diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10927 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997248968363136,
+  "global_step": 1817,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 3.5287,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 3.4725,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 3.5201,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 3.5255,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 3.4729,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 3.2834,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7e-06,
+      "loss": 3.1705,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 3.0604,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-06,
+      "loss": 2.9558,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1e-05,
+      "loss": 2.8512,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 2.7594,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2e-05,
+      "loss": 2.6949,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 2.6266,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4e-05,
+      "loss": 2.5429,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 2.5144,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.4491,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7e-05,
+      "loss": 2.4587,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-05,
+      "loss": 2.3918,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9e-05,
+      "loss": 2.3708,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.3268,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.3011,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.3302,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.3179,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.3356,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.2943,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.2598,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 2.2332,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.247,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2522,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.22,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1994,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2543,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2392,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2315,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1895,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2543,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.2079,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1862,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1478,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1977,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1521,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.151,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1495,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1904,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 2.1684,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1907,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1834,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.195,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.2203,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1773,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1711,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1323,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1371,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1781,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1321,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1583,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1496,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1668,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1371,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.13,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1201,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.144,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 2.1217,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1507,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1845,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1272,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1133,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1423,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1157,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1512,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1491,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1123,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.145,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1352,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1291,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1301,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1492,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.155,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1116,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1392,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 2.1583,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.111,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1102,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.0885,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.14,
+      "step": 85
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1353,
+      "step": 86
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1629,
+      "step": 87
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1658,
+      "step": 88
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1315,
+      "step": 89
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1145,
+      "step": 90
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1345,
+      "step": 91
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1452,
+      "step": 92
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.0712,
+      "step": 93
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.0912,
+      "step": 94
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1401,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1085,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1292,
+      "step": 97
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1161,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.1245,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1083,
+      "step": 100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1103,
+      "step": 101
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1587,
+      "step": 102
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0944,
+      "step": 103
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0868,
+      "step": 104
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1105,
+      "step": 105
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1135,
+      "step": 106
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1306,
+      "step": 107
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1449,
+      "step": 108
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0915,
+      "step": 109
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1306,
+      "step": 110
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1084,
+      "step": 111
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1282,
+      "step": 112
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1205,
+      "step": 113
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1185,
+      "step": 114
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.108,
+      "step": 115
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1245,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0727,
+      "step": 117
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1333,
+      "step": 118
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1307,
+      "step": 119
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1595,
+      "step": 120
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0831,
+      "step": 121
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0709,
+      "step": 122
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1195,
+      "step": 123
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0876,
+      "step": 124
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1316,
+      "step": 125
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1464,
+      "step": 126
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1128,
+      "step": 127
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1309,
+      "step": 128
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1152,
+      "step": 129
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0959,
+      "step": 130
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1144,
+      "step": 131
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0785,
+      "step": 132
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0977,
+      "step": 133
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.118,
+      "step": 134
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.1033,
+      "step": 135
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.139,
+      "step": 136
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1166,
+      "step": 137
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0851,
+      "step": 138
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0661,
+      "step": 139
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0855,
+      "step": 140
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0968,
+      "step": 141
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.107,
+      "step": 142
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1196,
+      "step": 143
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0773,
+      "step": 144
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0794,
+      "step": 145
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1284,
+      "step": 146
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.091,
+      "step": 147
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1145,
+      "step": 148
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.079,
+      "step": 149
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1201,
+      "step": 150
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0725,
+      "step": 151
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1544,
+      "step": 152
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.1328,
+      "step": 153
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0945,
+      "step": 154
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1158,
+      "step": 155
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1438,
+      "step": 156
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1196,
+      "step": 157
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.087,
+      "step": 158
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1075,
+      "step": 159
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.0641,
+      "step": 160
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1242,
+      "step": 161
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.0691,
+      "step": 162
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1236,
+      "step": 163
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1142,
+      "step": 164
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.0581,
+      "step": 165
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1504,
+      "step": 166
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1257,
+      "step": 167
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1065,
+      "step": 168
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.0895,
+      "step": 169
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1096,
+      "step": 170
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1409,
+      "step": 171
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 2.1074,
+      "step": 172
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0818,
+      "step": 173
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0799,
+      "step": 174
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0459,
+      "step": 175
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.1076,
+      "step": 176
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0621,
+      "step": 177
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0721,
+      "step": 178
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0924,
+      "step": 179
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0771,
+      "step": 180
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.1166,
+      "step": 181
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0964,
+      "step": 182
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.1025,
+      "step": 183
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.1119,
+      "step": 184
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0872,
+      "step": 185
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.109,
+      "step": 186
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0924,
+      "step": 187
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.1081,
+      "step": 188
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0919,
+      "step": 189
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 2.0652,
+      "step": 190
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0797,
+      "step": 191
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0738,
+      "step": 192
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.104,
+      "step": 193
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1573,
+      "step": 194
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0781,
+      "step": 195
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0546,
+      "step": 196
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0969,
+      "step": 197
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0815,
+      "step": 198
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.086,
+      "step": 199
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1331,
+      "step": 200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1172,
+      "step": 201
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0747,
+      "step": 202
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0772,
+      "step": 203
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1179,
+      "step": 204
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0742,
+      "step": 205
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0997,
+      "step": 206
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.0957,
+      "step": 207
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1015,
+      "step": 208
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 2.1207,
+      "step": 209
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0774,
+      "step": 210
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.1076,
+      "step": 211
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.1016,
+      "step": 212
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0763,
+      "step": 213
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0663,
+      "step": 214
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0953,
+      "step": 215
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0939,
+      "step": 216
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0831,
+      "step": 217
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.1048,
+      "step": 218
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0606,
+      "step": 219
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.1579,
+      "step": 220
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.1238,
+      "step": 221
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0841,
+      "step": 222
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0926,
+      "step": 223
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0391,
+      "step": 224
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0377,
+      "step": 225
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.0656,
+      "step": 226
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 2.098,
+      "step": 227
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0589,
+      "step": 228
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.081,
+      "step": 229
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.1036,
+      "step": 230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0911,
+      "step": 231
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.1309,
+      "step": 232
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0144,
+      "step": 233
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0428,
+      "step": 234
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0677,
+      "step": 235
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0468,
+      "step": 236
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0817,
+      "step": 237
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0463,
+      "step": 238
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0632,
+      "step": 239
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.08,
+      "step": 240
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.1104,
+      "step": 241
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 1.9874,
+      "step": 242
