diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8236 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.998722860791826,
+  "global_step": 1370,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0,
+      "loss": 1.0562,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 0.4675,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.3332,
+      "step": 3
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1e-05,
+      "loss": 0.4042,
+      "step": 4
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.4214,
+      "step": 5
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1e-05,
+      "loss": 0.3946,
+      "step": 6
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.3061,
+      "step": 7
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1e-05,
+      "loss": 0.2138,
+      "step": 8
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.3077,
+      "step": 9
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.3196,
+      "step": 10
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.235,
+      "step": 11
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1e-05,
+      "loss": 0.1768,
+      "step": 12
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.1698,
+      "step": 13
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1e-05,
+      "loss": 0.1664,
+      "step": 14
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.2661,
+      "step": 15
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1e-05,
+      "loss": 0.1765,
+      "step": 16
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.1538,
+      "step": 17
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1e-05,
+      "loss": 0.1342,
+      "step": 18
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0749,
+      "step": 19
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0845,
+      "step": 20
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.3328,
+      "step": 21
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 0.068,
+      "step": 22
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.1653,
+      "step": 23
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1e-05,
+      "loss": 0.1063,
+      "step": 24
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.1936,
+      "step": 25
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1e-05,
+      "loss": 0.1183,
+      "step": 26
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.2157,
+      "step": 27
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1e-05,
+      "loss": 0.1069,
+      "step": 28
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.3282,
+      "step": 29
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1e-05,
+      "loss": 0.1977,
+      "step": 30
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.1899,
+      "step": 31
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1e-05,
+      "loss": 0.1373,
+      "step": 32
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0571,
+      "step": 33
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1e-05,
+      "loss": 0.1905,
+      "step": 34
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.1854,
+      "step": 35
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1e-05,
+      "loss": 0.1395,
+      "step": 36
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.1495,
+      "step": 37
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1e-05,
+      "loss": 0.2867,
+      "step": 38
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.1169,
+      "step": 39
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1e-05,
+      "loss": 0.074,
+      "step": 40
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.1304,
+      "step": 41
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 0.2853,
+      "step": 42
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0653,
+      "step": 43
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1e-05,
+      "loss": 0.1973,
+      "step": 44
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.1049,
+      "step": 45
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1e-05,
+      "loss": 0.2182,
+      "step": 46
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0839,
+      "step": 47
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.1563,
+      "step": 48
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1e-05,
+      "loss": 0.1184,
+      "step": 49
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.1367,
+      "step": 50
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 51
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.1879,
+      "step": 52
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1e-05,
+      "loss": 0.1286,
+      "step": 53
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.112,
+      "step": 54
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0938,
+      "step": 55
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0198,
+      "step": 56
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1e-05,
+      "loss": 0.1999,
+      "step": 57
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0676,
+      "step": 58
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 0.1658,
+      "step": 59
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0914,
+      "step": 60
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1e-05,
+      "loss": 0.1006,
+      "step": 61
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0464,
+      "step": 62
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1e-05,
+      "loss": 0.1222,
+      "step": 63
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0143,
+      "step": 64
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1e-05,
+      "loss": 0.2137,
+      "step": 65
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.2755,
+      "step": 66
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1e-05,
+      "loss": 0.1399,
+      "step": 67
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.2166,
+      "step": 68
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1e-05,
+      "loss": 0.1916,
+      "step": 69
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.1414,
+      "step": 70
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0748,
+      "step": 71
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0892,
+      "step": 72
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0685,
+      "step": 73
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.1457,
+      "step": 74
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1e-05,
+      "loss": 0.1464,
+      "step": 75
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.101,
+      "step": 76
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1e-05,
+      "loss": 0.2092,
+      "step": 77
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.1977,
+      "step": 78
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1e-05,
+      "loss": 0.1607,
+      "step": 79
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2879,
+      "step": 80
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1e-05,
+      "loss": 0.2386,
+      "step": 81
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.2433,
+      "step": 82
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1e-05,
+      "loss": 0.159,
+      "step": 83
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.095,
+      "step": 84
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0821,
+      "step": 85
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0784,
+      "step": 86
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1e-05,
+      "loss": 0.1487,
+      "step": 87
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.1947,
+      "step": 88
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1e-05,
+      "loss": 0.1335,
+      "step": 89
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.3123,
+      "step": 90
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0989,
+      "step": 91
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0772,
+      "step": 92
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.1531,
+      "step": 93
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0714,
+      "step": 94
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.1081,
+      "step": 95
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0669,
+      "step": 96
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.1384,
+      "step": 97
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1e-05,
+      "loss": 0.1662,
+      "step": 98
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.1573,
+      "step": 99
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0713,
+      "step": 100
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2945,
+      "step": 101
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2392,
+      "step": 102
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0472,
+      "step": 103
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1e-05,
+      "loss": 0.1296,
+      "step": 104
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.1508,
+      "step": 105
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0673,
+      "step": 106
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.1289,
+      "step": 107
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1e-05,
+      "loss": 0.1599,
+      "step": 108
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.1136,
+      "step": 109
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1e-05,
+      "loss": 0.1192,
+      "step": 110
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.1378,
+      "step": 111
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0723,
+      "step": 112
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.095,
+      "step": 113
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1e-05,
+      "loss": 0.1105,
+      "step": 114
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.1114,
+      "step": 115
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1e-05,
+      "loss": 0.3623,
+      "step": 116
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.2149,
+      "step": 117
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1e-05,
+      "loss": 0.1858,
+      "step": 118
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.1264,
+      "step": 119
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1e-05,
+      "loss": 0.1113,
+      "step": 120
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.113,
+      "step": 121
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1e-05,
+      "loss": 0.1364,
+      "step": 122
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0383,
+      "step": 123
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0848,
+      "step": 124
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0921,
+      "step": 125
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1e-05,
+      "loss": 0.1009,
+      "step": 126
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0949,
+      "step": 127
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0877,
+      "step": 128
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.1961,
+      "step": 129
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0304,
+      "step": 130
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.1637,
+      "step": 131
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1e-05,
+      "loss": 0.1214,
+      "step": 132
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.13,
+      "step": 133
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0122,
+      "step": 134
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.1911,
+      "step": 135
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0672,
+      "step": 136
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2112,
+      "step": 137
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1e-05,
+      "loss": 0.2525,
+      "step": 138
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1e-05,
+      "loss": 0.1582,
+      "step": 139
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0819,
+      "step": 140
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0925,
+      "step": 141
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0635,
+      "step": 142
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1e-05,
+      "loss": 0.337,
+      "step": 143
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.2296,
+      "step": 144
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1e-05,
+      "loss": 0.1795,
+      "step": 145
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0803,
+      "step": 146
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1e-05,
+      "loss": 0.1189,
+      "step": 147
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.1089,
+      "step": 148
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1e-05,
+      "loss": 0.1216,
+      "step": 149
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.1148,
+      "step": 150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0407,
+      "step": 151
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.2256,
+      "step": 152
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0644,
+      "step": 153
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.1588,
+      "step": 154
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1e-05,
+      "loss": 0.3197,
+      "step": 155
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.2215,
+      "step": 156
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1e-05,
+      "loss": 0.151,
+      "step": 157
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0279,
+      "step": 158
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0064,
+      "step": 159
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.199,
+      "step": 160
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0465,
+      "step": 161
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.2614,
+      "step": 162
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0382,
+      "step": 163
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0665,
+      "step": 164
