diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,136 +1,5416 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.36272953978689637,
-  "global_step": 200,
+  "epoch": 2.9250457038391224,
+  "global_step": 1800,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 1.6531,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 1.7623,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 1.4946,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2e-05,
+      "loss": 1.8882,
+      "step": 8
+    },
     {
       "epoch": 0.02,
       "learning_rate": 2e-05,
-      "loss": 1.9524,
+      "loss": 1.7562,
       "step": 10
     },
     {
-      "epoch": 0.04,
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 1.875,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2e-05,
+      "loss": 1.9855,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 1.833,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
       "learning_rate": 2e-05,
-      "loss": 1.9249,
+      "loss": 1.7799,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 1.9502,
       "step": 20
     },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 1.8134,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 1.8352,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2e-05,
+      "loss": 1.9382,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.2001,
+      "step": 28
+    },
     {
       "epoch": 0.05,
       "learning_rate": 2e-05,
-      "loss": 1.9685,
+      "loss": 1.9664,
       "step": 30
     },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2e-05,
+      "loss": 2.0627,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0913,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.1183,
+      "step": 36
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2e-05,
+      "loss": 2.2824,
+      "step": 38
+    },
     {
       "epoch": 0.07,
       "learning_rate": 2e-05,
-      "loss": 2.1624,
+      "loss": 2.2368,
       "step": 40
     },
     {
-      "epoch": 0.09,
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.3417,
+      "step": 42
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.3669,
+      "step": 44
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2e-05,
+      "loss": 2.5686,
+      "step": 46
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 2.6004,
+      "step": 48
+    },
+    {
+      "epoch": 0.08,
       "learning_rate": 2e-05,
-      "loss": 2.3455,
+      "loss": 2.7763,
       "step": 50
     },
     {
-      "epoch": 0.11,
+      "epoch": 0.08,
+      "learning_rate": 2e-05,
+      "loss": 1.7123,
+      "step": 52
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.5842,
+      "step": 54
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.634,
+      "step": 56
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 1.6568,
+      "step": 58
+    },
+    {
+      "epoch": 0.1,
       "learning_rate": 2e-05,
-      "loss": 1.7288,
+      "loss": 1.7682,
       "step": 60
     },
     {
-      "epoch": 0.13,
+      "epoch": 0.1,
       "learning_rate": 2e-05,
-      "loss": 1.712,
+      "loss": 1.7746,
+      "step": 62
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 1.7256,
+      "step": 64
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 1.8528,
+      "step": 66
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 1.8566,
+      "step": 68
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2e-05,
+      "loss": 1.8693,
       "step": 70
     },
     {
-      "epoch": 0.15,
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 1.9164,
+      "step": 72
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2e-05,
+      "loss": 1.8215,
+      "step": 74
+    },
+    {
+      "epoch": 0.12,
       "learning_rate": 2e-05,
-      "loss": 1.8443,
+      "loss": 1.6364,
+      "step": 76
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 1.9358,
+      "step": 78
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 1.829,
       "step": 80
     },
     {
-      "epoch": 0.16,
+      "epoch": 0.13,
+      "learning_rate": 2e-05,
+      "loss": 2.0402,
+      "step": 82
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.046,
+      "step": 84
+    },
+    {
+      "epoch": 0.14,
       "learning_rate": 2e-05,
-      "loss": 1.9731,
+      "loss": 1.9946,
+      "step": 86
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0116,
+      "step": 88
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.1454,
       "step": 90
     },
     {
-      "epoch": 0.18,
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.3368,
+      "step": 92
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2e-05,
+      "loss": 2.3731,
+      "step": 94
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.3366,
+      "step": 96
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2e-05,
+      "loss": 2.3437,
+      "step": 98
+    },
+    {
+      "epoch": 0.16,
       "learning_rate": 2e-05,
-      "loss": 2.1967,
+      "loss": 2.524,
       "step": 100
     },
     {
-      "epoch": 0.2,
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.5793,
+      "step": 102
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2e-05,
+      "loss": 1.6267,
+      "step": 104
+    },
+    {
+      "epoch": 0.17,
       "learning_rate": 2e-05,
-      "loss": 1.6612,
+      "loss": 1.6976,
+      "step": 106
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.7402,
+      "step": 108
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.6939,
       "step": 110
     },
     {
-      "epoch": 0.22,
+      "epoch": 0.18,
+      "learning_rate": 2e-05,
+      "loss": 1.8356,
+      "step": 112
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.8122,
+      "step": 114
+    },
+    {
+      "epoch": 0.19,
       "learning_rate": 2e-05,
-      "loss": 1.6775,
+      "loss": 1.6891,
+      "step": 116
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2e-05,
+      "loss": 1.7671,
+      "step": 118
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.8497,
       "step": 120
     },
     {
-      "epoch": 0.24,
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.7755,
+      "step": 122
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.8383,
+      "step": 124
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2e-05,
+      "loss": 1.9323,
+      "step": 126
+    },
+    {
+      "epoch": 0.21,
       "learning_rate": 2e-05,
-      "loss": 1.7389,
+      "loss": 1.8231,
+      "step": 128
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 1.8705,
       "step": 130
     },
     {
-      "epoch": 0.25,
+      "epoch": 0.21,
+      "learning_rate": 2e-05,
+      "loss": 1.8914,
+      "step": 132
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.1717,
+      "step": 134
+    },
+    {
+      "epoch": 0.22,
       "learning_rate": 2e-05,
-      "loss": 1.9169,
+      "loss": 2.0413,
+      "step": 136
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 2.0281,
+      "step": 138
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.1997,
       "step": 140
     },
     {
-      "epoch": 0.27,
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.3263,
+      "step": 142
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2e-05,
+      "loss": 2.1469,
+      "step": 144
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.2804,
+      "step": 146
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2e-05,
+      "loss": 2.4742,
+      "step": 148
+    },
+    {
+      "epoch": 0.24,
       "learning_rate": 2e-05,
-      "loss": 2.2075,
+      "loss": 2.3649,
       "step": 150
     },
     {
-      "epoch": 0.29,
+      "epoch": 0.25,
       "learning_rate": 2e-05,
-      "loss": 1.6245,
-      "step": 160
+      "loss": 1.6274,
+      "step": 152
     },
     {
-      "epoch": 0.31,
+      "epoch": 0.25,
       "learning_rate": 2e-05,
-      "loss": 1.6433,
-      "step": 170
+      "loss": 1.6698,
+      "step": 154
     },
     {
-      "epoch": 0.33,
+      "epoch": 0.25,
       "learning_rate": 2e-05,
-      "loss": 1.7877,
-      "step": 180
+      "loss": 1.7127,
+      "step": 156
     },
     {
-      "epoch": 0.34,
+      "epoch": 0.26,
       "learning_rate": 2e-05,
-      "loss": 1.9041,
-      "step": 190
+      "loss": 1.6559,
+      "step": 158
     },
     {
-      "epoch": 0.36,
+      "epoch": 0.26,
       "learning_rate": 2e-05,
-      "loss": 2.1649,
-      "step": 200
+      "loss": 1.7203,
+      "step": 160
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2e-05,
+      "loss": 1.7946,
+      "step": 162
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.71,
+      "step": 164
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.8896,
+      "step": 166
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2e-05,
+      "loss": 1.9708,
+      "step": 168
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.8426,
+      "step": 170
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.7623,
+      "step": 172
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2e-05,
+      "loss": 1.8608,
+      "step": 174
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.8288,
+      "step": 176
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.8339,
+      "step": 178
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2e-05,
+      "loss": 1.955,
+      "step": 180
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0129,
+      "step": 182
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0187,
+      "step": 184
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0102,
+      "step": 186
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.0275,
+      "step": 188
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.2052,
+      "step": 190
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2e-05,
+      "loss": 2.2022,
+      "step": 192
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.207,
+      "step": 194
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.2069,
+      "step": 196
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2e-05,
+      "loss": 2.4695,
+      "step": 198
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 2.3821,
+      "step": 200
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.5257,
+      "step": 202
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.6402,
+      "step": 204
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2e-05,
+      "loss": 1.6208,
+      "step": 206
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.7039,
+      "step": 208
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.6589,
+      "step": 210
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2e-05,
+      "loss": 1.6637,
+      "step": 212
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.6445,
+      "step": 214
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.6848,
+      "step": 216
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2e-05,
+      "loss": 1.5876,
+      "step": 218
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.7515,
+      "step": 220
