
+      "loss": 1.5634,
+      "step": 137
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018797061524334252,
+      "loss": 1.2732,
+      "step": 138
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001878787878787879,
+      "loss": 1.1909,
+      "step": 139
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018778696051423326,
+      "loss": 1.4005,
+      "step": 140
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001876951331496786,
+      "loss": 2.0868,
+      "step": 141
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018760330578512397,
+      "loss": 2.0016,
+      "step": 142
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018751147842056934,
+      "loss": 1.5659,
+      "step": 143
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001874196510560147,
+      "loss": 1.8462,
+      "step": 144
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00018732782369146008,
+      "loss": 1.8284,
+      "step": 145
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00018723599632690542,
+      "loss": 1.5153,
+      "step": 146
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001871441689623508,
+      "loss": 2.9084,
+      "step": 147
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00018705234159779614,
+      "loss": 1.5477,
+      "step": 148
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00018696051423324153,
+      "loss": 2.8646,
+      "step": 149
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00018686868686868687,
+      "loss": 1.5144,
+      "step": 150
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00018677685950413224,
+      "loss": 1.84,
+      "step": 151
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00018668503213957759,
+      "loss": 3.2192,
+      "step": 152
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00018659320477502295,
+      "loss": 3.936,
+      "step": 153
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00018650137741046832,
+      "loss": 1.7515,
+      "step": 154
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001864095500459137,
+      "loss": 3.2783,
+      "step": 155
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00018631772268135906,
+      "loss": 2.1479,
+      "step": 156
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001862258953168044,
+      "loss": 2.8695,
+      "step": 157
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00018613406795224977,
+      "loss": 1.6558,
+      "step": 158
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00018604224058769514,
+      "loss": 1.7248,
+      "step": 159
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001859504132231405,
+      "loss": 1.1563,
+      "step": 160
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00018585858585858586,
+      "loss": 1.5922,
+      "step": 161
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00018576675849403122,
+      "loss": 2.3494,
+      "step": 162
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001856749311294766,
+      "loss": 1.5131,
+      "step": 163
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00018558310376492196,
+      "loss": 2.1479,
+      "step": 164
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00018549127640036733,
+      "loss": 2.1028,
+      "step": 165
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00018539944903581267,
+      "loss": 1.7847,
+      "step": 166
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00018530762167125804,
+      "loss": 1.2428,
+      "step": 167
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001852157943067034,
+      "loss": 2.1484,
+      "step": 168
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00018512396694214878,
+      "loss": 1.1372,
+      "step": 169
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00018503213957759412,
+      "loss": 0.6819,
+      "step": 170
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001849403122130395,
+      "loss": 1.0891,
+      "step": 171
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00018484848484848484,
+      "loss": 0.7477,
+      "step": 172
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00018475665748393023,
+      "loss": 1.4889,
+      "step": 173
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00018466483011937558,
+      "loss": 1.6568,
+      "step": 174
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00018457300275482094,
+      "loss": 0.8507,
+      "step": 175
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00018448117539026631,
+      "loss": 1.2445,
+      "step": 176
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00018438934802571166,
+      "loss": 0.9695,
+      "step": 177
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00018429752066115705,
+      "loss": 3.3201,
+      "step": 178
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0001842056932966024,
+      "loss": 1.7003,
+      "step": 179
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00018411386593204776,
+      "loss": 1.4902,
+      "step": 180
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0001840220385674931,
+      "loss": 0.8284,
+      "step": 181
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00018393021120293848,
+      "loss": 1.4903,
+      "step": 182
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00018383838383838384,
+      "loss": 1.7325,
+      "step": 183
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00018374655647382921,
+      "loss": 0.9902,
+      "step": 184
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00018365472910927458,
+      "loss": 0.6308,
+      "step": 185
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00018356290174471993,
+      "loss": 1.9786,
+      "step": 186
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00018347107438016532,
+      "loss": 1.2777,
+      "step": 187
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00018347107438016532,
+      "loss": 1.648,
+      "step": 188
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00018337924701561066,
+      "loss": 0.4818,
+      "step": 189
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00018337924701561066,
+      "loss": 3.6985,
+      "step": 190
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00018328741965105603,
+      "loss": 2.1847,
+      "step": 191
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00018319559228650138,
+      "loss": 1.4776,
+      "step": 192
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00018310376492194675,
+      "loss": 2.1751,
+      "step": 193
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00018301193755739211,
+      "loss": 1.3619,
+      "step": 194
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00018292011019283748,
+      "loss": 1.1032,
+      "step": 195
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00018282828282828283,
+      "loss": 1.0621,
+      "step": 196
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0001827364554637282,
+      "loss": 1.7267,
+      "step": 197
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00018264462809917356,
+      "loss": 1.8748,
+      "step": 198
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00018255280073461893,
+      "loss": 1.4783,
+      "step": 199
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0001824609733700643,
+      "loss": 0.9338,
+      "step": 200
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00018236914600550965,
+      "loss": 1.132,
+      "step": 201
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00018227731864095501,
+      "loss": 2.4952,
+      "step": 202
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00018218549127640036,
+      "loss": 0.4302,
+      "step": 203
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00018209366391184575,
+      "loss": 0.4397,
+      "step": 204
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0001820018365472911,
+      "loss": 2.4572,
+      "step": 205
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00018191000918273646,
+      "loss": 2.2511,
+      "step": 206
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 1.2697,
+      "step": 207
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00018172635445362718,
+      "loss": 2.5505,
+      "step": 208
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00018163452708907255,
+      "loss": 1.2977,
+      "step": 209
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00018154269972451792,
+      "loss": 1.8401,
+      "step": 210
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00018145087235996328,
+      "loss": 1.7906,
+      "step": 211
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00018135904499540863,
+      "loss": 2.4357,
+      "step": 212
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00018126721763085402,
+      "loss": 2.0001,
+      "step": 213
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00018117539026629937,
+      "loss": 2.5161,
+      "step": 214
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00018108356290174473,
+      "loss": 1.5329,
+      "step": 215
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00018099173553719008,
+      "loss": 1.408,
+      "step": 216
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00018089990817263545,
+      "loss": 1.4189,
+      "step": 217
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00018080808080808082,
+      "loss": 1.044,
+      "step": 218
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.3353658536585366,
+      "eval_loss": 2.1567869186401367,
+      "eval_runtime": 13.942,
+      "eval_samples_per_second": 11.763,
+      "eval_steps_per_second": 1.506,
+      "step": 218
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00018071625344352618,
+      "loss": 2.1143,
+      "step": 219
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00018062442607897155,
+      "loss": 1.235,
+      "step": 220
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0001805325987144169,
+      "loss": 2.2235,
+      "step": 221
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00018044077134986227,
+      "loss": 0.8353,
+      "step": 222
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00018034894398530764,
+      "loss": 0.523,
+      "step": 223
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.000180257116620753,
+      "loss": 1.9252,
+      "step": 224
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00018016528925619835,
+      "loss": 1.3179,
+      "step": 225
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00018007346189164372,
+      "loss": 1.4736,
+      "step": 226
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00017998163452708906,
+      "loss": 0.9673,
+      "step": 227
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00017988980716253445,
+      "loss": 0.9521,
+      "step": 228
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0001797979797979798,
+      "loss": 1.2891,
+      "step": 229
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00017970615243342517,
+      "loss": 1.2211,
+      "step": 230
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00017961432506887054,
+      "loss": 0.2571,
+      "step": 231
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00017952249770431588,
+      "loss": 0.543,
+      "step": 232
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00017943067033976127,
+      "loss": 0.8909,
+      "step": 233
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00017933884297520662,
+      "loss": 2.6065,
+      "step": 234
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00017924701561065199,
+      "loss": 0.1199,
+      "step": 235
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00017915518824609733,
+      "loss": 0.5213,
+      "step": 236
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00017906336088154272,
+      "loss": 2.405,
+      "step": 237
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00017897153351698807,
+      "loss": 0.6277,
+      "step": 238
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00017887970615243344,
+      "loss": 0.5324,
+      "step": 239
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0001787878787878788,
+      "loss": 1.1505,
+      "step": 240
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00017869605142332415,
+      "loss": 1.0073,
+      "step": 241
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00017860422405876954,
+      "loss": 1.0481,
+      "step": 242
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.00017851239669421489,
+      "loss": 1.7732,
+      "step": 243
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00017842056932966026,
+      "loss": 0.2245,
+      "step": 244
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0001783287419651056,
+      "loss": 1.2774,
+      "step": 245
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00017823691460055097,
+      "loss": 0.3739,
+      "step": 246
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00017814508723599634,
+      "loss": 0.5907,
+      "step": 247
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0001780532598714417,
+      "loss": 1.2472,
+      "step": 248
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00017796143250688705,
+      "loss": 0.9016,
+      "step": 249
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00017786960514233242,
+      "loss": 1.5772,
+      "step": 250
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00017777777777777779,
+      "loss": 2.0311,
+      "step": 251
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00017768595041322316,
+      "loss": 0.3377,
+      "step": 252
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00017759412304866852,
+      "loss": 0.6826,
+      "step": 253
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00017750229568411387,
+      "loss": 0.9807,
+      "step": 254
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00017741046831955924,
+      "loss": 0.741,
+      "step": 255
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00017731864095500458,
+      "loss": 0.4986,
+      "step": 256
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00017722681359044998,
+      "loss": 0.0701,
+      "step": 257
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00017713498622589532,
+      "loss": 0.0226,
+      "step": 258
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0001770431588613407,
+      "loss": 0.2058,
+      "step": 259
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00017695133149678606,
+      "loss": 0.17,
+      "step": 260
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00017685950413223143,
+      "loss": 0.7985,
+      "step": 261
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0001767676767676768,
+      "loss": 3.3993,
+      "step": 262
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00017667584940312214,
+      "loss": 1.2827,
+      "step": 263
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0001765840220385675,
+      "loss": 1.7463,
+      "step": 264
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00017649219467401285,
+      "loss": 1.1765,
+      "step": 265
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00017640036730945824,
+      "loss": 1.0938,
+      "step": 266
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0001763085399449036,
+      "loss": 0.7418,
+      "step": 267
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00017621671258034896,
+      "loss": 1.2793,
+      "step": 268
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0001761248852157943,
+      "loss": 2.1036,
+      "step": 269
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00017603305785123967,
+      "loss": 2.1484,
+      "step": 270
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00017594123048668504,
+      "loss": 2.6003,
+      "step": 271
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0001758494031221304,
+      "loss": 1.2115,
+      "step": 272
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00017575757575757578,
+      "loss": 1.5348,
+      "step": 273
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00017566574839302112,
+      "loss": 0.7968,
+      "step": 274
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0001755739210284665,
+      "loss": 1.3583,
+      "step": 275
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00017548209366391186,
+      "loss": 1.2731,
+      "step": 276
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00017539026629935723,
+      "loss": 0.4426,
+      "step": 277
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00017529843893480257,
+      "loss": 0.7437,
+      "step": 278
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00017520661157024794,
+      "loss": 0.4913,
+      "step": 279
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0001751147842056933,
+      "loss": 1.2527,
+      "step": 280
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00017502295684113868,
+      "loss": 1.1962,
+      "step": 281
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00017493112947658405,
+      "loss": 0.6869,
+      "step": 282
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0001748393021120294,
+      "loss": 1.3276,
+      "step": 283
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00017474747474747476,
+      "loss": 0.1047,
+      "step": 284
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00017465564738292013,
+      "loss": 0.9093,
+      "step": 285
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0001745638200183655,
+      "loss": 1.7248,
+      "step": 286
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00017447199265381084,
+      "loss": 0.5996,
+      "step": 287
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0001743801652892562,
+      "loss": 0.9297,
+      "step": 288
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00017428833792470155,
+      "loss": 0.2835,
+      "step": 289
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00017419651056014695,
+      "loss": 0.0966,
+      "step": 290
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0001741046831955923,
+      "loss": 0.7369,
+      "step": 291
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00017401285583103766,
+      "loss": 0.4588,
+      "step": 292
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00017392102846648303,
+      "loss": 2.184,
+      "step": 293
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00017382920110192837,
+      "loss": 0.2157,
+      "step": 294
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00017373737373737377,
+      "loss": 0.0372,
+      "step": 295
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0001736455463728191,
+      "loss": 1.6605,
+      "step": 296
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00017355371900826448,
+      "loss": 0.7851,
+      "step": 297
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00017346189164370982,
+      "loss": 0.8381,
+      "step": 298
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0001733700642791552,
+      "loss": 0.9296,
+      "step": 299
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00017327823691460056,
+      "loss": 1.0241,
+      "step": 300
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00017318640955004593,
+      "loss": 0.8458,
+      "step": 301
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0001730945821854913,
+      "loss": 0.3009,
+      "step": 302
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00017300275482093664,
+      "loss": 0.4724,
+      "step": 303
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.000172910927456382,
+      "loss": 0.0897,
+      "step": 304
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00017281910009182738,
+      "loss": 1.4927,
+      "step": 305
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00017272727272727275,
+      "loss": 4.2679,
+      "step": 306
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.0001726354453627181,
+      "loss": 0.8545,
+      "step": 307
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00017254361799816346,
+      "loss": 0.2054,
+      "step": 308
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00017245179063360883,
+      "loss": 1.3975,
+      "step": 309
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.0001723599632690542,
+      "loss": 2.4307,
+      "step": 310
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00017226813590449954,
+      "loss": 1.594,
+      "step": 311
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.0001721763085399449,
+      "loss": 0.0803,
+      "step": 312
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00017208448117539028,
+      "loss": 0.6297,
+      "step": 313
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00017199265381083565,
+      "loss": 1.3187,
+      "step": 314
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00017190082644628102,
+      "loss": 1.0996,
+      "step": 315
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00017180899908172636,
+      "loss": 1.5042,
+      "step": 316
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00017171717171717173,
+      "loss": 0.0772,
+      "step": 317
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00017162534435261707,
+      "loss": 0.9539,
+      "step": 318
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00017153351698806247,
+      "loss": 1.3146,
+      "step": 319
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.0001714416896235078,
+      "loss": 0.9767,
+      "step": 320
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00017134986225895318,
+      "loss": 0.4927,
+      "step": 321
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00017125803489439852,
+      "loss": 1.4654,
+      "step": 322
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.0001711662075298439,
+      "loss": 1.5264,
+      "step": 323
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00017107438016528926,
+      "loss": 1.1878,
+      "step": 324
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00017098255280073463,
+      "loss": 0.9639,
+      "step": 325
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00017089072543618,
+      "loss": 0.6191,
+      "step": 326
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00017079889807162534,
+      "loss": 0.3249,
+      "step": 327
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.7195121951219512,
+      "eval_loss": 0.797982931137085,
+      "eval_runtime": 13.9325,
+      "eval_samples_per_second": 11.771,
+      "eval_steps_per_second": 1.507,
+      "step": 327
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.0001707070707070707,
+      "loss": 0.119,
+      "step": 328
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00017061524334251608,
+      "loss": 0.1891,
+      "step": 329
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00017052341597796145,
+      "loss": 0.0806,
+      "step": 330
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.0001704315886134068,
+      "loss": 0.002,
+      "step": 331
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00017033976124885216,
+      "loss": 0.6022,
+      "step": 332
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00017024793388429753,
+      "loss": 0.1378,
+      "step": 333
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.0001701561065197429,
+      "loss": 0.202,
+      "step": 334
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00017006427915518827,
+      "loss": 0.3606,
+      "step": 335
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.0001699724517906336,
+      "loss": 0.376,
+      "step": 336
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00016988062442607898,
+      "loss": 1.4451,
+      "step": 337
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00016978879706152435,
+      "loss": 0.3627,
+      "step": 338
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00016969696969696972,
+      "loss": 0.2575,
+      "step": 339
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00016960514233241506,
+      "loss": 0.1301,
+      "step": 340
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00016951331496786043,
+      "loss": 1.6799,
+      "step": 341
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00016942148760330577,
+      "loss": 1.7604,
+      "step": 342
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00016932966023875117,
+      "loss": 0.0693,
+      "step": 343
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.0001692378328741965,
+      "loss": 0.1237,
+      "step": 344
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00016914600550964188,
+      "loss": 0.1062,
+      "step": 345
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00016905417814508725,
+      "loss": 0.0324,
+      "step": 346
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.0001689623507805326,
+      "loss": 0.0141,
+      "step": 347
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.000168870523415978,
+      "loss": 0.0107,
+      "step": 348
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00016877869605142333,
+      "loss": 0.1254,
+      "step": 349
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.0001686868686868687,
+      "loss": 0.0089,
+      "step": 350
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00016859504132231404,
+      "loss": 0.0041,
+      "step": 351
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0001685032139577594,
+      "loss": 0.5326,
+      "step": 352
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00016841138659320478,
+      "loss": 0.2984,
+      "step": 353
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00016831955922865015,
+      "loss": 0.2489,
+      "step": 354
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00016822773186409552,
+      "loss": 0.3279,
+      "step": 355
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00016813590449954086,
+      "loss": 1.0218,
+      "step": 356
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00016804407713498626,
+      "loss": 0.139,
+      "step": 357
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0001679522497704316,
+      "loss": 1.3741,
+      "step": 358
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00016786042240587697,
+      "loss": 0.5736,
+      "step": 359
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.0001677685950413223,
+      "loss": 0.2553,
+      "step": 360
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00016767676767676768,
+      "loss": 3.9418,
+      "step": 361
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00016758494031221305,
+      "loss": 0.0009,
+      "step": 362
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00016749311294765842,
+      "loss": 0.4919,
+      "step": 363
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00016740128558310376,
+      "loss": 1.4435,
+      "step": 364
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.00016730945821854913,
+      "loss": 0.6161,
+      "step": 365
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.0001672176308539945,
+      "loss": 0.7265,
+      "step": 366
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00016712580348943987,
+      "loss": 0.0839,
+      "step": 367
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00016703397612488524,
+      "loss": 0.0882,
+      "step": 368
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00016694214876033058,
+      "loss": 0.8354,
+      "step": 369
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00016685032139577595,
+      "loss": 0.0249,
+      "step": 370
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.0001667584940312213,
+      "loss": 0.9113,
+      "step": 371
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 0.3306,
+      "step": 372
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00016657483930211203,
+      "loss": 0.3899,
+      "step": 373
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.0001664830119375574,
+      "loss": 0.0084,
+      "step": 374
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00016639118457300277,
+      "loss": 0.0104,
+      "step": 375
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00016629935720844814,
+      "loss": 0.2984,
+      "step": 376
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.0001662075298438935,
+      "loss": 0.7124,
+      "step": 377
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00016611570247933885,
+      "loss": 0.7344,
+      "step": 378
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00016602387511478422,
+      "loss": 2.2644,
+      "step": 379
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00016593204775022956,
+      "loss": 0.0097,
+      "step": 380
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00016584022038567496,
+      "loss": 0.0001,
+      "step": 381
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.0001657483930211203,
+      "loss": 1.6672,
+      "step": 382
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00016565656565656567,
+      "loss": 0.5593,
+      "step": 383
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.000165564738292011,
+      "loss": 0.0007,
+      "step": 384
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00016547291092745638,
+      "loss": 0.001,
+      "step": 385
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00016538108356290175,
+      "loss": 0.8986,
+      "step": 386
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.00016528925619834712,
+      "loss": 0.7435,
+      "step": 387
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.0001651974288337925,
+      "loss": 0.0159,
+      "step": 388
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00016510560146923783,
+      "loss": 0.026,
+      "step": 389
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.0001650137741046832,
+      "loss": 0.0052,
+      "step": 390
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00016492194674012857,
+      "loss": 1.6697,
+      "step": 391
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00016483011937557394,
+      "loss": 3.5634,
+      "step": 392
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00016473829201101928,
+      "loss": 0.011,
+      "step": 393
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00016464646464646465,
+      "loss": 0.0614,
+      "step": 394
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00016455463728191002,
+      "loss": 0.0016,
+      "step": 395
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.0001644628099173554,
+      "loss": 0.0438,
+      "step": 396
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00016437098255280076,
+      "loss": 0.0009,
+      "step": 397