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1e-05,
+      "loss": 0.1416,
+      "step": 165
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.1066,
+      "step": 166
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0894,
+      "step": 167
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0602,
+      "step": 168
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1e-05,
+      "loss": 0.1677,
+      "step": 169
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0181,
+      "step": 170
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1e-05,
+      "loss": 0.4301,
+      "step": 171
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.1563,
+      "step": 172
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1e-05,
+      "loss": 0.1198,
+      "step": 173
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.1015,
+      "step": 174
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1e-05,
+      "loss": 0.1107,
+      "step": 175
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.1831,
+      "step": 176
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0354,
+      "step": 177
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.2364,
+      "step": 178
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1e-05,
+      "loss": 0.1992,
+      "step": 179
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.1165,
+      "step": 180
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1e-05,
+      "loss": 0.198,
+      "step": 181
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.2144,
+      "step": 182
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1e-05,
+      "loss": 0.1842,
+      "step": 183
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0275,
+      "step": 184
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0246,
+      "step": 185
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1e-05,
+      "loss": 0.1705,
+      "step": 186
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0736,
+      "step": 187
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0545,
+      "step": 188
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.2176,
+      "step": 189
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0435,
+      "step": 190
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0855,
+      "step": 191
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1e-05,
+      "loss": 0.1094,
+      "step": 192
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 193
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1e-05,
+      "loss": 0.1278,
+      "step": 194
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.1514,
+      "step": 195
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0519,
+      "step": 196
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0148,
+      "step": 197
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0307,
+      "step": 198
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.1395,
+      "step": 199
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0545,
+      "step": 200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0808,
+      "step": 201
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0461,
+      "step": 202
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0158,
+      "step": 203
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0845,
+      "step": 204
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1e-05,
+      "loss": 0.1393,
+      "step": 205
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0598,
+      "step": 206
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0928,
+      "step": 207
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0459,
+      "step": 208
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1e-05,
+      "loss": 0.046,
+      "step": 209
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 210
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0415,
+      "step": 211
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0789,
+      "step": 212
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0958,
+      "step": 213
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0824,
+      "step": 214
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0241,
+      "step": 215
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1e-05,
+      "loss": 0.05,
+      "step": 216
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0578,
+      "step": 217
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0381,
+      "step": 218
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0453,
+      "step": 219
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1e-05,
+      "loss": 0.024,
+      "step": 220
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0314,
+      "step": 221
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0575,
+      "step": 222
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1e-05,
+      "loss": 0.136,
+      "step": 223
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0502,
+      "step": 224
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1e-05,
+      "loss": 0.2071,
+      "step": 225
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0483,
+      "step": 226
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1e-05,
+      "loss": 0.1118,
+      "step": 227
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0858,
+      "step": 228
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0048,
+      "step": 229
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0405,
+      "step": 230
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0719,
+      "step": 231
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0146,
+      "step": 232
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1e-05,
+      "loss": 0.083,
+      "step": 233
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0878,
+      "step": 234
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0432,
+      "step": 235
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0508,
+      "step": 236
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0574,
+      "step": 237
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0288,
+      "step": 238
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0995,
+      "step": 239
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1e-05,
+      "loss": 0.1162,
+      "step": 240
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0281,
+      "step": 241
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0097,
+      "step": 242
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0669,
+      "step": 243
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0366,
+      "step": 244
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0755,
+      "step": 245
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1e-05,
+      "loss": 0.038,
+      "step": 246
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0146,
+      "step": 247
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0297,
+      "step": 248
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0166,
+      "step": 249
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0809,
+      "step": 250
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0964,
+      "step": 251
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0212,
+      "step": 252
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0685,
+      "step": 253
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0599,
+      "step": 254
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0527,
+      "step": 255
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0657,
+      "step": 256
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0047,
+      "step": 257
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1e-05,
+      "loss": 0.1389,
+      "step": 258
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1e-05,
+      "loss": 0.041,
+      "step": 259
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0826,
+      "step": 260
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1e-05,
+      "loss": 0.172,
+      "step": 261
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0584,
+      "step": 262
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0729,
+      "step": 263
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0758,
+      "step": 264
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0384,
+      "step": 265
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0507,
+      "step": 266
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0573,
+      "step": 267
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0237,
+      "step": 268
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0828,
+      "step": 269
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0392,
+      "step": 270
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1e-05,
+      "loss": 0.032,
+      "step": 271
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0119,
+      "step": 272
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0365,
+      "step": 273
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0832,
+      "step": 274
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0032,
+      "step": 275
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1e-05,
+      "loss": 0.028,
+      "step": 276
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1e-05,
+      "loss": 0.1069,
+      "step": 277
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1e-05,
+      "loss": 0.1684,
+      "step": 278
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0604,
+      "step": 279
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0674,
+      "step": 280
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1e-05,
+      "loss": 0.1875,
+      "step": 281
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0029,
+      "step": 282
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0255,
+      "step": 283
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0997,
+      "step": 284
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0452,
+      "step": 285
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0746,
+      "step": 286
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0116,
+      "step": 287
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0664,
+      "step": 288
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1e-05,
+      "loss": 0.1026,
+      "step": 289
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0351,
+      "step": 290
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0463,
+      "step": 291
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1e-05,
+      "loss": 0.2448,
+      "step": 292
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1e-05,
+      "loss": 0.1813,
+      "step": 293
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0877,
+      "step": 294
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1e-05,
+      "loss": 0.074,
+      "step": 295
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0288,
+      "step": 296
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1e-05,
+      "loss": 0.021,
+      "step": 297
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0533,
+      "step": 298
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1e-05,
+      "loss": 0.042,
+      "step": 299
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1e-05,
+      "loss": 0.1755,
+      "step": 300
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 301
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0291,
+      "step": 302
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0373,
+      "step": 303
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1e-05,
+      "loss": 0.1668,
+      "step": 304
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1e-05,
+      "loss": 0.1088,
+      "step": 305
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0273,
+      "step": 306
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1e-05,
+      "loss": 0.1134,
+      "step": 307
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0703,
+      "step": 308
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0572,
+      "step": 309
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0153,
+      "step": 310
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0403,
+      "step": 311
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0547,
+      "step": 312
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0689,
+      "step": 313
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0714,
+      "step": 314
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0256,
+      "step": 315
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0505,
+      "step": 316
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0283,
+      "step": 317
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1e-05,
+      "loss": 0.1025,
+      "step": 318
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0101,
+      "step": 319
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0805,
+      "step": 320
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0834,
+      "step": 321
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0599,
+      "step": 322
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0204,
+      "step": 323
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0145,
+      "step": 324
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1e-05,
+      "loss": 0.1251,
+      "step": 325
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0421,
+      "step": 326
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0871,
+      "step": 327
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1e-05,
+      "loss": 0.096,
+      "step": 328
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0823,
+      "step": 329
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1e-05,
+      "loss": 0.1059,
+      "step": 330
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1e-05,
+      "loss": 0.039,
+      "step": 331
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0183,
+      "step": 332
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0215,