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.7731,
+      "step": 222
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2e-05,
+      "loss": 1.8234,
+      "step": 224
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.7628,
+      "step": 226
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.8525,
+      "step": 228
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2e-05,
+      "loss": 1.8992,
+      "step": 230
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 1.8982,
+      "step": 232
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 2.0687,
+      "step": 234
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2e-05,
+      "loss": 2.0344,
+      "step": 236
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.2035,
+      "step": 238
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.1326,
+      "step": 240
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2e-05,
+      "loss": 2.3343,
+      "step": 242
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.331,
+      "step": 244
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.1588,
+      "step": 246
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2e-05,
+      "loss": 2.4497,
+      "step": 248
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 2.3209,
+      "step": 250
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.4431,
+      "step": 252
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2e-05,
+      "loss": 1.5372,
+      "step": 254
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.5978,
+      "step": 256
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.7773,
+      "step": 258
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2e-05,
+      "loss": 1.8033,
+      "step": 260
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.7154,
+      "step": 262
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.7038,
+      "step": 264
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2e-05,
+      "loss": 1.9145,
+      "step": 266
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.7382,
+      "step": 268
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.712,
+      "step": 270
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2e-05,
+      "loss": 1.6205,
+      "step": 272
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.8052,
+      "step": 274
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.7942,
+      "step": 276
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2e-05,
+      "loss": 1.941,
+      "step": 278
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.8929,
+      "step": 280
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9285,
+      "step": 282
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 2.0238,
+      "step": 284
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2e-05,
+      "loss": 2.0938,
+      "step": 286
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0362,
+      "step": 288
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.2545,
+      "step": 290
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2e-05,
+      "loss": 2.1432,
+      "step": 292
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.3454,
+      "step": 294
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.359,
+      "step": 296
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2e-05,
+      "loss": 2.4653,
+      "step": 298
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 2.2918,
+      "step": 300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.748,
+      "step": 302
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2e-05,
+      "loss": 1.6338,
+      "step": 304
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.5935,
+      "step": 306
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.6318,
+      "step": 308
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2e-05,
+      "loss": 1.7291,
+      "step": 310
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.8349,
+      "step": 312
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.6695,
+      "step": 314
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2e-05,
+      "loss": 1.8873,
+      "step": 316
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.7472,
+      "step": 318
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.8549,
+      "step": 320
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2e-05,
+      "loss": 1.7141,
+      "step": 322
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.7487,
+      "step": 324
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.8681,
+      "step": 326
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2e-05,
+      "loss": 1.9445,
+      "step": 328
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.8244,
+      "step": 330
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.8444,
+      "step": 332
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9026,
+      "step": 334
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 2.0311,
+      "step": 336
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9602,
+      "step": 338
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2e-05,
+      "loss": 2.1369,
+      "step": 340
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9471,
+      "step": 342
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 2.3028,
+      "step": 344
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2e-05,
+      "loss": 2.1903,
+      "step": 346
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 2.4402,
+      "step": 348
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 2.5361,
+      "step": 350
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2e-05,
+      "loss": 1.5716,
+      "step": 352
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.5262,
+      "step": 354
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.5327,
+      "step": 356
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2e-05,
+      "loss": 1.5762,
+      "step": 358
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7237,
+      "step": 360
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.6061,
+      "step": 362
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.6987,
+      "step": 364
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7565,
+      "step": 366
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.8116,
+      "step": 368
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.9457,
+      "step": 370
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2e-05,
+      "loss": 1.6663,
+      "step": 372
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.8635,
+      "step": 374
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.7305,
+      "step": 376
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2e-05,
+      "loss": 1.8786,
+      "step": 378
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.87,
+      "step": 380
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.8725,
+      "step": 382
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2e-05,
+      "loss": 1.9773,
+      "step": 384
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 2.0312,
+      "step": 386
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 2.0488,
+      "step": 388
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2e-05,
+      "loss": 2.3269,
+      "step": 390
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 2.1795,
+      "step": 392
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 2.3712,
+      "step": 394
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2e-05,
+      "loss": 2.4111,
+      "step": 396
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 2.1541,
+      "step": 398
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 2.549,
+      "step": 400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2e-05,
+      "loss": 1.6017,
+      "step": 402
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.5165,
+      "step": 404
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.682,
+      "step": 406
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2e-05,
+      "loss": 1.7208,
+      "step": 408
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.6538,
+      "step": 410
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.7341,
+      "step": 412
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2e-05,
+      "loss": 1.6098,
+      "step": 414
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.518,
+      "step": 416
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.7724,
+      "step": 418
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2e-05,
+      "loss": 1.9174,
+      "step": 420
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.8182,
+      "step": 422
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.7367,
+      "step": 424
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2e-05,
+      "loss": 1.643,
+      "step": 426
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.8868,
+      "step": 428
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9231,
+      "step": 430
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9085,
+      "step": 432
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 1.8751,
+      "step": 434
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 2.0751,
+      "step": 436
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2e-05,
+      "loss": 2.1407,
+      "step": 438
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 1.9238,
+      "step": 440
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 2.3091,
+      "step": 442
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 2.2479,
+      "step": 444
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2e-05,
+      "loss": 2.3011,
+      "step": 446
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 2.185,
+      "step": 448
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 2.2973,
+      "step": 450
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2e-05,
+      "loss": 1.7113,
+      "step": 452
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.5509,
+      "step": 454
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.6546,
+      "step": 456
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2e-05,
+      "loss": 1.6055,
+      "step": 458
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.7226,
+      "step": 460
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.7313,
+      "step": 462
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2e-05,
+      "loss": 1.7648,
+      "step": 464
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.6838,
+      "step": 466
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7465,
+      "step": 468
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7302,
+      "step": 470
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.7998,
+      "step": 472
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.8177,
+      "step": 474
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2e-05,
+      "loss": 1.6845,
