diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6471 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9944341372912802,
+  "eval_steps": 800,
+  "global_step": 4300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0,
+      "loss": 4.1961,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0,
+      "loss": 3.4713,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 3.6452,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 3.2857,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 3.8003,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.541201032390638e-07,
+      "loss": 3.0254,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6908759378408314e-06,
+      "loss": 1.2014,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1729399483804686e-06,
+      "loss": 0.888,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.490467394684078e-06,
+      "loss": 0.6906,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7276317724425993e-06,
+      "loss": 0.9432,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917003870770575e-06,
+      "loss": 0.761,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.074652113699448e-06,
+      "loss": 0.7311,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.209695782982236e-06,
+      "loss": 0.6145,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3278117189435736e-06,
+      "loss": 0.7238,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.432774886884842e-06,
+      "loss": 0.8738,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.527223229285846e-06,
+      "loss": 0.6586,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6130729169433793e-06,
+      "loss": 0.9213,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6917597935218728e-06,
+      "loss": 0.8081,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.746747919084026e-06,
+      "loss": 0.5199,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.8154106182475455e-06,
+      "loss": 0.8188,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.8794145195304064e-06,
+      "loss": 0.707,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.9393518556017876e-06,
+      "loss": 0.8889,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.995708648306083e-06,
+      "loss": 0.6421,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.0488886794862905e-06,
+      "loss": 0.9915,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.099231058976525e-06,
+      "loss": 0.7983,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.1470233384070705e-06,
+      "loss": 0.7454,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.192511460845915e-06,
+      "loss": 0.6397,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.235907420412398e-06,
+      "loss": 0.8665,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.27739523661862e-06,
+      "loss": 0.7032,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.317135669700268e-06,
+      "loss": 0.6611,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.355269982485126e-06,
+      "loss": 1.0662,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.391922971188186e-06,
+      "loss": 0.8082,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.427205429264097e-06,
+      "loss": 0.8027,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.461216167003915e-06,
+      "loss": 0.7987,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.494043679663919e-06,
+      "loss": 0.7789,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.525767535063022e-06,
+      "loss": 0.7888,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.556459535424214e-06,
+      "loss": 0.669,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.5861846961499975e-06,
+      "loss": 0.5449,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.607879808611406e-06,
+      "loss": 0.785,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.636051925421702e-06,
+      "loss": 0.7444,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.663407343064547e-06,
+      "loss": 0.6551,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.689992082159791e-06,
+      "loss": 0.4599,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.715848379822425e-06,
+      "loss": 0.6218,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.74101509336297e-06,
+      "loss": 0.9368,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.76552805154028e-06,
+      "loss": 0.663,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.789420361336724e-06,
+      "loss": 1.0055,
+      "step": 184
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.812722676847563e-06,
+      "loss": 0.738,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.835463435763974e-06,
+      "loss": 0.9295,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.857669068026358e-06,
+      "loss": 0.8095,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.879364180487766e-06,
+      "loss": 0.8501,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.900571720823068e-06,
+      "loss": 0.7591,
+      "step": 204
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.921313123421507e-06,
+      "loss": 0.9022,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.941608439588058e-06,
+      "loss": 1.0702,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9614764540361516e-06,
+      "loss": 0.8529,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.980934789368156e-06,
+      "loss": 1.0398,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-06,
+      "loss": 0.8779,
+      "step": 224
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.997606892150607e-06,
+      "loss": 0.6527,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.994416081684748e-06,
+      "loss": 1.2818,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.99122527121889e-06,
+      "loss": 0.8251,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.988034460753032e-06,
+      "loss": 0.9644,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.984843650287174e-06,
+      "loss": 0.7095,
+      "step": 244
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.981652839821315e-06,