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0387,
+      "step": 243
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0399,
+      "step": 244
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0234,
+      "step": 245
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0354,
+      "step": 246
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.062,
+      "step": 247
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0754,
+      "step": 248
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0509,
+      "step": 249
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0889,
+      "step": 250
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0585,
+      "step": 251
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0295,
+      "step": 252
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0718,
+      "step": 253
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0158,
+      "step": 254
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0573,
+      "step": 255
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0418,
+      "step": 256
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0592,
+      "step": 257
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0632,
+      "step": 258
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.075,
+      "step": 259
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0248,
+      "step": 260
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0812,
+      "step": 261
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.039,
+      "step": 262
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.006,
+      "step": 263
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0641,
+      "step": 264
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0758,
+      "step": 265
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0592,
+      "step": 266
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0131,
+      "step": 267
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 1.9898,
+      "step": 268
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0015,
+      "step": 269
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0264,
+      "step": 270
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0775,
+      "step": 271
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0201,
+      "step": 272
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.022,
+      "step": 273
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0361,
+      "step": 274
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0346,
+      "step": 275
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 1.9848,
+      "step": 276
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0037,
+      "step": 277
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0069,
+      "step": 278
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0366,
+      "step": 279
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0058,
+      "step": 280
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0307,
+      "step": 281
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0217,
+      "step": 282
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0178,
+      "step": 283
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0607,
+      "step": 284
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0516,
+      "step": 285
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 1.9862,
+      "step": 286
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0468,
+      "step": 287
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0202,
+      "step": 288
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0077,
+      "step": 289
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0277,
+      "step": 290
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0222,
+      "step": 291
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0107,
+      "step": 292
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0258,
+      "step": 293
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0141,
+      "step": 294
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.036,
+      "step": 295
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0142,
+      "step": 296
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0292,
+      "step": 297
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0052,
+      "step": 298
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 1.9944,
+      "step": 299
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0269,
+      "step": 300
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0423,
+      "step": 301
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.9734,
+      "step": 302
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.011,
+      "step": 303
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0527,
+      "step": 304
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.9949,
+      "step": 305
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.9886,
+      "step": 306
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0108,
+      "step": 307
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0021,
+      "step": 308
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.97,
+      "step": 309
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.9827,
+      "step": 310
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0207,
+      "step": 311
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0363,
+      "step": 312
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.017,
+      "step": 313
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0769,
+      "step": 314
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0107,
+      "step": 315
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.017,
+      "step": 316
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0316,
+      "step": 317
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 2.0552,
+      "step": 318
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0109,
+      "step": 319
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0351,
+      "step": 320
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.066,
+      "step": 321
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.9991,
+      "step": 322
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0569,
+      "step": 323
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.9812,
+      "step": 324
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0256,
+      "step": 325
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0228,
+      "step": 326
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0238,
+      "step": 327
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0219,
+      "step": 328
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.9757,
+      "step": 329
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.968,
+      "step": 330
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0322,
+      "step": 331
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0505,
+      "step": 332
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.9867,
+      "step": 333
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0157,
+      "step": 334
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 2.0101,
+      "step": 335
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.9784,
+      "step": 336
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0062,
+      "step": 337
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0132,
+      "step": 338
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9828,
+      "step": 339
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9784,
+      "step": 340
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0258,
+      "step": 341
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0497,
+      "step": 342
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9959,
+      "step": 343
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9886,
+      "step": 344
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.988,
+      "step": 345
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0362,
+      "step": 346
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9911,
+      "step": 347
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0171,
+      "step": 348
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.963,
+      "step": 349
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0294,
+      "step": 350
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0204,
+      "step": 351
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.9821,
+      "step": 352
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0163,
+      "step": 353
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 2.0196,
+      "step": 354
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.9599,
+      "step": 355
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0219,
+      "step": 356
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0059,
+      "step": 357
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0409,
+      "step": 358
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.022,
+      "step": 359
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0202,
+      "step": 360
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.9953,
+      "step": 361
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0186,
+      "step": 362
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0083,
+      "step": 363
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0242,
+      "step": 364
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.9946,
+      "step": 365
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0194,
+      "step": 366
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.016,
+      "step": 367
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0064,
+      "step": 368
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0174,
+      "step": 369
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0517,
+      "step": 370
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.9899,
+      "step": 371
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 2.0398,
+      "step": 372
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0075,
+      "step": 373
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0197,
+      "step": 374
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0232,
+      "step": 375
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0181,
+      "step": 376
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 1.9957,
+      "step": 377
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 1.9796,
+      "step": 378
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 1.9974,
+      "step": 379
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.018,
+      "step": 380
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0084,
+      "step": 381
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0074,
+      "step": 382
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0338,
+      "step": 383
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0229,
+      "step": 384
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0035,
+      "step": 385
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.001,
+      "step": 386
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0004,
+      "step": 387
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0068,
+      "step": 388
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0231,
+      "step": 389
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0099,
+      "step": 390
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0138,
+      "step": 391
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0237,
+      "step": 392
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9771,
+      "step": 393
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0113,
+      "step": 394
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0111,
+      "step": 395
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0193,
+      "step": 396
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9848,
+      "step": 397
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9846,
+      "step": 398
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0186,
+      "step": 399
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9788,
+      "step": 400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9959,
+      "step": 401
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.98,
+      "step": 402
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0022,
+      "step": 403
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9651,
+      "step": 404
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.993,
+      "step": 405
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0139,
+      "step": 406
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0236,
+      "step": 407
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0535,
+      "step": 408
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9998,
+      "step": 409
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0123,
+      "step": 410
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9617,
+      "step": 411
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.978,
+      "step": 412
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9763,
+      "step": 413
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0003,
+      "step": 414
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.003,
+      "step": 415
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9956,
+      "step": 416
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9868,
+      "step": 417
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9651,
+      "step": 418
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9764,
+      "step": 419
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0123,
+      "step": 420
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9654,
+      "step": 421
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0037,
+      "step": 422
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 1.963,
+      "step": 423
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0099,
+      "step": 424
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0487,
+      "step": 425
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0296,
+      "step": 426
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0028,
+      "step": 427
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0248,
+      "step": 428
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0234,
+      "step": 429
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9808,
+      "step": 430
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.018,
+      "step": 431
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0178,
+      "step": 432
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0056,
+      "step": 433
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9948,
+      "step": 434
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0177,
+      "step": 435
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9955,
+      "step": 436
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0092,
+      "step": 437
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0045,
+      "step": 438
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0012,
+      "step": 439
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0066,
+      "step": 440
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9919,
+      "step": 441
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9933,
+      "step": 442
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9325,
+      "step": 443
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0078,
+      "step": 444
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 1.9853,
+      "step": 445
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9903,
+      "step": 446
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9643,
+      "step": 447
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.994,
+      "step": 448
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.985,
+      "step": 449
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0437,
+      "step": 450
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9905,
+      "step": 451
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9768,
+      "step": 452
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0112,
+      "step": 453
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9895,
+      "step": 454
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9836,
+      "step": 455
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9899,
+      "step": 456
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0276,
+      "step": 457
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0286,
+      "step": 458
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0153,
+      "step": 459
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9821,
+      "step": 460
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9867,
+      "step": 461
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 1.9603,
+      "step": 462
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2e-05,
+      "loss": 2.0133,
+      "step": 463
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0002,
+      "step": 464
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9688,
+      "step": 465
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0163,