+      "step": 476
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.7208,
+      "step": 478
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9087,
+      "step": 480
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2e-05,
+      "loss": 1.7116,
+      "step": 482
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 2.0417,
+      "step": 484
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9781,
+      "step": 486
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2e-05,
+      "loss": 2.1279,
+      "step": 488
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 1.9915,
+      "step": 490
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 2.1453,
+      "step": 492
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2e-05,
+      "loss": 2.1639,
+      "step": 494
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 2.2218,
+      "step": 496
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 2.2928,
+      "step": 498
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2e-05,
+      "loss": 2.1805,
+      "step": 500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.6048,
+      "step": 502
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.6373,
+      "step": 504
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2e-05,
+      "loss": 1.6171,
+      "step": 506
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.6666,
+      "step": 508
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.5267,
+      "step": 510
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2e-05,
+      "loss": 1.8299,
+      "step": 512
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.6865,
+      "step": 514
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.899,
+      "step": 516
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2e-05,
+      "loss": 1.7581,
+      "step": 518
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.7595,
+      "step": 520
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.7733,
+      "step": 522
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.8707,
+      "step": 524
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2e-05,
+      "loss": 1.773,
+      "step": 526
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.8177,
+      "step": 528
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.8523,
+      "step": 530
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9317,
+      "step": 532
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 2.0409,
+      "step": 534
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 2.059,
+      "step": 536
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2e-05,
+      "loss": 2.1391,
+      "step": 538
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 2.2914,
+      "step": 540
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 2.1879,
+      "step": 542
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2e-05,
+      "loss": 2.3531,
+      "step": 544
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 2.2852,
+      "step": 546
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 2.3946,
+      "step": 548
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2e-05,
+      "loss": 2.1933,
+      "step": 550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.587,
+      "step": 552
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.5762,
+      "step": 554
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2e-05,
+      "loss": 1.4593,
+      "step": 556
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.6386,
+      "step": 558
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.759,
+      "step": 560
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2e-05,
+      "loss": 1.6945,
+      "step": 562
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.6703,
+      "step": 564
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.7655,
+      "step": 566
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2e-05,
+      "loss": 1.7459,
+      "step": 568
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.8065,
+      "step": 570
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.7765,
+      "step": 572
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2e-05,
+      "loss": 1.7079,
+      "step": 574
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.8376,
+      "step": 576
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 2.0,
+      "step": 578
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2e-05,
+      "loss": 1.876,
+      "step": 580
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.8461,
+      "step": 582
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9589,
+      "step": 584
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2e-05,
+      "loss": 1.9262,
+      "step": 586
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 2.0984,
+      "step": 588
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 2.1432,
+      "step": 590
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2e-05,
+      "loss": 2.2204,
+      "step": 592
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 2.3121,
+      "step": 594
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 2.3432,
+      "step": 596
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2e-05,
+      "loss": 2.2153,
+      "step": 598
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 2.4357,
+      "step": 600
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.6654,
+      "step": 602
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.5651,
+      "step": 604
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2e-05,
+      "loss": 1.8015,
+      "step": 606
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.8798,
+      "step": 608
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 1.8118,
+      "step": 610
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2e-05,
+      "loss": 2.101,
+      "step": 612
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 2.2339,
+      "step": 614
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 2.0958,
+      "step": 616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2e-05,
+      "loss": 1.5011,
+      "step": 618
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2e-05,
+      "loss": 1.6487,
+      "step": 620
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2e-05,
+      "loss": 1.663,
+      "step": 622
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2e-05,
+      "loss": 1.7526,
+      "step": 624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2e-05,
+      "loss": 1.6576,
+      "step": 626
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2e-05,
+      "loss": 1.8096,
+      "step": 628
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2e-05,
+      "loss": 1.6533,
+      "step": 630
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2e-05,
+      "loss": 1.8164,
+      "step": 632
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2e-05,
+      "loss": 1.7376,
+      "step": 634
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2e-05,
+      "loss": 1.7395,
+      "step": 636
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2e-05,
+      "loss": 1.7583,
+      "step": 638
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2e-05,
+      "loss": 1.8188,
+      "step": 640
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2e-05,
+      "loss": 1.8144,
+      "step": 642
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2e-05,
+      "loss": 1.8275,
+      "step": 644
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2e-05,
+      "loss": 1.8898,
+      "step": 646
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2e-05,
+      "loss": 1.831,
+      "step": 648
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2e-05,
+      "loss": 2.017,
+      "step": 650
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0152,
+      "step": 652
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0229,
+      "step": 654
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0969,
+      "step": 656
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2e-05,
+      "loss": 2.2473,
+      "step": 658
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2e-05,
+      "loss": 2.2124,
+      "step": 660
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2e-05,
+      "loss": 2.4132,
+      "step": 662
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2e-05,
+      "loss": 2.2386,
+      "step": 664
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2e-05,
+      "loss": 1.9457,
+      "step": 666
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2e-05,
+      "loss": 1.5253,
+      "step": 668
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2e-05,
+      "loss": 1.6223,
+      "step": 670
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2e-05,
+      "loss": 1.5866,
+      "step": 672
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2e-05,
+      "loss": 1.639,
+      "step": 674
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2e-05,
+      "loss": 1.7485,
+      "step": 676
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2e-05,
+      "loss": 1.7922,
+      "step": 678
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2e-05,
+      "loss": 1.6644,
+      "step": 680
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2e-05,
+      "loss": 1.7133,
+      "step": 682
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2e-05,
+      "loss": 1.7215,
+      "step": 684
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2e-05,
+      "loss": 1.6665,
+      "step": 686
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2e-05,
+      "loss": 1.809,
+      "step": 688
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2e-05,
+      "loss": 1.8063,
+      "step": 690
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2e-05,
+      "loss": 1.8331,
+      "step": 692
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2e-05,
+      "loss": 1.6631,
+      "step": 694
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2e-05,
+      "loss": 1.7219,
+      "step": 696
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2e-05,
+      "loss": 1.793,
+      "step": 698
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2e-05,
+      "loss": 1.9089,
+      "step": 700
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0203,
+      "step": 702
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2e-05,
+      "loss": 2.0187,
+      "step": 704
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2e-05,
+      "loss": 2.2132,
+      "step": 706
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2e-05,
+      "loss": 2.3242,
+      "step": 708
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2e-05,
+      "loss": 2.3611,
+      "step": 710
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2e-05,
+      "loss": 2.1036,
+      "step": 712
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2e-05,
+      "loss": 2.1611,
+      "step": 714
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2e-05,