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.0001642791551882461,
+      "loss": 0.6832,
+      "step": 398
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00016418732782369147,
+      "loss": 0.8306,
+      "step": 399
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00016409550045913684,
+      "loss": 0.8695,
+      "step": 400
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.0001640036730945822,
+      "loss": 0.0572,
+      "step": 401
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00016391184573002755,
+      "loss": 0.1185,
+      "step": 402
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00016382001836547292,
+      "loss": 0.011,
+      "step": 403
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00016372819100091826,
+      "loss": 0.6307,
+      "step": 404
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 0.8132,
+      "step": 405
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.000163544536271809,
+      "loss": 0.2316,
+      "step": 406
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00016345270890725437,
+      "loss": 0.2169,
+      "step": 407
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00016336088154269974,
+      "loss": 0.0111,
+      "step": 408
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00016326905417814508,
+      "loss": 1.446,
+      "step": 409
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00016317722681359048,
+      "loss": 0.0002,
+      "step": 410
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00016308539944903582,
+      "loss": 2.1242,
+      "step": 411
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.0001629935720844812,
+      "loss": 2.9539,
+      "step": 412
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00016290174471992653,
+      "loss": 3.8165,
+      "step": 413
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.0001628099173553719,
+      "loss": 2.3232,
+      "step": 414
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00016271808999081727,
+      "loss": 1.8742,
+      "step": 415
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00016262626262626264,
+      "loss": 2.6036,
+      "step": 416
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00016253443526170798,
+      "loss": 2.7972,
+      "step": 417
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00016244260789715335,
+      "loss": 2.6851,
+      "step": 418
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.00016235078053259872,
+      "loss": 1.8555,
+      "step": 419
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.0001622589531680441,
+      "loss": 2.5249,
+      "step": 420
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.00016216712580348946,
+      "loss": 2.0005,
+      "step": 421
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.0001620752984389348,
+      "loss": 2.6406,
+      "step": 422
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00016198347107438017,
+      "loss": 3.7138,
+      "step": 423
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00016189164370982554,
+      "loss": 2.0045,
+      "step": 424
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.0001617998163452709,
+      "loss": 3.0576,
+      "step": 425
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00016170798898071625,
+      "loss": 3.2314,
+      "step": 426
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00016161616161616162,
+      "loss": 2.6701,
+      "step": 427
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.000161524334251607,
+      "loss": 2.3032,
+      "step": 428
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00016143250688705236,
+      "loss": 2.3721,
+      "step": 429
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00016134067952249773,
+      "loss": 2.0273,
+      "step": 430
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00016124885215794307,
+      "loss": 2.7288,
+      "step": 431
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00016115702479338844,
+      "loss": 2.0249,
+      "step": 432
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00016106519742883378,
+      "loss": 1.9295,
+      "step": 433
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00016097337006427918,
+      "loss": 1.9697,
+      "step": 434
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00016088154269972452,
+      "loss": 1.6958,
+      "step": 435
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.0001607897153351699,
+      "loss": 1.3202,
+      "step": 436
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.17682926829268292,
+      "eval_loss": 2.2678847312927246,
+      "eval_runtime": 13.9111,
+      "eval_samples_per_second": 11.789,
+      "eval_steps_per_second": 1.51,
+      "step": 436
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.00016069788797061523,
+      "loss": 2.0627,
+      "step": 437
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.0001606060606060606,
+      "loss": 3.5309,
+      "step": 438
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00016051423324150597,
+      "loss": 2.1681,
+      "step": 439
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.00016042240587695134,
+      "loss": 2.6779,
+      "step": 440
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.0001603305785123967,
+      "loss": 1.8395,
+      "step": 441
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00016023875114784205,
+      "loss": 2.5511,
+      "step": 442
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00016014692378328742,
+      "loss": 2.4688,
+      "step": 443
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.0001600550964187328,
+      "loss": 2.3184,
+      "step": 444
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 0.00015996326905417816,
+      "loss": 1.9437,
+      "step": 445
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.0001598714416896235,
+      "loss": 1.3691,
+      "step": 446
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.00015977961432506887,
+      "loss": 2.7536,
+      "step": 447
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00015968778696051424,
+      "loss": 2.1437,
+      "step": 448
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 0.0001595959595959596,
+      "loss": 2.1953,
+      "step": 449
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00015950413223140498,
+      "loss": 2.3601,
+      "step": 450
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.00015941230486685032,
+      "loss": 1.7018,
+      "step": 451
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 0.0001593204775022957,
+      "loss": 1.534,
+      "step": 452
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.00015922865013774106,
+      "loss": 1.1632,
+      "step": 453
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00015913682277318643,
+      "loss": 0.3199,
+      "step": 454
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00015904499540863177,
+      "loss": 2.0837,
+      "step": 455
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.00015895316804407714,
+      "loss": 0.8584,
+      "step": 456
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.00015886134067952248,
+      "loss": 0.5752,
+      "step": 457
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 0.00015876951331496788,
+      "loss": 0.0831,
+      "step": 458
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.00015867768595041322,
+      "loss": 0.0017,
+      "step": 459
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 0.0001585858585858586,
+      "loss": 0.0099,
+      "step": 460
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.00015849403122130396,
+      "loss": 0.0073,
+      "step": 461
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.0001584022038567493,
+      "loss": 1.7517,
+      "step": 462
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 0.0001583103764921947,
+      "loss": 0.4021,
+      "step": 463
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00015821854912764004,
+      "loss": 0.5455,
+      "step": 464
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.0001581267217630854,
+      "loss": 0.0007,
+      "step": 465
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00015803489439853075,
+      "loss": 0.0105,
+      "step": 466
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00015794306703397612,
+      "loss": 1.0384,
+      "step": 467
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.0001578512396694215,
+      "loss": 0.0301,
+      "step": 468
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 0.00015775941230486686,
+      "loss": 0.1302,
+      "step": 469
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00015766758494031223,
+      "loss": 0.0822,
+      "step": 470
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 0.00015757575757575757,
+      "loss": 0.8087,
+      "step": 471
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.00015748393021120297,
+      "loss": 2.0351,
+      "step": 472
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.0001573921028466483,
+      "loss": 0.0215,
+      "step": 473
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.00015730027548209368,
+      "loss": 0.0335,
+      "step": 474
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 0.00015720844811753902,
+      "loss": 0.1199,
+      "step": 475
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.0001571166207529844,
+      "loss": 0.0553,
+      "step": 476
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 0.00015702479338842976,
+      "loss": 0.0092,
+      "step": 477
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00015693296602387513,
+      "loss": 0.474,
+      "step": 478
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00015684113865932047,
+      "loss": 2.1964,
+      "step": 479
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.00015674931129476584,
+      "loss": 1.224,
+      "step": 480
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.0001566574839302112,
+      "loss": 0.9762,
+      "step": 481
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.00015656565656565658,
+      "loss": 0.0182,
+      "step": 482
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.00015647382920110195,
+      "loss": 0.0007,
+      "step": 483
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 0.0001563820018365473,
+      "loss": 1.0608,
+      "step": 484
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00015629017447199266,
+      "loss": 0.5263,
+      "step": 485
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 0.000156198347107438,
+      "loss": 0.5466,
+      "step": 486
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.0001561065197428834,
+      "loss": 0.3039,
+      "step": 487
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 0.00015601469237832874,
+      "loss": 0.5474,
+      "step": 488
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.0001559228650137741,
+      "loss": 0.0048,
+      "step": 489
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 0.00015583103764921948,
+      "loss": 0.3158,
+      "step": 490
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 0.00015573921028466482,
+      "loss": 2.2812,
+      "step": 491
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.00015564738292011022,
+      "loss": 1.8101,
+      "step": 492
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.00015555555555555556,
+      "loss": 3.4362,
+      "step": 493
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 0.00015546372819100093,
+      "loss": 2.2707,
+      "step": 494
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.00015537190082644627,
+      "loss": 1.967,
+      "step": 495
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 0.00015528007346189167,
+      "loss": 1.9928,
+      "step": 496
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.000155188246097337,
+      "loss": 2.7732,
+      "step": 497
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.00015509641873278238,
+      "loss": 3.6444,
+      "step": 498
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.00015500459136822772,
+      "loss": 3.2376,
+      "step": 499
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0001549127640036731,
+      "loss": 2.2956,
+      "step": 500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00015482093663911846,
+      "loss": 2.1408,
+      "step": 501
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.00015472910927456383,
+      "loss": 3.2886,
+      "step": 502
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.0001546372819100092,
+      "loss": 1.957,
+      "step": 503
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.00015454545454545454,
+      "loss": 2.2583,
+      "step": 504
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.00015445362718089991,
+      "loss": 1.6257,
+      "step": 505
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 0.00015436179981634528,
+      "loss": 1.2913,
+      "step": 506
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.00015426997245179065,
+      "loss": 2.321,
+      "step": 507
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.000154178145087236,
+      "loss": 2.4237,
+      "step": 508
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 0.00015408631772268136,
+      "loss": 2.063,
+      "step": 509
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.00015399449035812673,
+      "loss": 1.9174,
+      "step": 510
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.0001539026629935721,
+      "loss": 2.3132,
+      "step": 511
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.00015381083562901744,
+      "loss": 1.9715,
+      "step": 512
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 0.00015371900826446281,
+      "loss": 1.8311,
+      "step": 513
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00015362718089990818,
+      "loss": 2.1849,
+      "step": 514
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00015353535353535353,
+      "loss": 2.0573,
+      "step": 515
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 0.00015344352617079892,
+      "loss": 2.5415,
+      "step": 516
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.00015335169880624426,
+      "loss": 1.9154,
+      "step": 517
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.00015325987144168963,
+      "loss": 1.7083,
+      "step": 518
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 0.00015316804407713498,
+      "loss": 2.1989,
+      "step": 519
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00015307621671258037,
+      "loss": 1.5602,
+      "step": 520
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 0.00015298438934802571,
+      "loss": 2.3966,
+      "step": 521
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 0.00015289256198347108,
+      "loss": 1.4966,
+      "step": 522
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00015280073461891645,
+      "loss": 1.9301,
+      "step": 523
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 0.0001527089072543618,
+      "loss": 1.842,
+      "step": 524
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.0001526170798898072,
+      "loss": 1.8986,
+      "step": 525
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.00015252525252525253,
+      "loss": 3.696,
+      "step": 526
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.0001524334251606979,
+      "loss": 2.1863,
+      "step": 527
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.00015234159779614325,
+      "loss": 2.1146,
+      "step": 528
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 0.00015224977043158861,
+      "loss": 2.3613,
+      "step": 529
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00015215794306703398,
+      "loss": 3.5017,
+      "step": 530
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.00015206611570247935,
+      "loss": 1.8296,
+      "step": 531
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 0.0001519742883379247,
+      "loss": 1.4259,
+      "step": 532
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 0.00015188246097337006,
+      "loss": 1.7603,
+      "step": 533
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.00015179063360881543,
+      "loss": 2.0712,
+      "step": 534
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 0.0001516988062442608,
+      "loss": 1.7131,
+      "step": 535
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00015160697887970617,
+      "loss": 1.6436,
+      "step": 536
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 0.0898,
+      "step": 537
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.00015142332415059688,
+      "loss": 0.0005,
+      "step": 538
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.00015133149678604223,
+      "loss": 1.6559,
+      "step": 539
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00015123966942148762,
+      "loss": 0.0107,
+      "step": 540
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00015114784205693297,
+      "loss": 0.0001,
+      "step": 541
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 0.00015105601469237833,
+      "loss": 0.4699,
+      "step": 542
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.0001509641873278237,
+      "loss": 0.0016,
+      "step": 543
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.00015087235996326907,
+      "loss": 0.0092,
+      "step": 544
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.00015078053259871444,
+      "loss": 1.325,
+      "step": 545
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.823170731707317,
+      "eval_loss": 0.9486518502235413,
+      "eval_runtime": 13.9491,
+      "eval_samples_per_second": 11.757,
+      "eval_steps_per_second": 1.505,
+      "step": 545
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00015068870523415978,
+      "loss": 0.0007,
+      "step": 546
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.00015059687786960515,
+      "loss": 0.0023,
+      "step": 547
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.0001505050505050505,
+      "loss": 0.0,
+      "step": 548
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0001504132231404959,
+      "loss": 0.0245,
+      "step": 549
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.00015032139577594123,
+      "loss": 0.0082,
+      "step": 550
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.0001502295684113866,
+      "loss": 0.9097,
+      "step": 551
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.00015013774104683195,
+      "loss": 0.0088,
+      "step": 552
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00015004591368227732,
+      "loss": 0.0001,
+      "step": 553
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.00014995408631772269,
+      "loss": 0.0,
+      "step": 554
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.00014986225895316805,
+      "loss": 0.8129,
+      "step": 555
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.00014977043158861342,
+      "loss": 0.0001,
+      "step": 556
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.00014967860422405877,
+      "loss": 0.0001,
+      "step": 557
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.00014958677685950414,
+      "loss": 0.002,
+      "step": 558
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.0001494949494949495,
+      "loss": 0.1438,
+      "step": 559
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00014940312213039487,
+      "loss": 0.0017,
+      "step": 560
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.00014931129476584022,
+      "loss": 0.0002,
+      "step": 561
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00014921946740128559,
+      "loss": 0.008,
+      "step": 562
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00014912764003673095,
+      "loss": 0.0018,
+      "step": 563
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.00014903581267217632,
+      "loss": 0.2074,
+      "step": 564
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.0001489439853076217,
+      "loss": 0.0,
+      "step": 565
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.00014885215794306704,
+      "loss": 0.0007,
+      "step": 566
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.0001487603305785124,
+      "loss": 1.2486,
+      "step": 567
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.00014866850321395777,
+      "loss": 0.0019,
+      "step": 568
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.00014857667584940314,
+      "loss": 0.1593,
+      "step": 569
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00014848484848484849,
+      "loss": 0.0005,
+      "step": 570
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.00014839302112029386,
+      "loss": 0.0636,
+      "step": 571
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.0001483011937557392,
+      "loss": 0.0011,
+      "step": 572
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.0001482093663911846,
+      "loss": 1.0408,
+      "step": 573
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.00014811753902662994,
+      "loss": 1.8716,
+      "step": 574
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.0001480257116620753,
+      "loss": 0.7952,
+      "step": 575
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.00014793388429752067,
+      "loss": 3.4657,
+      "step": 576
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00014784205693296602,
+      "loss": 1.2134,
+      "step": 577
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.0001477502295684114,
+      "loss": 1.0284,
+      "step": 578
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00014765840220385676,
+      "loss": 2.3283,
+      "step": 579
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.00014765840220385676,
+      "loss": 1.7083,
+      "step": 580
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.00014756657483930212,
+      "loss": 0.9105,
+      "step": 581
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.00014747474747474747,
+      "loss": 3.5487,
+      "step": 582
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00014738292011019284,
+      "loss": 1.6785,
+      "step": 583
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.0001472910927456382,
+      "loss": 1.2918,
+      "step": 584
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.00014719926538108357,
+      "loss": 1.078,
+      "step": 585
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.00014710743801652894,
+      "loss": 1.0432,
+      "step": 586
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.0001470156106519743,
+      "loss": 1.1971,
+      "step": 587
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.00014692378328741966,
+      "loss": 1.1088,
+      "step": 588
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.00014683195592286503,
+      "loss": 0.0321,
+      "step": 589
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.0001467401285583104,
+      "loss": 0.7571,
+      "step": 590
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.00014664830119375574,
+      "loss": 0.0104,
+      "step": 591
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.0001465564738292011,
+      "loss": 0.0356,
+      "step": 592
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.00014646464646464648,
+      "loss": 0.0022,
+      "step": 593
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 0.00014637281910009184,
+      "loss": 0.0675,
+      "step": 594
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.0001462809917355372,
+      "loss": 0.0208,
+      "step": 595
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 0.00014618916437098256,
+      "loss": 0.008,
+      "step": 596
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00014609733700642793,
+      "loss": 0.0047,
+      "step": 597
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.0001460055096418733,
+      "loss": 0.5652,
+      "step": 598
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.00014591368227731866,
+      "loss": 0.14,
+      "step": 599
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.000145821854912764,
+      "loss": 0.0,
+      "step": 600
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.00014573002754820938,
+      "loss": 0.0002,
+      "step": 601
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.00014563820018365472,
+      "loss": 0.0003,
+      "step": 602
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.00014554637281910011,
+      "loss": 0.048,
+      "step": 603
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 2.1273,
+      "step": 604
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00014536271808999083,
+      "loss": 0.0002,
+      "step": 605
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.0001452708907254362,
+      "loss": 0.0,
+      "step": 606
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00014517906336088154,
+      "loss": 0.0,
+      "step": 607
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.0001450872359963269,
+      "loss": 0.0,
+      "step": 608
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00014499540863177228,
+      "loss": 0.0056,
+      "step": 609
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00014490358126721765,
+      "loss": 0.0041,
+      "step": 610
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.000144811753902663,
+      "loss": 0.527,
+      "step": 611
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00014471992653810836,
+      "loss": 0.0001,
+      "step": 612
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.00014462809917355373,
+      "loss": 0.0268,
+      "step": 613
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.0001445362718089991,
+      "loss": 0.014,
+      "step": 614
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00014444444444444444,
+      "loss": 0.2602,
+      "step": 615
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.0001443526170798898,
+      "loss": 0.0074,
+      "step": 616
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00014426078971533518,
+      "loss": 0.9447,
+      "step": 617
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.00014416896235078055,
+      "loss": 0.0,
+      "step": 618
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 0.00014407713498622592,
+      "loss": 0.0,
+      "step": 619
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.00014398530762167126,
+      "loss": 0.0001,
+      "step": 620
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00014389348025711663,
+      "loss": 0.0,
+      "step": 621
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.000143801652892562,
+      "loss": 0.0,
+      "step": 622
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00014370982552800737,
+      "loss": 0.0003,
+      "step": 623
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.0001436179981634527,
+      "loss": 0.0,
+      "step": 624
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.00014352617079889808,
+      "loss": 0.0013,
+      "step": 625
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00014343434343434342,
+      "loss": 2.1797,
+      "step": 626
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 0.00014334251606978882,
+      "loss": 2.0018,
+      "step": 627
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00014325068870523416,
+      "loss": 1.1224,
+      "step": 628
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.00014315886134067953,
+      "loss": 0.3622,
+      "step": 629
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.0001430670339761249,
+      "loss": 0.3867,
+      "step": 630
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00014297520661157024,
+      "loss": 0.0003,
+      "step": 631
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.00014288337924701563,
+      "loss": 0.4459,
+      "step": 632
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00014279155188246098,
+      "loss": 1.258,
+      "step": 633
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00014269972451790635,
+      "loss": 0.4443,
+      "step": 634
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.0001426078971533517,
+      "loss": 0.0004,
+      "step": 635
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00014251606978879706,
+      "loss": 0.0141,
+      "step": 636
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00014242424242424243,
+      "loss": 0.0043,
+      "step": 637
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 0.0001423324150596878,
+      "loss": 0.1493,
+      "step": 638
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.00014224058769513317,
+      "loss": 0.0011,
+      "step": 639
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.0001421487603305785,
+      "loss": 0.0017,
+      "step": 640
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.0001420569329660239,
+      "loss": 0.0027,
+      "step": 641
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 0.00014196510560146925,
+      "loss": 0.004,
+      "step": 642
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.00014187327823691462,
+      "loss": 0.0002,
+      "step": 643
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.00014178145087235996,
+      "loss": 0.0028,
+      "step": 644
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.00014168962350780533,
+      "loss": 0.0,
+      "step": 645
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.0001415977961432507,
+      "loss": 0.155,
+      "step": 646
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.00014150596877869607,
+      "loss": 0.0248,
+      "step": 647
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.0001414141414141414,
+      "loss": 0.0034,
+      "step": 648
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00014132231404958678,
+      "loss": 0.0012,
+      "step": 649
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.00014123048668503215,
+      "loss": 0.0,
+      "step": 650
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.00014113865932047752,
+      "loss": 0.0456,
+      "step": 651
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00014104683195592289,
+      "loss": 0.0008,
+      "step": 652
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.00014095500459136823,
+      "loss": 0.3352,
+      "step": 653
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.0001408631772268136,
+      "loss": 0.0001,
+      "step": 654
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8170731707317073,
+      "eval_loss": 1.3516559600830078,
+      "eval_runtime": 13.932,
+      "eval_samples_per_second": 11.771,
+      "eval_steps_per_second": 1.507,
+      "step": 654