+      "step": 333
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0335,
+      "step": 334
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0089,
+      "step": 335
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1e-05,
+      "loss": 0.001,
+      "step": 336
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1e-05,
+      "loss": 0.1457,
+      "step": 337
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0318,
+      "step": 338
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1e-05,
+      "loss": 0.1633,
+      "step": 339
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1e-05,
+      "loss": 0.1226,
+      "step": 340
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0449,
+      "step": 341
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0969,
+      "step": 342
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0949,
+      "step": 343
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0887,
+      "step": 344
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0147,
+      "step": 345
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1e-05,
+      "loss": 0.021,
+      "step": 346
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0226,
+      "step": 347
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0279,
+      "step": 348
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0476,
+      "step": 349
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0382,
+      "step": 350
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0442,
+      "step": 351
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0845,
+      "step": 352
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0842,
+      "step": 353
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0967,
+      "step": 354
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0387,
+      "step": 355
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0396,
+      "step": 356
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0364,
+      "step": 357
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0366,
+      "step": 358
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0676,
+      "step": 359
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1e-05,
+      "loss": 0.054,
+      "step": 360
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1e-05,
+      "loss": 0.025,
+      "step": 361
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1e-05,
+      "loss": 0.1253,
+      "step": 362
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0427,
+      "step": 363
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0504,
+      "step": 364
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1e-05,
+      "loss": 0.1029,
+      "step": 365
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0917,
+      "step": 366
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1e-05,
+      "loss": 0.1112,
+      "step": 367
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1e-05,
+      "loss": 0.068,
+      "step": 368
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0719,
+      "step": 369
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0385,
+      "step": 370
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1e-05,
+      "loss": 0.1012,
+      "step": 371
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0363,
+      "step": 372
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0378,
+      "step": 373
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1e-05,
+      "loss": 0.1173,
+      "step": 374
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1e-05,
+      "loss": 0.1131,
+      "step": 375
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0358,
+      "step": 376
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1e-05,
+      "loss": 0.1565,
+      "step": 377
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0629,
+      "step": 378
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0342,
+      "step": 379
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1e-05,
+      "loss": 0.1085,
+      "step": 380
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0749,
+      "step": 381
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1e-05,
+      "loss": 0.1015,
+      "step": 382
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0954,
+      "step": 383
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0926,
+      "step": 384
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1e-05,
+      "loss": 0.135,
+      "step": 385
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0751,
+      "step": 386
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0873,
+      "step": 387
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0328,
+      "step": 388
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0938,
+      "step": 389
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0299,
+      "step": 390
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1e-05,
+      "loss": 0.25,
+      "step": 391
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0366,
+      "step": 392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0182,
+      "step": 393
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0399,
+      "step": 394
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0875,
+      "step": 395
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0208,
+      "step": 396
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0359,
+      "step": 397
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0998,
+      "step": 398
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0102,
+      "step": 399
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0126,
+      "step": 400
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0814,
+      "step": 401
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0727,
+      "step": 402
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1e-05,
+      "loss": 0.022,
+      "step": 403
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1e-05,
+      "loss": 0.056,
+      "step": 404
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0691,
+      "step": 405
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 406
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0156,
+      "step": 407
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0529,
+      "step": 408
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1e-05,
+      "loss": 0.031,
+      "step": 409
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0159,
+      "step": 410
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0461,
+      "step": 411
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0773,
+      "step": 412
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0845,
+      "step": 413
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0074,
+      "step": 414
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0143,
+      "step": 415
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0104,
+      "step": 416
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0436,
+      "step": 417
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0151,
+      "step": 418
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0272,
+      "step": 419
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0761,
+      "step": 420
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0123,
+      "step": 421
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0681,
+      "step": 422
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0303,
+      "step": 423
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0555,
+      "step": 424
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0878,
+      "step": 425
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0801,
+      "step": 426
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0284,
+      "step": 427
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1e-05,
+      "loss": 0.041,
+      "step": 428
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1e-05,
+      "loss": 0.1293,
+      "step": 429
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0437,
+      "step": 430
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1e-05,
+      "loss": 0.07,
+      "step": 431
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1e-05,
+      "loss": 0.057,
+      "step": 432
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0213,
+      "step": 433
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0186,
+      "step": 434
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0132,
+      "step": 435
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0487,
+      "step": 436
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0066,
+      "step": 437
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0793,
+      "step": 438
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0406,
+      "step": 439
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0216,
+      "step": 440
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0269,
+      "step": 441
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0583,
+      "step": 442
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0421,
+      "step": 443
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0305,
+      "step": 444
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0476,
+      "step": 445
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0188,
+      "step": 446
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0147,
+      "step": 447
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0272,
+      "step": 448
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0036,
+      "step": 449
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0176,
+      "step": 450
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0116,
+      "step": 451
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0749,
+      "step": 452
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0338,
+      "step": 453
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1e-05,
+      "loss": 0.051,
+      "step": 454
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0291,
+      "step": 455
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0544,
+      "step": 456
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0319,
+      "step": 457
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1e-05,
+      "loss": 0.1306,
+      "step": 458
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0187,
+      "step": 459
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0082,
+      "step": 460
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0135,
+      "step": 461
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0405,
+      "step": 462
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0217,
+      "step": 463
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0575,
+      "step": 464
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0798,
+      "step": 465
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0174,
+      "step": 466
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0632,
+      "step": 467
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0061,
+      "step": 468
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0144,
+      "step": 469
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0324,
+      "step": 470
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1e-05,
+      "loss": 0.0593,
+      "step": 471
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1e-05,
+      "loss": 0.0274,
+      "step": 472
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0252,
+      "step": 473
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0239,
+      "step": 474
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0207,
+      "step": 475
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1e-05,
+      "loss": 0.046,
+      "step": 476
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0387,
+      "step": 477
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0267,
+      "step": 478
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0253,
+      "step": 479
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 480
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1e-05,
+      "loss": 0.081,
+      "step": 481
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0126,
+      "step": 482
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0183,
+      "step": 483
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0195,
+      "step": 484
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0609,
+      "step": 485
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0382,
+      "step": 486
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0207,
+      "step": 487
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1e-05,
+      "loss": 0.022,
+      "step": 488
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1e-05,
+      "loss": 0.1224,
+      "step": 489
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 490
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0294,
+      "step": 491
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1e-05,
+      "loss": 0.061,
+      "step": 492
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0948,
+      "step": 493
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0544,
+      "step": 494
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1e-05,
+      "loss": 0.063,
+      "step": 495
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0178,
+      "step": 496
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0957,
+      "step": 497
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "step": 498
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 499
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0004,
+      "step": 500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0074,
+      "step": 501
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0368,
+      "step": 502
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0316,
+      "step": 503
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0738,
+      "step": 504