+      "loss": 2.084,
+      "step": 716
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2e-05,
+      "loss": 1.5422,
+      "step": 718
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2e-05,
+      "loss": 1.5357,
+      "step": 720
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2e-05,
+      "loss": 1.5656,
+      "step": 722
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2e-05,
+      "loss": 1.665,
+      "step": 724
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2e-05,
+      "loss": 1.6541,
+      "step": 726
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2e-05,
+      "loss": 1.7007,
+      "step": 728
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2e-05,
+      "loss": 1.6715,
+      "step": 730
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2e-05,
+      "loss": 1.7646,
+      "step": 732
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2e-05,
+      "loss": 1.7159,
+      "step": 734
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2e-05,
+      "loss": 1.8008,
+      "step": 736
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2e-05,
+      "loss": 1.7396,
+      "step": 738
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2e-05,
+      "loss": 1.5893,
+      "step": 740
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2e-05,
+      "loss": 1.7379,
+      "step": 742
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2e-05,
+      "loss": 1.9598,
+      "step": 744
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2e-05,
+      "loss": 1.767,
+      "step": 746
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2e-05,
+      "loss": 1.6946,
+      "step": 748
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2e-05,
+      "loss": 1.9439,
+      "step": 750
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2e-05,
+      "loss": 1.809,
+      "step": 752
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2e-05,
+      "loss": 1.9658,
+      "step": 754
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2e-05,
+      "loss": 2.3269,
+      "step": 756
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2e-05,
+      "loss": 2.1706,
+      "step": 758
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2e-05,
+      "loss": 2.1369,
+      "step": 760
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2e-05,
+      "loss": 2.2149,
+      "step": 762
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2e-05,
+      "loss": 2.2046,
+      "step": 764
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2e-05,
+      "loss": 2.0422,
+      "step": 766
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2e-05,
+      "loss": 1.5257,
+      "step": 768
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2e-05,
+      "loss": 1.5342,
+      "step": 770
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2e-05,
+      "loss": 1.6119,
+      "step": 772
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2e-05,
+      "loss": 1.4974,
+      "step": 774
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2e-05,
+      "loss": 1.6021,
+      "step": 776
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2e-05,
+      "loss": 1.7111,
+      "step": 778
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2e-05,
+      "loss": 1.7096,
+      "step": 780
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2e-05,
+      "loss": 1.7536,
+      "step": 782
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2e-05,
+      "loss": 1.7301,
+      "step": 784
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2e-05,
+      "loss": 1.7513,
+      "step": 786
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2e-05,
+      "loss": 1.6764,
+      "step": 788
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2e-05,
+      "loss": 1.6348,
+      "step": 790
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2e-05,
+      "loss": 1.7046,
+      "step": 792
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2e-05,
+      "loss": 1.8024,
+      "step": 794
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2e-05,
+      "loss": 1.8867,
+      "step": 796
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2e-05,
+      "loss": 1.8189,
+      "step": 798
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2e-05,
+      "loss": 2.0048,
+      "step": 800
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2e-05,
+      "loss": 1.7588,
+      "step": 802
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9648,
+      "step": 804
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2e-05,
+      "loss": 2.1632,
+      "step": 806
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2e-05,
+      "loss": 2.1084,
+      "step": 808
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2e-05,
+      "loss": 2.1,
+      "step": 810
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2e-05,
+      "loss": 2.183,
+      "step": 812
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2e-05,
+      "loss": 2.2913,
+      "step": 814
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2e-05,
+      "loss": 1.8887,
+      "step": 816
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2e-05,
+      "loss": 1.578,
+      "step": 818
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2e-05,
+      "loss": 1.4916,
+      "step": 820
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2e-05,
+      "loss": 1.5374,
+      "step": 822
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2e-05,
+      "loss": 1.625,
+      "step": 824
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2e-05,
+      "loss": 1.7883,
+      "step": 826
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2e-05,
+      "loss": 1.7382,
+      "step": 828
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2e-05,
+      "loss": 1.6878,
+      "step": 830
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2e-05,
+      "loss": 1.5995,
+      "step": 832
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2e-05,
+      "loss": 1.8318,
+      "step": 834
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2e-05,
+      "loss": 1.7415,
+      "step": 836
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2e-05,
+      "loss": 1.8562,
+      "step": 838
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2e-05,
+      "loss": 1.7258,
+      "step": 840
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2e-05,
+      "loss": 1.8745,
+      "step": 842
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2e-05,
+      "loss": 1.6258,
+      "step": 844
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2e-05,
+      "loss": 1.8662,
+      "step": 846
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2e-05,
+      "loss": 1.8824,
+      "step": 848
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2e-05,
+      "loss": 1.8848,
+      "step": 850
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2e-05,
+      "loss": 2.0283,
+      "step": 852
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9087,
+      "step": 854
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2e-05,
+      "loss": 2.1342,
+      "step": 856
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2e-05,
+      "loss": 2.3005,
+      "step": 858
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2e-05,
+      "loss": 2.4106,
+      "step": 860
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2e-05,
+      "loss": 2.4499,
+      "step": 862
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2e-05,
+      "loss": 2.3442,
+      "step": 864
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2e-05,
+      "loss": 2.0918,
+      "step": 866
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2e-05,
+      "loss": 1.6702,
+      "step": 868
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2e-05,
+      "loss": 1.4252,
+      "step": 870
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2e-05,
+      "loss": 1.4626,
+      "step": 872
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2e-05,
+      "loss": 1.5953,
+      "step": 874
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2e-05,
+      "loss": 1.7418,
+      "step": 876
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2e-05,
+      "loss": 1.7604,
+      "step": 878
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2e-05,
+      "loss": 1.6774,
+      "step": 880
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2e-05,
+      "loss": 1.6192,
+      "step": 882
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2e-05,
+      "loss": 1.7959,
+      "step": 884
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2e-05,
+      "loss": 1.691,
+      "step": 886
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2e-05,
+      "loss": 1.6482,
+      "step": 888
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2e-05,
+      "loss": 1.7933,
+      "step": 890
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2e-05,
+      "loss": 1.7287,
+      "step": 892
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2e-05,
+      "loss": 1.7073,
+      "step": 894
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2e-05,
+      "loss": 1.9,
+      "step": 896
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2e-05,
+      "loss": 1.8793,
+      "step": 898
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2e-05,
+      "loss": 1.7948,
+      "step": 900
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2e-05,
+      "loss": 1.9856,
+      "step": 902
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0757,
+      "step": 904
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0352,
+      "step": 906
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2e-05,
+      "loss": 2.2688,
+      "step": 908
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2e-05,
+      "loss": 2.041,
+      "step": 910
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2e-05,
+      "loss": 2.265,
+      "step": 912
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2e-05,
+      "loss": 2.1646,
+      "step": 914
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2e-05,
+      "loss": 1.8594,
+      "step": 916
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2e-05,
+      "loss": 1.5378,
+      "step": 918
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2e-05,
+      "loss": 1.6376,
+      "step": 920
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2e-05,
+      "loss": 1.522,
+      "step": 922
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2e-05,
+      "loss": 1.5827,
+      "step": 924
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2e-05,
+      "loss": 1.633,
+      "step": 926
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2e-05,
+      "loss": 1.6128,
+      "step": 928
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2e-05,
+      "loss": 1.7496,
+      "step": 930
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2e-05,
+      "loss": 1.7611,
+      "step": 932
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2e-05,