+      "loss": 0.7467,
+      "step": 248
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.978462029355457e-06,
+      "loss": 0.689,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.975271218889599e-06,
+      "loss": 0.9441,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.97208040842374e-06,
+      "loss": 0.8769,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.968889597957882e-06,
+      "loss": 0.9131,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.965698787492023e-06,
+      "loss": 0.9625,
+      "step": 268
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.962507977026166e-06,
+      "loss": 0.8598,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.959317166560307e-06,
+      "loss": 0.7809,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.956126356094449e-06,
+      "loss": 0.7042,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.95293554562859e-06,
+      "loss": 0.7617,
+      "step": 284
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9497447351627315e-06,
+      "loss": 0.6065,
+      "step": 288
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.946553924696874e-06,
+      "loss": 0.9993,
+      "step": 292
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.943363114231015e-06,
+      "loss": 0.8739,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.940172303765157e-06,
+      "loss": 0.7481,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.936981493299298e-06,
+      "loss": 0.9159,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9337906828334406e-06,
+      "loss": 0.765,
+      "step": 308
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.930599872367582e-06,
+      "loss": 0.8167,
+      "step": 312
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9274090619017235e-06,
+      "loss": 1.1514,
+      "step": 316
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.924218251435865e-06,
+      "loss": 0.7979,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9210274409700065e-06,
+      "loss": 0.8139,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.917836630504149e-06,
+      "loss": 0.7969,
+      "step": 328
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.91464582003829e-06,
+      "loss": 0.9692,
+      "step": 332
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.911455009572432e-06,
+      "loss": 1.2616,
+      "step": 336
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.908264199106573e-06,
+      "loss": 0.8522,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9050733886407155e-06,
+      "loss": 0.7732,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.901882578174857e-06,
+      "loss": 0.7179,
+      "step": 348
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.8986917677089984e-06,
+      "loss": 0.7246,
+      "step": 352
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.89550095724314e-06,
+      "loss": 1.1594,
+      "step": 356
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.892310146777281e-06,
+      "loss": 0.7413,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.889119336311424e-06,
+      "loss": 0.8847,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.885928525845565e-06,
+      "loss": 0.8232,
+      "step": 368
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.882737715379707e-06,
+      "loss": 0.7903,
+      "step": 372
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.879546904913848e-06,
+      "loss": 0.988,
+      "step": 376
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.87635609444799e-06,
+      "loss": 0.6777,
+      "step": 380
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.873165283982132e-06,
+      "loss": 0.7942,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.869974473516273e-06,
+      "loss": 0.5827,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.866783663050415e-06,
+      "loss": 0.9409,
+      "step": 392
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.863592852584556e-06,
+      "loss": 0.839,
+      "step": 396
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.860402042118699e-06,
+      "loss": 0.7943,
+      "step": 400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.85721123165284e-06,
+      "loss": 0.5618,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8540204211869816e-06,
+      "loss": 0.892,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.850829610721123e-06,
+      "loss": 0.6751,
+      "step": 412
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.847638800255265e-06,
+      "loss": 0.8687,
+      "step": 416
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.844447989789407e-06,
+      "loss": 1.3904,
+      "step": 420
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.841257179323548e-06,
+      "loss": 1.1064,
+      "step": 424
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.83806636885769e-06,
+      "loss": 1.1111,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.834875558391831e-06,
+      "loss": 0.6871,
+      "step": 432
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8316847479259735e-06,
+      "loss": 0.7987,
+      "step": 436
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.828493937460115e-06,
+      "loss": 0.5025,
+      "step": 440
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8253031269942565e-06,
+      "loss": 0.8831,
+      "step": 444
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.822112316528399e-06,
+      "loss": 0.7851,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.81892150606254e-06,
+      "loss": 0.7234,
+      "step": 452
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.815730695596682e-06,
+      "loss": 1.1858,
+      "step": 456
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.812539885130823e-06,
+      "loss": 0.9807,
+      "step": 460
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8093490746649655e-06,
+      "loss": 0.8921,
+      "step": 464
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.806158264199107e-06,
+      "loss": 0.7873,
+      "step": 468
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8029674537332485e-06,