+      "step": 466
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0064,
+      "step": 467
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0301,
+      "step": 468
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9967,
+      "step": 469
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.021,
+      "step": 470
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9933,
+      "step": 471
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0313,
+      "step": 472
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0122,
+      "step": 473
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0004,
+      "step": 474
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.992,
+      "step": 475
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9558,
+      "step": 476
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9604,
+      "step": 477
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.9854,
+      "step": 478
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0098,
+      "step": 479
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0002,
+      "step": 480
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 2.0435,
+      "step": 481
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9963,
+      "step": 482
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0213,
+      "step": 483
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0013,
+      "step": 484
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9831,
+      "step": 485
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9873,
+      "step": 486
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9825,
+      "step": 487
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9943,
+      "step": 488
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9701,
+      "step": 489
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0015,
+      "step": 490
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0218,
+      "step": 491
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0145,
+      "step": 492
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9713,
+      "step": 493
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0122,
+      "step": 494
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0223,
+      "step": 495
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0324,
+      "step": 496
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9814,
+      "step": 497
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 2.0176,
+      "step": 498
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9991,
+      "step": 499
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9567,
+      "step": 500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9889,
+      "step": 501
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9452,
+      "step": 502
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9722,
+      "step": 503
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9763,
+      "step": 504
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 2.0067,
+      "step": 505
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 2.0024,
+      "step": 506
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9976,
+      "step": 507
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9999,
+      "step": 508
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 2.0095,
+      "step": 509
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.963,
+      "step": 510
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9813,
+      "step": 511
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9428,
+      "step": 512
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9961,
+      "step": 513
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9598,
+      "step": 514
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.9851,
+      "step": 515
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 2.0683,
+      "step": 516
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 2.0087,
+      "step": 517
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9929,
+      "step": 518
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0066,
+      "step": 519
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9806,
+      "step": 520
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.029,
+      "step": 521
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9634,
+      "step": 522
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9723,
+      "step": 523
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9784,
+      "step": 524
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0017,
+      "step": 525
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9686,
+      "step": 526
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0267,
+      "step": 527
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0175,
+      "step": 528
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0089,
+      "step": 529
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9551,
+      "step": 530
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9662,
+      "step": 531
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 2.0205,
+      "step": 532
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9934,
+      "step": 533
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9414,
+      "step": 534
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9823,
+      "step": 535
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.9876,
+      "step": 536
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9811,
+      "step": 537
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9469,
+      "step": 538
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0184,
+      "step": 539
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.009,
+      "step": 540
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9928,
+      "step": 541
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9609,
+      "step": 542
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9833,
+      "step": 543
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9614,
+      "step": 544
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0246,
+      "step": 545
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9965,
+      "step": 546
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9736,
+      "step": 547
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9948,
+      "step": 548
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0021,
+      "step": 549
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.983,
+      "step": 550
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0173,
+      "step": 551
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9767,
+      "step": 552
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0192,
+      "step": 553
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9983,
+      "step": 554
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9545,
+      "step": 555
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.977,
+      "step": 556
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9959,
+      "step": 557
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9843,
+      "step": 558
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9505,
+      "step": 559
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9841,
+      "step": 560
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.0031,
+      "step": 561
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9383,
+      "step": 562
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9909,
+      "step": 563
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9459,
+      "step": 564
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9689,
+      "step": 565
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.0106,
+      "step": 566
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9954,
+      "step": 567
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.001,
+      "step": 568
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9896,
+      "step": 569
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.0215,
+      "step": 570
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9864,
+      "step": 571
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9825,
+      "step": 572
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.0395,
+      "step": 573
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.0083,
+      "step": 574
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.0203,
+      "step": 575
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9917,
+      "step": 576
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9785,
+      "step": 577
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.027,
+      "step": 578
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9722,
+      "step": 579
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.0056,
+      "step": 580
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9664,
+      "step": 581
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9692,
+      "step": 582
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9794,
+      "step": 583
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9989,
+      "step": 584
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9534,
+      "step": 585
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9966,
+      "step": 586
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9426,
+      "step": 587
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.983,
+      "step": 588
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.0124,
+      "step": 589
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 1.9605,
+      "step": 590
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9862,
+      "step": 591
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9819,
+      "step": 592
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9707,
+      "step": 593
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9605,
+      "step": 594
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9798,
+      "step": 595
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 2.0164,
+      "step": 596
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9666,
+      "step": 597
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9713,
+      "step": 598
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9979,
+      "step": 599
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9807,
+      "step": 600
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9836,
+      "step": 601
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9766,
+      "step": 602
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9555,
+      "step": 603
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 2.0136,
+      "step": 604
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9896,
+      "step": 605
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 2.0367,
+      "step": 606
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9585,
+      "step": 607
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.9671,
+      "step": 608
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9508,
+      "step": 609
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9972,
+      "step": 610
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9648,
+      "step": 611
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9671,
+      "step": 612
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 2.0069,
+      "step": 613
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9872,
+      "step": 614
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9988,
+      "step": 615
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9799,
+      "step": 616
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9674,
+      "step": 617
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9516,
+      "step": 618
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9774,
+      "step": 619
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9866,
+      "step": 620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9472,
+      "step": 621
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 2.0026,
+      "step": 622
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9345,
+      "step": 623
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9691,
+      "step": 624
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9647,
+      "step": 625
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 2.0133,
+      "step": 626
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.9869,
+      "step": 627
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 2.0003,
+      "step": 628
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9777,
+      "step": 629
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 2.0139,
+      "step": 630
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9789,
+      "step": 631
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9576,
+      "step": 632
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 2.0034,
+      "step": 633
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9473,
+      "step": 634
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9738,
+      "step": 635
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9728,
+      "step": 636
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9815,
+      "step": 637
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9845,
+      "step": 638
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 2.0307,
+      "step": 639
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9984,
+      "step": 640
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9782,
+      "step": 641
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9966,
+      "step": 642
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9771,
+      "step": 643
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.9676,
+      "step": 644
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 2.0106,
+      "step": 645
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 2.0026,
+      "step": 646
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9878,
+      "step": 647
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9981,
+      "step": 648
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9698,
+      "step": 649
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9925,
+      "step": 650
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9074,
+      "step": 651
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9719,
+      "step": 652
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9706,
+      "step": 653
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9781,
+      "step": 654
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9557,
+      "step": 655
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9842,
+      "step": 656
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.999,
+      "step": 657
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9658,
+      "step": 658
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9775,
+      "step": 659
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 2.0151,
+      "step": 660
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 2.009,
+      "step": 661
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.9574,
+      "step": 662
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 2.0367,
+      "step": 663
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9856,
+      "step": 664
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9556,
+      "step": 665
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9806,
+      "step": 666
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9974,
+      "step": 667
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9528,
+      "step": 668
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 2.0069,
+      "step": 669
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9542,
+      "step": 670
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9736,
+      "step": 671
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9897,
+      "step": 672
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9429,
+      "step": 673
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9455,
+      "step": 674
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9618,
+      "step": 675
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9561,
+      "step": 676
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9981,
+      "step": 677
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 2.0017,
+      "step": 678
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9967,
+      "step": 679
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9826,
+      "step": 680
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.9834,
+      "step": 681
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9451,
+      "step": 682
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9786,
+      "step": 683
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 2.004,
+      "step": 684
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9618,
+      "step": 685
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9945,
+      "step": 686
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9456,
+      "step": 687
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9934,
+      "step": 688
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9465,
+      "step": 689