+      "loss": 1.7722,
+      "step": 934
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2e-05,
+      "loss": 1.8811,
+      "step": 936
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2e-05,
+      "loss": 1.8488,
+      "step": 938
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2e-05,
+      "loss": 1.6723,
+      "step": 940
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2e-05,
+      "loss": 1.6077,
+      "step": 942
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2e-05,
+      "loss": 1.7119,
+      "step": 944
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9478,
+      "step": 946
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2e-05,
+      "loss": 1.9045,
+      "step": 948
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2e-05,
+      "loss": 1.8904,
+      "step": 950
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2e-05,
+      "loss": 2.1644,
+      "step": 952
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2e-05,
+      "loss": 2.1249,
+      "step": 954
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9582,
+      "step": 956
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2e-05,
+      "loss": 1.9816,
+      "step": 958
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2e-05,
+      "loss": 2.2362,
+      "step": 960
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2e-05,
+      "loss": 2.1076,
+      "step": 962
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2e-05,
+      "loss": 2.2971,
+      "step": 964
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2e-05,
+      "loss": 2.007,
+      "step": 966
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2e-05,
+      "loss": 1.6316,
+      "step": 968
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2e-05,
+      "loss": 1.5545,
+      "step": 970
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2e-05,
+      "loss": 1.6186,
+      "step": 972
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2e-05,
+      "loss": 1.6096,
+      "step": 974
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7449,
+      "step": 976
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7391,
+      "step": 978
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7077,
+      "step": 980
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2e-05,
+      "loss": 1.6408,
+      "step": 982
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2e-05,
+      "loss": 1.7383,
+      "step": 984
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2e-05,
+      "loss": 1.7688,
+      "step": 986
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2e-05,
+      "loss": 1.7425,
+      "step": 988
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2e-05,
+      "loss": 1.6938,
+      "step": 990
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2e-05,
+      "loss": 1.7232,
+      "step": 992
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2e-05,
+      "loss": 1.7924,
+      "step": 994
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2e-05,
+      "loss": 1.8015,
+      "step": 996
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2e-05,
+      "loss": 2.0395,
+      "step": 998
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2e-05,
+      "loss": 1.946,
+      "step": 1000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9046,
+      "step": 1002
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2e-05,
+      "loss": 2.0198,
+      "step": 1004
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2e-05,
+      "loss": 2.1411,
+      "step": 1006
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2e-05,
+      "loss": 1.8638,
+      "step": 1008
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2e-05,
+      "loss": 2.3288,
+      "step": 1010
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2e-05,
+      "loss": 2.1452,
+      "step": 1012
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2e-05,
+      "loss": 2.4029,
+      "step": 1014
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2e-05,
+      "loss": 1.721,
+      "step": 1016
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2e-05,
+      "loss": 1.7016,
+      "step": 1018
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2e-05,
+      "loss": 1.6516,
+      "step": 1020
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2e-05,
+      "loss": 1.5379,
+      "step": 1022
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2e-05,
+      "loss": 1.5188,
+      "step": 1024
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2e-05,
+      "loss": 1.6925,
+      "step": 1026
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2e-05,
+      "loss": 1.6326,
+      "step": 1028
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2e-05,
+      "loss": 1.7466,
+      "step": 1030
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2e-05,
+      "loss": 1.7166,
+      "step": 1032
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2e-05,
+      "loss": 1.8074,
+      "step": 1034
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2e-05,
+      "loss": 1.7404,
+      "step": 1036
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2e-05,
+      "loss": 1.6528,
+      "step": 1038
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2e-05,
+      "loss": 1.8054,
+      "step": 1040
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2e-05,
+      "loss": 1.9392,
+      "step": 1042
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2e-05,
+      "loss": 1.8748,
+      "step": 1044
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9289,
+      "step": 1046
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9047,
+      "step": 1048
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2e-05,
+      "loss": 2.0054,
+      "step": 1050
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9701,
+      "step": 1052
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2e-05,
+      "loss": 2.0893,
+      "step": 1054
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2e-05,
+      "loss": 2.073,
+      "step": 1056
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2e-05,
+      "loss": 2.2494,
+      "step": 1058
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2e-05,
+      "loss": 2.3391,
+      "step": 1060
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2e-05,
+      "loss": 2.39,
+      "step": 1062
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2e-05,
+      "loss": 2.2983,
+      "step": 1064
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2e-05,
+      "loss": 1.9709,
+      "step": 1066
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2e-05,
+      "loss": 1.7431,
+      "step": 1068
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2e-05,
+      "loss": 1.5658,
+      "step": 1070
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2e-05,
+      "loss": 1.8086,
+      "step": 1072
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6591,
+      "step": 1074
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6473,
+      "step": 1076
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6038,
+      "step": 1078
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7007,
+      "step": 1080
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7281,
+      "step": 1082
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2e-05,
+      "loss": 1.6404,
+      "step": 1084
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7932,
+      "step": 1086
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2e-05,
+      "loss": 1.7749,
+      "step": 1088
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2e-05,
+      "loss": 1.7351,
+      "step": 1090
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2e-05,
+      "loss": 1.6254,
+      "step": 1092
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9305,
+      "step": 1094
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2e-05,
+      "loss": 1.8313,
+      "step": 1096
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2e-05,
+      "loss": 1.8954,
+      "step": 1098
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9984,
+      "step": 1100
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2e-05,
+      "loss": 2.0387,
+      "step": 1102
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2e-05,
+      "loss": 2.0381,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2e-05,
+      "loss": 2.248,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2e-05,
+      "loss": 2.094,
+      "step": 1108
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2e-05,
+      "loss": 2.0483,
+      "step": 1110
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2e-05,
+      "loss": 2.2801,
+      "step": 1112
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2e-05,
+      "loss": 2.1668,
+      "step": 1114
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2e-05,
+      "loss": 1.9295,
+      "step": 1116
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2e-05,
+      "loss": 1.5339,
+      "step": 1118
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2e-05,
+      "loss": 1.4249,
+      "step": 1120
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2e-05,
+      "loss": 1.6543,
+      "step": 1122
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2e-05,
+      "loss": 1.5864,
+      "step": 1124
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2e-05,
+      "loss": 1.728,
+      "step": 1126
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2e-05,
+      "loss": 1.6257,
+      "step": 1128
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2e-05,
+      "loss": 1.7116,
+      "step": 1130
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2e-05,
+      "loss": 1.7008,
+      "step": 1132
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2e-05,
+      "loss": 1.8602,
+      "step": 1134
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2e-05,
+      "loss": 1.6803,
+      "step": 1136
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2e-05,
+      "loss": 1.5637,
+      "step": 1138
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2e-05,
+      "loss": 1.8221,
+      "step": 1140
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9582,
+      "step": 1142
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9902,
+      "step": 1144
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2e-05,
+      "loss": 1.9478,
+      "step": 1146
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2e-05,
+      "loss": 2.0016,
+      "step": 1148
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9449,
+      "step": 1150
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2e-05,
+      "loss": 1.9599,
+      "step": 1152
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2e-05,
+      "loss": 2.1458,
+      "step": 1154
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2e-05,
+      "loss": 2.0979,
+      "step": 1156
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2e-05,
+      "loss": 2.1437,