+      "loss": 0.6174,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.79977664326739e-06,
+      "loss": 0.5353,
+      "step": 476
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.796585832801532e-06,
+      "loss": 0.8381,
+      "step": 480
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.793395022335674e-06,
+      "loss": 0.7193,
+      "step": 484
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.790204211869815e-06,
+      "loss": 0.9925,
+      "step": 488
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.7870134014039575e-06,
+      "loss": 0.9972,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.783822590938099e-06,
+      "loss": 1.0455,
+      "step": 496
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.7806317804722405e-06,
+      "loss": 0.6904,
+      "step": 500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.777440970006382e-06,
+      "loss": 0.823,
+      "step": 504
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.774250159540524e-06,
+      "loss": 0.8133,
+      "step": 508
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.771059349074666e-06,
+      "loss": 0.9556,
+      "step": 512
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.767868538608807e-06,
+      "loss": 0.8012,
+      "step": 516
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.764677728142949e-06,
+      "loss": 0.5765,
+      "step": 520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.761486917677091e-06,
+      "loss": 0.7058,
+      "step": 524
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.7582961072112324e-06,
+      "loss": 0.7267,
+      "step": 528
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.755105296745374e-06,
+      "loss": 0.6016,
+      "step": 532
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.751914486279515e-06,
+      "loss": 0.6567,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.748723675813657e-06,
+      "loss": 0.8572,
+      "step": 540
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.745532865347799e-06,
+      "loss": 0.9085,
+      "step": 544
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.742342054881941e-06,
+      "loss": 0.96,
+      "step": 548
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.739151244416082e-06,
+      "loss": 0.7202,
+      "step": 552
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.735960433950224e-06,
+      "loss": 0.6218,
+      "step": 556
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.732769623484366e-06,
+      "loss": 0.7925,
+      "step": 560
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.729578813018507e-06,
+      "loss": 0.9587,
+      "step": 564
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.726388002552649e-06,
+      "loss": 0.8156,
+      "step": 568
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.72319719208679e-06,
+      "loss": 0.5733,
+      "step": 572
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.720006381620932e-06,
+      "loss": 0.5818,
+      "step": 576
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.716815571155074e-06,
+      "loss": 1.0231,
+      "step": 580
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7136247606892156e-06,
+      "loss": 0.7375,
+      "step": 584
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.710433950223357e-06,
+      "loss": 0.7752,
+      "step": 588
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7072431397574985e-06,
+      "loss": 0.8014,
+      "step": 592
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.704052329291641e-06,
+      "loss": 0.6724,
+      "step": 596
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.700861518825782e-06,
+      "loss": 0.8042,
+      "step": 600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.697670708359924e-06,
+      "loss": 1.0964,
+      "step": 604
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.694479897894065e-06,
+      "loss": 0.9096,
+      "step": 608
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.691289087428207e-06,
+      "loss": 0.5177,
+      "step": 612
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.688098276962349e-06,
+      "loss": 0.8655,
+      "step": 616
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6849074664964905e-06,
+      "loss": 0.8737,
+      "step": 620
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.681716656030632e-06,
+      "loss": 0.5623,
+      "step": 624
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6785258455647734e-06,
+      "loss": 0.4819,
+      "step": 628
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.675335035098916e-06,
+      "loss": 0.8688,
+      "step": 632
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.672144224633057e-06,
+      "loss": 0.6851,
+      "step": 636
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.668953414167199e-06,
+      "loss": 0.7127,
+      "step": 640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.66576260370134e-06,
+      "loss": 1.0049,
+      "step": 644
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.662571793235482e-06,
+      "loss": 0.5841,
+      "step": 648
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.659380982769624e-06,
+      "loss": 0.7495,
+      "step": 652
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.656190172303765e-06,
+      "loss": 0.8183,
+      "step": 656
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.652999361837907e-06,
+      "loss": 0.5477,
+      "step": 660
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.649808551372048e-06,
+      "loss": 0.8446,
+      "step": 664
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.646617740906191e-06,
+      "loss": 0.9351,
+      "step": 668
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.643426930440332e-06,
+      "loss": 0.8057,
+      "step": 672
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.640236119974474e-06,
+      "loss": 0.9213,
+      "step": 676
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.637045309508615e-06,
+      "loss": 0.5699,
+      "step": 680