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0463,
+      "step": 505
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0237,
+      "step": 506
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0076,
+      "step": 507
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0393,
+      "step": 508
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1e-05,
+      "loss": 0.1051,
+      "step": 509
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1e-05,
+      "loss": 0.1106,
+      "step": 510
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0295,
+      "step": 511
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0935,
+      "step": 512
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0326,
+      "step": 513
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 514
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1e-05,
+      "loss": 0.015,
+      "step": 515
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0312,
+      "step": 516
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0144,
+      "step": 517
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0474,
+      "step": 518
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0682,
+      "step": 519
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0238,
+      "step": 520
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0927,
+      "step": 521
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0074,
+      "step": 522
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0222,
+      "step": 523
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0841,
+      "step": 524
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1e-05,
+      "loss": 0.1028,
+      "step": 525
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1e-05,
+      "loss": 0.1052,
+      "step": 526
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0465,
+      "step": 527
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0116,
+      "step": 528
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0257,
+      "step": 529
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0691,
+      "step": 530
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0721,
+      "step": 531
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1e-05,
+      "loss": 0.072,
+      "step": 532
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0938,
+      "step": 533
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0141,
+      "step": 534
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1e-05,
+      "loss": 0.064,
+      "step": 535
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0194,
+      "step": 536
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0304,
+      "step": 537
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0347,
+      "step": 538
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0282,
+      "step": 539
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0635,
+      "step": 540
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0376,
+      "step": 541
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0686,
+      "step": 542
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0183,
+      "step": 543
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1e-05,
+      "loss": 0.1212,
+      "step": 544
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1e-05,
+      "loss": 0.018,
+      "step": 545
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0446,
+      "step": 546
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0031,
+      "step": 547
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0174,
+      "step": 548
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0256,
+      "step": 549
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1e-05,
+      "loss": 0.1656,
+      "step": 550
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0738,
+      "step": 551
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0873,
+      "step": 552
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0372,
+      "step": 553
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0425,
+      "step": 554
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0291,
+      "step": 555
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0214,
+      "step": 556
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0401,
+      "step": 557
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0679,
+      "step": 558
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0271,
+      "step": 559
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0211,
+      "step": 560
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0704,
+      "step": 561
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1e-05,
+      "loss": 0.1727,
+      "step": 562
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0594,
+      "step": 563
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0852,
+      "step": 564
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0152,
+      "step": 565
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1e-05,
+      "loss": 0.02,
+      "step": 566
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0298,
+      "step": 567
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 568
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0354,
+      "step": 569
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0093,
+      "step": 570
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0266,
+      "step": 571
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0313,
+      "step": 572
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1e-05,
+      "loss": 0.1279,
+      "step": 573
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1e-05,
+      "loss": 0.1133,
+      "step": 574
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0026,
+      "step": 575
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0782,
+      "step": 576
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0416,
+      "step": 577
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0081,
+      "step": 578
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0213,
+      "step": 579
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0061,
+      "step": 580
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0564,
+      "step": 581
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0517,
+      "step": 582
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0276,
+      "step": 583
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0617,
+      "step": 584
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0656,
+      "step": 585
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1e-05,
+      "loss": 0.035,
+      "step": 586
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0153,
+      "step": 587
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0274,
+      "step": 588
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0145,
+      "step": 589
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0075,
+      "step": 590
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 591
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0537,
+      "step": 592
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0137,
+      "step": 593
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 594
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0129,
+      "step": 595
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0151,
+      "step": 596
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0347,
+      "step": 597
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1e-05,
+      "loss": 0.1182,
+      "step": 598
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0445,
+      "step": 599
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1e-05,
+      "loss": 0.1153,
+      "step": 600
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0183,
+      "step": 601
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1e-05,
+      "loss": 0.1174,
+      "step": 602
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0447,
+      "step": 603
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0211,
+      "step": 604
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0001,
+      "step": 605
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0681,
+      "step": 606
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 607
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0516,
+      "step": 608
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0177,
+      "step": 609
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0015,
+      "step": 610
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 611
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0019,
+      "step": 612
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0158,
+      "step": 613
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0087,
+      "step": 614
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0066,
+      "step": 615
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0015,
+      "step": 616
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0435,
+      "step": 617
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0325,
+      "step": 618
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0163,
+      "step": 619
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0404,
+      "step": 620
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0107,
+      "step": 621
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0217,
+      "step": 622
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1e-05,
+      "loss": 0.1039,
+      "step": 623
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 624
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0186,
+      "step": 625
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1e-05,
+      "loss": 0.017,
+      "step": 626
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0025,
+      "step": 627
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0041,
+      "step": 628
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1e-05,
+      "loss": 0.016,
+      "step": 629
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0354,
+      "step": 630
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0111,
+      "step": 631
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1e-05,
+      "loss": 0.084,
+      "step": 632
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1e-05,
+      "loss": 0.028,
+      "step": 633
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0048,
+      "step": 634
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0096,
+      "step": 635
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 636
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1e-05,
+      "loss": 0.1489,
+      "step": 637
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0037,
+      "step": 638
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0878,
+      "step": 639
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0076,
+      "step": 640
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0123,
+      "step": 641
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0499,
+      "step": 642
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0102,
+      "step": 643
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0265,
+      "step": 644
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0144,
+      "step": 645
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0245,
+      "step": 646
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0482,
+      "step": 647
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0146,
+      "step": 648
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0268,
+      "step": 649
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0232,
+      "step": 650
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0168,
+      "step": 651
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0155,
+      "step": 652
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0496,
+      "step": 653
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0169,
+      "step": 654
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 655
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0529,
+      "step": 656
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0111,
+      "step": 657
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0954,
+      "step": 658
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0222,
+      "step": 659
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1e-05,
+      "loss": 0.019,
+      "step": 660
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0166,
+      "step": 661
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1e-05,
+      "loss": 0.024,
+      "step": 662
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0682,
+      "step": 663
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0676,
+      "step": 664
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0017,
+      "step": 665
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0849,
+      "step": 666
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 667
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1e-05,
+      "loss": 0.1424,
+      "step": 668
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1e-05,
+      "loss": 0.1053,
+      "step": 669
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0417,
+      "step": 670
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0578,
+      "step": 671
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0364,
+      "step": 672
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0583,
+      "step": 673
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0337,
+      "step": 674
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0048,
+      "step": 675
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0901,
+      "step": 676
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0958,
+      "step": 677
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0605,
+      "step": 678
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0221,
+      "step": 679
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1e-05,