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 0.00014077134986225894,
+      "loss": 0.0,
+      "step": 655
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00014067952249770434,
+      "loss": 0.0009,
+      "step": 656
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 0.00014058769513314968,
+      "loss": 0.0,
+      "step": 657
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00014049586776859505,
+      "loss": 1.9818,
+      "step": 658
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.00014040404040404042,
+      "loss": 0.0001,
+      "step": 659
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00014031221303948576,
+      "loss": 0.005,
+      "step": 660
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00014022038567493116,
+      "loss": 0.0001,
+      "step": 661
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.0001401285583103765,
+      "loss": 0.0,
+      "step": 662
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.00014003673094582187,
+      "loss": 0.0,
+      "step": 663
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.0001399449035812672,
+      "loss": 0.0,
+      "step": 664
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.0001398530762167126,
+      "loss": 0.0102,
+      "step": 665
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00013976124885215795,
+      "loss": 0.0,
+      "step": 666
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.00013966942148760332,
+      "loss": 0.0,
+      "step": 667
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.00013957759412304866,
+      "loss": 0.5508,
+      "step": 668
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00013948576675849403,
+      "loss": 0.0002,
+      "step": 669
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.0001393939393939394,
+      "loss": 0.0,
+      "step": 670
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.00013930211202938477,
+      "loss": 0.0022,
+      "step": 671
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00013921028466483014,
+      "loss": 0.4246,
+      "step": 672
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00013911845730027548,
+      "loss": 0.0086,
+      "step": 673
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00013902662993572085,
+      "loss": 0.0032,
+      "step": 674
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 0.00013893480257116622,
+      "loss": 0.0,
+      "step": 675
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.0001388429752066116,
+      "loss": 0.0002,
+      "step": 676
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00013875114784205693,
+      "loss": 0.4911,
+      "step": 677
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.0001386593204775023,
+      "loss": 0.0791,
+      "step": 678
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.00013856749311294767,
+      "loss": 0.001,
+      "step": 679
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00013847566574839304,
+      "loss": 0.0052,
+      "step": 680
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.0001383838383838384,
+      "loss": 0.0014,
+      "step": 681
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00013829201101928375,
+      "loss": 0.0001,
+      "step": 682
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.00013820018365472912,
+      "loss": 0.0,
+      "step": 683
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00013810835629017446,
+      "loss": 0.0007,
+      "step": 684
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00013801652892561986,
+      "loss": 0.0022,
+      "step": 685
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.0001379247015610652,
+      "loss": 0.0,
+      "step": 686
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00013783287419651057,
+      "loss": 0.0042,
+      "step": 687
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 0.0001377410468319559,
+      "loss": 0.0289,
+      "step": 688
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.0001376492194674013,
+      "loss": 0.1011,
+      "step": 689
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.00013755739210284665,
+      "loss": 0.004,
+      "step": 690
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00013746556473829202,
+      "loss": 0.1793,
+      "step": 691
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.0001373737373737374,
+      "loss": 0.0,
+      "step": 692
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.00013728191000918273,
+      "loss": 0.0,
+      "step": 693
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.00013719008264462813,
+      "loss": 0.0,
+      "step": 694
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00013709825528007347,
+      "loss": 0.0,
+      "step": 695
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00013700642791551884,
+      "loss": 0.0008,
+      "step": 696
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00013691460055096418,
+      "loss": 0.0117,
+      "step": 697
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00013682277318640955,
+      "loss": 0.0,
+      "step": 698
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.00013673094582185492,
+      "loss": 0.0,
+      "step": 699
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.0001366391184573003,
+      "loss": 0.0046,
+      "step": 700
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.00013654729109274566,
+      "loss": 0.0,
+      "step": 701
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.000136455463728191,
+      "loss": 0.1436,
+      "step": 702
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 0.0021,
+      "step": 703
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00013627180899908174,
+      "loss": 0.001,
+      "step": 704
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.0001361799816345271,
+      "loss": 0.0001,
+      "step": 705
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.00013608815426997245,
+      "loss": 1.1613,
+      "step": 706
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.00013599632690541782,
+      "loss": 0.5292,
+      "step": 707
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00013590449954086316,
+      "loss": 0.0,
+      "step": 708
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00013581267217630856,
+      "loss": 0.0,
+      "step": 709
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.0001357208448117539,
+      "loss": 0.0,
+      "step": 710
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.00013562901744719927,
+      "loss": 0.0,
+      "step": 711
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00013553719008264464,
+      "loss": 0.0006,
+      "step": 712
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 0.00013544536271809,
+      "loss": 2.0593,
+      "step": 713
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.00013535353535353538,
+      "loss": 0.0409,
+      "step": 714
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00013526170798898072,
+      "loss": 0.0373,
+      "step": 715
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.0001351698806244261,
+      "loss": 2.3203,
+      "step": 716
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00013507805325987143,
+      "loss": 0.0028,
+      "step": 717
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.00013498622589531683,
+      "loss": 0.0004,
+      "step": 718
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00013489439853076217,
+      "loss": 0.0135,
+      "step": 719
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00013480257116620754,
+      "loss": 0.0008,
+      "step": 720
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00013471074380165288,
+      "loss": 0.0001,
+      "step": 721
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00013461891643709825,
+      "loss": 0.4002,
+      "step": 722
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00013452708907254362,
+      "loss": 0.0,
+      "step": 723
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.000134435261707989,
+      "loss": 0.0031,
+      "step": 724
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.00013434343434343436,
+      "loss": 0.2869,
+      "step": 725
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.0001342516069788797,
+      "loss": 1.0244,
+      "step": 726
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.00013415977961432507,
+      "loss": 0.1245,
+      "step": 727
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00013406795224977044,
+      "loss": 0.6639,
+      "step": 728
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.0001339761248852158,
+      "loss": 0.0012,
+      "step": 729
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00013388429752066115,
+      "loss": 0.2594,
+      "step": 730
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00013379247015610652,
+      "loss": 0.5933,
+      "step": 731
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.0001337006427915519,
+      "loss": 0.0208,
+      "step": 732
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.00013360881542699726,
+      "loss": 0.0,
+      "step": 733
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.00013351698806244263,
+      "loss": 0.0,
+      "step": 734
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00013342516069788797,
+      "loss": 0.0011,
+      "step": 735
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.034,
+      "step": 736
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.0001332415059687787,
+      "loss": 0.0002,
+      "step": 737
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.00013314967860422408,
+      "loss": 0.0,
+      "step": 738
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.00013305785123966942,
+      "loss": 0.0024,
+      "step": 739
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.0001329660238751148,
+      "loss": 0.2918,
+      "step": 740
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.00013287419651056013,
+      "loss": 0.0003,
+      "step": 741
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.00013278236914600553,
+      "loss": 0.015,
+      "step": 742
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00013269054178145087,
+      "loss": 0.0,
+      "step": 743
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.00013259871441689624,
+      "loss": 0.0,
+      "step": 744
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.0001325068870523416,
+      "loss": 0.0,
+      "step": 745
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 0.00013241505968778695,
+      "loss": 0.3426,
+      "step": 746
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00013232323232323235,
+      "loss": 0.0,
+      "step": 747
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 0.0001322314049586777,
+      "loss": 0.0013,
+      "step": 748
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.00013213957759412306,
+      "loss": 0.0128,
+      "step": 749
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.0001320477502295684,
+      "loss": 0.0005,
+      "step": 750
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.00013195592286501377,
+      "loss": 0.0,
+      "step": 751
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00013186409550045914,
+      "loss": 0.0928,
+      "step": 752
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.0001317722681359045,
+      "loss": 0.0,
+      "step": 753
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00013168044077134988,
+      "loss": 0.0,
+      "step": 754
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 0.00013158861340679522,
+      "loss": 0.0,
+      "step": 755
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00013149678604224062,
+      "loss": 0.013,
+      "step": 756
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00013140495867768596,
+      "loss": 0.0,
+      "step": 757
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00013131313131313133,
+      "loss": 0.0,
+      "step": 758
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 0.00013122130394857667,
+      "loss": 0.0048,
+      "step": 759
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00013112947658402204,
+      "loss": 0.0,
+      "step": 760
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.0001310376492194674,
+      "loss": 0.3698,
+      "step": 761
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.00013094582185491278,
+      "loss": 0.0,
+      "step": 762
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00013085399449035812,
+      "loss": 1.8235,
+      "step": 763
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.7621951219512195,
+      "eval_loss": 1.5762226581573486,
+      "eval_runtime": 13.9363,
+      "eval_samples_per_second": 11.768,
+      "eval_steps_per_second": 1.507,
+      "step": 763
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.0001307621671258035,
+      "loss": 1.0492,
+      "step": 764
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 0.00013067033976124886,
+      "loss": 0.4626,
+      "step": 765
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.00013057851239669423,
+      "loss": 0.0307,
+      "step": 766
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.0001304866850321396,
+      "loss": 0.0,
+      "step": 767
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.00013039485766758494,
+      "loss": 0.0,
+      "step": 768
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.0001303030303030303,
+      "loss": 0.0,
+      "step": 769
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.00013021120293847565,
+      "loss": 0.0,
+      "step": 770
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00013011937557392105,
+      "loss": 0.0,
+      "step": 771
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.0001300275482093664,
+      "loss": 0.0,
+      "step": 772
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00012993572084481176,
+      "loss": 0.0,
+      "step": 773
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 0.00012984389348025713,
+      "loss": 0.0001,
+      "step": 774
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.00012975206611570247,
+      "loss": 0.0018,
+      "step": 775
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.00012966023875114787,
+      "loss": 0.4323,
+      "step": 776
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.0001295684113865932,
+      "loss": 0.0,
+      "step": 777
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 0.00012947658402203858,
+      "loss": 0.0,
+      "step": 778
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.00012938475665748392,
+      "loss": 0.0112,
+      "step": 779
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.00012929292929292932,
+      "loss": 0.0002,
+      "step": 780
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.00012920110192837466,
+      "loss": 0.0017,
+      "step": 781
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.00012910927456382003,
+      "loss": 0.009,
+      "step": 782
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00012901744719926537,
+      "loss": 0.3859,
+      "step": 783
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00012892561983471074,
+      "loss": 0.5658,
+      "step": 784
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 0.0001288337924701561,
+      "loss": 0.0039,
+      "step": 785
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.00012874196510560148,
+      "loss": 0.0,
+      "step": 786
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00012865013774104685,
+      "loss": 0.0,
+      "step": 787
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.0001285583103764922,
+      "loss": 0.0002,
+      "step": 788
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 0.00012846648301193756,
+      "loss": 0.0015,
+      "step": 789
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00012837465564738293,
+      "loss": 0.0,
+      "step": 790
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.0001282828282828283,
+      "loss": 0.0004,
+      "step": 791
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00012819100091827364,
+      "loss": 0.0,
+      "step": 792
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 0.000128099173553719,
+      "loss": 0.0006,
+      "step": 793
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 0.00012800734618916438,
+      "loss": 0.4329,
+      "step": 794
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00012791551882460975,
+      "loss": 0.0019,
+      "step": 795
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.00012782369146005512,
+      "loss": 0.0,
+      "step": 796
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 0.00012773186409550046,
+      "loss": 0.0,
+      "step": 797
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00012764003673094583,
+      "loss": 0.0,
+      "step": 798
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.00012754820936639117,
+      "loss": 0.0009,
+      "step": 799
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 0.00012745638200183657,
+      "loss": 0.0001,
+      "step": 800
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.0001273645546372819,
+      "loss": 0.0,
+      "step": 801
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 1.599,
+      "step": 802
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00012718089990817262,
+      "loss": 0.0001,
+      "step": 803
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.00012708907254361802,
+      "loss": 0.0056,
+      "step": 804
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00012699724517906336,
+      "loss": 0.0,
+      "step": 805
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00012690541781450873,
+      "loss": 0.0,
+      "step": 806
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 0.0001268135904499541,
+      "loss": 0.0002,
+      "step": 807
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00012672176308539944,
+      "loss": 0.0001,
+      "step": 808
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 0.00012662993572084484,
+      "loss": 0.0,
+      "step": 809
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 0.00012653810835629018,
+      "loss": 0.0005,
+      "step": 810
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 0.00012644628099173555,
+      "loss": 0.0,
+      "step": 811
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 0.0001263544536271809,
+      "loss": 0.0,
+      "step": 812
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 0.00012626262626262626,
+      "loss": 0.0091,
+      "step": 813
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 0.00012617079889807163,
+      "loss": 0.0217,
+      "step": 814
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 0.000126078971533517,
+      "loss": 0.0,
+      "step": 815
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 0.00012598714416896234,
+      "loss": 0.0,
+      "step": 816
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.0001258953168044077,
+      "loss": 0.0,
+      "step": 817
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.00012580348943985308,
+      "loss": 0.0001,
+      "step": 818
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 0.00012571166207529845,
+      "loss": 0.0,
+      "step": 819
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 0.00012561983471074382,
+      "loss": 0.4714,
+      "step": 820
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 0.00012552800734618916,
+      "loss": 0.004,
+      "step": 821
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 0.00012543617998163453,
+      "loss": 0.0,
+      "step": 822
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 0.00012534435261707987,
+      "loss": 0.0,
+      "step": 823
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 0.00012525252525252527,
+      "loss": 0.0,
+      "step": 824
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.0001251606978879706,
+      "loss": 0.4662,
+      "step": 825
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 0.00012506887052341598,
+      "loss": 0.0636,
+      "step": 826
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 0.00012497704315886135,
+      "loss": 0.0,
+      "step": 827
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 0.00012488521579430672,
+      "loss": 0.0,
+      "step": 828
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.0001247933884297521,
+      "loss": 0.0,
+      "step": 829
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.00012470156106519743,
+      "loss": 0.0,
+      "step": 830
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 0.0001246097337006428,
+      "loss": 0.0246,
+      "step": 831
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 0.00012451790633608814,
+      "loss": 0.0005,
+      "step": 832
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 0.00012442607897153354,
+      "loss": 0.0,
+      "step": 833
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 0.00012433425160697888,
+      "loss": 0.0011,
+      "step": 834
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 0.00012424242424242425,
+      "loss": 0.0,
+      "step": 835
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 0.0001241505968778696,
+      "loss": 0.0024,
+      "step": 836
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 0.00012405876951331496,
+      "loss": 0.0,
+      "step": 837
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 0.00012396694214876033,
+      "loss": 1.0339,
+      "step": 838
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 0.0001238751147842057,
+      "loss": 0.0139,
+      "step": 839
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 0.00012378328741965107,
+      "loss": 0.0,
+      "step": 840
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 0.00012369146005509641,
+      "loss": 0.0,
+      "step": 841
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 0.00012359963269054178,
+      "loss": 0.8451,
+      "step": 842
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 0.00012350780532598715,
+      "loss": 0.0,
+      "step": 843
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 0.00012341597796143252,
+      "loss": 0.018,
+      "step": 844
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 0.00012332415059687786,
+      "loss": 0.1816,
+      "step": 845
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 0.00012323232323232323,
+      "loss": 0.0,
+      "step": 846
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 0.0001231404958677686,
+      "loss": 0.0,
+      "step": 847
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 0.00012304866850321397,
+      "loss": 0.0,
+      "step": 848
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 0.00012295684113865934,
+      "loss": 0.0008,
+      "step": 849
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 0.00012286501377410468,
+      "loss": 0.0,
+      "step": 850
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 0.00012277318640955005,
+      "loss": 0.0002,
+      "step": 851
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 0.00012268135904499542,
+      "loss": 0.0004,
+      "step": 852
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.0001225895316804408,
+      "loss": 0.0005,
+      "step": 853
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.00012249770431588613,
+      "loss": 0.0,
+      "step": 854
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 0.0001224058769513315,
+      "loss": 0.0001,
+      "step": 855
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 0.00012231404958677685,
+      "loss": 0.0,
+      "step": 856
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 0.00012222222222222224,
+      "loss": 0.0,
+      "step": 857
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 0.00012213039485766758,
+      "loss": 0.0,
+      "step": 858
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 0.00012203856749311295,
+      "loss": 0.1222,
+      "step": 859
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 0.00012194674012855831,
+      "loss": 0.0,
+      "step": 860
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 0.00012185491276400366,
+      "loss": 0.0,
+      "step": 861
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.00012176308539944905,
+      "loss": 0.007,
+      "step": 862
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 0.0001216712580348944,
+      "loss": 0.0,
+      "step": 863
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 0.00012157943067033977,
+      "loss": 0.0,
+      "step": 864
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 0.00012148760330578513,
+      "loss": 0.0,
+      "step": 865
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 0.00012139577594123048,
+      "loss": 0.0,
+      "step": 866
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00012130394857667587,
+      "loss": 0.0,
+      "step": 867
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 0.0656,
+      "step": 868
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 0.00012112029384756658,
+      "loss": 0.0009,
+      "step": 869
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 0.00012102846648301193,
+      "loss": 0.0,
+      "step": 870
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 0.00012093663911845729,
+      "loss": 0.0,
+      "step": 871
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 0.00012084481175390267,
+      "loss": 0.0001,
+      "step": 872
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8414634146341463,
+      "eval_loss": 1.5415407419204712,
+      "eval_runtime": 13.9333,
+      "eval_samples_per_second": 11.77,
+      "eval_steps_per_second": 1.507,
+      "step": 872
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 0.00012075298438934803,
+      "loss": 0.0,
+      "step": 873
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 0.0001206611570247934,
+      "loss": 3.1146,
+      "step": 874
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 0.00012056932966023875,
+      "loss": 0.0,
+      "step": 875
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 0.00012047750229568414,
+      "loss": 0.0,
+      "step": 876
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 0.00012038567493112949,
+      "loss": 0.0,
+      "step": 877
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 0.00012029384756657485,
+      "loss": 0.0,
+      "step": 878
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 0.0001202020202020202,
+      "loss": 0.0001,
+      "step": 879
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 0.00012011019283746556,
+      "loss": 0.0001,
+      "step": 880
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 0.00012001836547291094,
+      "loss": 0.0001,
+      "step": 881
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 0.0001199265381083563,
+      "loss": 0.0,
+      "step": 882
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 0.00011983471074380165,
+      "loss": 0.0,
+      "step": 883
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 0.00011974288337924702,
+      "loss": 0.7715,
+      "step": 884
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 0.00011965105601469238,
+      "loss": 0.0143,
+      "step": 885
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 0.00011955922865013776,
+      "loss": 0.0054,
+      "step": 886
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 0.00011946740128558312,
+      "loss": 0.0002,
+      "step": 887
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 0.00011937557392102847,
+      "loss": 0.0002,
+      "step": 888
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 0.00011928374655647383,
+      "loss": 0.0,
+      "step": 889
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 0.00011919191919191919,
+      "loss": 0.0223,
+      "step": 890
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 0.00011910009182736457,
+      "loss": 0.0,
+      "step": 891
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 0.00011900826446280992,
+      "loss": 0.0001,
+      "step": 892
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 0.00011891643709825528,
+      "loss": 0.0,
+      "step": 893
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 0.00011882460973370065,
+      "loss": 0.0,
+      "step": 894
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 0.000118732782369146,
+      "loss": 0.0002,
+      "step": 895
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 0.00011864095500459137,
+      "loss": 0.002,
+      "step": 896
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 0.00011854912764003674,
+      "loss": 0.0847,
+      "step": 897
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 0.0001184573002754821,
+      "loss": 0.0,
+      "step": 898
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 0.00011836547291092745,
+      "loss": 0.0001,
+      "step": 899
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 0.00011827364554637284,
+      "loss": 0.0,
+      "step": 900
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 0.0001181818181818182,
+      "loss": 0.0,
+      "step": 901
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 0.00011808999081726355,
+      "loss": 0.0,
+      "step": 902
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 0.0001179981634527089,
+      "loss": 0.1406,
+      "step": 903
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 0.00011790633608815427,
+      "loss": 0.0,
+      "step": 904
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 0.00011781450872359964,
+      "loss": 0.0,
+      "step": 905
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 0.000117722681359045,
+      "loss": 0.0,
+      "step": 906
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 0.00011763085399449037,
+      "loss": 0.0,
+      "step": 907
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 0.00011753902662993572,
+      "loss": 0.0,
+      "step": 908
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 0.00011744719926538108,
+      "loss": 0.0,
+      "step": 909
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 0.00011735537190082646,
+      "loss": 0.0,
+      "step": 910
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 0.00011726354453627182,
+      "loss": 0.0,
+      "step": 911
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 0.00011717171717171717,
+      "loss": 0.0,
+      "step": 912
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00011707988980716253,
+      "loss": 0.0,
+      "step": 913
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 0.00011698806244260789,
+      "loss": 0.0029,
+      "step": 914
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 0.00011689623507805327,