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.6338544990427565e-06,
+      "loss": 0.641,
+      "step": 684
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.630663688576899e-06,
+      "loss": 0.5552,
+      "step": 688
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.62747287811104e-06,
+      "loss": 0.6875,
+      "step": 692
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.624282067645182e-06,
+      "loss": 0.6743,
+      "step": 696
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.621091257179324e-06,
+      "loss": 0.6982,
+      "step": 700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.617900446713466e-06,
+      "loss": 0.8998,
+      "step": 704
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.614709636247607e-06,
+      "loss": 0.6612,
+      "step": 708
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6115188257817485e-06,
+      "loss": 0.678,
+      "step": 712
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.608328015315891e-06,
+      "loss": 0.7824,
+      "step": 716
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.605137204850032e-06,
+      "loss": 0.6569,
+      "step": 720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.601946394384174e-06,
+      "loss": 0.6318,
+      "step": 724
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.598755583918315e-06,
+      "loss": 0.7621,
+      "step": 728
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.595564773452458e-06,
+      "loss": 0.6681,
+      "step": 732
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.592373962986599e-06,
+      "loss": 0.7158,
+      "step": 736
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.5891831525207405e-06,
+      "loss": 0.8249,
+      "step": 740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.585992342054883e-06,
+      "loss": 0.6547,
+      "step": 744
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.582801531589024e-06,
+      "loss": 0.8209,
+      "step": 748
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.579610721123166e-06,
+      "loss": 0.55,
+      "step": 752
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.576419910657307e-06,
+      "loss": 0.6688,
+      "step": 756
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5732291001914496e-06,
+      "loss": 0.94,
+      "step": 760
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.570038289725591e-06,
+      "loss": 0.5695,
+      "step": 764
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5668474792597325e-06,
+      "loss": 0.5494,
+      "step": 768
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.563656668793874e-06,
+      "loss": 1.0452,
+      "step": 772
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.560465858328016e-06,
+      "loss": 0.8215,
+      "step": 776
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.557275047862158e-06,
+      "loss": 0.8754,
+      "step": 780
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.554084237396299e-06,
+      "loss": 0.7479,
+      "step": 784
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.550893426930441e-06,
+      "loss": 0.8235,
+      "step": 788
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.547702616464582e-06,
+      "loss": 0.8411,
+      "step": 792
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.5445118059987245e-06,
+      "loss": 0.7939,
+      "step": 796
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.541320995532866e-06,
+      "loss": 0.6333,
+      "step": 800
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.5381301850670074e-06,
+      "loss": 0.7022,
+      "step": 804
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.534939374601149e-06,
+      "loss": 0.6277,
+      "step": 808
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.531748564135291e-06,
+      "loss": 0.5253,
+      "step": 812
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.528557753669433e-06,
+      "loss": 0.6167,
+      "step": 816
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.525366943203574e-06,
+      "loss": 0.5005,
+      "step": 820
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.522176132737716e-06,
+      "loss": 0.8225,
+      "step": 824
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.518985322271857e-06,
+      "loss": 0.48,
+      "step": 828
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.515794511805999e-06,
+      "loss": 0.875,
+      "step": 832
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.512603701340141e-06,
+      "loss": 0.9044,
+      "step": 836
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.509412890874282e-06,
+      "loss": 0.7879,
+      "step": 840
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.506222080408424e-06,
+      "loss": 0.6177,
+      "step": 844
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.503031269942566e-06,
+      "loss": 0.5069,
+      "step": 848
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.499840459476708e-06,
+      "loss": 0.7872,
+      "step": 852
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.496649649010849e-06,
+      "loss": 0.9284,
+      "step": 856
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.4934588385449906e-06,
+      "loss": 0.8791,
+      "step": 860
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.490268028079132e-06,
+      "loss": 0.7776,
+      "step": 864
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.487077217613274e-06,
+      "loss": 1.0799,
+      "step": 868
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.483886407147416e-06,
+      "loss": 0.7379,
+      "step": 872
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.480695596681557e-06,
+      "loss": 0.9199,
+      "step": 876
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.477504786215699e-06,
+      "loss": 0.5865,
+      "step": 880
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.474313975749841e-06,
+      "loss": 0.725,
+      "step": 884
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.4711231652839825e-06,
+      "loss": 0.559,
+      "step": 888