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9444,
+      "step": 690
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9615,
+      "step": 691
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9621,
+      "step": 692
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9919,
+      "step": 693
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9401,
+      "step": 694
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9854,
+      "step": 695
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9581,
+      "step": 696
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9539,
+      "step": 697
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.949,
+      "step": 698
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9497,
+      "step": 699
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9634,
+      "step": 700
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9702,
+      "step": 701
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9539,
+      "step": 702
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9984,
+      "step": 703
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9787,
+      "step": 704
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9762,
+      "step": 705
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.0105,
+      "step": 706
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9766,
+      "step": 707
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.0061,
+      "step": 708
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.994,
+      "step": 709
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.991,
+      "step": 710
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9699,
+      "step": 711
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9643,
+      "step": 712
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9603,
+      "step": 713
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9736,
+      "step": 714
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.0032,
+      "step": 715
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9929,
+      "step": 716
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9716,
+      "step": 717
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.8876,
+      "step": 718
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0171,
+      "step": 719
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0292,
+      "step": 720
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.991,
+      "step": 721
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9829,
+      "step": 722
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0048,
+      "step": 723
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0083,
+      "step": 724
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9769,
+      "step": 725
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9831,
+      "step": 726
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9475,
+      "step": 727
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 728
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9482,
+      "step": 729
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9726,
+      "step": 730
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9631,
+      "step": 731
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.962,
+      "step": 732
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9751,
+      "step": 733
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9694,
+      "step": 734
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.977,
+      "step": 735
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 1.9355,
+      "step": 736
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9973,
+      "step": 737
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9846,
+      "step": 738
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9664,
+      "step": 739
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9849,
+      "step": 740
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0251,
+      "step": 741
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 742
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9603,
+      "step": 743
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9669,
+      "step": 744
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9512,
+      "step": 745
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0142,
+      "step": 746
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0266,
+      "step": 747
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9906,
+      "step": 748
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0338,
+      "step": 749
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.001,
+      "step": 750
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9747,
+      "step": 751
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9423,
+      "step": 752
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.9633,
+      "step": 753
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0005,
+      "step": 754
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 2.0185,
+      "step": 755
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9393,
+      "step": 756
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 2.0104,
+      "step": 757
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 2.009,
+      "step": 758
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9949,
+      "step": 759
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.987,
+      "step": 760
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9566,
+      "step": 761
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.976,
+      "step": 762
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9725,
+      "step": 763
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9725,
+      "step": 764
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9668,
+      "step": 765
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9689,
+      "step": 766
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 2.0105,
+      "step": 767
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9802,
+      "step": 768
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 2.0073,
+      "step": 769
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.982,
+      "step": 770
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9894,
+      "step": 771
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.9993,
+      "step": 772
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9629,
+      "step": 773
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9879,
+      "step": 774
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9535,
+      "step": 775
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9619,
+      "step": 776
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9367,
+      "step": 777
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9935,
+      "step": 778
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9768,
+      "step": 779
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9276,
+      "step": 780
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9913,
+      "step": 781
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9684,
+      "step": 782
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9259,
+      "step": 783
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.983,
+      "step": 784
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 2.0079,
+      "step": 785
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9729,
+      "step": 786
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 2.0146,
+      "step": 787
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9836,
+      "step": 788
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9821,
+      "step": 789
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9501,
+      "step": 790
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9634,
+      "step": 791
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9501,
+      "step": 792
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9786,
+      "step": 793
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9975,
+      "step": 794
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9398,
+      "step": 795
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9401,
+      "step": 796
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 2.01,
+      "step": 797
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9682,
+      "step": 798
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9885,
+      "step": 799
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9347,
+      "step": 800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9986,
+      "step": 801
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9811,
+      "step": 802
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 2.0155,
+      "step": 803
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9464,
+      "step": 804
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9897,
+      "step": 805
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.946,
+      "step": 806
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9938,
+      "step": 807
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.9742,
+      "step": 808
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9465,
+      "step": 809
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9625,
+      "step": 810
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9393,
+      "step": 811
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9998,
+      "step": 812
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9365,
+      "step": 813
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9959,
+      "step": 814
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9903,
+      "step": 815
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9538,
+      "step": 816
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 2.0054,
+      "step": 817
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9783,
+      "step": 818
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9626,
+      "step": 819
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9331,
+      "step": 820
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9712,
+      "step": 821
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.975,
+      "step": 822
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9482,
+      "step": 823
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9712,
+      "step": 824
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.981,
+      "step": 825
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.9692,
+      "step": 826
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9964,
+      "step": 827
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.995,
+      "step": 828
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9522,
+      "step": 829
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.975,
+      "step": 830
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9659,
+      "step": 831
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9519,
+      "step": 832
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9378,
+      "step": 833
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9765,
+      "step": 834
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 2.0045,
+      "step": 835
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9481,
+      "step": 836
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9544,
+      "step": 837
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9974,
+      "step": 838
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9789,
+      "step": 839
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 2.0092,
+      "step": 840
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9757,
+      "step": 841
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9612,
+      "step": 842
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9567,
+      "step": 843
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9645,
+      "step": 844
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9953,
+      "step": 845
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0171,
+      "step": 846
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9891,
+      "step": 847
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0213,
+      "step": 848
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9723,
+      "step": 849
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9544,
+      "step": 850
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9532,
+      "step": 851
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0062,
+      "step": 852
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9749,
+      "step": 853
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9579,
+      "step": 854
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9354,
+      "step": 855
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9749,
+      "step": 856
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9715,
+      "step": 857
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9725,
+      "step": 858
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9265,
+      "step": 859
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9943,
+      "step": 860
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9474,
+      "step": 861
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9771,
+      "step": 862
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9538,
+      "step": 863
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9416,
+      "step": 864
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9498,
+      "step": 865
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9942,
+      "step": 866
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.0022,
+      "step": 867
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9752,
+      "step": 868
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9977,
+      "step": 869
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.971,
+      "step": 870
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9676,
+      "step": 871
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9332,
+      "step": 872
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9235,
+      "step": 873
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.0083,
+      "step": 874
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9661,
+      "step": 875
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9607,
+      "step": 876
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.0184,
+      "step": 877
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9937,
+      "step": 878
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9893,
+      "step": 879
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9791,
+      "step": 880
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9863,
+      "step": 881
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9645,
+      "step": 882
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9671,
+      "step": 883
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9752,
+      "step": 884
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9515,
+      "step": 885
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 2.0034,
+      "step": 886
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9442,
+      "step": 887
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9631,
+      "step": 888
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9569,
+      "step": 889
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9553,
+      "step": 890
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9717,
+      "step": 891
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9773,
+      "step": 892
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9593,
+      "step": 893
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.956,
+      "step": 894
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9669,
+      "step": 895
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9984,
+      "step": 896
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9854,
+      "step": 897
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9345,
+      "step": 898
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.9344,
+      "step": 899
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9556,
+      "step": 900
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9963,
+      "step": 901
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9278,
+      "step": 902
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.957,
+      "step": 903
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9462,
+      "step": 904
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9837,
+      "step": 905
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9876,
+      "step": 906
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.959,
+      "step": 907
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9735,
+      "step": 908
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9185,
+      "step": 909
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9494,
+      "step": 910
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9698,
+      "step": 911
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9685,
+      "step": 912
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9628,
+      "step": 913
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9759,