+      "loss": 0.0,
+      "step": 915
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 0.00011680440771349863,
+      "loss": 1.1478,
+      "step": 916
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 0.000116712580348944,
+      "loss": 0.0,
+      "step": 917
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 0.00011662075298438935,
+      "loss": 0.0,
+      "step": 918
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 0.0001165289256198347,
+      "loss": 0.0,
+      "step": 919
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 0.00011643709825528009,
+      "loss": 0.0,
+      "step": 920
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 0.00011634527089072544,
+      "loss": 0.0,
+      "step": 921
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 0.0001162534435261708,
+      "loss": 0.0002,
+      "step": 922
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 0.00011616161616161616,
+      "loss": 0.0007,
+      "step": 923
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 0.00011606978879706154,
+      "loss": 0.9369,
+      "step": 924
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 0.0001159779614325069,
+      "loss": 0.0007,
+      "step": 925
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 0.00011588613406795225,
+      "loss": 0.0704,
+      "step": 926
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 0.00011579430670339762,
+      "loss": 0.0,
+      "step": 927
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 0.00011570247933884298,
+      "loss": 0.0,
+      "step": 928
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 0.00011561065197428836,
+      "loss": 0.0,
+      "step": 929
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 0.00011551882460973371,
+      "loss": 0.0,
+      "step": 930
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 0.00011542699724517907,
+      "loss": 0.0,
+      "step": 931
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 0.00011533516988062443,
+      "loss": 0.0,
+      "step": 932
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 0.00011524334251606978,
+      "loss": 0.0,
+      "step": 933
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 0.00011515151515151516,
+      "loss": 0.0,
+      "step": 934
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 0.00011505968778696052,
+      "loss": 0.0004,
+      "step": 935
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 0.00011496786042240588,
+      "loss": 0.0,
+      "step": 936
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 0.00011487603305785125,
+      "loss": 0.0002,
+      "step": 937
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 0.0001147842056932966,
+      "loss": 0.0,
+      "step": 938
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 0.00011469237832874198,
+      "loss": 0.0,
+      "step": 939
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 0.00011460055096418734,
+      "loss": 0.0,
+      "step": 940
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 0.0001145087235996327,
+      "loss": 0.0,
+      "step": 941
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 0.00011441689623507805,
+      "loss": 0.0015,
+      "step": 942
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 0.00011432506887052341,
+      "loss": 0.0,
+      "step": 943
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 0.00011423324150596879,
+      "loss": 0.0,
+      "step": 944
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 0.00011414141414141415,
+      "loss": 0.0,
+      "step": 945
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 0.0001140495867768595,
+      "loss": 0.0,
+      "step": 946
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 0.00011395775941230487,
+      "loss": 0.0,
+      "step": 947
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 0.00011386593204775024,
+      "loss": 0.3561,
+      "step": 948
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 0.00011377410468319561,
+      "loss": 0.0,
+      "step": 949
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 0.00011368227731864097,
+      "loss": 0.0,
+      "step": 950
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 0.00011359044995408632,
+      "loss": 0.0,
+      "step": 951
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 0.00011349862258953168,
+      "loss": 0.0,
+      "step": 952
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 0.00011340679522497706,
+      "loss": 0.0,
+      "step": 953
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 0.00011331496786042242,
+      "loss": 0.0,
+      "step": 954
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 0.00011322314049586777,
+      "loss": 0.0002,
+      "step": 955
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 0.00011313131313131313,
+      "loss": 0.0,
+      "step": 956
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 0.0001130394857667585,
+      "loss": 0.0,
+      "step": 957
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 0.00011294765840220387,
+      "loss": 0.0,
+      "step": 958
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 0.00011285583103764923,
+      "loss": 0.0,
+      "step": 959
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 0.00011276400367309459,
+      "loss": 0.0006,
+      "step": 960
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 0.00011267217630853995,
+      "loss": 0.0,
+      "step": 961
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 0.0001125803489439853,
+      "loss": 0.0002,
+      "step": 962
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 0.00011248852157943069,
+      "loss": 0.0,
+      "step": 963
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 0.00011239669421487604,
+      "loss": 0.0,
+      "step": 964
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 0.0001123048668503214,
+      "loss": 1.7818,
+      "step": 965
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 0.00011221303948576675,
+      "loss": 0.0001,
+      "step": 966
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 0.00011212121212121212,
+      "loss": 0.001,
+      "step": 967
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 0.00011202938475665749,
+      "loss": 0.0,
+      "step": 968
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 0.00011193755739210286,
+      "loss": 0.0,
+      "step": 969
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 0.00011184573002754822,
+      "loss": 0.0,
+      "step": 970
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 0.00011175390266299357,
+      "loss": 0.0,
+      "step": 971
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 0.00011166207529843895,
+      "loss": 0.0001,
+      "step": 972
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 0.00011157024793388431,
+      "loss": 0.0,
+      "step": 973
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 0.00011147842056932967,
+      "loss": 0.0781,
+      "step": 974
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 0.00011138659320477502,
+      "loss": 0.0002,
+      "step": 975
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 0.00011129476584022038,
+      "loss": 0.0,
+      "step": 976
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 0.00011120293847566576,
+      "loss": 0.0,
+      "step": 977
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 0.00011111111111111112,
+      "loss": 0.0,
+      "step": 978
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 0.00011101928374655649,
+      "loss": 0.001,
+      "step": 979
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 0.00011092745638200184,
+      "loss": 0.0,
+      "step": 980
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 0.0001108356290174472,
+      "loss": 0.0,
+      "step": 981
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.8109756097560976,
+      "eval_loss": 1.1195001602172852,
+      "eval_runtime": 13.9461,
+      "eval_samples_per_second": 11.76,
+      "eval_steps_per_second": 1.506,
+      "step": 981
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 0.00011074380165289258,
+      "loss": 0.0,
+      "step": 982
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 0.00011065197428833794,
+      "loss": 0.0,
+      "step": 983
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 0.00011056014692378329,
+      "loss": 0.0,
+      "step": 984
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 0.00011046831955922865,
+      "loss": 0.0,
+      "step": 985
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 0.000110376492194674,
+      "loss": 0.0,
+      "step": 986
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 0.00011028466483011939,
+      "loss": 0.0,
+      "step": 987
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 0.00011019283746556474,
+      "loss": 0.0,
+      "step": 988
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 0.00011010101010101011,
+      "loss": 0.0013,
+      "step": 989
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 0.00011000918273645547,
+      "loss": 0.0,
+      "step": 990
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 0.00010991735537190082,
+      "loss": 0.0,
+      "step": 991
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 0.0001098255280073462,
+      "loss": 0.0,
+      "step": 992
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 0.00010973370064279156,
+      "loss": 0.0,
+      "step": 993
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 0.00010964187327823692,
+      "loss": 0.0024,
+      "step": 994
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 0.00010955004591368227,
+      "loss": 0.0,
+      "step": 995
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 0.00010945821854912766,
+      "loss": 0.0,
+      "step": 996
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 0.00010936639118457301,
+      "loss": 0.0,
+      "step": 997
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 0.00010927456382001837,
+      "loss": 0.346,
+      "step": 998
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00010918273645546374,
+      "loss": 0.0,
+      "step": 999
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 0.0,
+      "step": 1000
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 0.00010899908172635446,
+      "loss": 0.0,
+      "step": 1001
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 0.00010890725436179983,
+      "loss": 0.0,
+      "step": 1002
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 0.00010881542699724519,
+      "loss": 0.0,
+      "step": 1003
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 0.00010872359963269054,
+      "loss": 0.0,
+      "step": 1004
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 0.0001086317722681359,
+      "loss": 0.0,
+      "step": 1005
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 0.00010853994490358128,
+      "loss": 0.0,
+      "step": 1006
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 0.00010844811753902664,
+      "loss": 0.0,
+      "step": 1007
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 0.00010835629017447199,
+      "loss": 0.0,
+      "step": 1008
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 0.00010826446280991735,
+      "loss": 0.0,
+      "step": 1009
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 0.00010817263544536272,
+      "loss": 0.0,
+      "step": 1010
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 0.00010808080808080809,
+      "loss": 0.0,
+      "step": 1011
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 0.00010798898071625346,
+      "loss": 0.0002,
+      "step": 1012
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 0.00010789715335169881,
+      "loss": 1.099,
+      "step": 1013
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 0.00010780532598714417,
+      "loss": 0.0,
+      "step": 1014
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 0.00010771349862258952,
+      "loss": 0.0,
+      "step": 1015
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 0.0001076216712580349,
+      "loss": 0.0,
+      "step": 1016
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 0.00010752984389348026,
+      "loss": 0.0,
+      "step": 1017
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 0.00010743801652892562,
+      "loss": 0.0001,
+      "step": 1018
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 0.00010734618916437097,
+      "loss": 0.0,
+      "step": 1019
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 0.00010725436179981636,
+      "loss": 0.0016,
+      "step": 1020
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 0.00010716253443526171,
+      "loss": 0.0,
+      "step": 1021
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 0.00010707070707070708,
+      "loss": 0.0,
+      "step": 1022
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 0.00010697887970615244,
+      "loss": 0.0001,
+      "step": 1023
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 0.0001068870523415978,
+      "loss": 0.0,
+      "step": 1024
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 0.00010679522497704318,
+      "loss": 0.0,
+      "step": 1025
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 0.00010670339761248853,
+      "loss": 0.0,
+      "step": 1026
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 0.00010661157024793389,
+      "loss": 0.0,
+      "step": 1027
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 0.00010651974288337924,
+      "loss": 0.0,
+      "step": 1028
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 0.0001064279155188246,
+      "loss": 0.0001,
+      "step": 1029
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 0.00010633608815426998,
+      "loss": 0.0,
+      "step": 1030
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 0.00010624426078971534,
+      "loss": 0.0,
+      "step": 1031
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 0.00010615243342516071,
+      "loss": 0.0003,
+      "step": 1032
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 0.0,
+      "step": 1033
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 0.00010596877869605142,
+      "loss": 0.0,
+      "step": 1034
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 0.0001058769513314968,
+      "loss": 0.0,
+      "step": 1035
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 0.00010578512396694216,
+      "loss": 0.0,
+      "step": 1036
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 0.00010569329660238751,
+      "loss": 0.0,
+      "step": 1037
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 0.00010560146923783287,
+      "loss": 0.0,
+      "step": 1038
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 0.00010550964187327822,
+      "loss": 0.0,
+      "step": 1039
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 0.00010541781450872361,
+      "loss": 0.0,
+      "step": 1040
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 0.00010532598714416896,
+      "loss": 0.0,
+      "step": 1041
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 0.00010523415977961433,
+      "loss": 0.8288,
+      "step": 1042
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 0.00010514233241505969,
+      "loss": 0.0,
+      "step": 1043
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 0.00010505050505050507,
+      "loss": 0.0,
+      "step": 1044
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 0.00010495867768595043,
+      "loss": 0.0,
+      "step": 1045
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 0.00010486685032139578,
+      "loss": 0.0,
+      "step": 1046
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 0.00010477502295684114,
+      "loss": 0.0,
+      "step": 1047
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 0.0001046831955922865,
+      "loss": 0.0,
+      "step": 1048
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 0.00010459136822773188,
+      "loss": 0.0,
+      "step": 1049
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 0.00010449954086317723,
+      "loss": 0.0,
+      "step": 1050
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 0.00010440771349862259,
+      "loss": 0.0,
+      "step": 1051
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 0.00010431588613406796,
+      "loss": 0.0,
+      "step": 1052
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 0.00010422405876951331,
+      "loss": 0.0,
+      "step": 1053
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 0.0001041322314049587,
+      "loss": 0.0,
+      "step": 1054
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 0.00010404040404040405,
+      "loss": 0.0006,
+      "step": 1055
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 0.00010394857667584941,
+      "loss": 0.0,
+      "step": 1056
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 0.00010385674931129476,
+      "loss": 0.0,
+      "step": 1057
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 0.00010376492194674012,
+      "loss": 0.0,
+      "step": 1058
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 0.0001036730945821855,
+      "loss": 0.0,
+      "step": 1059
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 0.00010358126721763086,
+      "loss": 0.0,
+      "step": 1060
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 0.00010348943985307621,
+      "loss": 0.0,
+      "step": 1061
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 0.00010339761248852158,
+      "loss": 0.0,
+      "step": 1062
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 0.00010330578512396694,
+      "loss": 0.0,
+      "step": 1063
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 0.00010321395775941232,
+      "loss": 0.0,
+      "step": 1064
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 0.00010312213039485768,
+      "loss": 0.0005,
+      "step": 1065
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 0.00010303030303030303,
+      "loss": 0.0,
+      "step": 1066
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 0.00010293847566574839,
+      "loss": 0.0,
+      "step": 1067
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 0.00010284664830119377,
+      "loss": 0.0,
+      "step": 1068
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 0.00010275482093663913,
+      "loss": 0.0,
+      "step": 1069
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 0.00010266299357208448,
+      "loss": 0.0001,
+      "step": 1070
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 0.00010257116620752984,
+      "loss": 0.0,
+      "step": 1071
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 0.00010247933884297521,
+      "loss": 0.0,
+      "step": 1072
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 0.00010238751147842058,
+      "loss": 0.0033,
+      "step": 1073
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 0.00010229568411386595,
+      "loss": 0.0,
+      "step": 1074
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 0.0001022038567493113,
+      "loss": 0.0,
+      "step": 1075
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 0.00010211202938475666,
+      "loss": 0.0002,
+      "step": 1076
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 0.00010202020202020202,
+      "loss": 0.0,
+      "step": 1077
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 0.0001019283746556474,
+      "loss": 0.0,
+      "step": 1078
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 0.00010183654729109275,
+      "loss": 0.0,
+      "step": 1079
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 0.00010174471992653811,
+      "loss": 0.0,
+      "step": 1080
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 0.00010165289256198347,
+      "loss": 0.0,
+      "step": 1081
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 0.00010156106519742883,
+      "loss": 0.0,
+      "step": 1082
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 0.0001014692378328742,
+      "loss": 0.0517,
+      "step": 1083
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 0.00010137741046831957,
+      "loss": 0.0,
+      "step": 1084
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 0.00010128558310376493,
+      "loss": 0.0,
+      "step": 1085
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 0.00010119375573921028,
+      "loss": 0.0,
+      "step": 1086
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 0.00010110192837465564,
+      "loss": 0.0,
+      "step": 1087
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 0.00010101010101010102,
+      "loss": 0.0,
+      "step": 1088
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 0.00010091827364554638,
+      "loss": 0.0,
+      "step": 1089
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.00010082644628099174,
+      "loss": 0.0,
+      "step": 1090
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.823170731707317,
+      "eval_loss": 1.2256759405136108,
+      "eval_runtime": 13.9488,
+      "eval_samples_per_second": 11.757,
+      "eval_steps_per_second": 1.506,
+      "step": 1090
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 0.00010073461891643709,
+      "loss": 0.0,
+      "step": 1091
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 0.00010064279155188247,
+      "loss": 0.0,
+      "step": 1092
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 0.00010055096418732783,
+      "loss": 0.0,
+      "step": 1093
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 0.0001004591368227732,
+      "loss": 0.0,
+      "step": 1094
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 0.00010036730945821855,
+      "loss": 0.0,
+      "step": 1095
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 0.00010027548209366391,
+      "loss": 0.0001,
+      "step": 1096
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 0.0001001836547291093,
+      "loss": 0.0,
+      "step": 1097
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 0.00010009182736455465,
+      "loss": 0.0,
+      "step": 1098
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 1099
+    },
+    {
+      "epoch": 10.09,
+      "learning_rate": 9.990817263544536e-05,
+      "loss": 0.0,
+      "step": 1100
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 9.981634527089073e-05,
+      "loss": 0.0,
+      "step": 1101
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 9.972451790633609e-05,
+      "loss": 0.0,
+      "step": 1102
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 9.963269054178145e-05,
+      "loss": 0.0,
+      "step": 1103
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 9.954086317722681e-05,
+      "loss": 0.0,
+      "step": 1104
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 9.944903581267218e-05,
+      "loss": 0.0,
+      "step": 1105
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 9.935720844811755e-05,
+      "loss": 0.0,
+      "step": 1106
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 9.92653810835629e-05,
+      "loss": 0.0,
+      "step": 1107
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 9.917355371900827e-05,
+      "loss": 0.0,
+      "step": 1108
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 9.908172635445363e-05,
+      "loss": 0.0,
+      "step": 1109
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 9.8989898989899e-05,
+      "loss": 0.0002,
+      "step": 1110
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 9.889807162534436e-05,
+      "loss": 0.0,
+      "step": 1111
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 9.880624426078971e-05,
+      "loss": 0.0,
+      "step": 1112
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 9.871441689623508e-05,
+      "loss": 0.0,
+      "step": 1113
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 9.862258953168044e-05,
+      "loss": 0.0,
+      "step": 1114
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 9.85307621671258e-05,
+      "loss": 0.0,
+      "step": 1115
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 9.843893480257117e-05,
+      "loss": 0.0,
+      "step": 1116
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 9.834710743801654e-05,
+      "loss": 0.0,
+      "step": 1117
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 9.82552800734619e-05,
+      "loss": 0.0,
+      "step": 1118
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 9.816345270890726e-05,
+      "loss": 0.0,
+      "step": 1119
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 9.807162534435263e-05,
+      "loss": 0.0062,
+      "step": 1120
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 9.797979797979798e-05,
+      "loss": 0.0,
+      "step": 1121
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 9.788797061524335e-05,
+      "loss": 0.0,
+      "step": 1122
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 9.77961432506887e-05,
+      "loss": 0.0,
+      "step": 1123
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 9.770431588613406e-05,
+      "loss": 0.0,
+      "step": 1124
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 9.761248852157943e-05,
+      "loss": 0.0,
+      "step": 1125
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 9.75206611570248e-05,
+      "loss": 0.0,
+      "step": 1126
+    },
+    {
+      "epoch": 10.34,
+      "learning_rate": 9.742883379247017e-05,
+      "loss": 0.0,
+      "step": 1127
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 9.733700642791553e-05,
+      "loss": 0.0,
+      "step": 1128
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 9.72451790633609e-05,
+      "loss": 0.0,
+      "step": 1129
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 9.715335169880625e-05,
+      "loss": 0.0,
+      "step": 1130
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 9.70615243342516e-05,
+      "loss": 0.0,
+      "step": 1131
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 9.696969696969698e-05,
+      "loss": 0.0,
+      "step": 1132
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 9.687786960514233e-05,
+      "loss": 0.0,
+      "step": 1133
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 9.67860422405877e-05,
+      "loss": 0.0,
+      "step": 1134
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 9.669421487603306e-05,
+      "loss": 0.0,
+      "step": 1135
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 9.660238751147843e-05,
+      "loss": 0.0005,
+      "step": 1136
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 9.65105601469238e-05,
+      "loss": 0.0,
+      "step": 1137
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 9.641873278236915e-05,
+      "loss": 0.0001,
+      "step": 1138
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 9.632690541781452e-05,
+      "loss": 0.0,
+      "step": 1139
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 9.623507805325988e-05,
+      "loss": 0.0,
+      "step": 1140
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 9.614325068870525e-05,
+      "loss": 0.0,
+      "step": 1141
+    },
+    {
+      "epoch": 10.48,
+      "learning_rate": 9.60514233241506e-05,
+      "loss": 0.0,
+      "step": 1142
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 9.595959595959596e-05,
+      "loss": 0.0001,
+      "step": 1143
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 9.586776859504133e-05,
+      "loss": 0.0,
+      "step": 1144
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 9.577594123048668e-05,
+      "loss": 0.0,
+      "step": 1145
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 9.568411386593205e-05,
+      "loss": 0.0001,
+      "step": 1146
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 9.559228650137742e-05,
+      "loss": 0.0,
+      "step": 1147
+    },
+    {
+      "epoch": 10.53,
+      "learning_rate": 9.550045913682278e-05,
+      "loss": 0.0,
+      "step": 1148
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 9.540863177226815e-05,
+      "loss": 0.0,
+      "step": 1149
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 9.53168044077135e-05,
+      "loss": 0.0,
+      "step": 1150
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 9.522497704315887e-05,
+      "loss": 0.0,
+      "step": 1151
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 9.513314967860423e-05,
+      "loss": 0.0,
+      "step": 1152
+    },
+    {
+      "epoch": 10.58,
+      "learning_rate": 9.50413223140496e-05,
+      "loss": 0.0,
+      "step": 1153
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 9.494949494949495e-05,
+      "loss": 0.0,
+      "step": 1154
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 9.485766758494031e-05,
+      "loss": 0.0,
+      "step": 1155
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 9.476584022038568e-05,
+      "loss": 0.0,
+      "step": 1156
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 9.467401285583105e-05,
+      "loss": 0.0,
+      "step": 1157
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 9.458218549127642e-05,
+      "loss": 0.0,
+      "step": 1158
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 9.449035812672177e-05,
+      "loss": 0.0007,
+      "step": 1159
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 9.439853076216713e-05,
+      "loss": 0.0,
+      "step": 1160
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 9.43067033976125e-05,
+      "loss": 0.0,
+      "step": 1161
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 9.421487603305785e-05,
+      "loss": 0.0,
+      "step": 1162
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 9.412304866850322e-05,
+      "loss": 0.0,
+      "step": 1163
+    },
+    {
+      "epoch": 10.68,
+      "learning_rate": 9.403122130394858e-05,
+      "loss": 0.0,
+      "step": 1164
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 9.393939393939395e-05,
+      "loss": 0.0,
+      "step": 1165
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 9.38475665748393e-05,
+      "loss": 0.0,
+      "step": 1166
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 9.375573921028467e-05,
+      "loss": 0.0001,
+      "step": 1167
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 9.366391184573004e-05,
+      "loss": 0.0,
+      "step": 1168
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 9.35720844811754e-05,
+      "loss": 0.0,
+      "step": 1169
+    },
+    {
+      "epoch": 10.73,
+      "learning_rate": 9.348025711662077e-05,
+      "loss": 0.0,
+      "step": 1170
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 9.338842975206612e-05,
+      "loss": 0.0,
+      "step": 1171
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 9.329660238751148e-05,
+      "loss": 0.0,
+      "step": 1172
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 9.320477502295685e-05,
+      "loss": 0.0,
+      "step": 1173