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.467932354818124e-06,
+      "loss": 0.5129,
+      "step": 892
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4647415443522655e-06,
+      "loss": 0.8001,
+      "step": 896
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.461550733886407e-06,
+      "loss": 0.6969,
+      "step": 900
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.458359923420549e-06,
+      "loss": 0.6153,
+      "step": 904
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.455169112954691e-06,
+      "loss": 0.5391,
+      "step": 908
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.451978302488832e-06,
+      "loss": 0.913,
+      "step": 912
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.448787492022974e-06,
+      "loss": 0.9261,
+      "step": 916
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.445596681557116e-06,
+      "loss": 0.4837,
+      "step": 920
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4424058710912575e-06,
+      "loss": 0.8113,
+      "step": 924
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.439215060625399e-06,
+      "loss": 0.6893,
+      "step": 928
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.43602425015954e-06,
+      "loss": 0.7842,
+      "step": 932
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.432833439693682e-06,
+      "loss": 0.7386,
+      "step": 936
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.429642629227824e-06,
+      "loss": 0.7332,
+      "step": 940
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.426451818761966e-06,
+      "loss": 0.5775,
+      "step": 944
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.423261008296107e-06,
+      "loss": 0.8146,
+      "step": 948
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4200701978302494e-06,
+      "loss": 0.7466,
+      "step": 952
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.416879387364391e-06,
+      "loss": 0.6786,
+      "step": 956
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.413688576898532e-06,
+      "loss": 0.5748,
+      "step": 960
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.410497766432674e-06,
+      "loss": 0.5671,
+      "step": 964
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.407306955966816e-06,
+      "loss": 0.6996,
+      "step": 968
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.404116145500958e-06,
+      "loss": 0.9416,
+      "step": 972
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.400925335035099e-06,
+      "loss": 0.5928,
+      "step": 976
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.397734524569241e-06,
+      "loss": 0.6034,
+      "step": 980
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.394543714103383e-06,
+      "loss": 0.8989,
+      "step": 984
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.391352903637524e-06,
+      "loss": 0.8094,
+      "step": 988
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.388162093171666e-06,
+      "loss": 0.5301,
+      "step": 992
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.384971282705808e-06,
+      "loss": 1.0705,
+      "step": 996
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.38178047223995e-06,
+      "loss": 0.5879,
+      "step": 1000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.378589661774091e-06,
+      "loss": 0.8233,
+      "step": 1004
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3753988513082326e-06,
+      "loss": 0.476,
+      "step": 1008
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.372208040842375e-06,
+      "loss": 0.5304,
+      "step": 1012
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.369017230376516e-06,
+      "loss": 0.4795,
+      "step": 1016
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.365826419910658e-06,
+      "loss": 0.6065,
+      "step": 1020
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.362635609444799e-06,
+      "loss": 0.6391,
+      "step": 1024
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.359444798978942e-06,
+      "loss": 0.5775,
+      "step": 1028
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.356253988513083e-06,
+      "loss": 0.6313,
+      "step": 1032
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3530631780472246e-06,
+      "loss": 0.8725,
+      "step": 1036
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.349872367581366e-06,
+      "loss": 0.8685,
+      "step": 1040
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3466815571155075e-06,
+      "loss": 0.7983,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.34349074664965e-06,
+      "loss": 0.7019,
+      "step": 1048
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.340299936183791e-06,
+      "loss": 0.8012,
+      "step": 1052
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.337109125717933e-06,
+      "loss": 0.7772,
+      "step": 1056
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.333918315252074e-06,
+      "loss": 0.7787,
+      "step": 1060
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3307275047862165e-06,
+      "loss": 0.6103,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.327536694320358e-06,
+      "loss": 0.8516,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3243458838544995e-06,
+      "loss": 0.7831,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.321155073388641e-06,
+      "loss": 0.5453,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.317964262922782e-06,
+      "loss": 0.8411,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.314773452456925e-06,
+      "loss": 0.5653,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.311582641991066e-06,
+      "loss": 0.4892,
+      "step": 1088
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.308391831525208e-06,
+      "loss": 0.8213,
+      "step": 1092
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.305201021059349e-06,
+      "loss": 0.7465,
+      "step": 1096
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.3020102105934915e-06,
+      "loss": 0.671,
+      "step": 1100