+      "step": 914
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.959,
+      "step": 915
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9703,
+      "step": 916
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.9498,
+      "step": 917
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.971,
+      "step": 918
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.975,
+      "step": 919
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9929,
+      "step": 920
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9918,
+      "step": 921
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9806,
+      "step": 922
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9595,
+      "step": 923
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9342,
+      "step": 924
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 2.0064,
+      "step": 925
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9305,
+      "step": 926
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.974,
+      "step": 927
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9324,
+      "step": 928
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9743,
+      "step": 929
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9971,
+      "step": 930
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.951,
+      "step": 931
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 2.0086,
+      "step": 932
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9811,
+      "step": 933
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9919,
+      "step": 934
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 2.0172,
+      "step": 935
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.9544,
+      "step": 936
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9474,
+      "step": 937
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9841,
+      "step": 938
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 2.0136,
+      "step": 939
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9691,
+      "step": 940
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9625,
+      "step": 941
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9833,
+      "step": 942
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9982,
+      "step": 943
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9753,
+      "step": 944
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9471,
+      "step": 945
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.993,
+      "step": 946
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9222,
+      "step": 947
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.951,
+      "step": 948
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9396,
+      "step": 949
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9433,
+      "step": 950
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9966,
+      "step": 951
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9783,
+      "step": 952
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9574,
+      "step": 953
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.9657,
+      "step": 954
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9555,
+      "step": 955
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9684,
+      "step": 956
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9793,
+      "step": 957
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 958
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9898,
+      "step": 959
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9528,
+      "step": 960
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.978,
+      "step": 961
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9926,
+      "step": 962
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.915,
+      "step": 963
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9913,
+      "step": 964
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9773,
+      "step": 965
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.912,
+      "step": 966
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9541,
+      "step": 967
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9307,
+      "step": 968
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 2.0023,
+      "step": 969
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9692,
+      "step": 970
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9561,
+      "step": 971
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 2.0004,
+      "step": 972
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9708,
+      "step": 973
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9454,
+      "step": 974
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.93,
+      "step": 975
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9831,
+      "step": 976
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.985,
+      "step": 977
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9554,
+      "step": 978
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9872,
+      "step": 979
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9894,
+      "step": 980
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9461,
+      "step": 981
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9465,
+      "step": 982
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.972,
+      "step": 983
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.987,
+      "step": 984
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9511,
+      "step": 985
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9597,
+      "step": 986
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9476,
+      "step": 987
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9602,
+      "step": 988
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9189,
+      "step": 989
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9427,
+      "step": 990
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9757,
+      "step": 991
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 2.0115,
+      "step": 992
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9937,
+      "step": 993
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9522,
+      "step": 994
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9583,
+      "step": 995
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9599,
+      "step": 996
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9581,
+      "step": 997
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.951,
+      "step": 998
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9543,
+      "step": 999
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 2.0377,
+      "step": 1000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9785,
+      "step": 1001
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9405,
+      "step": 1002
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 2.0009,
+      "step": 1003
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9956,
+      "step": 1004
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9728,
+      "step": 1005
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9543,
+      "step": 1006
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9709,
+      "step": 1007
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9797,
+      "step": 1008
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9563,
+      "step": 1009
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9753,
+      "step": 1010
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9719,
+      "step": 1011
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 2.0196,
+      "step": 1012
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9415,
+      "step": 1013
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9958,
+      "step": 1014
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9758,
+      "step": 1015
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9429,
+      "step": 1016
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9601,
+      "step": 1017
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 2.0083,
+      "step": 1018
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9823,
+      "step": 1019
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9394,
+      "step": 1020
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9874,
+      "step": 1021
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9641,
+      "step": 1022
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9694,
+      "step": 1023
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9999,
+      "step": 1024
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9781,
+      "step": 1025
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 2.0217,
+      "step": 1026
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9722,
+      "step": 1027
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9668,
+      "step": 1028
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9442,
+      "step": 1029
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9357,
+      "step": 1030
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9505,
+      "step": 1031
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9565,
+      "step": 1032
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9518,
+      "step": 1033
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9369,
+      "step": 1034
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9487,
+      "step": 1035
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.953,
+      "step": 1036
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9588,
+      "step": 1037
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9829,
+      "step": 1038
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.996,
+      "step": 1039
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9603,
+      "step": 1040
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.942,
+      "step": 1041
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9845,
+      "step": 1042
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9937,
+      "step": 1043
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 2.0366,
+      "step": 1044
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.9659,
+      "step": 1045
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9294,
+      "step": 1046
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9926,
+      "step": 1047
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9459,
+      "step": 1048
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9718,
+      "step": 1049
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9319,
+      "step": 1050
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9642,
+      "step": 1051
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9738,
+      "step": 1052
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9532,
+      "step": 1053
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9421,
+      "step": 1054
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9764,
+      "step": 1055
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9817,
+      "step": 1056
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9465,
+      "step": 1057
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9566,
+      "step": 1058
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9675,
+      "step": 1059
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9362,
+      "step": 1060
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9146,
+      "step": 1061
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.9857,
+      "step": 1062
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 2.0044,
+      "step": 1063
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9675,
+      "step": 1064
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.965,
+      "step": 1065
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9627,
+      "step": 1066
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9808,
+      "step": 1067
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9659,
+      "step": 1068
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9775,
+      "step": 1069
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9887,
+      "step": 1070
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9902,
+      "step": 1071
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9531,
+      "step": 1072
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9487,
+      "step": 1073
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 2.0081,
+      "step": 1074
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9595,
+      "step": 1075
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9815,
+      "step": 1076
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9812,
+      "step": 1077
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9878,
+      "step": 1078
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9458,
+      "step": 1079
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.9701,
+      "step": 1080
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 2.0107,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.952,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9399,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9782,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9619,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.988,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9477,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9861,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9684,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9817,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9477,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9869,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9442,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9419,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9815,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9589,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9442,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9372,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9461,
+      "step": 1099
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9523,
+      "step": 1100
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 2.0082,
+      "step": 1101
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9879,
+      "step": 1102
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9878,
+      "step": 1103
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 1104
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9808,
+      "step": 1105
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 2.0165,
+      "step": 1106
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9494,
+      "step": 1107
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9505,
+      "step": 1108
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9186,
+      "step": 1109
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9396,
+      "step": 1110
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9975,
+      "step": 1111
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9369,
+      "step": 1112
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9517,
+      "step": 1113
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9749,
+      "step": 1114
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9541,
+      "step": 1115
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 2.0052,
+      "step": 1116
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.9866,
+      "step": 1117
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9784,
+      "step": 1118
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9065,
+      "step": 1119
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9159,
+      "step": 1120
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9455,
+      "step": 1121
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9953,
+      "step": 1122
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9497,
+      "step": 1123
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9664,
+      "step": 1124
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.985,
+      "step": 1125
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9818,
+      "step": 1126
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9583,
+      "step": 1127
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9482,
+      "step": 1128
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9786,
+      "step": 1129
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9882,
+      "step": 1130
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9729,
+      "step": 1131
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9619,
+      "step": 1132
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9516,
+      "step": 1133
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9539,
+      "step": 1134
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.97,
+      "step": 1135
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9464,
+      "step": 1136
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9977,
+      "step": 1137
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9302,
+      "step": 1138
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9606,
+      "step": 1139
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9507,
+      "step": 1140
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9183,
+      "step": 1141