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 9.31129476584022e-05,
+      "loss": 0.0,
+      "step": 1174
+    },
+    {
+      "epoch": 10.78,
+      "learning_rate": 9.302112029384757e-05,
+      "loss": 0.0,
+      "step": 1175
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 9.292929292929293e-05,
+      "loss": 0.0,
+      "step": 1176
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 9.28374655647383e-05,
+      "loss": 0.0,
+      "step": 1177
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 9.274563820018367e-05,
+      "loss": 0.0,
+      "step": 1178
+    },
+    {
+      "epoch": 10.82,
+      "learning_rate": 9.265381083562902e-05,
+      "loss": 0.0,
+      "step": 1179
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 9.256198347107439e-05,
+      "loss": 0.0,
+      "step": 1180
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 9.247015610651975e-05,
+      "loss": 0.0,
+      "step": 1181
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 9.237832874196512e-05,
+      "loss": 0.0,
+      "step": 1182
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 9.228650137741047e-05,
+      "loss": 0.0,
+      "step": 1183
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 9.219467401285583e-05,
+      "loss": 0.0,
+      "step": 1184
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 9.21028466483012e-05,
+      "loss": 0.0001,
+      "step": 1185
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 9.201101928374655e-05,
+      "loss": 0.0,
+      "step": 1186
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 9.191919191919192e-05,
+      "loss": 0.0,
+      "step": 1187
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 9.182736455463729e-05,
+      "loss": 0.0,
+      "step": 1188
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 9.173553719008266e-05,
+      "loss": 0.0,
+      "step": 1189
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 9.164370982552802e-05,
+      "loss": 0.0,
+      "step": 1190
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 9.155188246097337e-05,
+      "loss": 0.0002,
+      "step": 1191
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 9.146005509641874e-05,
+      "loss": 0.0,
+      "step": 1192
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 9.13682277318641e-05,
+      "loss": 0.0,
+      "step": 1193
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 9.127640036730947e-05,
+      "loss": 0.0,
+      "step": 1194
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 9.118457300275482e-05,
+      "loss": 0.0,
+      "step": 1195
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 9.109274563820018e-05,
+      "loss": 0.0,
+      "step": 1196
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 9.100091827364555e-05,
+      "loss": 0.0,
+      "step": 1197
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 0.0,
+      "step": 1198
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 9.081726354453627e-05,
+      "loss": 0.0,
+      "step": 1199
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.8170731707317073,
+      "eval_loss": 1.3680109977722168,
+      "eval_runtime": 13.9462,
+      "eval_samples_per_second": 11.76,
+      "eval_steps_per_second": 1.506,
+      "step": 1199
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 9.072543617998164e-05,
+      "loss": 0.0,
+      "step": 1200
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 9.063360881542701e-05,
+      "loss": 0.0,
+      "step": 1201
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 9.054178145087237e-05,
+      "loss": 0.0,
+      "step": 1202
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 9.044995408631772e-05,
+      "loss": 0.0,
+      "step": 1203
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 9.035812672176309e-05,
+      "loss": 0.0,
+      "step": 1204
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 9.026629935720845e-05,
+      "loss": 0.0,
+      "step": 1205
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 9.017447199265382e-05,
+      "loss": 0.0,
+      "step": 1206
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 9.008264462809917e-05,
+      "loss": 0.0,
+      "step": 1207
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 8.999081726354453e-05,
+      "loss": 0.0,
+      "step": 1208
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 8.98989898989899e-05,
+      "loss": 0.0,
+      "step": 1209
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 8.980716253443527e-05,
+      "loss": 0.0,
+      "step": 1210
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 8.971533516988064e-05,
+      "loss": 0.0,
+      "step": 1211
+    },
+    {
+      "epoch": 11.12,
+      "learning_rate": 8.962350780532599e-05,
+      "loss": 0.0,
+      "step": 1212
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 8.953168044077136e-05,
+      "loss": 0.0,
+      "step": 1213
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 8.943985307621672e-05,
+      "loss": 0.0,
+      "step": 1214
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 8.934802571166207e-05,
+      "loss": 0.0001,
+      "step": 1215
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 8.925619834710744e-05,
+      "loss": 0.0,
+      "step": 1216
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 8.91643709825528e-05,
+      "loss": 0.0,
+      "step": 1217
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 8.907254361799817e-05,
+      "loss": 0.0,
+      "step": 1218
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 8.898071625344352e-05,
+      "loss": 0.0,
+      "step": 1219
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 0.0,
+      "step": 1220
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 8.879706152433426e-05,
+      "loss": 0.0,
+      "step": 1221
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 8.870523415977962e-05,
+      "loss": 0.0,
+      "step": 1222
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 8.861340679522499e-05,
+      "loss": 0.0,
+      "step": 1223
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 8.852157943067034e-05,
+      "loss": 0.0,
+      "step": 1224
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 8.842975206611571e-05,
+      "loss": 0.0,
+      "step": 1225
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 8.833792470156107e-05,
+      "loss": 0.0,
+      "step": 1226
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 8.824609733700642e-05,
+      "loss": 0.0,
+      "step": 1227
+    },
+    {
+      "epoch": 11.27,
+      "learning_rate": 8.81542699724518e-05,
+      "loss": 0.0,
+      "step": 1228
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 8.806244260789715e-05,
+      "loss": 0.0,
+      "step": 1229
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 8.797061524334252e-05,
+      "loss": 0.0,
+      "step": 1230
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 8.787878787878789e-05,
+      "loss": 0.0001,
+      "step": 1231
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 8.778696051423324e-05,
+      "loss": 0.0,
+      "step": 1232
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 8.769513314967861e-05,
+      "loss": 0.0,
+      "step": 1233
+    },
+    {
+      "epoch": 11.32,
+      "learning_rate": 8.760330578512397e-05,
+      "loss": 0.0,
+      "step": 1234
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 8.751147842056934e-05,
+      "loss": 0.0,
+      "step": 1235
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 8.74196510560147e-05,
+      "loss": 0.0,
+      "step": 1236
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 8.732782369146006e-05,
+      "loss": 0.0,
+      "step": 1237
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 8.723599632690542e-05,
+      "loss": 0.0,
+      "step": 1238
+    },
+    {
+      "epoch": 11.37,
+      "learning_rate": 8.714416896235077e-05,
+      "loss": 0.0,
+      "step": 1239
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 8.705234159779614e-05,
+      "loss": 0.0,
+      "step": 1240
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 8.696051423324151e-05,
+      "loss": 0.0,
+      "step": 1241
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 8.686868686868688e-05,
+      "loss": 0.0,
+      "step": 1242
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 8.677685950413224e-05,
+      "loss": 0.0,
+      "step": 1243
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 8.66850321395776e-05,
+      "loss": 0.0,
+      "step": 1244
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 8.659320477502296e-05,
+      "loss": 0.0,
+      "step": 1245
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 8.650137741046832e-05,
+      "loss": 0.0,
+      "step": 1246
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 8.640955004591369e-05,
+      "loss": 0.0,
+      "step": 1247
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 8.631772268135904e-05,
+      "loss": 0.0,
+      "step": 1248
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 8.622589531680441e-05,
+      "loss": 0.0,
+      "step": 1249
+    },
+    {
+      "epoch": 11.47,
+      "learning_rate": 8.613406795224977e-05,
+      "loss": 0.0,
+      "step": 1250
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 8.604224058769514e-05,
+      "loss": 0.0,
+      "step": 1251
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 8.595041322314051e-05,
+      "loss": 0.0002,
+      "step": 1252
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 8.585858585858586e-05,
+      "loss": 0.0,
+      "step": 1253
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 8.576675849403123e-05,
+      "loss": 0.0,
+      "step": 1254
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 8.567493112947659e-05,
+      "loss": 0.0,
+      "step": 1255
+    },
+    {
+      "epoch": 11.52,
+      "learning_rate": 8.558310376492194e-05,
+      "loss": 0.0,
+      "step": 1256
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 8.549127640036731e-05,
+      "loss": 0.0,
+      "step": 1257
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 8.539944903581267e-05,
+      "loss": 0.0,
+      "step": 1258
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 8.530762167125804e-05,
+      "loss": 0.0,
+      "step": 1259
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 8.52157943067034e-05,
+      "loss": 0.0,
+      "step": 1260
+    },
+    {
+      "epoch": 11.57,
+      "learning_rate": 8.512396694214876e-05,
+      "loss": 0.0,
+      "step": 1261
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 8.503213957759413e-05,
+      "loss": 0.0,
+      "step": 1262
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 8.494031221303949e-05,
+      "loss": 0.0,
+      "step": 1263
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 8.484848484848486e-05,
+      "loss": 0.0,
+      "step": 1264
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 8.475665748393021e-05,
+      "loss": 0.0,
+      "step": 1265
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 8.466483011937558e-05,
+      "loss": 0.0,
+      "step": 1266
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 8.457300275482094e-05,
+      "loss": 0.0,
+      "step": 1267
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 8.44811753902663e-05,
+      "loss": 0.0,
+      "step": 1268
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 8.438934802571166e-05,
+      "loss": 0.0,
+      "step": 1269
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 8.429752066115702e-05,
+      "loss": 0.0,
+      "step": 1270
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 8.420569329660239e-05,
+      "loss": 0.0,
+      "step": 1271
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 8.411386593204776e-05,
+      "loss": 0.0,
+      "step": 1272
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 8.402203856749313e-05,
+      "loss": 0.0,
+      "step": 1273
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 8.393021120293848e-05,
+      "loss": 0.0,
+      "step": 1274
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 8.383838383838384e-05,
+      "loss": 0.0,
+      "step": 1275
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 8.374655647382921e-05,
+      "loss": 0.0,
+      "step": 1276
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 8.365472910927457e-05,
+      "loss": 0.0,
+      "step": 1277
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 8.356290174471993e-05,
+      "loss": 0.0,
+      "step": 1278
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 8.347107438016529e-05,
+      "loss": 0.0,
+      "step": 1279
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 8.337924701561065e-05,
+      "loss": 0.0,
+      "step": 1280
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 8.328741965105602e-05,
+      "loss": 0.0,
+      "step": 1281
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 8.319559228650138e-05,
+      "loss": 0.0,
+      "step": 1282
+    },
+    {
+      "epoch": 11.77,
+      "learning_rate": 8.310376492194675e-05,
+      "loss": 0.0,
+      "step": 1283
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 8.301193755739211e-05,
+      "loss": 0.0,
+      "step": 1284
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 8.292011019283748e-05,
+      "loss": 0.0,
+      "step": 1285
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 8.282828282828283e-05,
+      "loss": 0.0,
+      "step": 1286
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 8.273645546372819e-05,
+      "loss": 0.0,
+      "step": 1287
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 8.264462809917356e-05,
+      "loss": 0.0,
+      "step": 1288
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 8.255280073461892e-05,
+      "loss": 0.0,
+      "step": 1289
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 8.246097337006428e-05,
+      "loss": 0.0,
+      "step": 1290
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 8.236914600550964e-05,
+      "loss": 0.0,
+      "step": 1291
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 8.227731864095501e-05,
+      "loss": 0.0,
+      "step": 1292
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 8.218549127640038e-05,
+      "loss": 0.0,
+      "step": 1293
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 8.209366391184574e-05,
+      "loss": 0.0,
+      "step": 1294
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 8.20018365472911e-05,
+      "loss": 0.0,
+      "step": 1295
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 8.191000918273646e-05,
+      "loss": 0.0,
+      "step": 1296
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 0.0,
+      "step": 1297
+    },
+    {
+      "epoch": 11.91,
+      "learning_rate": 8.172635445362719e-05,
+      "loss": 0.0,
+      "step": 1298
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 8.163452708907254e-05,
+      "loss": 0.0,
+      "step": 1299
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 8.154269972451791e-05,
+      "loss": 0.0,
+      "step": 1300
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 8.145087235996327e-05,
+      "loss": 0.0,
+      "step": 1301
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 8.135904499540864e-05,
+      "loss": 0.0,
+      "step": 1302
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 8.126721763085399e-05,
+      "loss": 0.0,
+      "step": 1303
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 8.117539026629936e-05,
+      "loss": 0.0,
+      "step": 1304
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 8.108356290174473e-05,
+      "loss": 0.0,
+      "step": 1305
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 8.099173553719009e-05,
+      "loss": 0.0,
+      "step": 1306
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 8.089990817263545e-05,
+      "loss": 0.0,
+      "step": 1307
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 8.080808080808081e-05,
+      "loss": 0.0,
+      "step": 1308
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.8170731707317073,
+      "eval_loss": 1.3485139608383179,
+      "eval_runtime": 13.9503,
+      "eval_samples_per_second": 11.756,
+      "eval_steps_per_second": 1.505,
+      "step": 1308
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 8.071625344352618e-05,
+      "loss": 0.0,
+      "step": 1309
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 8.062442607897154e-05,
+      "loss": 0.0,
+      "step": 1310
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 8.053259871441689e-05,
+      "loss": 0.0,
+      "step": 1311
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 8.044077134986226e-05,
+      "loss": 0.0,
+      "step": 1312
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 8.034894398530762e-05,
+      "loss": 0.0,
+      "step": 1313
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 8.025711662075299e-05,
+      "loss": 0.0,
+      "step": 1314
+    },
+    {
+      "epoch": 12.06,
+      "learning_rate": 8.016528925619836e-05,
+      "loss": 0.0,
+      "step": 1315
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 8.007346189164371e-05,
+      "loss": 0.0,
+      "step": 1316
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 7.998163452708908e-05,
+      "loss": 0.0,
+      "step": 1317
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 7.988980716253444e-05,
+      "loss": 0.0,
+      "step": 1318
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 7.97979797979798e-05,
+      "loss": 0.0,
+      "step": 1319
+    },
+    {
+      "epoch": 12.11,
+      "learning_rate": 7.970615243342516e-05,
+      "loss": 0.0,
+      "step": 1320
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 7.961432506887053e-05,
+      "loss": 0.0,
+      "step": 1321
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 7.952249770431589e-05,
+      "loss": 0.0,
+      "step": 1322
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 7.943067033976124e-05,
+      "loss": 0.0,
+      "step": 1323
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 7.933884297520661e-05,
+      "loss": 0.0,
+      "step": 1324
+    },
+    {
+      "epoch": 12.16,
+      "learning_rate": 7.924701561065198e-05,
+      "loss": 0.0,
+      "step": 1325
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 7.915518824609735e-05,
+      "loss": 0.0,
+      "step": 1326
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 7.90633608815427e-05,
+      "loss": 0.0,
+      "step": 1327
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 7.897153351698806e-05,
+      "loss": 0.0,
+      "step": 1328
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 7.887970615243343e-05,
+      "loss": 0.0,
+      "step": 1329
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 7.878787878787879e-05,
+      "loss": 0.0,
+      "step": 1330
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 7.869605142332416e-05,
+      "loss": 0.0,
+      "step": 1331
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 7.860422405876951e-05,
+      "loss": 0.0,
+      "step": 1332
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 7.851239669421488e-05,
+      "loss": 0.0,
+      "step": 1333
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 7.842056932966024e-05,
+      "loss": 0.0,
+      "step": 1334
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 7.83287419651056e-05,
+      "loss": 0.0,
+      "step": 1335
+    },
+    {
+      "epoch": 12.26,
+      "learning_rate": 7.823691460055098e-05,
+      "loss": 0.0,
+      "step": 1336
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 7.814508723599633e-05,
+      "loss": 0.0,
+      "step": 1337
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 7.80532598714417e-05,
+      "loss": 0.0,
+      "step": 1338
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 7.796143250688706e-05,
+      "loss": 0.0,
+      "step": 1339
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 7.786960514233241e-05,
+      "loss": 0.0,
+      "step": 1340
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.0,
+      "step": 1341
+    },
+    {
+      "epoch": 12.31,
+      "learning_rate": 7.768595041322314e-05,
+      "loss": 0.0,
+      "step": 1342
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 7.75941230486685e-05,
+      "loss": 0.0,
+      "step": 1343
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 7.750229568411386e-05,
+      "loss": 0.0,
+      "step": 1344
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 7.741046831955923e-05,
+      "loss": 0.0,
+      "step": 1345
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 7.73186409550046e-05,
+      "loss": 0.0,
+      "step": 1346
+    },
+    {
+      "epoch": 12.36,
+      "learning_rate": 7.722681359044996e-05,
+      "loss": 0.0,
+      "step": 1347
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 7.713498622589533e-05,
+      "loss": 0.0,
+      "step": 1348
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 7.704315886134068e-05,
+      "loss": 0.0,
+      "step": 1349
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 7.695133149678605e-05,
+      "loss": 0.0,
+      "step": 1350
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 7.685950413223141e-05,
+      "loss": 0.0,
+      "step": 1351
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 7.676767676767676e-05,
+      "loss": 0.0,
+      "step": 1352
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 7.667584940312213e-05,
+      "loss": 0.0,
+      "step": 1353
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 7.658402203856749e-05,
+      "loss": 0.0,
+      "step": 1354
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 7.649219467401286e-05,
+      "loss": 0.0,
+      "step": 1355
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 7.640036730945823e-05,
+      "loss": 0.0,
+      "step": 1356
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 7.63085399449036e-05,
+      "loss": 0.0,
+      "step": 1357
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 7.621671258034895e-05,
+      "loss": 0.0,
+      "step": 1358
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 7.612488521579431e-05,
+      "loss": 0.0,
+      "step": 1359
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 7.603305785123968e-05,
+      "loss": 0.0,
+      "step": 1360
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 7.594123048668503e-05,
+      "loss": 0.0,
+      "step": 1361
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 7.58494031221304e-05,
+      "loss": 0.0,
+      "step": 1362
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 0.0,
+      "step": 1363
+    },
+    {
+      "epoch": 12.51,
+      "learning_rate": 7.566574839302111e-05,
+      "loss": 0.0,
+      "step": 1364
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 7.557392102846648e-05,
+      "loss": 0.0,
+      "step": 1365
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 7.548209366391185e-05,
+      "loss": 0.0,
+      "step": 1366
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 7.539026629935722e-05,
+      "loss": 0.0,
+      "step": 1367
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 7.529843893480258e-05,
+      "loss": 0.0,
+      "step": 1368
+    },
+    {
+      "epoch": 12.56,
+      "learning_rate": 7.520661157024795e-05,
+      "loss": 0.0,
+      "step": 1369
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 7.51147842056933e-05,
+      "loss": 0.0,
+      "step": 1370
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 7.502295684113866e-05,
+      "loss": 0.0,
+      "step": 1371
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 7.493112947658403e-05,
+      "loss": 0.0,
+      "step": 1372
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 7.483930211202938e-05,
+      "loss": 0.0,
+      "step": 1373
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 7.474747474747475e-05,
+      "loss": 0.0,
+      "step": 1374
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 7.465564738292011e-05,
+      "loss": 0.0,
+      "step": 1375
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 7.456382001836548e-05,
+      "loss": 0.0,
+      "step": 1376
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 7.447199265381085e-05,
+      "loss": 0.0,
+      "step": 1377
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 7.43801652892562e-05,
+      "loss": 0.0,
+      "step": 1378
+    },
+    {
+      "epoch": 12.65,
+      "learning_rate": 7.428833792470157e-05,
+      "loss": 0.0,
+      "step": 1379
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 7.419651056014693e-05,
+      "loss": 0.0,
+      "step": 1380
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 7.41046831955923e-05,
+      "loss": 0.0,
+      "step": 1381
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 7.401285583103765e-05,
+      "loss": 0.0,
+      "step": 1382
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 7.392102846648301e-05,
+      "loss": 0.0,
+      "step": 1383
+    },
+    {
+      "epoch": 12.7,
+      "learning_rate": 7.382920110192838e-05,
+      "loss": 0.0,
+      "step": 1384
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 7.373737373737373e-05,
+      "loss": 0.0,
+      "step": 1385
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 7.36455463728191e-05,
+      "loss": 0.0,
+      "step": 1386
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 7.355371900826447e-05,
+      "loss": 0.0,
+      "step": 1387
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 7.346189164370983e-05,
+      "loss": 0.0,
+      "step": 1388
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 7.33700642791552e-05,
+      "loss": 0.0,
+      "step": 1389
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 7.327823691460055e-05,
+      "loss": 0.0,
+      "step": 1390
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 7.318640955004592e-05,
+      "loss": 0.0,
+      "step": 1391
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 7.309458218549128e-05,
+      "loss": 0.0,
+      "step": 1392
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 7.300275482093665e-05,
+      "loss": 0.0,
+      "step": 1393
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 7.2910927456382e-05,
+      "loss": 0.0,
+      "step": 1394
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 7.281910009182736e-05,
+      "loss": 0.0,
+      "step": 1395
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 0.0,
+      "step": 1396
+    },
+    {
+      "epoch": 12.82,
+      "learning_rate": 7.26354453627181e-05,
+      "loss": 0.0,
+      "step": 1397
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 7.254361799816345e-05,
+      "loss": 0.0,
+      "step": 1398
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 7.245179063360882e-05,
+      "loss": 0.0,
+      "step": 1399
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 7.235996326905418e-05,
+      "loss": 0.0,
+      "step": 1400
+    },
+    {
+      "epoch": 12.85,
+      "learning_rate": 7.226813590449955e-05,
+      "loss": 0.0,
+      "step": 1401
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 7.21763085399449e-05,
+      "loss": 0.0,
+      "step": 1402
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 7.208448117539027e-05,
+      "loss": 0.0,
+      "step": 1403
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 7.199265381083563e-05,
+      "loss": 0.0,
+      "step": 1404
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 7.1900826446281e-05,
+      "loss": 0.0,
+      "step": 1405
+    },
+    {
+      "epoch": 12.9,
+      "learning_rate": 7.180899908172635e-05,
+      "loss": 0.0,
+      "step": 1406
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 7.171717171717171e-05,
+      "loss": 0.0,
+      "step": 1407
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 7.162534435261708e-05,
+      "loss": 0.0,
+      "step": 1408
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 7.153351698806245e-05,
+      "loss": 0.0,
+      "step": 1409
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 7.144168962350782e-05,
+      "loss": 0.0,
+      "step": 1410
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 7.134986225895317e-05,
+      "loss": 0.0,
+      "step": 1411
+    },
+    {
+      "epoch": 12.95,
+      "learning_rate": 7.125803489439853e-05,
+      "loss": 0.0,
+      "step": 1412
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 7.11662075298439e-05,
+      "loss": 0.0,
+      "step": 1413
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 7.107438016528925e-05,
+      "loss": 0.0,
+      "step": 1414
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 7.098255280073462e-05,
+      "loss": 0.0,
+      "step": 1415
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 7.089072543617998e-05,
+      "loss": 0.0,
+      "step": 1416
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 7.079889807162535e-05,
+      "loss": 0.0,
+      "step": 1417
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.8170731707317073,
+      "eval_loss": 1.3482344150543213,
+      "eval_runtime": 13.9481,
+      "eval_samples_per_second": 11.758,
+      "eval_steps_per_second": 1.506,
+      "step": 1417
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 7.07070707070707e-05,
+      "loss": 0.0,
+      "step": 1418
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 7.061524334251607e-05,
+      "loss": 0.0,
+      "step": 1419
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 7.052341597796144e-05,
+      "loss": 0.0,
+      "step": 1420
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 7.04315886134068e-05,
+      "loss": 0.0,
+      "step": 1421
+    },
+    {
+      "epoch": 13.05,
+      "learning_rate": 7.033976124885217e-05,
+      "loss": 0.0,
+      "step": 1422
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 7.024793388429752e-05,
+      "loss": 0.0,
+      "step": 1423
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 7.015610651974288e-05,
+      "loss": 0.0,
+      "step": 1424
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 7.006427915518825e-05,
+      "loss": 0.0,
+      "step": 1425
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 6.99724517906336e-05,
+      "loss": 0.0,
+      "step": 1426
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 6.988062442607897e-05,
+      "loss": 0.0,
+      "step": 1427
+    },
+    {
+      "epoch": 13.1,
+      "learning_rate": 6.978879706152433e-05,
+      "loss": 0.0,
+      "step": 1428
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 6.96969696969697e-05,
+      "loss": 0.0,
+      "step": 1429
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 6.960514233241507e-05,
+      "loss": 0.0,
+      "step": 1430