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.298819400127633e-06,
+      "loss": 0.8927,
+      "step": 1104
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.295628589661774e-06,
+      "loss": 0.7458,
+      "step": 1108
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.292437779195916e-06,
+      "loss": 0.5961,
+      "step": 1112
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.289246968730057e-06,
+      "loss": 0.711,
+      "step": 1116
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.2860561582642e-06,
+      "loss": 0.705,
+      "step": 1120
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.282865347798341e-06,
+      "loss": 0.5282,
+      "step": 1124
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.279674537332483e-06,
+      "loss": 0.7639,
+      "step": 1128
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.276483726866624e-06,
+      "loss": 0.8014,
+      "step": 1132
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.273292916400766e-06,
+      "loss": 0.6999,
+      "step": 1136
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.270102105934908e-06,
+      "loss": 0.6685,
+      "step": 1140
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.266911295469049e-06,
+      "loss": 0.6752,
+      "step": 1144
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.263720485003191e-06,
+      "loss": 0.7828,
+      "step": 1148
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.260529674537332e-06,
+      "loss": 0.8872,
+      "step": 1152
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.257338864071475e-06,
+      "loss": 0.9956,
+      "step": 1156
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.254148053605616e-06,
+      "loss": 0.6747,
+      "step": 1160
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.2509572431397575e-06,
+      "loss": 0.6513,
+      "step": 1164
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.247766432673899e-06,
+      "loss": 0.515,
+      "step": 1168
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.244575622208041e-06,
+      "loss": 0.8237,
+      "step": 1172
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.241384811742183e-06,
+      "loss": 0.7621,
+      "step": 1176
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.238194001276324e-06,
+      "loss": 0.5645,
+      "step": 1180
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.235003190810466e-06,
+      "loss": 0.508,
+      "step": 1184
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.231812380344608e-06,
+      "loss": 0.7818,
+      "step": 1188
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2286215698787495e-06,
+      "loss": 0.4548,
+      "step": 1192
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.225430759412891e-06,
+      "loss": 0.4979,
+      "step": 1196
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2222399489470325e-06,
+      "loss": 0.8764,
+      "step": 1200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.219049138481175e-06,
+      "loss": 0.7185,
+      "step": 1204
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.215858328015316e-06,
+      "loss": 0.8167,
+      "step": 1208
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.212667517549458e-06,
+      "loss": 0.8222,
+      "step": 1212
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.209476707083599e-06,
+      "loss": 0.6022,
+      "step": 1216
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2062858966177415e-06,
+      "loss": 0.8809,
+      "step": 1220
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.203095086151883e-06,
+      "loss": 0.3133,
+      "step": 1224
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1999042756860244e-06,
+      "loss": 0.8026,
+      "step": 1228
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.196713465220166e-06,
+      "loss": 0.6442,
+      "step": 1232
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.193522654754308e-06,
+      "loss": 0.4915,
+      "step": 1236
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.19033184428845e-06,
+      "loss": 0.8688,
+      "step": 1240
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.187141033822591e-06,
+      "loss": 0.6234,
+      "step": 1244
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1839502233567335e-06,
+      "loss": 0.6482,
+      "step": 1248
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.180759412890875e-06,
+      "loss": 0.7072,
+      "step": 1252
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.177568602425016e-06,
+      "loss": 0.7761,
+      "step": 1256
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.174377791959158e-06,
+      "loss": 0.7278,
+      "step": 1260
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.1711869814933e-06,
+      "loss": 0.8857,
+      "step": 1264
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.167996171027442e-06,
+      "loss": 0.4587,
+      "step": 1268
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.164805360561583e-06,
+      "loss": 0.8217,
+      "step": 1272
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.161614550095725e-06,
+      "loss": 0.5097,
+      "step": 1276
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.158423739629867e-06,
+      "loss": 0.8389,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.155232929164008e-06,
+      "loss": 0.6538,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.15204211869815e-06,
+      "loss": 0.6541,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.148851308232291e-06,
+      "loss": 0.8788,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.145660497766433e-06,
+      "loss": 0.4737,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.142469687300575e-06,
+      "loss": 0.6639,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.139278876834717e-06,
+      "loss": 0.768,
+      "step": 1304
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.136088066368858e-06,
+      "loss": 0.4519,
+      "step": 1308
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1328972559029995e-06,
+      "loss": 0.5798,