+      "loss": 0.009,
+      "step": 680
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0236,
+      "step": 681
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 682
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0099,
+      "step": 683
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0159,
+      "step": 684
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0645,
+      "step": 685
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0404,
+      "step": 686
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0136,
+      "step": 687
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0299,
+      "step": 688
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0017,
+      "step": 689
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1e-05,
+      "loss": 0.1964,
+      "step": 690
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 691
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0227,
+      "step": 692
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0574,
+      "step": 693
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1e-05,
+      "loss": 0.1469,
+      "step": 694
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0734,
+      "step": 695
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1e-05,
+      "loss": 0.1699,
+      "step": 696
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0328,
+      "step": 697
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0255,
+      "step": 698
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0131,
+      "step": 699
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0203,
+      "step": 700
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 701
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0289,
+      "step": 702
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0152,
+      "step": 703
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1e-05,
+      "loss": 0.04,
+      "step": 704
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0675,
+      "step": 705
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1e-05,
+      "loss": 0.1065,
+      "step": 706
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0459,
+      "step": 707
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0168,
+      "step": 708
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1e-05,
+      "loss": 0.1417,
+      "step": 709
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0311,
+      "step": 710
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0499,
+      "step": 711
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0463,
+      "step": 712
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0196,
+      "step": 713
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0913,
+      "step": 714
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0069,
+      "step": 715
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0614,
+      "step": 716
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0156,
+      "step": 717
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0663,
+      "step": 718
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 719
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0366,
+      "step": 720
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0066,
+      "step": 721
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0245,
+      "step": 722
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0113,
+      "step": 723
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0165,
+      "step": 724
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0322,
+      "step": 725
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0445,
+      "step": 726
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0187,
+      "step": 727
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0121,
+      "step": 728
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0906,
+      "step": 729
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1e-05,
+      "loss": 1.1047,
+      "step": 730
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0414,
+      "step": 731
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0593,
+      "step": 732
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0972,
+      "step": 733
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0155,
+      "step": 734
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 735
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0085,
+      "step": 736
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0704,
+      "step": 737
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0245,
+      "step": 738
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0404,
+      "step": 739
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0119,
+      "step": 740
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1e-05,
+      "loss": 0.015,
+      "step": 741
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1e-05,
+      "loss": 0.023,
+      "step": 742
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 743
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0196,
+      "step": 744
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0361,
+      "step": 745
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0726,
+      "step": 746
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0749,
+      "step": 747
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1e-05,
+      "loss": 0.076,
+      "step": 748
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "step": 749
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0346,
+      "step": 750
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0294,
+      "step": 751
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0949,
+      "step": 752
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0242,
+      "step": 753
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0437,
+      "step": 754
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 755
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0189,
+      "step": 756
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1e-05,
+      "loss": 0.03,
+      "step": 757
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0082,
+      "step": 758
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0612,
+      "step": 759
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0174,
+      "step": 760
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0124,
+      "step": 761
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0276,
+      "step": 762
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0406,
+      "step": 763
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0645,
+      "step": 764
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1e-05,
+      "loss": 0.003,
+      "step": 765
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0307,
+      "step": 766
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0154,
+      "step": 767
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0919,
+      "step": 768
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0569,
+      "step": 769
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0263,
+      "step": 770
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0175,
+      "step": 771
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0174,
+      "step": 772
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0307,
+      "step": 773
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1e-05,
+      "loss": 0.3598,
+      "step": 774
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0602,
+      "step": 775
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0278,
+      "step": 776
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0101,
+      "step": 777
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0334,
+      "step": 778
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0468,
+      "step": 779
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0199,
+      "step": 780
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0427,
+      "step": 781
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0092,
+      "step": 782
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 783
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0011,
+      "step": 784
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0364,
+      "step": 785
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 786
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0388,
+      "step": 787
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0147,
+      "step": 788
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0057,
+      "step": 789
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0173,
+      "step": 790
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0499,
+      "step": 791
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0119,
+      "step": 792
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0192,
+      "step": 793
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 794
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0248,
+      "step": 795
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0222,
+      "step": 796
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0024,
+      "step": 797
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 798
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 799
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0657,
+      "step": 800
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0138,
+      "step": 801
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0732,
+      "step": 802
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0032,
+      "step": 803
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1e-05,
+      "loss": 0.008,
+      "step": 804
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0206,
+      "step": 805
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1e-05,
+      "loss": 0.013,
+      "step": 806
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0155,
+      "step": 807
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 808
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0331,
+      "step": 809
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0148,
+      "step": 810
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0173,
+      "step": 811
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0033,
+      "step": 812
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0217,
+      "step": 813
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0332,
+      "step": 814
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0121,
+      "step": 815
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0106,
+      "step": 816
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0241,
+      "step": 817
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0271,
+      "step": 818
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "step": 819
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 820
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "step": 821
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1e-05,
+      "loss": 0.008,
+      "step": 822
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 823
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0108,
+      "step": 824
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0131,
+      "step": 825
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0346,
+      "step": 826
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0625,
+      "step": 827
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0916,
+      "step": 828
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0556,
+      "step": 829
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1e-05,
+      "loss": 0.053,
+      "step": 830
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 831
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0294,
+      "step": 832
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0031,
+      "step": 833
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0077,
+      "step": 834
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0109,
+      "step": 835
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0169,
+      "step": 836
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0113,
+      "step": 837
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0093,
+      "step": 838
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0102,
+      "step": 839
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0387,
+      "step": 840
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0135,
+      "step": 841
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0272,
+      "step": 842
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1e-05,
+      "loss": 0.1128,
+      "step": 843
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1e-05,
+      "loss": 0.022,
+      "step": 844
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0061,
+      "step": 845
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0134,
+      "step": 846
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0036,
+      "step": 847
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0145,
+      "step": 848
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 849
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0025,
+      "step": 850
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0036,
+      "step": 851
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0317,
+      "step": 852
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0179,
+      "step": 853
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0224,
+      "step": 854
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0264,
+      "step": 855
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1e-05,
+      "loss": 0.1031,
+      "step": 856
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0061,
+      "step": 857
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0178,
+      "step": 858
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1e-05,
+      "loss": 0.1004,
+      "step": 859