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9691,
+      "step": 1142
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9924,
+      "step": 1143
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9581,
+      "step": 1144
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9309,
+      "step": 1145
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9911,
+      "step": 1146
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9437,
+      "step": 1147
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9671,
+      "step": 1148
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9545,
+      "step": 1149
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9489,
+      "step": 1150
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9866,
+      "step": 1151
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9503,
+      "step": 1152
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9415,
+      "step": 1153
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9252,
+      "step": 1154
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9729,
+      "step": 1155
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9308,
+      "step": 1156
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9788,
+      "step": 1157
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.974,
+      "step": 1158
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9525,
+      "step": 1159
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.948,
+      "step": 1160
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9382,
+      "step": 1161
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9243,
+      "step": 1162
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9732,
+      "step": 1163
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9224,
+      "step": 1164
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9767,
+      "step": 1165
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9362,
+      "step": 1166
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9897,
+      "step": 1167
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9981,
+      "step": 1168
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9607,
+      "step": 1169
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9658,
+      "step": 1170
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9631,
+      "step": 1171
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 1.9692,
+      "step": 1172
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9519,
+      "step": 1173
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 2.0003,
+      "step": 1174
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9983,
+      "step": 1175
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9654,
+      "step": 1176
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9811,
+      "step": 1177
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9309,
+      "step": 1178
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.958,
+      "step": 1179
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9762,
+      "step": 1180
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9504,
+      "step": 1181
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9585,
+      "step": 1182
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.947,
+      "step": 1183
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9667,
+      "step": 1184
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9972,
+      "step": 1185
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9379,
+      "step": 1186
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9701,
+      "step": 1187
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9608,
+      "step": 1188
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 2.0189,
+      "step": 1189
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.9436,
+      "step": 1190
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9464,
+      "step": 1191
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9618,
+      "step": 1192
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9761,
+      "step": 1193
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9441,
+      "step": 1194
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9459,
+      "step": 1195
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9998,
+      "step": 1196
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9549,
+      "step": 1197
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9581,
+      "step": 1198
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.984,
+      "step": 1199
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9494,
+      "step": 1200
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9693,
+      "step": 1201
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 2.0063,
+      "step": 1202
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9594,
+      "step": 1203
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9159,
+      "step": 1204
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9076,
+      "step": 1205
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9102,
+      "step": 1206
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9435,
+      "step": 1207
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.9631,
+      "step": 1208
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9511,
+      "step": 1209
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9348,
+      "step": 1210
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9758,
+      "step": 1211
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9727,
+      "step": 1212
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9496,
+      "step": 1213
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9334,
+      "step": 1214
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9323,
+      "step": 1215
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9289,
+      "step": 1216
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9497,
+      "step": 1217
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.958,
+      "step": 1218
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.972,
+      "step": 1219
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9512,
+      "step": 1220
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 2.0025,
+      "step": 1221
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9698,
+      "step": 1222
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9319,
+      "step": 1223
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9224,
+      "step": 1224
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.9705,
+      "step": 1225
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.8952,
+      "step": 1226
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9625,
+      "step": 1227
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9322,
+      "step": 1228
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9348,
+      "step": 1229
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9253,
+      "step": 1230
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.966,
+      "step": 1231
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9329,
+      "step": 1232
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9711,
+      "step": 1233
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9538,
+      "step": 1234
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9335,
+      "step": 1235
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9519,
+      "step": 1236
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9666,
+      "step": 1237
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9724,
+      "step": 1238
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9748,
+      "step": 1239
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 2.0175,
+      "step": 1240
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9358,
+      "step": 1241
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9364,
+      "step": 1242
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.8931,
+      "step": 1243
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9637,
+      "step": 1244
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9885,
+      "step": 1245
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9786,
+      "step": 1246
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9639,
+      "step": 1247
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9419,
+      "step": 1248
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9519,
+      "step": 1249
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9512,
+      "step": 1250
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9252,
+      "step": 1251
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9866,
+      "step": 1252
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9646,
+      "step": 1253
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9686,
+      "step": 1254
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9499,
+      "step": 1255
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9318,
+      "step": 1256
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9568,
+      "step": 1257
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.955,
+      "step": 1258
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9427,
+      "step": 1259
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9725,
+      "step": 1260
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9683,
+      "step": 1261
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9458,
+      "step": 1262
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9527,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9644,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.928,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9467,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9095,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.967,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9296,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9729,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9123,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9355,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9428,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9188,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9111,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9676,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9481,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9485,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9673,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9362,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9348,
+      "step": 1281
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.952,
+      "step": 1282
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 2.0017,
+      "step": 1283
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9335,
+      "step": 1284
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9391,
+      "step": 1285
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9596,
+      "step": 1286
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9639,
+      "step": 1287
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9491,
+      "step": 1288
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9373,
+      "step": 1289
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9557,
+      "step": 1290
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9825,
+      "step": 1291
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9576,
+      "step": 1292
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.993,
+      "step": 1293
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9335,
+      "step": 1294
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 2.0085,
+      "step": 1295
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.978,
+      "step": 1296
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.97,
+      "step": 1297
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9333,
+      "step": 1298
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9418,
+      "step": 1299
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9761,
+      "step": 1300
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9846,
+      "step": 1301
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9894,
+      "step": 1302
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9548,
+      "step": 1303
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.912,
+      "step": 1304
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.935,
+      "step": 1305
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9667,
+      "step": 1306
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9635,
+      "step": 1307
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.941,
+      "step": 1308
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9214,
+      "step": 1309
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9534,
+      "step": 1310
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9464,
+      "step": 1311
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9559,
+      "step": 1312
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9543,
+      "step": 1313
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9337,
+      "step": 1314
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9771,
+      "step": 1315
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.949,
+      "step": 1316
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9749,
+      "step": 1317
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9307,
+      "step": 1318
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 2.0091,
+      "step": 1319
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9638,
+      "step": 1320
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9539,
+      "step": 1321
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9612,
+      "step": 1322
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9371,
+      "step": 1323
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9648,
+      "step": 1324
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9423,
+      "step": 1325
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9682,
+      "step": 1326
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9439,
+      "step": 1327
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9734,
+      "step": 1328
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9149,
+      "step": 1329
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9127,
+      "step": 1330
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9714,
+      "step": 1331
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9513,
+      "step": 1332
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9702,
+      "step": 1333
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9189,
+      "step": 1334
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9554,
+      "step": 1335
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9707,
+      "step": 1336
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9289,
+      "step": 1337
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9407,
+      "step": 1338
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9752,
+      "step": 1339
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.927,
+      "step": 1340
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.953,
+      "step": 1341
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9668,
+      "step": 1342
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9403,
+      "step": 1343
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9634,
+      "step": 1344
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9428,
+      "step": 1345
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9359,
+      "step": 1346
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9855,
+      "step": 1347
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9883,
+      "step": 1348
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9233,
+      "step": 1349
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9156,
+      "step": 1350
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9714,
+      "step": 1351
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9593,
+      "step": 1352
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9876,
+      "step": 1353
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.9481,
+      "step": 1354
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9265,
+      "step": 1355
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9489,
+      "step": 1356
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.965,
+      "step": 1357
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9293,
+      "step": 1358
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9623,
+      "step": 1359
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9121,
+      "step": 1360
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9747,
+      "step": 1361
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9333,
+      "step": 1362
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9638,
+      "step": 1363
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.919,
+      "step": 1364
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9463,
+      "step": 1365
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9372,
+      "step": 1366
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9426,
+      "step": 1367