+      "step": 1312
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.129706445437142e-06,
+      "loss": 0.617,
+      "step": 1316
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.126515634971283e-06,
+      "loss": 0.5391,
+      "step": 1320
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.123324824505425e-06,
+      "loss": 0.6149,
+      "step": 1324
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.120134014039566e-06,
+      "loss": 0.7053,
+      "step": 1328
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.116943203573708e-06,
+      "loss": 0.6092,
+      "step": 1332
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.11375239310785e-06,
+      "loss": 0.6277,
+      "step": 1336
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.1105615826419915e-06,
+      "loss": 0.6903,
+      "step": 1340
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.107370772176133e-06,
+      "loss": 0.8365,
+      "step": 1344
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.1041799617102745e-06,
+      "loss": 0.752,
+      "step": 1348
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.100989151244417e-06,
+      "loss": 0.7835,
+      "step": 1352
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.097798340778558e-06,
+      "loss": 0.7287,
+      "step": 1356
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0946075303127e-06,
+      "loss": 0.6843,
+      "step": 1360
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.091416719846841e-06,
+      "loss": 0.7026,
+      "step": 1364
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.088225909380983e-06,
+      "loss": 0.5323,
+      "step": 1368
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.085035098915125e-06,
+      "loss": 0.5855,
+      "step": 1372
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0818442884492665e-06,
+      "loss": 0.6391,
+      "step": 1376
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.078653477983408e-06,
+      "loss": 0.5578,
+      "step": 1380
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.075462667517549e-06,
+      "loss": 0.5891,
+      "step": 1384
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.072271857051692e-06,
+      "loss": 1.0298,
+      "step": 1388
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.069081046585833e-06,
+      "loss": 0.5537,
+      "step": 1392
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.065890236119975e-06,
+      "loss": 0.3564,
+      "step": 1396
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.062699425654116e-06,
+      "loss": 0.5444,
+      "step": 1400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.059508615188258e-06,
+      "loss": 0.5904,
+      "step": 1404
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.0563178047224e-06,
+      "loss": 0.4737,
+      "step": 1408
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.053126994256541e-06,
+      "loss": 0.8174,
+      "step": 1412
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.049936183790683e-06,
+      "loss": 0.5257,
+      "step": 1416
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.046745373324824e-06,
+      "loss": 0.5681,
+      "step": 1420
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.043554562858967e-06,
+      "loss": 0.714,
+      "step": 1424
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.040363752393108e-06,
+      "loss": 0.6049,
+      "step": 1428
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.0371729419272496e-06,
+      "loss": 0.8058,
+      "step": 1432
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.033982131461391e-06,
+      "loss": 0.6571,
+      "step": 1436
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.030791320995533e-06,
+      "loss": 0.6082,
+      "step": 1440
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.027600510529675e-06,
+      "loss": 0.8361,
+      "step": 1444
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.024409700063816e-06,
+      "loss": 0.69,
+      "step": 1448
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.021218889597958e-06,
+      "loss": 0.4518,
+      "step": 1452
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0180280791321e-06,
+      "loss": 0.747,
+      "step": 1456
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0148372686662416e-06,
+      "loss": 0.7835,
+      "step": 1460
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.011646458200383e-06,
+      "loss": 0.4455,
+      "step": 1464
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0084556477345245e-06,
+      "loss": 0.6174,
+      "step": 1468
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.005264837268667e-06,
+      "loss": 0.624,
+      "step": 1472
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.002074026802808e-06,
+      "loss": 0.5624,
+      "step": 1476
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.99888321633695e-06,
+      "loss": 0.5388,
+      "step": 1480
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.995692405871091e-06,
+      "loss": 0.5208,
+      "step": 1484
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.9925015954052335e-06,
+      "loss": 0.4847,
+      "step": 1488
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.989310784939375e-06,
+      "loss": 0.6944,
+      "step": 1492
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.9861199744735165e-06,
+      "loss": 0.5778,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.982929164007659e-06,
+      "loss": 0.7586,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.9797383535418e-06,
+      "loss": 0.7339,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.976547543075942e-06,
+      "loss": 0.7787,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.973356732610083e-06,
+      "loss": 0.6295,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.9701659221442255e-06,
+      "loss": 0.5411,
+      "step": 1516
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.966975111678367e-06,
+      "loss": 0.6872,
+      "step": 1520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9637843012125085e-06,