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 6.951331496786042e-05,
+      "loss": 0.0,
+      "step": 1431
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 6.94214876033058e-05,
+      "loss": 0.0,
+      "step": 1432
+    },
+    {
+      "epoch": 13.15,
+      "learning_rate": 6.932966023875115e-05,
+      "loss": 0.0,
+      "step": 1433
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 6.923783287419652e-05,
+      "loss": 0.0,
+      "step": 1434
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 6.914600550964187e-05,
+      "loss": 0.0,
+      "step": 1435
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 6.905417814508723e-05,
+      "loss": 0.0,
+      "step": 1436
+    },
+    {
+      "epoch": 13.18,
+      "learning_rate": 6.89623507805326e-05,
+      "loss": 0.0,
+      "step": 1437
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 6.887052341597796e-05,
+      "loss": 0.0,
+      "step": 1438
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 6.877869605142332e-05,
+      "loss": 0.0,
+      "step": 1439
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 6.86868686868687e-05,
+      "loss": 0.0,
+      "step": 1440
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 6.859504132231406e-05,
+      "loss": 0.0,
+      "step": 1441
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 6.850321395775942e-05,
+      "loss": 0.0,
+      "step": 1442
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 6.841138659320477e-05,
+      "loss": 0.0,
+      "step": 1443
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 6.831955922865014e-05,
+      "loss": 0.0,
+      "step": 1444
+    },
+    {
+      "epoch": 13.26,
+      "learning_rate": 6.82277318640955e-05,
+      "loss": 0.0,
+      "step": 1445
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 6.813590449954087e-05,
+      "loss": 0.0,
+      "step": 1446
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 6.804407713498622e-05,
+      "loss": 0.0,
+      "step": 1447
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 6.795224977043158e-05,
+      "loss": 0.0,
+      "step": 1448
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 6.786042240587695e-05,
+      "loss": 0.0,
+      "step": 1449
+    },
+    {
+      "epoch": 13.3,
+      "learning_rate": 6.776859504132232e-05,
+      "loss": 0.0,
+      "step": 1450
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 6.767676767676769e-05,
+      "loss": 0.0,
+      "step": 1451
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 6.758494031221304e-05,
+      "loss": 0.0,
+      "step": 1452
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 6.749311294765841e-05,
+      "loss": 0.0,
+      "step": 1453
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 6.740128558310377e-05,
+      "loss": 0.0,
+      "step": 1454
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 6.730945821854913e-05,
+      "loss": 0.0,
+      "step": 1455
+    },
+    {
+      "epoch": 13.36,
+      "learning_rate": 6.72176308539945e-05,
+      "loss": 0.0,
+      "step": 1456
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 6.712580348943985e-05,
+      "loss": 0.0,
+      "step": 1457
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 6.703397612488522e-05,
+      "loss": 0.0,
+      "step": 1458
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 6.694214876033058e-05,
+      "loss": 0.0,
+      "step": 1459
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 6.685032139577594e-05,
+      "loss": 0.0,
+      "step": 1460
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 6.675849403122131e-05,
+      "loss": 0.0,
+      "step": 1461
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.0,
+      "step": 1462
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 6.657483930211204e-05,
+      "loss": 0.0,
+      "step": 1463
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 6.64830119375574e-05,
+      "loss": 0.0,
+      "step": 1464
+    },
+    {
+      "epoch": 13.44,
+      "learning_rate": 6.639118457300276e-05,
+      "loss": 0.0,
+      "step": 1465
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 6.629935720844812e-05,
+      "loss": 0.0,
+      "step": 1466
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 6.620752984389348e-05,
+      "loss": 0.0,
+      "step": 1467
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 6.611570247933885e-05,
+      "loss": 0.0,
+      "step": 1468
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 6.60238751147842e-05,
+      "loss": 0.0,
+      "step": 1469
+    },
+    {
+      "epoch": 13.49,
+      "learning_rate": 6.593204775022957e-05,
+      "loss": 0.0,
+      "step": 1470
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 6.584022038567494e-05,
+      "loss": 0.0,
+      "step": 1471
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 6.574839302112031e-05,
+      "loss": 0.0,
+      "step": 1472
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 6.565656565656566e-05,
+      "loss": 0.0,
+      "step": 1473
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 6.556473829201102e-05,
+      "loss": 0.0,
+      "step": 1474
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 6.547291092745639e-05,
+      "loss": 0.0,
+      "step": 1475
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 6.538108356290175e-05,
+      "loss": 0.0,
+      "step": 1476
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 6.528925619834711e-05,
+      "loss": 0.0,
+      "step": 1477
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 6.519742883379247e-05,
+      "loss": 0.0,
+      "step": 1478
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 6.510560146923783e-05,
+      "loss": 0.0,
+      "step": 1479
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 6.50137741046832e-05,
+      "loss": 0.0,
+      "step": 1480
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 6.492194674012857e-05,
+      "loss": 0.0,
+      "step": 1481
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 6.483011937557393e-05,
+      "loss": 0.0,
+      "step": 1482
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 6.473829201101929e-05,
+      "loss": 0.0,
+      "step": 1483
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 6.464646464646466e-05,
+      "loss": 0.0,
+      "step": 1484
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 6.455463728191002e-05,
+      "loss": 0.0,
+      "step": 1485
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 6.446280991735537e-05,
+      "loss": 0.0,
+      "step": 1486
+    },
+    {
+      "epoch": 13.64,
+      "learning_rate": 6.437098255280074e-05,
+      "loss": 0.0,
+      "step": 1487
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 6.42791551882461e-05,
+      "loss": 0.0,
+      "step": 1488
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 6.418732782369147e-05,
+      "loss": 0.0,
+      "step": 1489
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 6.409550045913682e-05,
+      "loss": 0.0,
+      "step": 1490
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 6.400367309458219e-05,
+      "loss": 0.0,
+      "step": 1491
+    },
+    {
+      "epoch": 13.69,
+      "learning_rate": 6.391184573002756e-05,
+      "loss": 0.0,
+      "step": 1492
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 6.382001836547292e-05,
+      "loss": 0.0,
+      "step": 1493
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 6.372819100091828e-05,
+      "loss": 0.0,
+      "step": 1494
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 0.0,
+      "step": 1495
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 6.354453627180901e-05,
+      "loss": 0.0,
+      "step": 1496
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 6.345270890725437e-05,
+      "loss": 0.0,
+      "step": 1497
+    },
+    {
+      "epoch": 13.74,
+      "learning_rate": 6.336088154269972e-05,
+      "loss": 0.0,
+      "step": 1498
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 6.326905417814509e-05,
+      "loss": 0.0,
+      "step": 1499
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 6.317722681359045e-05,
+      "loss": 0.0,
+      "step": 1500
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 6.308539944903582e-05,
+      "loss": 0.0,
+      "step": 1501
+    },
+    {
+      "epoch": 13.78,
+      "learning_rate": 6.299357208448117e-05,
+      "loss": 0.0,
+      "step": 1502
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 6.290174471992654e-05,
+      "loss": 0.0,
+      "step": 1503
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 6.280991735537191e-05,
+      "loss": 0.0,
+      "step": 1504
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 6.271808999081727e-05,
+      "loss": 0.0,
+      "step": 1505
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 6.262626262626264e-05,
+      "loss": 0.0,
+      "step": 1506
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 6.253443526170799e-05,
+      "loss": 0.0,
+      "step": 1507
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 6.244260789715336e-05,
+      "loss": 0.0,
+      "step": 1508
+    },
+    {
+      "epoch": 13.84,
+      "learning_rate": 6.235078053259872e-05,
+      "loss": 0.0,
+      "step": 1509
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 6.225895316804407e-05,
+      "loss": 0.0,
+      "step": 1510
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 6.216712580348944e-05,
+      "loss": 0.0,
+      "step": 1511
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 6.20752984389348e-05,
+      "loss": 0.0,
+      "step": 1512
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 6.198347107438017e-05,
+      "loss": 0.0,
+      "step": 1513
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 6.189164370982554e-05,
+      "loss": 0.0,
+      "step": 1514
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 6.179981634527089e-05,
+      "loss": 0.0,
+      "step": 1515
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 6.170798898071626e-05,
+      "loss": 0.0,
+      "step": 1516
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 6.161616161616162e-05,
+      "loss": 0.0,
+      "step": 1517
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 6.152433425160699e-05,
+      "loss": 0.0,
+      "step": 1518
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 6.143250688705234e-05,
+      "loss": 0.0,
+      "step": 1519
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 6.134067952249771e-05,
+      "loss": 0.0,
+      "step": 1520
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 6.124885215794307e-05,
+      "loss": 0.0,
+      "step": 1521
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 6.115702479338842e-05,
+      "loss": 0.0,
+      "step": 1522
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 6.106519742883379e-05,
+      "loss": 0.0,
+      "step": 1523
+    },
+    {
+      "epoch": 13.98,
+      "learning_rate": 6.0973370064279154e-05,
+      "loss": 0.0,
+      "step": 1524
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 6.0881542699724524e-05,
+      "loss": 0.0,
+      "step": 1525
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 6.0789715335169886e-05,
+      "loss": 0.0,
+      "step": 1526
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.8170731707317073,
+      "eval_loss": 1.3480682373046875,
+      "eval_runtime": 13.9431,
+      "eval_samples_per_second": 11.762,
+      "eval_steps_per_second": 1.506,
+      "step": 1526
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 6.069788797061524e-05,
+      "loss": 0.0,
+      "step": 1527
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 0.0,
+      "step": 1528
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 6.051423324150597e-05,
+      "loss": 0.0,
+      "step": 1529
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 6.0422405876951337e-05,
+      "loss": 0.0,
+      "step": 1530
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 6.03305785123967e-05,
+      "loss": 0.0,
+      "step": 1531
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 6.023875114784207e-05,
+      "loss": 0.0,
+      "step": 1532
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 6.0146923783287424e-05,
+      "loss": 0.0,
+      "step": 1533
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 6.005509641873278e-05,
+      "loss": 0.0,
+      "step": 1534
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 5.996326905417815e-05,
+      "loss": 0.0,
+      "step": 1535
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 5.987144168962351e-05,
+      "loss": 0.0,
+      "step": 1536
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 5.977961432506888e-05,
+      "loss": 0.002,
+      "step": 1537
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 5.968778696051424e-05,
+      "loss": 0.0,
+      "step": 1538
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 5.959595959595959e-05,
+      "loss": 0.0,
+      "step": 1539
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 5.950413223140496e-05,
+      "loss": 0.0,
+      "step": 1540
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 5.9412304866850325e-05,
+      "loss": 0.0,
+      "step": 1541
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 5.932047750229569e-05,
+      "loss": 0.0,
+      "step": 1542
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 5.922865013774105e-05,
+      "loss": 0.0,
+      "step": 1543
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 5.913682277318642e-05,
+      "loss": 0.0,
+      "step": 1544
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 5.9044995408631775e-05,
+      "loss": 0.0,
+      "step": 1545
+    },
+    {
+      "epoch": 14.18,
+      "learning_rate": 5.895316804407714e-05,
+      "loss": 0.0,
+      "step": 1546
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 5.88613406795225e-05,
+      "loss": 0.0,
+      "step": 1547
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 5.876951331496786e-05,
+      "loss": 0.0,
+      "step": 1548
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 5.867768595041323e-05,
+      "loss": 0.0,
+      "step": 1549
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 5.858585858585859e-05,
+      "loss": 0.0,
+      "step": 1550
+    },
+    {
+      "epoch": 14.23,
+      "learning_rate": 5.849403122130394e-05,
+      "loss": 0.0,
+      "step": 1551
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 5.840220385674931e-05,
+      "loss": 0.0,
+      "step": 1552
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 5.8310376492194675e-05,
+      "loss": 0.0,
+      "step": 1553
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 5.8218549127640044e-05,
+      "loss": 0.0,
+      "step": 1554
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 5.81267217630854e-05,
+      "loss": 0.0,
+      "step": 1555
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 5.803489439853077e-05,
+      "loss": 0.0,
+      "step": 1556
+    },
+    {
+      "epoch": 14.28,
+      "learning_rate": 5.7943067033976125e-05,
+      "loss": 0.0,
+      "step": 1557
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 5.785123966942149e-05,
+      "loss": 0.0,
+      "step": 1558
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 5.775941230486686e-05,
+      "loss": 0.0,
+      "step": 1559
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 5.766758494031221e-05,
+      "loss": 0.0,
+      "step": 1560
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 5.757575757575758e-05,
+      "loss": 0.0,
+      "step": 1561
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 5.748393021120294e-05,
+      "loss": 0.0,
+      "step": 1562
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 5.73921028466483e-05,
+      "loss": 0.0,
+      "step": 1563
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 5.730027548209367e-05,
+      "loss": 0.0,
+      "step": 1564
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 5.7208448117539026e-05,
+      "loss": 0.0,
+      "step": 1565
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 5.7116620752984395e-05,
+      "loss": 0.0,
+      "step": 1566
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 5.702479338842975e-05,
+      "loss": 0.0,
+      "step": 1567
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 5.693296602387512e-05,
+      "loss": 0.0,
+      "step": 1568
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 5.684113865932048e-05,
+      "loss": 0.0,
+      "step": 1569
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 5.674931129476584e-05,
+      "loss": 0.0,
+      "step": 1570
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 5.665748393021121e-05,
+      "loss": 0.0,
+      "step": 1571
+    },
+    {
+      "epoch": 14.42,
+      "learning_rate": 5.6565656565656563e-05,
+      "loss": 0.0001,
+      "step": 1572
+    },
+    {
+      "epoch": 14.43,
+      "learning_rate": 5.647382920110193e-05,
+      "loss": 0.0,
+      "step": 1573
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 5.6382001836547295e-05,
+      "loss": 0.0,
+      "step": 1574
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 5.629017447199265e-05,
+      "loss": 0.0,
+      "step": 1575
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 5.619834710743802e-05,
+      "loss": 0.0,
+      "step": 1576
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 5.6106519742883376e-05,
+      "loss": 0.0,
+      "step": 1577
+    },
+    {
+      "epoch": 14.48,
+      "learning_rate": 5.6014692378328745e-05,
+      "loss": 0.0008,
+      "step": 1578
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 5.592286501377411e-05,
+      "loss": 0.0,
+      "step": 1579
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 5.583103764921948e-05,
+      "loss": 0.0,
+      "step": 1580
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 5.573921028466483e-05,
+      "loss": 0.0,
+      "step": 1581
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 5.564738292011019e-05,
+      "loss": 0.0,
+      "step": 1582
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.0,
+      "step": 1583
+    },
+    {
+      "epoch": 14.53,
+      "learning_rate": 5.546372819100092e-05,
+      "loss": 0.0,
+      "step": 1584
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 5.537190082644629e-05,
+      "loss": 0.0,
+      "step": 1585
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 5.5280073461891646e-05,
+      "loss": 0.0,
+      "step": 1586
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 5.5188246097337e-05,
+      "loss": 0.0,
+      "step": 1587
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 5.509641873278237e-05,
+      "loss": 0.0,
+      "step": 1588
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 5.5004591368227733e-05,
+      "loss": 0.0,
+      "step": 1589
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 5.49127640036731e-05,
+      "loss": 0.0,
+      "step": 1590
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 5.482093663911846e-05,
+      "loss": 0.0,
+      "step": 1591
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 5.472910927456383e-05,
+      "loss": 0.0,
+      "step": 1592
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 5.4637281910009184e-05,
+      "loss": 0.0,
+      "step": 1593
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 0.0,
+      "step": 1594
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 5.4453627180899916e-05,
+      "loss": 0.0,
+      "step": 1595
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 5.436179981634527e-05,
+      "loss": 0.0,
+      "step": 1596
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 5.426997245179064e-05,
+      "loss": 0.0,
+      "step": 1597
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 5.4178145087235996e-05,
+      "loss": 0.0,
+      "step": 1598
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 5.408631772268136e-05,
+      "loss": 0.0,
+      "step": 1599
+    },
+    {
+      "epoch": 14.68,
+      "learning_rate": 5.399449035812673e-05,
+      "loss": 0.0,
+      "step": 1600
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 5.3902662993572084e-05,
+      "loss": 0.0,
+      "step": 1601
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 5.381083562901745e-05,
+      "loss": 0.0,
+      "step": 1602
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 5.371900826446281e-05,
+      "loss": 0.0,
+      "step": 1603
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 5.362718089990818e-05,
+      "loss": 0.0,
+      "step": 1604
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 5.353535353535354e-05,
+      "loss": 0.0,
+      "step": 1605
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 5.34435261707989e-05,
+      "loss": 0.0,
+      "step": 1606
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 5.3351698806244266e-05,
+      "loss": 0.0,
+      "step": 1607
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 5.325987144168962e-05,
+      "loss": 0.0,
+      "step": 1608
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 5.316804407713499e-05,
+      "loss": 0.0,
+      "step": 1609
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 5.3076216712580354e-05,
+      "loss": 0.0,
+      "step": 1610
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 5.298438934802571e-05,
+      "loss": 0.0,
+      "step": 1611
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 5.289256198347108e-05,
+      "loss": 0.0,
+      "step": 1612
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 5.2800734618916435e-05,
+      "loss": 0.0,
+      "step": 1613
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 5.2708907254361804e-05,
+      "loss": 0.0,
+      "step": 1614
+    },
+    {
+      "epoch": 14.82,
+      "learning_rate": 5.2617079889807166e-05,
+      "loss": 0.0,
+      "step": 1615
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 5.2525252525252536e-05,
+      "loss": 0.0,
+      "step": 1616
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 5.243342516069789e-05,
+      "loss": 0.0,
+      "step": 1617
+    },
+    {
+      "epoch": 14.84,
+      "learning_rate": 5.234159779614325e-05,
+      "loss": 0.0,
+      "step": 1618
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 5.224977043158862e-05,
+      "loss": 0.0,
+      "step": 1619
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 5.215794306703398e-05,
+      "loss": 0.0,
+      "step": 1620
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 5.206611570247935e-05,
+      "loss": 0.0,
+      "step": 1621
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 5.1974288337924704e-05,
+      "loss": 0.0,
+      "step": 1622
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 5.188246097337006e-05,
+      "loss": 0.0,
+      "step": 1623
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 5.179063360881543e-05,
+      "loss": 0.0,
+      "step": 1624
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 5.169880624426079e-05,
+      "loss": 0.0,
+      "step": 1625
+    },
+    {
+      "epoch": 14.92,
+      "learning_rate": 5.160697887970616e-05,
+      "loss": 0.0,
+      "step": 1626
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 5.151515151515152e-05,
+      "loss": 0.0,
+      "step": 1627
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 5.1423324150596886e-05,
+      "loss": 0.0,
+      "step": 1628
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 5.133149678604224e-05,
+      "loss": 0.0,
+      "step": 1629
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 5.1239669421487605e-05,
+      "loss": 0.0,
+      "step": 1630
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 5.1147842056932974e-05,
+      "loss": 0.0,
+      "step": 1631
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 5.105601469237833e-05,
+      "loss": 0.0,
+      "step": 1632
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 5.09641873278237e-05,
+      "loss": 0.0,
+      "step": 1633
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 5.0872359963269055e-05,
+      "loss": 0.0,
+      "step": 1634
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 5.078053259871442e-05,
+      "loss": 0.0,
+      "step": 1635
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.8414634146341463,
+      "eval_loss": 1.36283540725708,
+      "eval_runtime": 13.9446,
+      "eval_samples_per_second": 11.761,
+      "eval_steps_per_second": 1.506,
+      "step": 1635
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 5.068870523415979e-05,
+      "loss": 0.0,
+      "step": 1636
+    },
+    {
+      "epoch": 15.02,
+      "learning_rate": 5.059687786960514e-05,
+      "loss": 0.0,
+      "step": 1637
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 5.050505050505051e-05,
+      "loss": 0.0,
+      "step": 1638
+    },
+    {
+      "epoch": 15.04,
+      "learning_rate": 5.041322314049587e-05,
+      "loss": 0.0,
+      "step": 1639
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 5.032139577594124e-05,
+      "loss": 0.0,
+      "step": 1640
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 5.02295684113866e-05,
+      "loss": 0.0,
+      "step": 1641
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 5.0137741046831955e-05,
+      "loss": 0.0,
+      "step": 1642
+    },
+    {
+      "epoch": 15.07,
+      "learning_rate": 5.0045913682277325e-05,
+      "loss": 0.0,
+      "step": 1643
+    },
+    {
+      "epoch": 15.08,
+      "learning_rate": 4.995408631772268e-05,
+      "loss": 0.0,
+      "step": 1644
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 4.986225895316804e-05,
+      "loss": 0.0,
+      "step": 1645
+    },
+    {
+      "epoch": 15.1,
+      "learning_rate": 4.9770431588613405e-05,
+      "loss": 0.0,
+      "step": 1646
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 4.9678604224058775e-05,
+      "loss": 0.0,
+      "step": 1647
+    },
+    {
+      "epoch": 15.12,
+      "learning_rate": 4.958677685950414e-05,
+      "loss": 0.0,
+      "step": 1648
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 4.94949494949495e-05,
+      "loss": 0.0,
+      "step": 1649
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 4.9403122130394856e-05,
+      "loss": 0.0,
+      "step": 1650
+    },
+    {
+      "epoch": 15.15,
+      "learning_rate": 4.931129476584022e-05,
+      "loss": 0.0,
+      "step": 1651
+    },
+    {
+      "epoch": 15.16,
+      "learning_rate": 4.921946740128559e-05,
+      "loss": 0.0,
+      "step": 1652
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 4.912764003673095e-05,
+      "loss": 0.0,
+      "step": 1653
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 4.903581267217631e-05,
+      "loss": 0.0,
+      "step": 1654
+    },
+    {
+      "epoch": 15.18,
+      "learning_rate": 4.8943985307621675e-05,
+      "loss": 0.0,
+      "step": 1655
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 4.885215794306703e-05,
+      "loss": 0.0,
+      "step": 1656
+    },
+    {
+      "epoch": 15.2,
+      "learning_rate": 4.87603305785124e-05,
+      "loss": 0.0,
+      "step": 1657
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 4.866850321395776e-05,
+      "loss": 0.0,
+      "step": 1658
+    },
+    {
+      "epoch": 15.22,
+      "learning_rate": 4.8576675849403125e-05,
+      "loss": 0.0,
+      "step": 1659
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.0,
+      "step": 1660
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 4.839302112029385e-05,
+      "loss": 0.0,
+      "step": 1661
+    },
+    {
+      "epoch": 15.25,
+      "learning_rate": 4.830119375573921e-05,
+      "loss": 0.0,
+      "step": 1662
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 4.8209366391184575e-05,
+      "loss": 0.0,
+      "step": 1663
+    },
+    {
+      "epoch": 15.27,
+      "learning_rate": 4.811753902662994e-05,
+      "loss": 0.0,
+      "step": 1664
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 4.80257116620753e-05,
+      "loss": 0.0,
+      "step": 1665
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 4.793388429752066e-05,
+      "loss": 0.0,
+      "step": 1666
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 4.7842056932966026e-05,
+      "loss": 0.0,
+      "step": 1667
+    },
+    {
+      "epoch": 15.3,
+      "learning_rate": 4.775022956841139e-05,
+      "loss": 0.0,
+      "step": 1668
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 4.765840220385675e-05,
+      "loss": 0.0,
+      "step": 1669
+    },
+    {
+      "epoch": 15.32,
+      "learning_rate": 4.756657483930211e-05,
+      "loss": 0.0,
+      "step": 1670
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 4.7474747474747476e-05,
+      "loss": 0.0,
+      "step": 1671
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 4.738292011019284e-05,
+      "loss": 0.0,
+      "step": 1672
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 4.729109274563821e-05,
+      "loss": 0.0,
+      "step": 1673
+    },
+    {
+      "epoch": 15.36,
+      "learning_rate": 4.7199265381083563e-05,
+      "loss": 0.0,
+      "step": 1674
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 4.7107438016528926e-05,
+      "loss": 0.0,
+      "step": 1675
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 4.701561065197429e-05,
+      "loss": 0.0,
+      "step": 1676
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 4.692378328741965e-05,
+      "loss": 0.0,
+      "step": 1677
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 4.683195592286502e-05,
+      "loss": 0.0,
+      "step": 1678
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 4.674012855831038e-05,
+      "loss": 0.0,
+      "step": 1679
+    },
+    {
+      "epoch": 15.41,
+      "learning_rate": 4.664830119375574e-05,
+      "loss": 0.0,
+      "step": 1680
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 4.65564738292011e-05,
+      "loss": 0.0,
+      "step": 1681
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 4.6464646464646464e-05,
+      "loss": 0.0,
+      "step": 1682
+    },
+    {
+      "epoch": 15.44,
+      "learning_rate": 4.637281910009183e-05,
+      "loss": 0.0,
+      "step": 1683
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 4.6280991735537196e-05,
+      "loss": 0.0,
+      "step": 1684
+    },
+    {
+      "epoch": 15.46,
+      "learning_rate": 4.618916437098256e-05,
+      "loss": 0.0,
+      "step": 1685
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 4.6097337006427914e-05,
+      "loss": 0.0,
+      "step": 1686
+    },
+    {
+      "epoch": 15.48,
+      "learning_rate": 4.6005509641873277e-05,
+      "loss": 0.0,
+      "step": 1687
+    },
+    {
+      "epoch": 15.49,
+      "learning_rate": 4.5913682277318646e-05,
+      "loss": 0.0,
+      "step": 1688