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9632,
+      "step": 1368
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9518,
+      "step": 1369
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9305,
+      "step": 1370
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.9668,
+      "step": 1371
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.946,
+      "step": 1372
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9834,
+      "step": 1373
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9597,
+      "step": 1374
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9305,
+      "step": 1375
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9662,
+      "step": 1376
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9823,
+      "step": 1377
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.945,
+      "step": 1378
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9076,
+      "step": 1379
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9289,
+      "step": 1380
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9365,
+      "step": 1381
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9217,
+      "step": 1382
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9734,
+      "step": 1383
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9751,
+      "step": 1384
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9794,
+      "step": 1385
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9601,
+      "step": 1386
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9415,
+      "step": 1387
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9085,
+      "step": 1388
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9722,
+      "step": 1389
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.9853,
+      "step": 1390
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9194,
+      "step": 1391
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9374,
+      "step": 1392
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9422,
+      "step": 1393
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9726,
+      "step": 1394
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9405,
+      "step": 1395
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9256,
+      "step": 1396
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9827,
+      "step": 1397
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9387,
+      "step": 1398
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9201,
+      "step": 1399
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9355,
+      "step": 1400
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9313,
+      "step": 1401
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9491,
+      "step": 1402
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9612,
+      "step": 1403
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9594,
+      "step": 1404
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9816,
+      "step": 1405
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9145,
+      "step": 1406
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9318,
+      "step": 1407
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.9207,
+      "step": 1408
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9408,
+      "step": 1409
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9493,
+      "step": 1410
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9268,
+      "step": 1411
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9371,
+      "step": 1412
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.983,
+      "step": 1413
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9651,
+      "step": 1414
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9367,
+      "step": 1415
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9315,
+      "step": 1416
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.944,
+      "step": 1417
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9432,
+      "step": 1418
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9515,
+      "step": 1419
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9241,
+      "step": 1420
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9492,
+      "step": 1421
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9603,
+      "step": 1422
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9429,
+      "step": 1423
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9443,
+      "step": 1424
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9802,
+      "step": 1425
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9408,
+      "step": 1426
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9568,
+      "step": 1427
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9468,
+      "step": 1428
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9663,
+      "step": 1429
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9449,
+      "step": 1430
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9216,
+      "step": 1431
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9371,
+      "step": 1432
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9429,
+      "step": 1433
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9361,
+      "step": 1434
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9754,
+      "step": 1435
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9266,
+      "step": 1436
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9657,
+      "step": 1437
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9301,
+      "step": 1438
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9392,
+      "step": 1439
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9369,
+      "step": 1440
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9421,
+      "step": 1441
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.975,
+      "step": 1442
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9505,
+      "step": 1443
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9545,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9608,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9654,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9134,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9519,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9746,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9786,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9716,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9905,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9722,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9418,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9533,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9225,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9255,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9669,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9749,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9393,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9517,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9562,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9418,
+      "step": 1463
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.973,
+      "step": 1464
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9463,
+      "step": 1465
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9654,
+      "step": 1466
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9486,
+      "step": 1467
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9811,
+      "step": 1468
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9696,
+      "step": 1469
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.961,
+      "step": 1470
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9677,
+      "step": 1471
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9598,
+      "step": 1472
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9635,
+      "step": 1473
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9252,
+      "step": 1474
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9777,
+      "step": 1475
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9515,
+      "step": 1476
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9486,
+      "step": 1477
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9036,
+      "step": 1478
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9866,
+      "step": 1479
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9849,
+      "step": 1480
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9508,
+      "step": 1481
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9806,
+      "step": 1482
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9348,
+      "step": 1483
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9509,
+      "step": 1484
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9826,
+      "step": 1485
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9559,
+      "step": 1486
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9558,
+      "step": 1487
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9355,
+      "step": 1488
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.959,
+      "step": 1489
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9574,
+      "step": 1490
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9232,
+      "step": 1491
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.919,
+      "step": 1492
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9435,
+      "step": 1493
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9554,
+      "step": 1494
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9281,
+      "step": 1495
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9621,
+      "step": 1496
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9556,
+      "step": 1497
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9025,
+      "step": 1498
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.9741,
+      "step": 1499
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9579,
+      "step": 1500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9489,
+      "step": 1501
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.931,
+      "step": 1502
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9511,
+      "step": 1503
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9413,
+      "step": 1504
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.8989,
+      "step": 1505
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9363,
+      "step": 1506
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9473,
+      "step": 1507
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 2.0073,
+      "step": 1508
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9349,
+      "step": 1509
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.915,
+      "step": 1510
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9369,
+      "step": 1511
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9816,
+      "step": 1512
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 2.0049,
+      "step": 1513
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 2.007,
+      "step": 1514
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9044,
+      "step": 1515
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9309,
+      "step": 1516
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.9485,
+      "step": 1517
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.939,
+      "step": 1518
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9681,
+      "step": 1519
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9871,
+      "step": 1520
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.967,
+      "step": 1521
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9579,
+      "step": 1522
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9485,
+      "step": 1523
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9642,
+      "step": 1524
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9724,
+      "step": 1525
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9082,
+      "step": 1526
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9201,
+      "step": 1527
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.8962,
+      "step": 1528
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9771,
+      "step": 1529
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9797,
+      "step": 1530
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9227,
+      "step": 1531
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9449,
+      "step": 1532
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9556,
+      "step": 1533
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9486,
+      "step": 1534
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.9248,
+      "step": 1535
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9596,
+      "step": 1536
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.935,
+      "step": 1537
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9269,
+      "step": 1538
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9017,
+      "step": 1539
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9447,
+      "step": 1540
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9651,
+      "step": 1541
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9111,
+      "step": 1542
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9547,
+      "step": 1543
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9165,
+      "step": 1544
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9695,
+      "step": 1545
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.94,
+      "step": 1546
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9366,
+      "step": 1547
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9271,
+      "step": 1548
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.8816,
+      "step": 1549
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9247,
+      "step": 1550
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.9393,
+      "step": 1551
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.946,
+      "step": 1552
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.8909,
+      "step": 1553
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9237,
+      "step": 1554
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9379,
+      "step": 1555
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9362,
+      "step": 1556
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9542,
+      "step": 1557
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9258,
+      "step": 1558
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9789,
+      "step": 1559
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9366,
+      "step": 1560
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9336,
+      "step": 1561
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9119,
+      "step": 1562
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9661,
+      "step": 1563
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9895,
+      "step": 1564
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9251,
+      "step": 1565
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.945,
+      "step": 1566
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9673,
+      "step": 1567
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9089,
+      "step": 1568
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9385,
+      "step": 1569
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.8947,
+      "step": 1570
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9441,
+      "step": 1571
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9647,
+      "step": 1572
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9686,
+      "step": 1573
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9959,
+      "step": 1574
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.979,
+      "step": 1575
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9049,
+      "step": 1576
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9736,
+      "step": 1577
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9475,
+      "step": 1578
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9493,
+      "step": 1579
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9823,
+      "step": 1580
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9548,
+      "step": 1581
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9156,
+      "step": 1582
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9532,
+      "step": 1583
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9841,
+      "step": 1584
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9968,
+      "step": 1585
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9267,
+      "step": 1586
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9477,
+      "step": 1587
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.918,
+      "step": 1588
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9655,
+      "step": 1589
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9412,
+      "step": 1590
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.8958,
+      "step": 1591
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9471,
+      "step": 1592
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9887,
+      "step": 1593
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9178,