+      "step": 1158
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2e-05,
+      "loss": 2.1176,
+      "step": 1160
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2e-05,
+      "loss": 2.0171,
+      "step": 1162
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2e-05,
+      "loss": 2.1096,
+      "step": 1164
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2e-05,
+      "loss": 1.9555,
+      "step": 1166
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2e-05,
+      "loss": 1.5505,
+      "step": 1168
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2e-05,
+      "loss": 1.5644,
+      "step": 1170
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2e-05,
+      "loss": 1.6082,
+      "step": 1172
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2e-05,
+      "loss": 1.675,
+      "step": 1174
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2e-05,
+      "loss": 1.7006,
+      "step": 1176
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2e-05,
+      "loss": 1.6374,
+      "step": 1178
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2e-05,
+      "loss": 1.5624,
+      "step": 1180
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2e-05,
+      "loss": 1.875,
+      "step": 1182
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2e-05,
+      "loss": 1.7416,
+      "step": 1184
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2e-05,
+      "loss": 1.7585,
+      "step": 1186
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2e-05,
+      "loss": 1.6764,
+      "step": 1188
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2e-05,
+      "loss": 1.7482,
+      "step": 1190
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2e-05,
+      "loss": 1.7605,
+      "step": 1192
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2e-05,
+      "loss": 1.8099,
+      "step": 1194
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2e-05,
+      "loss": 1.7902,
+      "step": 1196
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2e-05,
+      "loss": 1.7466,
+      "step": 1198
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2e-05,
+      "loss": 1.8851,
+      "step": 1200
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2e-05,
+      "loss": 1.8968,
+      "step": 1202
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2e-05,
+      "loss": 2.1639,
+      "step": 1204
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2e-05,
+      "loss": 2.1004,
+      "step": 1206
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2e-05,
+      "loss": 2.068,
+      "step": 1208
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2e-05,
+      "loss": 2.1009,
+      "step": 1210
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2e-05,
+      "loss": 2.2122,
+      "step": 1212
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2e-05,
+      "loss": 2.2953,
+      "step": 1214
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2e-05,
+      "loss": 1.9851,
+      "step": 1216
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2e-05,
+      "loss": 1.5092,
+      "step": 1218
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2e-05,
+      "loss": 1.58,
+      "step": 1220
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2e-05,
+      "loss": 1.6222,
+      "step": 1222
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2e-05,
+      "loss": 1.7341,
+      "step": 1224
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2e-05,
+      "loss": 1.9206,
+      "step": 1226
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2e-05,
+      "loss": 1.9424,
+      "step": 1228
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2e-05,
+      "loss": 2.1776,
+      "step": 1230
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2e-05,
+      "loss": 2.021,
+      "step": 1232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2e-05,
+      "loss": 1.5074,
+      "step": 1234
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2e-05,
+      "loss": 1.441,
+      "step": 1236
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2e-05,
+      "loss": 1.619,
+      "step": 1238
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2e-05,
+      "loss": 1.6916,
+      "step": 1240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2e-05,
+      "loss": 1.5746,
+      "step": 1242
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2e-05,
+      "loss": 1.7335,
+      "step": 1244
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2e-05,
+      "loss": 1.7544,
+      "step": 1246
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2e-05,
+      "loss": 1.6698,
+      "step": 1248
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2e-05,
+      "loss": 1.6689,
+      "step": 1250
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2e-05,
+      "loss": 1.8378,
+      "step": 1252
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2e-05,
+      "loss": 1.807,
+      "step": 1254
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2e-05,
+      "loss": 1.6988,
+      "step": 1256
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2e-05,
+      "loss": 1.5926,
+      "step": 1258
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2e-05,
+      "loss": 1.7638,
+      "step": 1260
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2e-05,
+      "loss": 1.7585,
+      "step": 1262
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2e-05,
+      "loss": 1.969,
+      "step": 1264
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2e-05,
+      "loss": 2.0372,
+      "step": 1266
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2e-05,
+      "loss": 1.8681,
+      "step": 1268
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2e-05,
+      "loss": 1.9413,
+      "step": 1270
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0071,
+      "step": 1272
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2e-05,
+      "loss": 2.0781,
+      "step": 1274
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2e-05,
+      "loss": 2.2626,
+      "step": 1276
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2e-05,
+      "loss": 2.3026,
+      "step": 1278
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2e-05,
+      "loss": 2.0699,
+      "step": 1280
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2e-05,
+      "loss": 1.6375,
+      "step": 1282
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2e-05,
+      "loss": 1.5935,
+      "step": 1284
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2e-05,
+      "loss": 1.6579,
+      "step": 1286
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2e-05,
+      "loss": 1.5962,
+      "step": 1288
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2e-05,
+      "loss": 1.5371,
+      "step": 1290
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2e-05,
+      "loss": 1.6297,
+      "step": 1292
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2e-05,
+      "loss": 1.5664,
+      "step": 1294
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2e-05,
+      "loss": 1.6135,
+      "step": 1296
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2e-05,
+      "loss": 1.7271,
+      "step": 1298
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2e-05,
+      "loss": 1.7224,
+      "step": 1300
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2e-05,
+      "loss": 1.6938,
+      "step": 1302
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2e-05,
+      "loss": 1.6161,
+      "step": 1304
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2e-05,
+      "loss": 1.7213,
+      "step": 1306
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2e-05,
+      "loss": 1.7919,
+      "step": 1308
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2e-05,
+      "loss": 1.6923,
+      "step": 1310
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2e-05,
+      "loss": 1.9948,
+      "step": 1312
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2e-05,
+      "loss": 1.8132,
+      "step": 1314
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2e-05,
+      "loss": 1.9876,
+      "step": 1316
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2e-05,
+      "loss": 1.9332,
+      "step": 1318
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2e-05,
+      "loss": 2.0376,
+      "step": 1320
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2e-05,
+      "loss": 2.1336,
+      "step": 1322
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2e-05,
+      "loss": 2.1492,
+      "step": 1324
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2e-05,
+      "loss": 2.2404,
+      "step": 1326
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2e-05,
+      "loss": 2.0328,
+      "step": 1328
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2e-05,
+      "loss": 2.2208,
+      "step": 1330
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2e-05,
+      "loss": 1.6264,
+      "step": 1332
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2e-05,
+      "loss": 1.6469,
+      "step": 1334
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2e-05,
+      "loss": 1.4819,
+      "step": 1336
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2e-05,
+      "loss": 1.6929,
+      "step": 1338
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2e-05,
+      "loss": 1.5354,
+      "step": 1340
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2e-05,
+      "loss": 1.5995,
+      "step": 1342
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2e-05,
+      "loss": 1.7638,
+      "step": 1344
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2e-05,
+      "loss": 1.7488,
+      "step": 1346
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2e-05,
+      "loss": 1.8151,
+      "step": 1348
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2e-05,
+      "loss": 1.7435,
+      "step": 1350
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2e-05,
+      "loss": 1.7724,
+      "step": 1352
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2e-05,
+      "loss": 1.7307,
+      "step": 1354
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2e-05,
+      "loss": 1.6689,
+      "step": 1356
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2e-05,
+      "loss": 1.8689,
+      "step": 1358
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2e-05,
+      "loss": 1.6998,
+      "step": 1360
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2e-05,
+      "loss": 2.0046,
+      "step": 1362
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2e-05,
+      "loss": 1.7219,
+      "step": 1364
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2e-05,
+      "loss": 1.8812,
+      "step": 1366
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2e-05,
+      "loss": 1.8674,
+      "step": 1368
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2e-05,
+      "loss": 2.0153,
+      "step": 1370
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2e-05,
+      "loss": 2.1036,
+      "step": 1372