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0144,
+      "step": 860
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0335,
+      "step": 861
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0311,
+      "step": 862
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 863
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1e-05,
+      "loss": 0.027,
+      "step": 864
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0005,
+      "step": 865
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0173,
+      "step": 866
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0117,
+      "step": 867
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0224,
+      "step": 868
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0062,
+      "step": 869
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0258,
+      "step": 870
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0675,
+      "step": 871
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0137,
+      "step": 872
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0789,
+      "step": 873
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0033,
+      "step": 874
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0181,
+      "step": 875
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0307,
+      "step": 876
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 877
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0058,
+      "step": 878
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0089,
+      "step": 879
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0346,
+      "step": 880
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1e-05,
+      "loss": 0.009,
+      "step": 881
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0081,
+      "step": 882
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0059,
+      "step": 883
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0404,
+      "step": 884
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0088,
+      "step": 885
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0362,
+      "step": 886
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0671,
+      "step": 887
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 888
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0077,
+      "step": 889
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0563,
+      "step": 890
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0199,
+      "step": 891
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0098,
+      "step": 892
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0044,
+      "step": 893
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0029,
+      "step": 894
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0105,
+      "step": 895
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0621,
+      "step": 896
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 897
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0059,
+      "step": 898
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1e-05,
+      "loss": 0.011,
+      "step": 899
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1e-05,
+      "loss": 0.009,
+      "step": 900
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0507,
+      "step": 901
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 902
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0893,
+      "step": 903
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1e-05,
+      "loss": 0.019,
+      "step": 904
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0131,
+      "step": 905
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0075,
+      "step": 906
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0055,
+      "step": 907
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0294,
+      "step": 908
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0904,
+      "step": 909
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1e-05,
+      "loss": 0.001,
+      "step": 910
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0133,
+      "step": 911
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0646,
+      "step": 912
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0442,
+      "step": 913
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0433,
+      "step": 914
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0256,
+      "step": 915
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0372,
+      "step": 916
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0172,
+      "step": 917
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 918
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0487,
+      "step": 919
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0081,
+      "step": 920
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0362,
+      "step": 921
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0231,
+      "step": 922
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0048,
+      "step": 923
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 924
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0153,
+      "step": 925
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0018,
+      "step": 926
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0264,
+      "step": 927
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0128,
+      "step": 928
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0013,
+      "step": 929
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0016,
+      "step": 930
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0252,
+      "step": 931
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0117,
+      "step": 932
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0239,
+      "step": 933
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0073,
+      "step": 934
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0411,
+      "step": 935
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0631,
+      "step": 936
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0257,
+      "step": 937
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0039,
+      "step": 938
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0455,
+      "step": 939
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0219,
+      "step": 940
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0954,
+      "step": 941
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0019,
+      "step": 942
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0508,
+      "step": 943
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0181,
+      "step": 944
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 945
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0039,
+      "step": 946
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0929,
+      "step": 947
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0421,
+      "step": 948
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0419,
+      "step": 949
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0062,
+      "step": 950
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0104,
+      "step": 951
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0133,
+      "step": 952
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1e-05,
+      "loss": 0.019,
+      "step": 953
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0215,
+      "step": 954
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0113,
+      "step": 955
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 956
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0223,
+      "step": 957
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0169,
+      "step": 958
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0088,
+      "step": 959
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0126,
+      "step": 960
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1e-05,
+      "loss": 0.033,
+      "step": 961
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0149,
+      "step": 962
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0033,
+      "step": 963
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0407,
+      "step": 964
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0089,
+      "step": 965
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0839,
+      "step": 966
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0378,
+      "step": 967
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0448,
+      "step": 968
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0478,
+      "step": 969
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 970
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0385,
+      "step": 971
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0097,
+      "step": 972
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0004,
+      "step": 973
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0188,
+      "step": 974
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0204,
+      "step": 975
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0363,
+      "step": 976
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0145,
+      "step": 977
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "step": 978
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0545,
+      "step": 979
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0344,
+      "step": 980
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0097,
+      "step": 981
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0035,
+      "step": 982
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0301,
+      "step": 983
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0112,
+      "step": 984
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1e-05,
+      "loss": 0.1391,
+      "step": 985
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0044,
+      "step": 986
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 987
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1e-05,
+      "loss": 0.008,
+      "step": 988
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0013,
+      "step": 989
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0014,
+      "step": 990
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0073,
+      "step": 991
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0062,
+      "step": 992
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0147,
+      "step": 993
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0074,
+      "step": 994
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0036,
+      "step": 995
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0099,
+      "step": 996
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0099,
+      "step": 997
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 998
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1e-05,
+      "loss": 0.007,
+      "step": 999
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "step": 1000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0064,
+      "step": 1001
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1e-05,
+      "loss": 0.002,
+      "step": 1002
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0969,
+      "step": 1003
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0139,
+      "step": 1004
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0659,
+      "step": 1005
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1e-05,
+      "loss": 0.008,
+      "step": 1006
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0148,
+      "step": 1007
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 1008
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0015,
+      "step": 1009
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0047,
+      "step": 1010
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0012,
+      "step": 1011
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1e-05,
+      "loss": 0.023,
+      "step": 1012
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0078,
+      "step": 1013
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 1014
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "step": 1015
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1e-05,
+      "loss": 0.0141,
+      "step": 1016
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0071,
+      "step": 1017
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 1018
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0018,
+      "step": 1019
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0133,
+      "step": 1020
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 1021
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1e-05,
+      "loss": 0.013,
+      "step": 1022
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0041,
+      "step": 1023
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0131,
+      "step": 1024
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 1025
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 1026
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0055,
+      "step": 1027
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0704,
+      "step": 1028
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 1029
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0325,
+      "step": 1030
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0106,
+      "step": 1031
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0135,
+      "step": 1032
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0225,
+      "step": 1033
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0205,
+      "step": 1034
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1e-05,
+      "loss": 0.019,
+      "step": 1035
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0324,
+      "step": 1036
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0061,
+      "step": 1037
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1038
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0042,
+      "step": 1039
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0108,
+      "step": 1040