+      "step": 1594
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9155,
+      "step": 1595
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9503,
+      "step": 1596
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9717,
+      "step": 1597
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9312,
+      "step": 1598
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9335,
+      "step": 1599
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9161,
+      "step": 1600
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9455,
+      "step": 1601
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9601,
+      "step": 1602
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9078,
+      "step": 1603
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9439,
+      "step": 1604
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9581,
+      "step": 1605
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9248,
+      "step": 1606
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9447,
+      "step": 1607
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9542,
+      "step": 1608
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9261,
+      "step": 1609
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9015,
+      "step": 1610
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9271,
+      "step": 1611
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9315,
+      "step": 1612
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9529,
+      "step": 1613
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9238,
+      "step": 1614
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9696,
+      "step": 1615
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9202,
+      "step": 1616
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.976,
+      "step": 1617
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.925,
+      "step": 1618
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.963,
+      "step": 1619
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.983,
+      "step": 1620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9517,
+      "step": 1621
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 2.0073,
+      "step": 1622
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9682,
+      "step": 1623
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9241,
+      "step": 1624
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9166,
+      "step": 1625
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9489,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9561,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9203,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9688,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9548,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9907,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9299,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.8796,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.981,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9253,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9263,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9587,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.973,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9443,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9571,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9915,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9792,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9175,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.9496,
+      "step": 1644
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9273,
+      "step": 1645
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.8904,
+      "step": 1646
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9906,
+      "step": 1647
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9284,
+      "step": 1648
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 1649
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9671,
+      "step": 1650
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.906,
+      "step": 1651
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9482,
+      "step": 1652
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9115,
+      "step": 1653
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9748,
+      "step": 1654
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9343,
+      "step": 1655
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9314,
+      "step": 1656
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9727,
+      "step": 1657
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9315,
+      "step": 1658
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9549,
+      "step": 1659
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9148,
+      "step": 1660
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9453,
+      "step": 1661
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9318,
+      "step": 1662
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.9355,
+      "step": 1663
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9241,
+      "step": 1664
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.934,
+      "step": 1665
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.915,
+      "step": 1666
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9212,
+      "step": 1667
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9922,
+      "step": 1668
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9525,
+      "step": 1669
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9324,
+      "step": 1670
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9535,
+      "step": 1671
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9825,
+      "step": 1672
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9202,
+      "step": 1673
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9862,
+      "step": 1674
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9444,
+      "step": 1675
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9692,
+      "step": 1676
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9413,
+      "step": 1677
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.942,
+      "step": 1678
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.9975,
+      "step": 1679
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.996,
+      "step": 1680
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.967,
+      "step": 1681
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 1682
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9427,
+      "step": 1683
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9108,
+      "step": 1684
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9564,
+      "step": 1685
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.966,
+      "step": 1686
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9273,
+      "step": 1687
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9471,
+      "step": 1688
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9592,
+      "step": 1689
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9845,
+      "step": 1690
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9225,
+      "step": 1691
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.925,
+      "step": 1692
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9759,
+      "step": 1693
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9353,
+      "step": 1694
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9128,
+      "step": 1695
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9766,
+      "step": 1696
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9312,
+      "step": 1697
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9662,
+      "step": 1698
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.9541,
+      "step": 1699
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9695,
+      "step": 1700
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9583,
+      "step": 1701
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.964,
+      "step": 1702
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9765,
+      "step": 1703
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9628,
+      "step": 1704
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9197,
+      "step": 1705
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9757,
+      "step": 1706
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9642,
+      "step": 1707
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9219,
+      "step": 1708
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9653,
+      "step": 1709
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9693,
+      "step": 1710
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9155,
+      "step": 1711
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9586,
+      "step": 1712
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9618,
+      "step": 1713
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9232,
+      "step": 1714
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.965,
+      "step": 1715
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9416,
+      "step": 1716
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.9773,
+      "step": 1717
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.952,
+      "step": 1718
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.945,
+      "step": 1719
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9409,
+      "step": 1720
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9516,
+      "step": 1721
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9223,
+      "step": 1722
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.926,
+      "step": 1723
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9423,
+      "step": 1724
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9529,
+      "step": 1725
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.949,
+      "step": 1726
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9365,
+      "step": 1727
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9154,
+      "step": 1728
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9478,
+      "step": 1729
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9363,
+      "step": 1730
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9152,
+      "step": 1731
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9756,
+      "step": 1732
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9471,
+      "step": 1733
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9019,
+      "step": 1734
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9666,
+      "step": 1735
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9614,
+      "step": 1736
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9348,
+      "step": 1737
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9535,
+      "step": 1738
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9534,
+      "step": 1739
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9392,
+      "step": 1740
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9696,
+      "step": 1741
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9371,
+      "step": 1742
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9019,
+      "step": 1743
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9819,
+      "step": 1744
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9306,
+      "step": 1745
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9395,
+      "step": 1746
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.97,
+      "step": 1747
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9287,
+      "step": 1748
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.8926,
+      "step": 1749
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9383,
+      "step": 1750
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9483,
+      "step": 1751
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9515,
+      "step": 1752
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 1.9361,
+      "step": 1753
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9457,
+      "step": 1754
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9408,
+      "step": 1755
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9729,
+      "step": 1756
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9375,
+      "step": 1757
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9444,
+      "step": 1758
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.954,
+      "step": 1759
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9605,
+      "step": 1760
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9422,
+      "step": 1761
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9371,
+      "step": 1762
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9462,
+      "step": 1763
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9492,
+      "step": 1764
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.938,
+      "step": 1765
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.961,
+      "step": 1766
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9411,
+      "step": 1767
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9464,
+      "step": 1768
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.923,
+      "step": 1769
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.8705,
+      "step": 1770
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.9212,
+      "step": 1771
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 1.897,
+      "step": 1772
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.934,
+      "step": 1773
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9101,
+      "step": 1774
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9915,
+      "step": 1775
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9606,
+      "step": 1776
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9642,
+      "step": 1777
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9375,
+      "step": 1778
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9409,
+      "step": 1779
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9369,
+      "step": 1780
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9716,
+      "step": 1781
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.952,
+      "step": 1782
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9257,
+      "step": 1783
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9681,
+      "step": 1784
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9396,
+      "step": 1785
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9592,
+      "step": 1786
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9107,
+      "step": 1787
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9446,
+      "step": 1788
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9618,
+      "step": 1789
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9327,
+      "step": 1790
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9276,
+      "step": 1791
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9477,
+      "step": 1792
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9335,
+      "step": 1793
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.8792,
+      "step": 1794
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9412,
+      "step": 1795
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.918,
+      "step": 1796
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.923,
+      "step": 1797
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9244,
+      "step": 1798
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.8912,
+      "step": 1799
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9602,
+      "step": 1800
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9275,
+      "step": 1801
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9541,
+      "step": 1802
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 2.0049,
+      "step": 1803
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9532,
+      "step": 1804
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9449,
+      "step": 1805
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9337,
+      "step": 1806
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9574,
+      "step": 1807
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9829,
+      "step": 1808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9195,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.8846,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9518,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9655,
+      "step": 1812
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9675,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9901,
+      "step": 1814
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9341,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9576,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.908,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0,
+      "step": 1817,
+      "total_flos": 2400083874938880.0,
+      "train_loss": 1.9985291261100875,
+      "train_runtime": 103339.7122,
+      "train_samples_per_second": 9.004,
+      "train_steps_per_second": 0.018
+    }
+  ],
+  "max_steps": 1817,
+  "num_train_epochs": 1,
+  "total_flos": 2400083874938880.0,
+  "trial_name": null,
+  "trial_params": null
+}