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 4.582185491276401e-05,
+      "loss": 0.0,
+      "step": 1689
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 4.573002754820937e-05,
+      "loss": 0.0,
+      "step": 1690
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 4.5638200183654733e-05,
+      "loss": 0.0,
+      "step": 1691
+    },
+    {
+      "epoch": 15.52,
+      "learning_rate": 4.554637281910009e-05,
+      "loss": 0.0,
+      "step": 1692
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.0,
+      "step": 1693
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 4.536271808999082e-05,
+      "loss": 0.0,
+      "step": 1694
+    },
+    {
+      "epoch": 15.55,
+      "learning_rate": 4.5270890725436184e-05,
+      "loss": 0.0,
+      "step": 1695
+    },
+    {
+      "epoch": 15.56,
+      "learning_rate": 4.5179063360881546e-05,
+      "loss": 0.0,
+      "step": 1696
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 4.508723599632691e-05,
+      "loss": 0.0,
+      "step": 1697
+    },
+    {
+      "epoch": 15.58,
+      "learning_rate": 4.4995408631772265e-05,
+      "loss": 0.0,
+      "step": 1698
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 4.4903581267217634e-05,
+      "loss": 0.0,
+      "step": 1699
+    },
+    {
+      "epoch": 15.6,
+      "learning_rate": 4.4811753902662996e-05,
+      "loss": 0.0,
+      "step": 1700
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 4.471992653810836e-05,
+      "loss": 0.0,
+      "step": 1701
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 4.462809917355372e-05,
+      "loss": 0.0,
+      "step": 1702
+    },
+    {
+      "epoch": 15.62,
+      "learning_rate": 4.4536271808999084e-05,
+      "loss": 0.0,
+      "step": 1703
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.0,
+      "step": 1704
+    },
+    {
+      "epoch": 15.64,
+      "learning_rate": 4.435261707988981e-05,
+      "loss": 0.0,
+      "step": 1705
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 4.426078971533517e-05,
+      "loss": 0.0,
+      "step": 1706
+    },
+    {
+      "epoch": 15.66,
+      "learning_rate": 4.4168962350780534e-05,
+      "loss": 0.0,
+      "step": 1707
+    },
+    {
+      "epoch": 15.67,
+      "learning_rate": 4.40771349862259e-05,
+      "loss": 0.0,
+      "step": 1708
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 4.398530762167126e-05,
+      "loss": 0.0,
+      "step": 1709
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 4.389348025711662e-05,
+      "loss": 0.0,
+      "step": 1710
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 4.3801652892561984e-05,
+      "loss": 0.0,
+      "step": 1711
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 4.370982552800735e-05,
+      "loss": 0.0,
+      "step": 1712
+    },
+    {
+      "epoch": 15.72,
+      "learning_rate": 4.361799816345271e-05,
+      "loss": 0.0,
+      "step": 1713
+    },
+    {
+      "epoch": 15.72,
+      "learning_rate": 4.352617079889807e-05,
+      "loss": 0.0,
+      "step": 1714
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 4.343434343434344e-05,
+      "loss": 0.0,
+      "step": 1715
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 4.33425160697888e-05,
+      "loss": 0.0,
+      "step": 1716
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 4.325068870523416e-05,
+      "loss": 0.0,
+      "step": 1717
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 4.315886134067952e-05,
+      "loss": 0.0,
+      "step": 1718
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 4.3067033976124885e-05,
+      "loss": 0.0,
+      "step": 1719
+    },
+    {
+      "epoch": 15.78,
+      "learning_rate": 4.2975206611570254e-05,
+      "loss": 0.0,
+      "step": 1720
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 4.288337924701562e-05,
+      "loss": 0.0,
+      "step": 1721
+    },
+    {
+      "epoch": 15.8,
+      "learning_rate": 4.279155188246097e-05,
+      "loss": 0.0,
+      "step": 1722
+    },
+    {
+      "epoch": 15.81,
+      "learning_rate": 4.2699724517906335e-05,
+      "loss": 0.0,
+      "step": 1723
+    },
+    {
+      "epoch": 15.82,
+      "learning_rate": 4.26078971533517e-05,
+      "loss": 0.0,
+      "step": 1724
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 4.251606978879707e-05,
+      "loss": 0.0,
+      "step": 1725
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.0,
+      "step": 1726
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 4.233241505968779e-05,
+      "loss": 0.0,
+      "step": 1727
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 4.224058769513315e-05,
+      "loss": 0.0,
+      "step": 1728
+    },
+    {
+      "epoch": 15.86,
+      "learning_rate": 4.214876033057851e-05,
+      "loss": 0.0,
+      "step": 1729
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 4.205693296602388e-05,
+      "loss": 0.0,
+      "step": 1730
+    },
+    {
+      "epoch": 15.88,
+      "learning_rate": 4.196510560146924e-05,
+      "loss": 0.0,
+      "step": 1731
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 4.1873278236914605e-05,
+      "loss": 0.0,
+      "step": 1732
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 4.178145087235997e-05,
+      "loss": 0.0,
+      "step": 1733
+    },
+    {
+      "epoch": 15.91,
+      "learning_rate": 4.168962350780532e-05,
+      "loss": 0.0,
+      "step": 1734
+    },
+    {
+      "epoch": 15.92,
+      "learning_rate": 4.159779614325069e-05,
+      "loss": 0.0,
+      "step": 1735
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 4.1505968778696055e-05,
+      "loss": 0.0,
+      "step": 1736
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 4.141414141414142e-05,
+      "loss": 0.0,
+      "step": 1737
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 4.132231404958678e-05,
+      "loss": 0.0,
+      "step": 1738
+    },
+    {
+      "epoch": 15.95,
+      "learning_rate": 4.123048668503214e-05,
+      "loss": 0.0,
+      "step": 1739
+    },
+    {
+      "epoch": 15.96,
+      "learning_rate": 4.1138659320477505e-05,
+      "loss": 0.0,
+      "step": 1740
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 4.104683195592287e-05,
+      "loss": 0.0,
+      "step": 1741
+    },
+    {
+      "epoch": 15.98,
+      "learning_rate": 4.095500459136823e-05,
+      "loss": 0.0,
+      "step": 1742
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 4.086317722681359e-05,
+      "loss": 0.0,
+      "step": 1743
+    },
+    {
+      "epoch": 16.0,
+      "learning_rate": 4.0771349862258955e-05,
+      "loss": 0.0,
+      "step": 1744
+    },
+    {
+      "epoch": 16.0,
+      "eval_accuracy": 0.8475609756097561,
+      "eval_loss": 1.3643109798431396,
+      "eval_runtime": 13.9466,
+      "eval_samples_per_second": 11.759,
+      "eval_steps_per_second": 1.506,
+      "step": 1744
+    },
+    {
+      "epoch": 16.01,
+      "learning_rate": 4.067952249770432e-05,
+      "loss": 0.0,
+      "step": 1745
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 4.058769513314968e-05,
+      "loss": 0.0,
+      "step": 1746
+    },
+    {
+      "epoch": 16.03,
+      "learning_rate": 4.049586776859504e-05,
+      "loss": 0.0,
+      "step": 1747
+    },
+    {
+      "epoch": 16.04,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 0.0,
+      "step": 1748
+    },
+    {
+      "epoch": 16.05,
+      "learning_rate": 4.031221303948577e-05,
+      "loss": 0.0,
+      "step": 1749
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 4.022038567493113e-05,
+      "loss": 0.0,
+      "step": 1750
+    },
+    {
+      "epoch": 16.06,
+      "learning_rate": 4.012855831037649e-05,
+      "loss": 0.0,
+      "step": 1751
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 4.0036730945821856e-05,
+      "loss": 0.0,
+      "step": 1752
+    },
+    {
+      "epoch": 16.08,
+      "learning_rate": 3.994490358126722e-05,
+      "loss": 0.0,
+      "step": 1753
+    },
+    {
+      "epoch": 16.09,
+      "learning_rate": 3.985307621671258e-05,
+      "loss": 0.0,
+      "step": 1754
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 3.976124885215794e-05,
+      "loss": 0.0,
+      "step": 1755
+    },
+    {
+      "epoch": 16.11,
+      "learning_rate": 3.9669421487603306e-05,
+      "loss": 0.0,
+      "step": 1756
+    },
+    {
+      "epoch": 16.12,
+      "learning_rate": 3.9577594123048675e-05,
+      "loss": 0.0,
+      "step": 1757
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 3.948576675849403e-05,
+      "loss": 0.0,
+      "step": 1758
+    },
+    {
+      "epoch": 16.14,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.0,
+      "step": 1759
+    },
+    {
+      "epoch": 16.15,
+      "learning_rate": 3.9302112029384756e-05,
+      "loss": 0.0,
+      "step": 1760
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 3.921028466483012e-05,
+      "loss": 0.0,
+      "step": 1761
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 3.911845730027549e-05,
+      "loss": 0.0,
+      "step": 1762
+    },
+    {
+      "epoch": 16.17,
+      "learning_rate": 3.902662993572085e-05,
+      "loss": 0.0,
+      "step": 1763
+    },
+    {
+      "epoch": 16.18,
+      "learning_rate": 3.8934802571166206e-05,
+      "loss": 0.0,
+      "step": 1764
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 3.884297520661157e-05,
+      "loss": 0.0,
+      "step": 1765
+    },
+    {
+      "epoch": 16.2,
+      "learning_rate": 3.875114784205693e-05,
+      "loss": 0.0,
+      "step": 1766
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 3.86593204775023e-05,
+      "loss": 0.0,
+      "step": 1767
+    },
+    {
+      "epoch": 16.22,
+      "learning_rate": 3.856749311294766e-05,
+      "loss": 0.0,
+      "step": 1768
+    },
+    {
+      "epoch": 16.23,
+      "learning_rate": 3.8475665748393026e-05,
+      "loss": 0.0,
+      "step": 1769
+    },
+    {
+      "epoch": 16.24,
+      "learning_rate": 3.838383838383838e-05,
+      "loss": 0.0,
+      "step": 1770
+    },
+    {
+      "epoch": 16.25,
+      "learning_rate": 3.8292011019283744e-05,
+      "loss": 0.0,
+      "step": 1771
+    },
+    {
+      "epoch": 16.26,
+      "learning_rate": 3.820018365472911e-05,
+      "loss": 0.0,
+      "step": 1772
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 3.8108356290174476e-05,
+      "loss": 0.0,
+      "step": 1773
+    },
+    {
+      "epoch": 16.28,
+      "learning_rate": 3.801652892561984e-05,
+      "loss": 0.0,
+      "step": 1774
+    },
+    {
+      "epoch": 16.28,
+      "learning_rate": 3.79247015610652e-05,
+      "loss": 0.0,
+      "step": 1775
+    },
+    {
+      "epoch": 16.29,
+      "learning_rate": 3.783287419651056e-05,
+      "loss": 0.0,
+      "step": 1776
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 3.7741046831955926e-05,
+      "loss": 0.0,
+      "step": 1777
+    },
+    {
+      "epoch": 16.31,
+      "learning_rate": 3.764921946740129e-05,
+      "loss": 0.0,
+      "step": 1778
+    },
+    {
+      "epoch": 16.32,
+      "learning_rate": 3.755739210284665e-05,
+      "loss": 0.0,
+      "step": 1779
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 3.7465564738292014e-05,
+      "loss": 0.0,
+      "step": 1780
+    },
+    {
+      "epoch": 16.34,
+      "learning_rate": 3.7373737373737376e-05,
+      "loss": 0.0,
+      "step": 1781
+    },
+    {
+      "epoch": 16.35,
+      "learning_rate": 3.728191000918274e-05,
+      "loss": 0.0,
+      "step": 1782
+    },
+    {
+      "epoch": 16.36,
+      "learning_rate": 3.71900826446281e-05,
+      "loss": 0.0,
+      "step": 1783
+    },
+    {
+      "epoch": 16.37,
+      "learning_rate": 3.7098255280073464e-05,
+      "loss": 0.0,
+      "step": 1784
+    },
+    {
+      "epoch": 16.38,
+      "learning_rate": 3.7006427915518826e-05,
+      "loss": 0.0,
+      "step": 1785
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 3.691460055096419e-05,
+      "loss": 0.0,
+      "step": 1786
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 3.682277318640955e-05,
+      "loss": 0.0,
+      "step": 1787
+    },
+    {
+      "epoch": 16.4,
+      "learning_rate": 3.6730945821854914e-05,
+      "loss": 0.0,
+      "step": 1788
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 3.6639118457300277e-05,
+      "loss": 0.0,
+      "step": 1789
+    },
+    {
+      "epoch": 16.42,
+      "learning_rate": 3.654729109274564e-05,
+      "loss": 0.0,
+      "step": 1790
+    },
+    {
+      "epoch": 16.43,
+      "learning_rate": 3.6455463728191e-05,
+      "loss": 0.0,
+      "step": 1791
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.0,
+      "step": 1792
+    },
+    {
+      "epoch": 16.45,
+      "learning_rate": 3.627180899908173e-05,
+      "loss": 0.0,
+      "step": 1793
+    },
+    {
+      "epoch": 16.46,
+      "learning_rate": 3.617998163452709e-05,
+      "loss": 0.0,
+      "step": 1794
+    },
+    {
+      "epoch": 16.47,
+      "learning_rate": 3.608815426997245e-05,
+      "loss": 0.0,
+      "step": 1795
+    },
+    {
+      "epoch": 16.48,
+      "learning_rate": 3.5996326905417814e-05,
+      "loss": 0.0,
+      "step": 1796
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 3.590449954086318e-05,
+      "loss": 0.0,
+      "step": 1797
+    },
+    {
+      "epoch": 16.5,
+      "learning_rate": 3.581267217630854e-05,
+      "loss": 0.0,
+      "step": 1798
+    },
+    {
+      "epoch": 16.5,
+      "learning_rate": 3.572084481175391e-05,
+      "loss": 0.0,
+      "step": 1799
+    },
+    {
+      "epoch": 16.51,
+      "learning_rate": 3.5629017447199265e-05,
+      "loss": 0.0,
+      "step": 1800
+    },
+    {
+      "epoch": 16.52,
+      "learning_rate": 3.553719008264463e-05,
+      "loss": 0.0,
+      "step": 1801
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 3.544536271808999e-05,
+      "loss": 0.0,
+      "step": 1802
+    },
+    {
+      "epoch": 16.54,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 0.0,
+      "step": 1803
+    },
+    {
+      "epoch": 16.55,
+      "learning_rate": 3.526170798898072e-05,
+      "loss": 0.0,
+      "step": 1804
+    },
+    {
+      "epoch": 16.56,
+      "learning_rate": 3.5169880624426084e-05,
+      "loss": 0.0,
+      "step": 1805
+    },
+    {
+      "epoch": 16.57,
+      "learning_rate": 3.507805325987144e-05,
+      "loss": 0.0,
+      "step": 1806
+    },
+    {
+      "epoch": 16.58,
+      "learning_rate": 3.49862258953168e-05,
+      "loss": 0.0,
+      "step": 1807
+    },
+    {
+      "epoch": 16.59,
+      "learning_rate": 3.4894398530762165e-05,
+      "loss": 0.0,
+      "step": 1808
+    },
+    {
+      "epoch": 16.6,
+      "learning_rate": 3.4802571166207534e-05,
+      "loss": 0.0,
+      "step": 1809
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 3.47107438016529e-05,
+      "loss": 0.0,
+      "step": 1810
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 3.461891643709826e-05,
+      "loss": 0.0,
+      "step": 1811
+    },
+    {
+      "epoch": 16.62,
+      "learning_rate": 3.4527089072543615e-05,
+      "loss": 0.0,
+      "step": 1812
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 3.443526170798898e-05,
+      "loss": 0.0,
+      "step": 1813
+    },
+    {
+      "epoch": 16.64,
+      "learning_rate": 3.434343434343435e-05,
+      "loss": 0.0,
+      "step": 1814
+    },
+    {
+      "epoch": 16.65,
+      "learning_rate": 3.425160697887971e-05,
+      "loss": 0.0,
+      "step": 1815
+    },
+    {
+      "epoch": 16.66,
+      "learning_rate": 3.415977961432507e-05,
+      "loss": 0.0,
+      "step": 1816
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 3.4067952249770435e-05,
+      "loss": 0.0,
+      "step": 1817
+    },
+    {
+      "epoch": 16.68,
+      "learning_rate": 3.397612488521579e-05,
+      "loss": 0.0,
+      "step": 1818
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 3.388429752066116e-05,
+      "loss": 0.0,
+      "step": 1819
+    },
+    {
+      "epoch": 16.7,
+      "learning_rate": 3.379247015610652e-05,
+      "loss": 0.0,
+      "step": 1820
+    },
+    {
+      "epoch": 16.71,
+      "learning_rate": 3.3700642791551885e-05,
+      "loss": 0.0,
+      "step": 1821
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 3.360881542699725e-05,
+      "loss": 0.0,
+      "step": 1822
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 3.351698806244261e-05,
+      "loss": 0.0,
+      "step": 1823
+    },
+    {
+      "epoch": 16.73,
+      "learning_rate": 3.342516069788797e-05,
+      "loss": 0.0,
+      "step": 1824
+    },
+    {
+      "epoch": 16.74,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.0,
+      "step": 1825
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 3.32415059687787e-05,
+      "loss": 0.0,
+      "step": 1826
+    },
+    {
+      "epoch": 16.76,
+      "learning_rate": 3.314967860422406e-05,
+      "loss": 0.0,
+      "step": 1827
+    },
+    {
+      "epoch": 16.77,
+      "learning_rate": 3.305785123966942e-05,
+      "loss": 0.0,
+      "step": 1828
+    },
+    {
+      "epoch": 16.78,
+      "learning_rate": 3.2966023875114785e-05,
+      "loss": 0.0,
+      "step": 1829
+    },
+    {
+      "epoch": 16.79,
+      "learning_rate": 3.2874196510560154e-05,
+      "loss": 0.0,
+      "step": 1830
+    },
+    {
+      "epoch": 16.8,
+      "learning_rate": 3.278236914600551e-05,
+      "loss": 0.0,
+      "step": 1831
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 3.269054178145087e-05,
+      "loss": 0.0,
+      "step": 1832
+    },
+    {
+      "epoch": 16.82,
+      "learning_rate": 3.2598714416896235e-05,
+      "loss": 0.0,
+      "step": 1833
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 3.25068870523416e-05,
+      "loss": 0.0,
+      "step": 1834
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 3.241505968778697e-05,
+      "loss": 0.0,
+      "step": 1835
+    },
+    {
+      "epoch": 16.84,
+      "learning_rate": 3.232323232323233e-05,
+      "loss": 0.0,
+      "step": 1836
+    },
+    {
+      "epoch": 16.85,
+      "learning_rate": 3.2231404958677685e-05,
+      "loss": 0.0,
+      "step": 1837
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 3.213957759412305e-05,
+      "loss": 0.0,
+      "step": 1838
+    },
+    {
+      "epoch": 16.87,
+      "learning_rate": 3.204775022956841e-05,
+      "loss": 0.0,
+      "step": 1839
+    },
+    {
+      "epoch": 16.88,
+      "learning_rate": 3.195592286501378e-05,
+      "loss": 0.0,
+      "step": 1840
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 3.186409550045914e-05,
+      "loss": 0.0,
+      "step": 1841
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 3.1772268135904505e-05,
+      "loss": 0.0,
+      "step": 1842
+    },
+    {
+      "epoch": 16.91,
+      "learning_rate": 3.168044077134986e-05,
+      "loss": 0.0,
+      "step": 1843
+    },
+    {
+      "epoch": 16.92,
+      "learning_rate": 3.158861340679522e-05,
+      "loss": 0.0,
+      "step": 1844
+    },
+    {
+      "epoch": 16.93,
+      "learning_rate": 3.1496786042240586e-05,
+      "loss": 0.0,
+      "step": 1845
+    },
+    {
+      "epoch": 16.94,
+      "learning_rate": 3.1404958677685955e-05,
+      "loss": 0.0,
+      "step": 1846
+    },
+    {
+      "epoch": 16.94,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 0.0,
+      "step": 1847
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 3.122130394857668e-05,
+      "loss": 0.0,
+      "step": 1848
+    },
+    {
+      "epoch": 16.96,
+      "learning_rate": 3.1129476584022036e-05,
+      "loss": 0.0,
+      "step": 1849
+    },
+    {
+      "epoch": 16.97,
+      "learning_rate": 3.10376492194674e-05,
+      "loss": 0.0,
+      "step": 1850
+    },
+    {
+      "epoch": 16.98,
+      "learning_rate": 3.094582185491277e-05,
+      "loss": 0.0,
+      "step": 1851
+    },
+    {
+      "epoch": 16.99,
+      "learning_rate": 3.085399449035813e-05,
+      "loss": 0.0,
+      "step": 1852
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 3.076216712580349e-05,
+      "loss": 0.0,
+      "step": 1853
+    },
+    {
+      "epoch": 17.0,
+      "eval_accuracy": 0.8475609756097561,
+      "eval_loss": 1.364932894706726,
+      "eval_runtime": 13.9485,
+      "eval_samples_per_second": 11.758,
+      "eval_steps_per_second": 1.506,
+      "step": 1853
+    },
+    {
+      "epoch": 17.01,
+      "learning_rate": 3.0670339761248856e-05,
+      "loss": 0.0,
+      "step": 1854
+    },
+    {
+      "epoch": 17.02,
+      "learning_rate": 3.057851239669421e-05,
+      "loss": 0.0,
+      "step": 1855
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 3.0486685032139577e-05,
+      "loss": 0.0,
+      "step": 1856
+    },
+    {
+      "epoch": 17.04,
+      "learning_rate": 3.0394857667584943e-05,
+      "loss": 0.0,
+      "step": 1857
+    },
+    {
+      "epoch": 17.05,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.0,
+      "step": 1858
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 3.0211202938475668e-05,
+      "loss": 0.0,
+      "step": 1859
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 3.0119375573921034e-05,
+      "loss": 0.0,
+      "step": 1860
+    },
+    {
+      "epoch": 17.07,
+      "learning_rate": 3.002754820936639e-05,
+      "loss": 0.0,
+      "step": 1861
+    },
+    {
+      "epoch": 17.08,
+      "learning_rate": 2.9935720844811756e-05,
+      "loss": 0.0,
+      "step": 1862
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 2.984389348025712e-05,
+      "loss": 0.0,
+      "step": 1863
+    },
+    {
+      "epoch": 17.1,
+      "learning_rate": 2.975206611570248e-05,
+      "loss": 0.0,
+      "step": 1864
+    },
+    {
+      "epoch": 17.11,
+      "learning_rate": 2.9660238751147844e-05,
+      "loss": 0.0,
+      "step": 1865
+    },
+    {
+      "epoch": 17.12,
+      "learning_rate": 2.956841138659321e-05,
+      "loss": 0.0,
+      "step": 1866
+    },
+    {
+      "epoch": 17.13,
+      "learning_rate": 2.947658402203857e-05,
+      "loss": 0.0,
+      "step": 1867
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 2.938475665748393e-05,
+      "loss": 0.0,
+      "step": 1868
+    },
+    {
+      "epoch": 17.15,
+      "learning_rate": 2.9292929292929294e-05,
+      "loss": 0.0,
+      "step": 1869
+    },
+    {
+      "epoch": 17.16,
+      "learning_rate": 2.9201101928374656e-05,
+      "loss": 0.0,
+      "step": 1870
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 2.9109274563820022e-05,
+      "loss": 0.0,
+      "step": 1871
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 2.9017447199265385e-05,
+      "loss": 0.0,
+      "step": 1872
+    },
+    {
+      "epoch": 17.18,
+      "learning_rate": 2.8925619834710744e-05,
+      "loss": 0.0,
+      "step": 1873
+    },
+    {
+      "epoch": 17.19,
+      "learning_rate": 2.8833792470156106e-05,
+      "loss": 0.0,
+      "step": 1874
+    },
+    {
+      "epoch": 17.2,
+      "learning_rate": 2.874196510560147e-05,
+      "loss": 0.0,
+      "step": 1875
+    },
+    {
+      "epoch": 17.21,
+      "learning_rate": 2.8650137741046835e-05,
+      "loss": 0.0,
+      "step": 1876
+    },
+    {
+      "epoch": 17.22,
+      "learning_rate": 2.8558310376492197e-05,
+      "loss": 0.0,
+      "step": 1877
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 2.846648301193756e-05,
+      "loss": 0.0,
+      "step": 1878
+    },
+    {
+      "epoch": 17.24,
+      "learning_rate": 2.837465564738292e-05,
+      "loss": 0.0,
+      "step": 1879
+    },
+    {
+      "epoch": 17.25,
+      "learning_rate": 2.8282828282828282e-05,
+      "loss": 0.0,
+      "step": 1880
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 2.8191000918273648e-05,
+      "loss": 0.0,
+      "step": 1881
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 2.809917355371901e-05,
+      "loss": 0.0,
+      "step": 1882
+    },
+    {
+      "epoch": 17.28,
+      "learning_rate": 2.8007346189164373e-05,
+      "loss": 0.0,
+      "step": 1883
+    },
+    {
+      "epoch": 17.28,
+      "learning_rate": 2.791551882460974e-05,
+      "loss": 0.0,
+      "step": 1884
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 2.7823691460055094e-05,
+      "loss": 0.0,
+      "step": 1885
+    },
+    {
+      "epoch": 17.3,
+      "learning_rate": 2.773186409550046e-05,
+      "loss": 0.0,
+      "step": 1886
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 2.7640036730945823e-05,
+      "loss": 0.0,
+      "step": 1887
+    },
+    {
+      "epoch": 17.32,
+      "learning_rate": 2.7548209366391185e-05,
+      "loss": 0.0,
+      "step": 1888
+    },
+    {
+      "epoch": 17.33,
+      "learning_rate": 2.745638200183655e-05,
+      "loss": 0.0,
+      "step": 1889
+    },
+    {
+      "epoch": 17.34,
+      "learning_rate": 2.7364554637281914e-05,
+      "loss": 0.0,
+      "step": 1890
+    },
+    {
+      "epoch": 17.35,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.0,
+      "step": 1891
+    },
+    {
+      "epoch": 17.36,
+      "learning_rate": 2.7180899908172636e-05,
+      "loss": 0.0,
+      "step": 1892
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 2.7089072543617998e-05,
+      "loss": 0.0,
+      "step": 1893
+    },
+    {
+      "epoch": 17.38,
+      "learning_rate": 2.6997245179063364e-05,
+      "loss": 0.0,
+      "step": 1894
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 2.6905417814508727e-05,
+      "loss": 0.0,
+      "step": 1895
+    },
+    {
+      "epoch": 17.39,
+      "learning_rate": 2.681359044995409e-05,
+      "loss": 0.0,
+      "step": 1896
+    },
+    {
+      "epoch": 17.4,
+      "learning_rate": 2.672176308539945e-05,
+      "loss": 0.0,
+      "step": 1897
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 2.662993572084481e-05,
+      "loss": 0.0,
+      "step": 1898
+    },
+    {
+      "epoch": 17.42,
+      "learning_rate": 2.6538108356290177e-05,
+      "loss": 0.0,
+      "step": 1899
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 2.644628099173554e-05,
+      "loss": 0.0,
+      "step": 1900
+    },
+    {
+      "epoch": 17.44,
+      "learning_rate": 2.6354453627180902e-05,
+      "loss": 0.0,
+      "step": 1901
+    },
+    {
+      "epoch": 17.45,
+      "learning_rate": 2.6262626262626268e-05,
+      "loss": 0.0,
+      "step": 1902
+    },
+    {
+      "epoch": 17.46,
+      "learning_rate": 2.6170798898071624e-05,
+      "loss": 0.0,
+      "step": 1903
+    },
+    {
+      "epoch": 17.47,
+      "learning_rate": 2.607897153351699e-05,
+      "loss": 0.0,
+      "step": 1904
+    },
+    {
+      "epoch": 17.48,
+      "learning_rate": 2.5987144168962352e-05,
+      "loss": 0.0,
+      "step": 1905
+    },
+    {
+      "epoch": 17.49,
+      "learning_rate": 2.5895316804407715e-05,
+      "loss": 0.0,
+      "step": 1906
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 2.580348943985308e-05,
+      "loss": 0.0,
+      "step": 1907
+    },
+    {
+      "epoch": 17.5,
+      "learning_rate": 2.5711662075298443e-05,
+      "loss": 0.0,
+      "step": 1908
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 2.5619834710743802e-05,
+      "loss": 0.0,
+      "step": 1909
+    },
+    {
+      "epoch": 17.52,
+      "learning_rate": 2.5528007346189165e-05,
+      "loss": 0.0,
+      "step": 1910
+    },
+    {
+      "epoch": 17.53,
+      "learning_rate": 2.5436179981634527e-05,
+      "loss": 0.0,
+      "step": 1911
+    },
+    {
+      "epoch": 17.54,
+      "learning_rate": 2.5344352617079893e-05,
+      "loss": 0.0,
+      "step": 1912
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 0.0,
+      "step": 1913
+    },
+    {
+      "epoch": 17.56,
+      "learning_rate": 2.516069788797062e-05,
+      "loss": 0.0,
+      "step": 1914
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 2.5068870523415978e-05,
+      "loss": 0.0,
+      "step": 1915
+    },
+    {
+      "epoch": 17.58,
+      "learning_rate": 2.497704315886134e-05,
+      "loss": 0.0,
+      "step": 1916
+    },
+    {
+      "epoch": 17.59,
+      "learning_rate": 2.4885215794306703e-05,
+      "loss": 0.0,
+      "step": 1917
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 2.479338842975207e-05,
+      "loss": 0.0,
+      "step": 1918
+    },
+    {
+      "epoch": 17.61,
+      "learning_rate": 2.4701561065197428e-05,
+      "loss": 0.0,
+      "step": 1919
+    },
+    {
+      "epoch": 17.61,
+      "learning_rate": 2.4609733700642794e-05,
+      "loss": 0.0,
+      "step": 1920
+    },
+    {
+      "epoch": 17.62,
+      "learning_rate": 2.4517906336088156e-05,
+      "loss": 0.0,
+      "step": 1921
+    },
+    {
+      "epoch": 17.63,
+      "learning_rate": 2.4426078971533515e-05,
+      "loss": 0.0,
+      "step": 1922
+    },
+    {
+      "epoch": 17.64,
+      "learning_rate": 2.433425160697888e-05,
+      "loss": 0.0,
+      "step": 1923
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.0,
+      "step": 1924
+    },
+    {
+      "epoch": 17.66,
+      "learning_rate": 2.4150596877869606e-05,
+      "loss": 0.0,
+      "step": 1925
+    },
+    {
+      "epoch": 17.67,
+      "learning_rate": 2.405876951331497e-05,
+      "loss": 0.0,
+      "step": 1926
+    },
+    {
+      "epoch": 17.68,
+      "learning_rate": 2.396694214876033e-05,
+      "loss": 0.0,
+      "step": 1927
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 2.3875114784205694e-05,
+      "loss": 0.0,
+      "step": 1928
+    },
+    {
+      "epoch": 17.7,
+      "learning_rate": 2.3783287419651057e-05,
+      "loss": 0.0,
+      "step": 1929
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 2.369146005509642e-05,
+      "loss": 0.0,
+      "step": 1930
+    },
+    {
+      "epoch": 17.72,
+      "learning_rate": 2.3599632690541782e-05,
+      "loss": 0.0,
+      "step": 1931
+    },
+    {
+      "epoch": 17.72,
+      "learning_rate": 2.3507805325987144e-05,
+      "loss": 0.0,
+      "step": 1932
+    },
+    {
+      "epoch": 17.73,
+      "learning_rate": 2.341597796143251e-05,
+      "loss": 0.0,
+      "step": 1933
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 2.332415059687787e-05,
+      "loss": 0.0,
+      "step": 1934
+    },
+    {
+      "epoch": 17.75,
+      "learning_rate": 2.3232323232323232e-05,
+      "loss": 0.0,
+      "step": 1935
+    },
+    {
+      "epoch": 17.76,
+      "learning_rate": 2.3140495867768598e-05,
+      "loss": 0.0,
+      "step": 1936
+    },
+    {
+      "epoch": 17.77,
+      "learning_rate": 2.3048668503213957e-05,
+      "loss": 0.0,
+      "step": 1937
+    },
+    {
+      "epoch": 17.78,
+      "learning_rate": 2.2956841138659323e-05,
+      "loss": 0.0,
+      "step": 1938
+    },
+    {
+      "epoch": 17.79,