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 1e-05,
+      "loss": 0.1134,
+      "step": 1041
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0139,
+      "step": 1042
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0156,
+      "step": 1043
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1e-05,
+      "loss": 0.012,
+      "step": 1044
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0542,
+      "step": 1045
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0213,
+      "step": 1046
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0164,
+      "step": 1047
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1048
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1e-05,
+      "loss": 0.065,
+      "step": 1049
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 1050
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1e-05,
+      "loss": 0.001,
+      "step": 1051
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0257,
+      "step": 1052
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 1053
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 1054
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "step": 1055
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0274,
+      "step": 1056
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0333,
+      "step": 1057
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0216,
+      "step": 1058
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 1e-05,
+      "loss": 0.018,
+      "step": 1059
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0134,
+      "step": 1060
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0326,
+      "step": 1061
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 1062
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0107,
+      "step": 1063
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0157,
+      "step": 1064
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0397,
+      "step": 1065
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0205,
+      "step": 1066
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0161,
+      "step": 1067
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 1e-05,
+      "loss": 0.003,
+      "step": 1068
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0025,
+      "step": 1069
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 1070
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0065,
+      "step": 1071
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 1072
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0526,
+      "step": 1073
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 1e-05,
+      "loss": 0.009,
+      "step": 1074
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0707,
+      "step": 1075
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0105,
+      "step": 1076
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0069,
+      "step": 1077
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0378,
+      "step": 1078
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0095,
+      "step": 1079
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0309,
+      "step": 1080
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1e-05,
+      "loss": 0.0301,
+      "step": 1081
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0011,
+      "step": 1082
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0093,
+      "step": 1083
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1084
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0052,
+      "step": 1085
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0114,
+      "step": 1086
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 1087
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0114,
+      "step": 1088
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0189,
+      "step": 1089
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0809,
+      "step": 1090
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0023,
+      "step": 1091
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 1092
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0166,
+      "step": 1093
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0096,
+      "step": 1094
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0078,
+      "step": 1095
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0093,
+      "step": 1096
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0687,
+      "step": 1097
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 1e-05,
+      "loss": 0.053,
+      "step": 1098
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "step": 1099
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0097,
+      "step": 1100
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1101
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0076,
+      "step": 1102
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0293,
+      "step": 1103
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 1104
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0183,
+      "step": 1105
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1e-05,
+      "loss": 0.026,
+      "step": 1106
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0249,
+      "step": 1107
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0359,
+      "step": 1108
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0166,
+      "step": 1109
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 1110
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0257,
+      "step": 1111
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0125,
+      "step": 1112
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0027,
+      "step": 1113
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0081,
+      "step": 1114
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0115,
+      "step": 1115
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0042,
+      "step": 1116
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0182,
+      "step": 1117
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0012,
+      "step": 1118
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0016,
+      "step": 1119
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0187,
+      "step": 1120
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0342,
+      "step": 1121
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 1122
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 1123
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1e-05,
+      "loss": 0.057,
+      "step": 1124
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0028,
+      "step": 1125
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0089,
+      "step": 1126
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0816,
+      "step": 1127
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0709,
+      "step": 1128
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0879,
+      "step": 1129
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0135,
+      "step": 1130
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0312,
+      "step": 1131
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0257,
+      "step": 1132
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0102,
+      "step": 1133
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0192,
+      "step": 1134
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0267,
+      "step": 1135
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0117,
+      "step": 1136
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0206,
+      "step": 1137
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 1138
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0033,
+      "step": 1139
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0248,
+      "step": 1140
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1e-05,
+      "loss": 0.049,
+      "step": 1141
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0078,
+      "step": 1142
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 1143
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0454,
+      "step": 1144
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0113,
+      "step": 1145
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0806,
+      "step": 1146
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0245,
+      "step": 1147
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0014,
+      "step": 1148
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 1149
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0726,
+      "step": 1150
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0083,
+      "step": 1151
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0581,
+      "step": 1152
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0611,
+      "step": 1153
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0195,
+      "step": 1154
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0024,
+      "step": 1155
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 1156
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0598,
+      "step": 1157
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0032,
+      "step": 1158
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1e-05,
+      "loss": 0.0341,
+      "step": 1159
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 1160
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 1e-05,
+      "loss": 0.014,
+      "step": 1161
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0188,
+      "step": 1162
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1e-05,
+      "loss": 0.006,
+      "step": 1163
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0111,
+      "step": 1164
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 1165
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0481,
+      "step": 1166
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0137,
+      "step": 1167
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0401,
+      "step": 1168
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0785,
+      "step": 1169
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0291,
+      "step": 1170
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0118,
+      "step": 1171
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0136,
+      "step": 1172
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0004,
+      "step": 1173
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0064,
+      "step": 1174
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0364,
+      "step": 1175
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0007,
+      "step": 1176
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 1e-05,
+      "loss": 0.0348,
+      "step": 1177
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1e-05,
+      "loss": 0.019,
+      "step": 1178
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 1179
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1e-05,
+      "loss": 0.0202,
+      "step": 1180
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1e-05,
+      "loss": 0.022,
+      "step": 1181
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0071,
+      "step": 1182
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1e-05,
+      "loss": 0.0112,
+      "step": 1183
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "step": 1184
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 1185
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0076,
+      "step": 1186
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1e-05,
+      "loss": 0.0039,
+      "step": 1187
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0122,
+      "step": 1188
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1e-05,
+      "loss": 0.0058,
+      "step": 1189
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "step": 1190
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1e-05,
+      "loss": 0.0107,
+      "step": 1191
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0152,
+      "step": 1192
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1e-05,
+      "loss": 0.0011,
+      "step": 1193
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0073,
+      "step": 1194
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1e-05,
+      "loss": 0.0155,
+      "step": 1195
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0165,
+      "step": 1196
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 1197
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1e-05,
+      "loss": 0.0039,
+      "step": 1198
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0152,
+      "step": 1199
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1200
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0251,
+      "step": 1201
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 1202
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0028,
+      "step": 1203
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1e-05,
+      "loss": 0.0093,
+      "step": 1204
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0586,
+      "step": 1205
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1e-05,
+      "loss": 0.0299,
+      "step": 1206
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0046,
+      "step": 1207
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1e-05,
+      "loss": 0.0342,
+      "step": 1208
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 1209
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1e-05,
+      "loss": 0.0197,
+      "step": 1210
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1e-05,
+      "loss": 0.021,
+      "step": 1211
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 1e-05,
+      "loss": 0.02,
+      "step": 1212
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0004,
+      "step": 1213
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 1214
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0209,
+      "step": 1215
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1e-05,
+      "loss": 0.0056,
+      "step": 1216
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0058,
+      "step": 1217