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2e-05,
+      "loss": 2.2218,
+      "step": 1374
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2e-05,
+      "loss": 2.2705,
+      "step": 1376
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2e-05,
+      "loss": 2.1251,
+      "step": 1378
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2e-05,
+      "loss": 2.27,
+      "step": 1380
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2e-05,
+      "loss": 1.7102,
+      "step": 1382
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2e-05,
+      "loss": 1.5466,
+      "step": 1384
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2e-05,
+      "loss": 1.7316,
+      "step": 1386
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2e-05,
+      "loss": 1.5031,
+      "step": 1388
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2e-05,
+      "loss": 1.608,
+      "step": 1390
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2e-05,
+      "loss": 1.5033,
+      "step": 1392
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2e-05,
+      "loss": 1.733,
+      "step": 1394
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2e-05,
+      "loss": 1.7289,
+      "step": 1396
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2e-05,
+      "loss": 1.7305,
+      "step": 1398
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2e-05,
+      "loss": 1.7743,
+      "step": 1400
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2e-05,
+      "loss": 1.8423,
+      "step": 1402
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2e-05,
+      "loss": 1.6997,
+      "step": 1404
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2e-05,
+      "loss": 1.8342,
+      "step": 1406
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2e-05,
+      "loss": 1.6843,
+      "step": 1408
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2e-05,
+      "loss": 1.8158,
+      "step": 1410
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2e-05,
+      "loss": 1.7417,
+      "step": 1412
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2e-05,
+      "loss": 1.8145,
+      "step": 1414
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9312,
+      "step": 1416
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2e-05,
+      "loss": 1.9542,
+      "step": 1418
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2e-05,
+      "loss": 2.0767,
+      "step": 1420
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2e-05,
+      "loss": 1.9053,
+      "step": 1422
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2e-05,
+      "loss": 2.4192,
+      "step": 1424
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2e-05,
+      "loss": 2.1551,
+      "step": 1426
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2e-05,
+      "loss": 2.1658,
+      "step": 1428
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2e-05,
+      "loss": 2.1308,
+      "step": 1430
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2e-05,
+      "loss": 1.6135,
+      "step": 1432
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2e-05,
+      "loss": 1.5897,
+      "step": 1434
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2e-05,
+      "loss": 1.4971,
+      "step": 1436
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2e-05,
+      "loss": 1.5434,
+      "step": 1438
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2e-05,
+      "loss": 1.6953,
+      "step": 1440
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2e-05,
+      "loss": 1.6321,
+      "step": 1442
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2e-05,
+      "loss": 1.572,
+      "step": 1444
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2e-05,
+      "loss": 1.5595,
+      "step": 1446
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2e-05,
+      "loss": 1.8133,
+      "step": 1448
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2e-05,
+      "loss": 1.641,
+      "step": 1450
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2e-05,
+      "loss": 1.6171,
+      "step": 1452
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2e-05,
+      "loss": 1.6589,
+      "step": 1454
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2e-05,
+      "loss": 1.786,
+      "step": 1456
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2e-05,
+      "loss": 1.7184,
+      "step": 1458
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2e-05,
+      "loss": 1.847,
+      "step": 1460
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2e-05,
+      "loss": 1.7423,
+      "step": 1462
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2e-05,
+      "loss": 2.0235,
+      "step": 1464
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2e-05,
+      "loss": 1.9874,
+      "step": 1466
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9473,
+      "step": 1468
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2e-05,
+      "loss": 2.0052,
+      "step": 1470
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2e-05,
+      "loss": 1.9972,
+      "step": 1472
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0354,
+      "step": 1474
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2e-05,
+      "loss": 2.1694,
+      "step": 1476
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2e-05,
+      "loss": 2.0919,
+      "step": 1478
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2e-05,
+      "loss": 2.1576,
+      "step": 1480
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2e-05,
+      "loss": 1.6458,
+      "step": 1482
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2e-05,
+      "loss": 1.5015,
+      "step": 1484
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2e-05,
+      "loss": 1.5359,
+      "step": 1486
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2e-05,
+      "loss": 1.7065,
+      "step": 1488
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2e-05,
+      "loss": 1.6893,
+      "step": 1490
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2e-05,
+      "loss": 1.7842,
+      "step": 1492
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2e-05,
+      "loss": 1.7229,
+      "step": 1494
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2e-05,
+      "loss": 1.6822,
+      "step": 1496
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2e-05,
+      "loss": 1.664,
+      "step": 1498
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2e-05,
+      "loss": 1.7587,
+      "step": 1500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2e-05,
+      "loss": 1.8268,
+      "step": 1502
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2e-05,
+      "loss": 1.6934,
+      "step": 1504
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2e-05,
+      "loss": 1.8765,
+      "step": 1506
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2e-05,
+      "loss": 1.845,
+      "step": 1508
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2e-05,
+      "loss": 1.8817,
+      "step": 1510
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2e-05,
+      "loss": 1.727,
+      "step": 1512
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2e-05,
+      "loss": 1.848,
+      "step": 1514
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2e-05,
+      "loss": 2.0766,
+      "step": 1516
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0203,
+      "step": 1518
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2e-05,
+      "loss": 2.0457,
+      "step": 1520
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2e-05,
+      "loss": 2.1893,
+      "step": 1522
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2e-05,
+      "loss": 1.9883,
+      "step": 1524
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2e-05,
+      "loss": 2.0896,
+      "step": 1526
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2e-05,
+      "loss": 2.1288,
+      "step": 1528
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2e-05,
+      "loss": 2.2434,
+      "step": 1530
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2e-05,
+      "loss": 1.8621,
+      "step": 1532
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2e-05,
+      "loss": 1.5902,
+      "step": 1534
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2e-05,
+      "loss": 1.6382,
+      "step": 1536
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2e-05,
+      "loss": 1.5176,
+      "step": 1538
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2e-05,
+      "loss": 1.7048,
+      "step": 1540
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2e-05,
+      "loss": 1.7238,
+      "step": 1542
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2e-05,
+      "loss": 1.7027,
+      "step": 1544
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2e-05,
+      "loss": 1.6627,
+      "step": 1546
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2e-05,
+      "loss": 1.6683,
+      "step": 1548
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2e-05,
+      "loss": 1.732,
+      "step": 1550
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2e-05,
+      "loss": 1.7039,
+      "step": 1552
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2e-05,
+      "loss": 1.5855,
+      "step": 1554
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2e-05,
+      "loss": 1.6969,
+      "step": 1556
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2e-05,
+      "loss": 1.7362,
+      "step": 1558
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2e-05,
+      "loss": 1.6448,
+      "step": 1560
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2e-05,
+      "loss": 1.7654,
+      "step": 1562
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2e-05,
+      "loss": 1.7655,
+      "step": 1564
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2e-05,
+      "loss": 1.8978,
+      "step": 1566
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9586,
+      "step": 1568
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2e-05,
+      "loss": 2.0546,
+      "step": 1570
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2e-05,
+      "loss": 1.9625,
+      "step": 1572
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2e-05,
+      "loss": 2.0044,
+      "step": 1574
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2e-05,
+      "loss": 2.0529,
+      "step": 1576
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2e-05,
+      "loss": 2.1048,
+      "step": 1578
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2e-05,
+      "loss": 2.0548,
+      "step": 1580
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2e-05,
+      "loss": 1.7534,
+      "step": 1582
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2e-05,
+      "loss": 1.4283,
+      "step": 1584
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2e-05,
+      "loss": 1.6125,
+      "step": 1586
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2e-05,
+      "loss": 1.5568,
+      "step": 1588