+      "learning_rate": 2.2865013774104685e-05,
+      "loss": 0.0,
+      "step": 1939
+    },
+    {
+      "epoch": 17.8,
+      "learning_rate": 2.2773186409550045e-05,
+      "loss": 0.0,
+      "step": 1940
+    },
+    {
+      "epoch": 17.81,
+      "learning_rate": 2.268135904499541e-05,
+      "loss": 0.0,
+      "step": 1941
+    },
+    {
+      "epoch": 17.82,
+      "learning_rate": 2.2589531680440773e-05,
+      "loss": 0.0,
+      "step": 1942
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 2.2497704315886132e-05,
+      "loss": 0.0,
+      "step": 1943
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 2.2405876951331498e-05,
+      "loss": 0.0,
+      "step": 1944
+    },
+    {
+      "epoch": 17.84,
+      "learning_rate": 2.231404958677686e-05,
+      "loss": 0.0,
+      "step": 1945
+    },
+    {
+      "epoch": 17.85,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.0,
+      "step": 1946
+    },
+    {
+      "epoch": 17.86,
+      "learning_rate": 2.2130394857667586e-05,
+      "loss": 0.0,
+      "step": 1947
+    },
+    {
+      "epoch": 17.87,
+      "learning_rate": 2.203856749311295e-05,
+      "loss": 0.0,
+      "step": 1948
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 2.194674012855831e-05,
+      "loss": 0.0,
+      "step": 1949
+    },
+    {
+      "epoch": 17.89,
+      "learning_rate": 2.1854912764003673e-05,
+      "loss": 0.0,
+      "step": 1950
+    },
+    {
+      "epoch": 17.9,
+      "learning_rate": 2.1763085399449036e-05,
+      "loss": 0.0,
+      "step": 1951
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 2.16712580348944e-05,
+      "loss": 0.0,
+      "step": 1952
+    },
+    {
+      "epoch": 17.92,
+      "learning_rate": 2.157943067033976e-05,
+      "loss": 0.0,
+      "step": 1953
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 2.1487603305785127e-05,
+      "loss": 0.0,
+      "step": 1954
+    },
+    {
+      "epoch": 17.94,
+      "learning_rate": 2.1395775941230486e-05,
+      "loss": 0.0,
+      "step": 1955
+    },
+    {
+      "epoch": 17.94,
+      "learning_rate": 2.130394857667585e-05,
+      "loss": 0.0,
+      "step": 1956
+    },
+    {
+      "epoch": 17.95,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.0,
+      "step": 1957
+    },
+    {
+      "epoch": 17.96,
+      "learning_rate": 2.1120293847566574e-05,
+      "loss": 0.0,
+      "step": 1958
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 2.102846648301194e-05,
+      "loss": 0.0,
+      "step": 1959
+    },
+    {
+      "epoch": 17.98,
+      "learning_rate": 2.0936639118457302e-05,
+      "loss": 0.0,
+      "step": 1960
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 2.084481175390266e-05,
+      "loss": 0.0,
+      "step": 1961
+    },
+    {
+      "epoch": 18.0,
+      "learning_rate": 2.0752984389348027e-05,
+      "loss": 0.0,
+      "step": 1962
+    },
+    {
+      "epoch": 18.0,
+      "eval_accuracy": 0.8475609756097561,
+      "eval_loss": 1.3659346103668213,
+      "eval_runtime": 13.9452,
+      "eval_samples_per_second": 11.76,
+      "eval_steps_per_second": 1.506,
+      "step": 1962
+    },
+    {
+      "epoch": 18.01,
+      "learning_rate": 2.066115702479339e-05,
+      "loss": 0.0,
+      "step": 1963
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 2.0569329660238752e-05,
+      "loss": 0.0,
+      "step": 1964
+    },
+    {
+      "epoch": 18.03,
+      "learning_rate": 2.0477502295684115e-05,
+      "loss": 0.0,
+      "step": 1965
+    },
+    {
+      "epoch": 18.04,
+      "learning_rate": 2.0385674931129478e-05,
+      "loss": 0.0,
+      "step": 1966
+    },
+    {
+      "epoch": 18.05,
+      "learning_rate": 2.029384756657484e-05,
+      "loss": 0.0,
+      "step": 1967
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 0.0,
+      "step": 1968
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 2.0110192837465565e-05,
+      "loss": 0.0,
+      "step": 1969
+    },
+    {
+      "epoch": 18.07,
+      "learning_rate": 2.0018365472910928e-05,
+      "loss": 0.0,
+      "step": 1970
+    },
+    {
+      "epoch": 18.08,
+      "learning_rate": 1.992653810835629e-05,
+      "loss": 0.0,
+      "step": 1971
+    },
+    {
+      "epoch": 18.09,
+      "learning_rate": 1.9834710743801653e-05,
+      "loss": 0.0,
+      "step": 1972
+    },
+    {
+      "epoch": 18.1,
+      "learning_rate": 1.9742883379247015e-05,
+      "loss": 0.0,
+      "step": 1973
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 1.9651056014692378e-05,
+      "loss": 0.0,
+      "step": 1974
+    },
+    {
+      "epoch": 18.12,
+      "learning_rate": 1.9559228650137744e-05,
+      "loss": 0.0,
+      "step": 1975
+    },
+    {
+      "epoch": 18.13,
+      "learning_rate": 1.9467401285583103e-05,
+      "loss": 0.0,
+      "step": 1976
+    },
+    {
+      "epoch": 18.14,
+      "learning_rate": 1.9375573921028466e-05,
+      "loss": 0.0,
+      "step": 1977
+    },
+    {
+      "epoch": 18.15,
+      "learning_rate": 1.928374655647383e-05,
+      "loss": 0.0,
+      "step": 1978
+    },
+    {
+      "epoch": 18.16,
+      "learning_rate": 1.919191919191919e-05,
+      "loss": 0.0,
+      "step": 1979
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 1.9100091827364557e-05,
+      "loss": 0.0,
+      "step": 1980
+    },
+    {
+      "epoch": 18.17,
+      "learning_rate": 1.900826446280992e-05,
+      "loss": 0.0,
+      "step": 1981
+    },
+    {
+      "epoch": 18.18,
+      "learning_rate": 1.891643709825528e-05,
+      "loss": 0.0,
+      "step": 1982
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 1.8824609733700644e-05,
+      "loss": 0.0,
+      "step": 1983
+    },
+    {
+      "epoch": 18.2,
+      "learning_rate": 1.8732782369146007e-05,
+      "loss": 0.0,
+      "step": 1984
+    },
+    {
+      "epoch": 18.21,
+      "learning_rate": 1.864095500459137e-05,
+      "loss": 0.0,
+      "step": 1985
+    },
+    {
+      "epoch": 18.22,
+      "learning_rate": 1.8549127640036732e-05,
+      "loss": 0.0,
+      "step": 1986
+    },
+    {
+      "epoch": 18.23,
+      "learning_rate": 1.8457300275482094e-05,
+      "loss": 0.0,
+      "step": 1987
+    },
+    {
+      "epoch": 18.24,
+      "learning_rate": 1.8365472910927457e-05,
+      "loss": 0.0,
+      "step": 1988
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 1.827364554637282e-05,
+      "loss": 0.0,
+      "step": 1989
+    },
+    {
+      "epoch": 18.26,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.0,
+      "step": 1990
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 1.8089990817263545e-05,
+      "loss": 0.0,
+      "step": 1991
+    },
+    {
+      "epoch": 18.28,
+      "learning_rate": 1.7998163452708907e-05,
+      "loss": 0.0,
+      "step": 1992
+    },
+    {
+      "epoch": 18.28,
+      "learning_rate": 1.790633608815427e-05,
+      "loss": 0.0,
+      "step": 1993
+    },
+    {
+      "epoch": 18.29,
+      "learning_rate": 1.7814508723599632e-05,
+      "loss": 0.0,
+      "step": 1994
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 1.7722681359044995e-05,
+      "loss": 0.0,
+      "step": 1995
+    },
+    {
+      "epoch": 18.31,
+      "learning_rate": 1.763085399449036e-05,
+      "loss": 0.0,
+      "step": 1996
+    },
+    {
+      "epoch": 18.32,
+      "learning_rate": 1.753902662993572e-05,
+      "loss": 0.0,
+      "step": 1997
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 1.7447199265381082e-05,
+      "loss": 0.0,
+      "step": 1998
+    },
+    {
+      "epoch": 18.34,
+      "learning_rate": 1.735537190082645e-05,
+      "loss": 0.0,
+      "step": 1999
+    },
+    {
+      "epoch": 18.35,
+      "learning_rate": 1.7263544536271808e-05,
+      "loss": 0.0,
+      "step": 2000
+    },
+    {
+      "epoch": 18.36,
+      "learning_rate": 1.7171717171717173e-05,
+      "loss": 0.0,
+      "step": 2001
+    },
+    {
+      "epoch": 18.37,
+      "learning_rate": 1.7079889807162536e-05,
+      "loss": 0.0,
+      "step": 2002
+    },
+    {
+      "epoch": 18.38,
+      "learning_rate": 1.6988062442607895e-05,
+      "loss": 0.0,
+      "step": 2003
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 1.689623507805326e-05,
+      "loss": 0.0,
+      "step": 2004
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 1.6804407713498624e-05,
+      "loss": 0.0,
+      "step": 2005
+    },
+    {
+      "epoch": 18.4,
+      "learning_rate": 1.6712580348943986e-05,
+      "loss": 0.0,
+      "step": 2006
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 1.662075298438935e-05,
+      "loss": 0.0,
+      "step": 2007
+    },
+    {
+      "epoch": 18.42,
+      "learning_rate": 1.652892561983471e-05,
+      "loss": 0.0,
+      "step": 2008
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 1.6437098255280077e-05,
+      "loss": 0.0,
+      "step": 2009
+    },
+    {
+      "epoch": 18.44,
+      "learning_rate": 1.6345270890725436e-05,
+      "loss": 0.0,
+      "step": 2010
+    },
+    {
+      "epoch": 18.45,
+      "learning_rate": 1.62534435261708e-05,
+      "loss": 0.0,
+      "step": 2011
+    },
+    {
+      "epoch": 18.46,
+      "learning_rate": 1.6161616161616165e-05,
+      "loss": 0.0,
+      "step": 2012
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 1.6069788797061524e-05,
+      "loss": 0.0,
+      "step": 2013
+    },
+    {
+      "epoch": 18.48,
+      "learning_rate": 1.597796143250689e-05,
+      "loss": 0.0,
+      "step": 2014
+    },
+    {
+      "epoch": 18.49,
+      "learning_rate": 1.5886134067952252e-05,
+      "loss": 0.0,
+      "step": 2015
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 1.579430670339761e-05,
+      "loss": 0.0,
+      "step": 2016
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 1.5702479338842978e-05,
+      "loss": 0.0,
+      "step": 2017
+    },
+    {
+      "epoch": 18.51,
+      "learning_rate": 1.561065197428834e-05,
+      "loss": 0.0,
+      "step": 2018
+    },
+    {
+      "epoch": 18.52,
+      "learning_rate": 1.55188246097337e-05,
+      "loss": 0.0,
+      "step": 2019
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 1.5426997245179065e-05,
+      "loss": 0.0,
+      "step": 2020
+    },
+    {
+      "epoch": 18.54,
+      "learning_rate": 1.5335169880624428e-05,
+      "loss": 0.0,
+      "step": 2021
+    },
+    {
+      "epoch": 18.55,
+      "learning_rate": 1.5243342516069789e-05,
+      "loss": 0.0,
+      "step": 2022
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.0,
+      "step": 2023
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 1.5059687786960517e-05,
+      "loss": 0.0,
+      "step": 2024
+    },
+    {
+      "epoch": 18.58,
+      "learning_rate": 1.4967860422405878e-05,
+      "loss": 0.0,
+      "step": 2025
+    },
+    {
+      "epoch": 18.59,
+      "learning_rate": 1.487603305785124e-05,
+      "loss": 0.0,
+      "step": 2026
+    },
+    {
+      "epoch": 18.6,
+      "learning_rate": 1.4784205693296605e-05,
+      "loss": 0.0,
+      "step": 2027
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 1.4692378328741966e-05,
+      "loss": 0.0,
+      "step": 2028
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 1.4600550964187328e-05,
+      "loss": 0.0,
+      "step": 2029
+    },
+    {
+      "epoch": 18.62,
+      "learning_rate": 1.4508723599632692e-05,
+      "loss": 0.0,
+      "step": 2030
+    },
+    {
+      "epoch": 18.63,
+      "learning_rate": 1.4416896235078053e-05,
+      "loss": 0.0,
+      "step": 2031
+    },
+    {
+      "epoch": 18.64,
+      "learning_rate": 1.4325068870523417e-05,
+      "loss": 0.0,
+      "step": 2032
+    },
+    {
+      "epoch": 18.65,
+      "learning_rate": 1.423324150596878e-05,
+      "loss": 0.0,
+      "step": 2033
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 1.4141414141414141e-05,
+      "loss": 0.0,
+      "step": 2034
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 1.4049586776859505e-05,
+      "loss": 0.0,
+      "step": 2035
+    },
+    {
+      "epoch": 18.68,
+      "learning_rate": 1.395775941230487e-05,
+      "loss": 0.0,
+      "step": 2036
+    },
+    {
+      "epoch": 18.69,
+      "learning_rate": 1.386593204775023e-05,
+      "loss": 0.0,
+      "step": 2037
+    },
+    {
+      "epoch": 18.7,
+      "learning_rate": 1.3774104683195593e-05,
+      "loss": 0.0,
+      "step": 2038
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 1.3682277318640957e-05,
+      "loss": 0.0,
+      "step": 2039
+    },
+    {
+      "epoch": 18.72,
+      "learning_rate": 1.3590449954086318e-05,
+      "loss": 0.0,
+      "step": 2040
+    },
+    {
+      "epoch": 18.72,
+      "learning_rate": 1.3498622589531682e-05,
+      "loss": 0.0,
+      "step": 2041
+    },
+    {
+      "epoch": 18.73,
+      "learning_rate": 1.3406795224977045e-05,
+      "loss": 0.0,
+      "step": 2042
+    },
+    {
+      "epoch": 18.74,
+      "learning_rate": 1.3314967860422405e-05,
+      "loss": 0.0,
+      "step": 2043
+    },
+    {
+      "epoch": 18.75,
+      "learning_rate": 1.322314049586777e-05,
+      "loss": 0.0,
+      "step": 2044
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 1.3131313131313134e-05,
+      "loss": 0.0,
+      "step": 2045
+    },
+    {
+      "epoch": 18.77,
+      "learning_rate": 1.3039485766758495e-05,
+      "loss": 0.0,
+      "step": 2046
+    },
+    {
+      "epoch": 18.78,
+      "learning_rate": 1.2947658402203857e-05,
+      "loss": 0.0,
+      "step": 2047
+    },
+    {
+      "epoch": 18.79,
+      "learning_rate": 1.2855831037649222e-05,
+      "loss": 0.0,
+      "step": 2048
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 1.2764003673094582e-05,
+      "loss": 0.0,
+      "step": 2049
+    },
+    {
+      "epoch": 18.81,
+      "learning_rate": 1.2672176308539947e-05,
+      "loss": 0.0,
+      "step": 2050
+    },
+    {
+      "epoch": 18.82,
+      "learning_rate": 1.258034894398531e-05,
+      "loss": 0.0,
+      "step": 2051
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 1.248852157943067e-05,
+      "loss": 0.0,
+      "step": 2052
+    },
+    {
+      "epoch": 18.83,
+      "learning_rate": 1.2396694214876034e-05,
+      "loss": 0.0,
+      "step": 2053
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 1.2304866850321397e-05,
+      "loss": 0.0,
+      "step": 2054
+    },
+    {
+      "epoch": 18.85,
+      "learning_rate": 1.2213039485766758e-05,
+      "loss": 0.0,
+      "step": 2055
+    },
+    {
+      "epoch": 18.86,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.0,
+      "step": 2056
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 1.2029384756657484e-05,
+      "loss": 0.0,
+      "step": 2057
+    },
+    {
+      "epoch": 18.88,
+      "learning_rate": 1.1937557392102847e-05,
+      "loss": 0.0,
+      "step": 2058
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 1.184573002754821e-05,
+      "loss": 0.0,
+      "step": 2059
+    },
+    {
+      "epoch": 18.9,
+      "learning_rate": 1.1753902662993572e-05,
+      "loss": 0.0,
+      "step": 2060
+    },
+    {
+      "epoch": 18.91,
+      "learning_rate": 1.1662075298438935e-05,
+      "loss": 0.0,
+      "step": 2061
+    },
+    {
+      "epoch": 18.92,
+      "learning_rate": 1.1570247933884299e-05,
+      "loss": 0.0,
+      "step": 2062
+    },
+    {
+      "epoch": 18.93,
+      "learning_rate": 1.1478420569329661e-05,
+      "loss": 0.0,
+      "step": 2063
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 1.1386593204775022e-05,
+      "loss": 0.0,
+      "step": 2064
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 1.1294765840220387e-05,
+      "loss": 0.0,
+      "step": 2065
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 1.1202938475665749e-05,
+      "loss": 0.0,
+      "step": 2066
+    },
+    {
+      "epoch": 18.96,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.0,
+      "step": 2067
+    },
+    {
+      "epoch": 18.97,
+      "learning_rate": 1.1019283746556474e-05,
+      "loss": 0.0,
+      "step": 2068
+    },
+    {
+      "epoch": 18.98,
+      "learning_rate": 1.0927456382001837e-05,
+      "loss": 0.0,
+      "step": 2069
+    },
+    {
+      "epoch": 18.99,
+      "learning_rate": 1.08356290174472e-05,
+      "loss": 0.0,
+      "step": 2070
+    },
+    {
+      "epoch": 19.0,
+      "learning_rate": 1.0743801652892564e-05,
+      "loss": 0.0,
+      "step": 2071
+    },
+    {
+      "epoch": 19.0,
+      "eval_accuracy": 0.8475609756097561,
+      "eval_loss": 1.36630117893219,
+      "eval_runtime": 13.944,
+      "eval_samples_per_second": 11.761,
+      "eval_steps_per_second": 1.506,
+      "step": 2071
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 1.0651974288337924e-05,
+      "loss": 0.0,
+      "step": 2072
+    },
+    {
+      "epoch": 19.02,
+      "learning_rate": 1.0560146923783287e-05,
+      "loss": 0.0,
+      "step": 2073
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 1.0468319559228651e-05,
+      "loss": 0.0,
+      "step": 2074
+    },
+    {
+      "epoch": 19.04,
+      "learning_rate": 1.0376492194674014e-05,
+      "loss": 0.0,
+      "step": 2075
+    },
+    {
+      "epoch": 19.05,
+      "learning_rate": 1.0284664830119376e-05,
+      "loss": 0.0,
+      "step": 2076
+    },
+    {
+      "epoch": 19.06,
+      "learning_rate": 1.0192837465564739e-05,
+      "loss": 0.0,
+      "step": 2077
+    },
+    {
+      "epoch": 19.06,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 0.0,
+      "step": 2078
+    },
+    {
+      "epoch": 19.07,
+      "learning_rate": 1.0009182736455464e-05,
+      "loss": 0.0,
+      "step": 2079
+    },
+    {
+      "epoch": 19.08,
+      "learning_rate": 9.917355371900826e-06,
+      "loss": 0.0,
+      "step": 2080
+    },
+    {
+      "epoch": 19.09,
+      "learning_rate": 9.825528007346189e-06,
+      "loss": 0.0,
+      "step": 2081
+    },
+    {
+      "epoch": 19.1,
+      "learning_rate": 9.733700642791552e-06,
+      "loss": 0.0,
+      "step": 2082
+    },
+    {
+      "epoch": 19.11,
+      "learning_rate": 9.641873278236916e-06,
+      "loss": 0.0,
+      "step": 2083
+    },
+    {
+      "epoch": 19.12,
+      "learning_rate": 9.550045913682278e-06,
+      "loss": 0.0,
+      "step": 2084
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 9.45821854912764e-06,
+      "loss": 0.0,
+      "step": 2085
+    },
+    {
+      "epoch": 19.14,
+      "learning_rate": 9.366391184573003e-06,
+      "loss": 0.0,
+      "step": 2086
+    },
+    {
+      "epoch": 19.15,
+      "learning_rate": 9.274563820018366e-06,
+      "loss": 0.0,
+      "step": 2087
+    },
+    {
+      "epoch": 19.16,
+      "learning_rate": 9.182736455463728e-06,
+      "loss": 0.0,
+      "step": 2088
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.0,
+      "step": 2089
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 8.999081726354454e-06,
+      "loss": 0.0,
+      "step": 2090
+    },
+    {
+      "epoch": 19.18,
+      "learning_rate": 8.907254361799816e-06,
+      "loss": 0.0,
+      "step": 2091
+    },
+    {
+      "epoch": 19.19,
+      "learning_rate": 8.81542699724518e-06,
+      "loss": 0.0,
+      "step": 2092
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 8.723599632690541e-06,
+      "loss": 0.0,
+      "step": 2093
+    },
+    {
+      "epoch": 19.21,
+      "learning_rate": 8.631772268135904e-06,
+      "loss": 0.0,
+      "step": 2094
+    },
+    {
+      "epoch": 19.22,
+      "learning_rate": 8.539944903581268e-06,
+      "loss": 0.0,
+      "step": 2095
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 8.44811753902663e-06,
+      "loss": 0.0,
+      "step": 2096
+    },
+    {
+      "epoch": 19.24,
+      "learning_rate": 8.356290174471993e-06,
+      "loss": 0.0,
+      "step": 2097
+    },
+    {
+      "epoch": 19.25,
+      "learning_rate": 8.264462809917356e-06,
+      "loss": 0.0,
+      "step": 2098
+    },
+    {
+      "epoch": 19.26,
+      "learning_rate": 8.172635445362718e-06,
+      "loss": 0.0,
+      "step": 2099
+    },
+    {
+      "epoch": 19.27,
+      "learning_rate": 8.080808080808082e-06,
+      "loss": 0.0,
+      "step": 2100
+    },
+    {
+      "epoch": 19.28,
+      "learning_rate": 7.988980716253445e-06,
+      "loss": 0.0,
+      "step": 2101
+    },
+    {
+      "epoch": 19.28,
+      "learning_rate": 7.897153351698806e-06,
+      "loss": 0.0,
+      "step": 2102
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 7.80532598714417e-06,
+      "loss": 0.0,
+      "step": 2103
+    },
+    {
+      "epoch": 19.3,
+      "learning_rate": 7.713498622589533e-06,
+      "loss": 0.0,
+      "step": 2104
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 7.621671258034894e-06,
+      "loss": 0.0,
+      "step": 2105
+    },
+    {
+      "epoch": 19.32,
+      "learning_rate": 7.5298438934802585e-06,
+      "loss": 0.0,
+      "step": 2106
+    },
+    {
+      "epoch": 19.33,
+      "learning_rate": 7.43801652892562e-06,
+      "loss": 0.0,
+      "step": 2107
+    },
+    {
+      "epoch": 19.34,
+      "learning_rate": 7.346189164370983e-06,
+      "loss": 0.0,
+      "step": 2108
+    },
+    {
+      "epoch": 19.35,
+      "learning_rate": 7.254361799816346e-06,
+      "loss": 0.0,
+      "step": 2109
+    },
+    {
+      "epoch": 19.36,
+      "learning_rate": 7.162534435261709e-06,
+      "loss": 0.0,
+      "step": 2110
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 7.0707070707070704e-06,
+      "loss": 0.0,
+      "step": 2111
+    },
+    {
+      "epoch": 19.38,
+      "learning_rate": 6.978879706152435e-06,
+      "loss": 0.0,
+      "step": 2112
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 6.887052341597796e-06,
+      "loss": 0.0,
+      "step": 2113
+    },
+    {
+      "epoch": 19.39,
+      "learning_rate": 6.795224977043159e-06,
+      "loss": 0.0,
+      "step": 2114
+    },
+    {
+      "epoch": 19.4,
+      "learning_rate": 6.703397612488522e-06,
+      "loss": 0.0,
+      "step": 2115
+    },
+    {
+      "epoch": 19.41,
+      "learning_rate": 6.611570247933885e-06,
+      "loss": 0.0,
+      "step": 2116
+    },
+    {
+      "epoch": 19.42,
+      "learning_rate": 6.519742883379247e-06,
+      "loss": 0.0,
+      "step": 2117
+    },
+    {
+      "epoch": 19.43,
+      "learning_rate": 6.427915518824611e-06,
+      "loss": 0.0,
+      "step": 2118
+    },
+    {
+      "epoch": 19.44,
+      "learning_rate": 6.336088154269973e-06,
+      "loss": 0.0,
+      "step": 2119
+    },
+    {
+      "epoch": 19.45,
+      "learning_rate": 6.244260789715335e-06,
+      "loss": 0.0,
+      "step": 2120
+    },
+    {
+      "epoch": 19.46,
+      "learning_rate": 6.1524334251606984e-06,
+      "loss": 0.0,
+      "step": 2121
+    },
+    {
+      "epoch": 19.47,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.0,
+      "step": 2122
+    },
+    {
+      "epoch": 19.48,
+      "learning_rate": 5.9687786960514235e-06,
+      "loss": 0.0,
+      "step": 2123
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 5.876951331496786e-06,
+      "loss": 0.0,
+      "step": 2124
+    },
+    {
+      "epoch": 19.5,
+      "learning_rate": 5.7851239669421495e-06,
+      "loss": 0.0,
+      "step": 2125
+    },
+    {
+      "epoch": 19.5,
+      "learning_rate": 5.693296602387511e-06,
+      "loss": 0.0,
+      "step": 2126
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 5.6014692378328745e-06,
+      "loss": 0.0,
+      "step": 2127
+    },
+    {
+      "epoch": 19.52,
+      "learning_rate": 5.509641873278237e-06,
+      "loss": 0.0,
+      "step": 2128
+    },
+    {
+      "epoch": 19.53,
+      "learning_rate": 5.4178145087236e-06,
+      "loss": 0.0,
+      "step": 2129
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 5.325987144168962e-06,
+      "loss": 0.0,
+      "step": 2130
+    },
+    {
+      "epoch": 19.55,
+      "learning_rate": 5.234159779614326e-06,
+      "loss": 0.0,
+      "step": 2131
+    },
+    {
+      "epoch": 19.56,
+      "learning_rate": 5.142332415059688e-06,
+      "loss": 0.0,
+      "step": 2132
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 0.0,
+      "step": 2133
+    },
+    {
+      "epoch": 19.58,
+      "learning_rate": 4.958677685950413e-06,
+      "loss": 0.0,
+      "step": 2134
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 4.866850321395776e-06,
+      "loss": 0.0,
+      "step": 2135
+    },
+    {
+      "epoch": 19.6,
+      "learning_rate": 4.775022956841139e-06,
+      "loss": 0.0,
+      "step": 2136
+    },
+    {
+      "epoch": 19.61,
+      "learning_rate": 4.683195592286502e-06,
+      "loss": 0.0,
+      "step": 2137
+    },
+    {
+      "epoch": 19.61,
+      "learning_rate": 4.591368227731864e-06,
+      "loss": 0.0,
+      "step": 2138
+    },
+    {
+      "epoch": 19.62,
+      "learning_rate": 4.499540863177227e-06,
+      "loss": 0.0,
+      "step": 2139
+    },
+    {
+      "epoch": 19.63,
+      "learning_rate": 4.40771349862259e-06,
+      "loss": 0.0,
+      "step": 2140
+    },
+    {
+      "epoch": 19.64,
+      "learning_rate": 4.315886134067952e-06,
+      "loss": 0.0,
+      "step": 2141
+    },
+    {
+      "epoch": 19.65,
+      "learning_rate": 4.224058769513315e-06,
+      "loss": 0.0,
+      "step": 2142
+    },
+    {
+      "epoch": 19.66,
+      "learning_rate": 4.132231404958678e-06,
+      "loss": 0.0,
+      "step": 2143
+    },
+    {
+      "epoch": 19.67,
+      "learning_rate": 4.040404040404041e-06,
+      "loss": 0.0,
+      "step": 2144
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 3.948576675849403e-06,
+      "loss": 0.0,
+      "step": 2145
+    },
+    {
+      "epoch": 19.69,
+      "learning_rate": 3.856749311294766e-06,
+      "loss": 0.0,
+      "step": 2146
+    },
+    {
+      "epoch": 19.7,
+      "learning_rate": 3.7649219467401293e-06,
+      "loss": 0.0,
+      "step": 2147
+    },
+    {
+      "epoch": 19.71,
+      "learning_rate": 3.6730945821854914e-06,
+      "loss": 0.0,
+      "step": 2148
+    },
+    {
+      "epoch": 19.72,
+      "learning_rate": 3.5812672176308544e-06,
+      "loss": 0.0,
+      "step": 2149
+    },
+    {
+      "epoch": 19.72,
+      "learning_rate": 3.4894398530762173e-06,
+      "loss": 0.0,
+      "step": 2150
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 3.3976124885215795e-06,
+      "loss": 0.0,
+      "step": 2151
+    },
+    {
+      "epoch": 19.74,
+      "learning_rate": 3.3057851239669424e-06,
+      "loss": 0.0,
+      "step": 2152
+    },
+    {
+      "epoch": 19.75,
+      "learning_rate": 3.2139577594123054e-06,
+      "loss": 0.0,
+      "step": 2153
+    },
+    {
+      "epoch": 19.76,
+      "learning_rate": 3.1221303948576675e-06,
+      "loss": 0.0,
+      "step": 2154
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.0,
+      "step": 2155
+    },
+    {
+      "epoch": 19.78,
+      "learning_rate": 2.938475665748393e-06,
+      "loss": 0.0,
+      "step": 2156
+    },
+    {
+      "epoch": 19.79,
+      "learning_rate": 2.8466483011937556e-06,
+      "loss": 0.0,
+      "step": 2157
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 2.7548209366391185e-06,
+      "loss": 0.0,
+      "step": 2158
+    },
+    {
+      "epoch": 19.81,
+      "learning_rate": 2.662993572084481e-06,
+      "loss": 0.0,
+      "step": 2159
+    },
+    {
+      "epoch": 19.82,
+      "learning_rate": 2.571166207529844e-06,
+      "loss": 0.0,
+      "step": 2160
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 2.4793388429752066e-06,
+      "loss": 0.0,
+      "step": 2161
+    },
+    {
+      "epoch": 19.83,
+      "learning_rate": 2.3875114784205696e-06,
+      "loss": 0.0,
+      "step": 2162
+    },
+    {
+      "epoch": 19.84,
+      "learning_rate": 2.295684113865932e-06,
+      "loss": 0.0,
+      "step": 2163
+    },
+    {
+      "epoch": 19.85,
+      "learning_rate": 2.203856749311295e-06,
+      "loss": 0.0,
+      "step": 2164
+    },
+    {
+      "epoch": 19.86,
+      "learning_rate": 2.1120293847566576e-06,
+      "loss": 0.0,
+      "step": 2165
+    },
+    {
+      "epoch": 19.87,
+      "learning_rate": 2.0202020202020206e-06,
+      "loss": 0.0,
+      "step": 2166
+    },
+    {
+      "epoch": 19.88,
+      "learning_rate": 1.928374655647383e-06,
+      "loss": 0.0,
+      "step": 2167
+    },
+    {
+      "epoch": 19.89,
+      "learning_rate": 1.8365472910927457e-06,
+      "loss": 0.0,
+      "step": 2168
+    },
+    {
+      "epoch": 19.9,
+      "learning_rate": 1.7447199265381087e-06,
+      "loss": 0.0,
+      "step": 2169
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 1.6528925619834712e-06,
+      "loss": 0.0,
+      "step": 2170
+    },
+    {
+      "epoch": 19.92,
+      "learning_rate": 1.5610651974288338e-06,
+      "loss": 0.0,
+      "step": 2171
+    },
+    {
+      "epoch": 19.93,
+      "learning_rate": 1.4692378328741965e-06,
+      "loss": 0.0,
+      "step": 2172
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 1.3774104683195593e-06,
+      "loss": 0.0,
+      "step": 2173
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 1.285583103764922e-06,
+      "loss": 0.0,
+      "step": 2174
+    },
+    {
+      "epoch": 19.95,
+      "learning_rate": 1.1937557392102848e-06,
+      "loss": 0.0,
+      "step": 2175
+    },
+    {
+      "epoch": 19.96,
+      "learning_rate": 1.1019283746556475e-06,
+      "loss": 0.0,
+      "step": 2176
+    },
+    {
+      "epoch": 19.97,
+      "learning_rate": 1.0101010101010103e-06,
+      "loss": 0.0,
+      "step": 2177
+    },
+    {
+      "epoch": 19.98,
+      "learning_rate": 9.182736455463728e-07,
+      "loss": 0.0,
+      "step": 2178
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 8.264462809917356e-07,
+      "loss": 0.0,
+      "step": 2179
+    },
+    {
+      "epoch": 20.0,
+      "learning_rate": 7.346189164370983e-07,
+      "loss": 0.0,
+      "step": 2180
+    },
+    {
+      "epoch": 20.0,
+      "eval_accuracy": 0.8475609756097561,
+      "eval_loss": 1.3666630983352661,
+      "eval_runtime": 13.9421,
+      "eval_samples_per_second": 11.763,
+      "eval_steps_per_second": 1.506,
+      "step": 2180
+    },
+    {
+      "epoch": 20.0,
+      "step": 2180,
+      "total_flos": 1.2663415269359616e+17,
+      "train_loss": 0.4154004427643197,
+      "train_runtime": 4412.8344,
+      "train_samples_per_second": 2.96,
+      "train_steps_per_second": 0.494
     }
   ],
   "logging_steps": 1,
-  "max_steps": 20,
-  "num_train_epochs": 1,
+  "max_steps": 2180,
+  "num_train_epochs": 20,
   "save_steps": 500,
-  "total_flos": 825752514723840.0,
+  "total_flos": 1.2663415269359616e+17,
   "trial_name": null,
   "trial_params": null
 }