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1e-05,
+      "loss": 0.0082,
+      "step": 1218
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1e-05,
+      "loss": 0.003,
+      "step": 1219
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1e-05,
+      "loss": 0.0,
+      "step": 1220
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0135,
+      "step": 1221
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1e-05,
+      "loss": 0.0469,
+      "step": 1222
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0295,
+      "step": 1223
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1e-05,
+      "loss": 0.0011,
+      "step": 1224
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0114,
+      "step": 1225
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1e-05,
+      "loss": 0.0084,
+      "step": 1226
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0075,
+      "step": 1227
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1e-05,
+      "loss": 0.0104,
+      "step": 1228
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1e-05,
+      "loss": 0.1115,
+      "step": 1229
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1e-05,
+      "loss": 0.0087,
+      "step": 1230
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1e-05,
+      "loss": 0.015,
+      "step": 1231
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1e-05,
+      "loss": 0.0154,
+      "step": 1232
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0923,
+      "step": 1233
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1e-05,
+      "loss": 0.0127,
+      "step": 1234
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0448,
+      "step": 1235
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1e-05,
+      "loss": 0.0694,
+      "step": 1236
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0096,
+      "step": 1237
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1e-05,
+      "loss": 0.0178,
+      "step": 1238
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 1239
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1e-05,
+      "loss": 0.0009,
+      "step": 1240
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1e-05,
+      "loss": 0.0699,
+      "step": 1241
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1242
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1e-05,
+      "loss": 0.0028,
+      "step": 1243
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0003,
+      "step": 1244
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1e-05,
+      "loss": 0.0103,
+      "step": 1245
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0163,
+      "step": 1246
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1e-05,
+      "loss": 0.0098,
+      "step": 1247
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0124,
+      "step": 1248
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 1249
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0127,
+      "step": 1250
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 1e-05,
+      "loss": 0.0122,
+      "step": 1251
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0092,
+      "step": 1252
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 1253
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1e-05,
+      "loss": 0.028,
+      "step": 1254
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1e-05,
+      "loss": 0.1754,
+      "step": 1255
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0275,
+      "step": 1256
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 1257
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0048,
+      "step": 1258
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1e-05,
+      "loss": 0.0122,
+      "step": 1259
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0232,
+      "step": 1260
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 1e-05,
+      "loss": 0.0046,
+      "step": 1261
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 1262
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1e-05,
+      "loss": 0.01,
+      "step": 1263
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1264
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1e-05,
+      "loss": 0.0062,
+      "step": 1265
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1e-05,
+      "loss": 0.01,
+      "step": 1266
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1e-05,
+      "loss": 0.0104,
+      "step": 1267
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1e-05,
+      "loss": 0.014,
+      "step": 1268
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1e-05,
+      "loss": 0.0052,
+      "step": 1269
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0028,
+      "step": 1270
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1e-05,
+      "loss": 0.0053,
+      "step": 1271
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0185,
+      "step": 1272
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1e-05,
+      "loss": 0.0128,
+      "step": 1273
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0225,
+      "step": 1274
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 1275
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1e-05,
+      "loss": 0.008,
+      "step": 1276
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1e-05,
+      "loss": 0.001,
+      "step": 1277
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0,
+      "step": 1278
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1e-05,
+      "loss": 0.0052,
+      "step": 1279
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0021,
+      "step": 1280
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1e-05,
+      "loss": 0.0157,
+      "step": 1281
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1e-05,
+      "loss": 0.027,
+      "step": 1282
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1e-05,
+      "loss": 0.0199,
+      "step": 1283
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0232,
+      "step": 1284
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1e-05,
+      "loss": 0.0078,
+      "step": 1285
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1e-05,
+      "loss": 0.0051,
+      "step": 1286
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1287
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0005,
+      "step": 1288
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1e-05,
+      "loss": 0.0018,
+      "step": 1289
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 1e-05,
+      "loss": 0.0222,
+      "step": 1290
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0001,
+      "step": 1291
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1e-05,
+      "loss": 0.0161,
+      "step": 1292
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0082,
+      "step": 1293
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1e-05,
+      "loss": 0.0554,
+      "step": 1294
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0011,
+      "step": 1295
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1e-05,
+      "loss": 0.0125,
+      "step": 1296
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0047,
+      "step": 1297
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1e-05,
+      "loss": 0.0014,
+      "step": 1298
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0075,
+      "step": 1299
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 1e-05,
+      "loss": 0.0247,
+      "step": 1300
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0288,
+      "step": 1301
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1e-05,
+      "loss": 0.0585,
+      "step": 1302
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0001,
+      "step": 1303
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1e-05,
+      "loss": 0.0151,
+      "step": 1304
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0008,
+      "step": 1305
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1e-05,
+      "loss": 0.0378,
+      "step": 1306
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0074,
+      "step": 1307
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1e-05,
+      "loss": 0.0054,
+      "step": 1308
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1e-05,
+      "loss": 0.024,
+      "step": 1309
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1e-05,
+      "loss": 0.0049,
+      "step": 1310
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0139,
+      "step": 1311
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1e-05,
+      "loss": 0.0012,
+      "step": 1312
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0228,
+      "step": 1313
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1e-05,
+      "loss": 0.0001,
+      "step": 1314
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0079,
+      "step": 1315
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1e-05,
+      "loss": 0.0036,
+      "step": 1316
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0169,
+      "step": 1317
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1e-05,
+      "loss": 0.0005,
+      "step": 1318
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0075,
+      "step": 1319
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 1320
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1321
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1e-05,
+      "loss": 0.0069,
+      "step": 1322
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1e-05,
+      "loss": 0.0179,
+      "step": 1323
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1e-05,
+      "loss": 0.05,
+      "step": 1324
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0004,
+      "step": 1325
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1e-05,
+      "loss": 0.0022,
+      "step": 1326
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0018,
+      "step": 1327
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 1e-05,
+      "loss": 0.0063,
+      "step": 1328
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1329
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1e-05,
+      "loss": 0.0161,
+      "step": 1330
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1e-05,
+      "loss": 0.0095,
+      "step": 1331
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1e-05,
+      "loss": 0.009,
+      "step": 1332
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1e-05,
+      "loss": 0.1165,
+      "step": 1333
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1e-05,
+      "loss": 0.015,
+      "step": 1334
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1e-05,
+      "loss": 0.0038,
+      "step": 1335
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1e-05,
+      "loss": 0.0077,
+      "step": 1336
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1e-05,
+      "loss": 0.014,
+      "step": 1337
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0068,
+      "step": 1338
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1e-05,
+      "loss": 0.0044,
+      "step": 1339
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0039,
+      "step": 1340
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1e-05,
+      "loss": 0.0214,
+      "step": 1341
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0045,
+      "step": 1342
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1e-05,
+      "loss": 0.0249,
+      "step": 1343
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0231,
+      "step": 1344
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1e-05,
+      "loss": 0.0014,
+      "step": 1345
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1e-05,
+      "loss": 0.003,
+      "step": 1346
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1e-05,
+      "loss": 0.0067,
+      "step": 1347
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0006,
+      "step": 1348
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1e-05,
+      "loss": 0.0092,
+      "step": 1349
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1e-05,
+      "loss": 0.0342,
+      "step": 1350
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1e-05,
+      "loss": 0.005,
+      "step": 1351
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0126,
+      "step": 1352
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1e-05,
+      "loss": 0.0105,
+      "step": 1353
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1e-05,
+      "loss": 0.044,
+      "step": 1354
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1e-05,
+      "loss": 0.002,
+      "step": 1355
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0113,
+      "step": 1356
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1e-05,
+      "loss": 0.0087,
+      "step": 1357
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0043,
+      "step": 1358
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1e-05,
+      "loss": 0.0155,
+      "step": 1359
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0086,
+      "step": 1360
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1e-05,
+      "loss": 0.0002,
+      "step": 1361
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0281,
+      "step": 1362
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1e-05,
+      "loss": 0.0123,
+      "step": 1363
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0271,
+      "step": 1364
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1e-05,
+      "loss": 0.0395,
+      "step": 1365
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0648,
+      "step": 1366
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 1e-05,
+      "loss": 0.0213,
+      "step": 1367
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0336,
+      "step": 1368
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1e-05,
+      "loss": 0.0156,
+      "step": 1369
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1e-05,
+      "loss": 0.0012,
+      "step": 1370
+    }
+  ],
+  "max_steps": 1560,
+  "num_train_epochs": 8,
+  "total_flos": 734226001428480.0,
+  "trial_name": null,
+  "trial_params": null
+}