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2e-05,
+      "loss": 1.6719,
+      "step": 1590
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2e-05,
+      "loss": 1.6142,
+      "step": 1592
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2e-05,
+      "loss": 1.6434,
+      "step": 1594
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2e-05,
+      "loss": 1.7194,
+      "step": 1596
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2e-05,
+      "loss": 1.7541,
+      "step": 1598
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2e-05,
+      "loss": 1.6681,
+      "step": 1600
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2e-05,
+      "loss": 1.81,
+      "step": 1602
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2e-05,
+      "loss": 1.6948,
+      "step": 1604
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2e-05,
+      "loss": 1.7533,
+      "step": 1606
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2e-05,
+      "loss": 1.8256,
+      "step": 1608
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2e-05,
+      "loss": 1.8959,
+      "step": 1610
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2e-05,
+      "loss": 1.7526,
+      "step": 1612
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2e-05,
+      "loss": 1.8277,
+      "step": 1614
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9229,
+      "step": 1616
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2e-05,
+      "loss": 1.9832,
+      "step": 1618
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2e-05,
+      "loss": 1.903,
+      "step": 1620
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2e-05,
+      "loss": 2.1626,
+      "step": 1622
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2e-05,
+      "loss": 2.2053,
+      "step": 1624
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2e-05,
+      "loss": 2.176,
+      "step": 1626
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2e-05,
+      "loss": 2.1653,
+      "step": 1628
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2e-05,
+      "loss": 2.179,
+      "step": 1630
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2e-05,
+      "loss": 1.7573,
+      "step": 1632
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2e-05,
+      "loss": 1.5932,
+      "step": 1634
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2e-05,
+      "loss": 1.5588,
+      "step": 1636
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2e-05,
+      "loss": 1.6959,
+      "step": 1638
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2e-05,
+      "loss": 1.598,
+      "step": 1640
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2e-05,
+      "loss": 1.748,
+      "step": 1642
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2e-05,
+      "loss": 1.7074,
+      "step": 1644
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2e-05,
+      "loss": 1.8215,
+      "step": 1646
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2e-05,
+      "loss": 1.6872,
+      "step": 1648
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2e-05,
+      "loss": 1.734,
+      "step": 1650
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2e-05,
+      "loss": 1.6693,
+      "step": 1652
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2e-05,
+      "loss": 1.7714,
+      "step": 1654
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2e-05,
+      "loss": 1.7195,
+      "step": 1656
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2e-05,
+      "loss": 1.7976,
+      "step": 1658
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2e-05,
+      "loss": 1.8392,
+      "step": 1660
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2e-05,
+      "loss": 1.7095,
+      "step": 1662
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2e-05,
+      "loss": 1.9426,
+      "step": 1664
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2e-05,
+      "loss": 1.9583,
+      "step": 1666
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2e-05,
+      "loss": 2.028,
+      "step": 1668
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2e-05,
+      "loss": 2.1083,
+      "step": 1670
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2e-05,
+      "loss": 2.1001,
+      "step": 1672
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2e-05,
+      "loss": 2.2984,
+      "step": 1674
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2e-05,
+      "loss": 2.0889,
+      "step": 1676
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2e-05,
+      "loss": 2.3576,
+      "step": 1678
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2e-05,
+      "loss": 2.0517,
+      "step": 1680
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2e-05,
+      "loss": 1.6826,
+      "step": 1682
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2e-05,
+      "loss": 1.5038,
+      "step": 1684
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2e-05,
+      "loss": 1.6221,
+      "step": 1686
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2e-05,
+      "loss": 1.5329,
+      "step": 1688
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6428,
+      "step": 1690
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6739,
+      "step": 1692
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2e-05,
+      "loss": 1.6846,
+      "step": 1694
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7064,
+      "step": 1696
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2e-05,
+      "loss": 1.6818,
+      "step": 1698
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2e-05,
+      "loss": 1.7676,
+      "step": 1700
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2e-05,
+      "loss": 1.6708,
+      "step": 1702
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2e-05,
+      "loss": 1.7152,
+      "step": 1704
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2e-05,
+      "loss": 1.4785,
+      "step": 1706
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9122,
+      "step": 1708
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2e-05,
+      "loss": 1.9274,
+      "step": 1710
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2e-05,
+      "loss": 1.982,
+      "step": 1712
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2e-05,
+      "loss": 1.9373,
+      "step": 1714
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2e-05,
+      "loss": 1.8395,
+      "step": 1716
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2e-05,
+      "loss": 1.8692,
+      "step": 1718
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2e-05,
+      "loss": 2.1762,
+      "step": 1720
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2e-05,
+      "loss": 2.0499,
+      "step": 1722
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2e-05,
+      "loss": 2.1598,
+      "step": 1724
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2e-05,
+      "loss": 2.2524,
+      "step": 1726
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2e-05,
+      "loss": 2.2295,
+      "step": 1728
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2e-05,
+      "loss": 2.4519,
+      "step": 1730
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2e-05,
+      "loss": 1.7711,
+      "step": 1732
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2e-05,
+      "loss": 1.5325,
+      "step": 1734
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2e-05,
+      "loss": 1.6951,
+      "step": 1736
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2e-05,
+      "loss": 1.5366,
+      "step": 1738
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2e-05,
+      "loss": 1.5947,
+      "step": 1740
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2e-05,
+      "loss": 1.6359,
+      "step": 1742
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2e-05,
+      "loss": 1.6945,
+      "step": 1744
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2e-05,
+      "loss": 1.7055,
+      "step": 1746
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2e-05,
+      "loss": 1.5715,
+      "step": 1748
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2e-05,
+      "loss": 1.6131,
+      "step": 1750
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2e-05,
+      "loss": 1.7353,
+      "step": 1752
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2e-05,
+      "loss": 1.6901,
+      "step": 1754
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2e-05,
+      "loss": 1.6287,
+      "step": 1756
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2e-05,
+      "loss": 1.7596,
+      "step": 1758
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2e-05,
+      "loss": 1.6615,
+      "step": 1760
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2e-05,
+      "loss": 1.7761,
+      "step": 1762
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2e-05,
+      "loss": 1.7353,
+      "step": 1764
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2e-05,
+      "loss": 2.0409,
+      "step": 1766
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2e-05,
+      "loss": 1.8417,
+      "step": 1768
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2e-05,
+      "loss": 1.9822,
+      "step": 1770
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2e-05,
+      "loss": 1.935,
+      "step": 1772
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2e-05,
+      "loss": 2.191,
+      "step": 1774
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2e-05,
+      "loss": 2.1619,
+      "step": 1776
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2e-05,
+      "loss": 2.179,
+      "step": 1778
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2e-05,
+      "loss": 2.4226,
+      "step": 1780
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2e-05,
+      "loss": 1.752,
+      "step": 1782
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2e-05,
+      "loss": 1.4133,
+      "step": 1784
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2e-05,
+      "loss": 1.3559,
+      "step": 1786
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2e-05,
+      "loss": 1.4966,
+      "step": 1788
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2e-05,
+      "loss": 1.5845,
+      "step": 1790
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2e-05,
+      "loss": 1.6532,
+      "step": 1792
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2e-05,
+      "loss": 1.6712,
+      "step": 1794
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2e-05,
+      "loss": 1.7321,
+      "step": 1796
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2e-05,
+      "loss": 1.4819,
+      "step": 1798
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2e-05,
+      "loss": 1.7174,
+      "step": 1800
     }
   ],
-  "max_steps": 1875,
-  "num_train_epochs": 4,
-  "total_flos": 4967090265784320.0,
+  "max_steps": 1845,
+  "num_train_epochs": 3,
+  "total_flos": 3.908152132295885e+16,
   "trial_name": null,
   "trial_params": null
 }