+      "loss": 0.8006,
+      "step": 1524
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.96059349074665e-06,
+      "loss": 0.5383,
+      "step": 1528
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.957402680280792e-06,
+      "loss": 0.5238,
+      "step": 1532
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.954211869814934e-06,
+      "loss": 0.4508,
+      "step": 1536
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.951021059349075e-06,
+      "loss": 0.7398,
+      "step": 1540
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.947830248883217e-06,
+      "loss": 0.3801,
+      "step": 1544
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.944639438417358e-06,
+      "loss": 0.5496,
+      "step": 1548
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9414486279515005e-06,
+      "loss": 0.6008,
+      "step": 1552
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.938257817485642e-06,
+      "loss": 0.6282,
+      "step": 1556
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.935067007019783e-06,
+      "loss": 0.4993,
+      "step": 1560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.931876196553925e-06,
+      "loss": 0.5208,
+      "step": 1564
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.928685386088067e-06,
+      "loss": 0.5643,
+      "step": 1568
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.925494575622209e-06,
+      "loss": 0.3732,
+      "step": 1572
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.92230376515635e-06,
+      "loss": 0.5719,
+      "step": 1576
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.919112954690492e-06,
+      "loss": 0.5204,
+      "step": 1580
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.915922144224633e-06,
+      "loss": 0.4546,
+      "step": 1584
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.912731333758775e-06,
+      "loss": 0.7574,
+      "step": 1588
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.909540523292917e-06,
+      "loss": 0.6268,
+      "step": 1592
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.906349712827058e-06,
+      "loss": 0.6145,
+      "step": 1596
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.9031589023612e-06,
+      "loss": 0.6054,
+      "step": 1600
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.899968091895342e-06,
+      "loss": 0.5141,
+      "step": 1604
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.896777281429484e-06,
+      "loss": 0.5058,
+      "step": 1608
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.893586470963625e-06,
+      "loss": 0.6534,
+      "step": 1612
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8903956604977665e-06,
+      "loss": 0.7528,
+      "step": 1616
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.887204850031908e-06,
+      "loss": 0.7047,
+      "step": 1620
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.88401403956605e-06,
+      "loss": 0.533,
+      "step": 1624
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.880823229100192e-06,
+      "loss": 0.8042,
+      "step": 1628
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.877632418634333e-06,
+      "loss": 0.6895,
+      "step": 1632
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.874441608168475e-06,
+      "loss": 0.806,
+      "step": 1636
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.871250797702617e-06,
+      "loss": 0.807,
+      "step": 1640
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8680599872367585e-06,
+      "loss": 0.4688,
+      "step": 1644
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8648691767709e-06,
+      "loss": 0.3543,
+      "step": 1648
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.8616783663050414e-06,
+      "loss": 0.7808,
+      "step": 1652
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.858487555839183e-06,
+      "loss": 0.5017,
+      "step": 1656
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.855296745373325e-06,
+      "loss": 0.5953,
+      "step": 1660
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.852105934907467e-06,
+      "loss": 0.44,
+      "step": 1664
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.848915124441608e-06,
+      "loss": 0.8274,
+      "step": 1668
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.84572431397575e-06,
+      "loss": 0.8961,
+      "step": 1672
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.842533503509892e-06,
+      "loss": 0.4973,
+      "step": 1676
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.8393426930440334e-06,
+      "loss": 0.5691,
+      "step": 1680
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.836151882578175e-06,
+      "loss": 0.7323,
+      "step": 1684
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.832961072112316e-06,
+      "loss": 0.4802,
+      "step": 1688
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.829770261646459e-06,
+      "loss": 0.535,
+      "step": 1692
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.8265794511806e-06,
+      "loss": 0.6412,
+      "step": 1696
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.824186343331207e-06,
+      "loss": 0.5543,
+      "step": 1700
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.820995532865348e-06,
+      "loss": 0.6086,
+      "step": 1704
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.81780472239949e-06,
+      "loss": 0.4478,
+      "step": 1708
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.814613911933632e-06,
+      "loss": 0.754,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.8114231014677734e-06,
+      "loss": 0.9988,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.808232291001915e-06,
+      "loss": 0.6436,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.8050414805360563e-06,
+      "loss": 0.7755,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.801850670070198e-06,
+      "loss": 0.7917,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.79865985960434e-06,
+      "loss": 0.5693,
+      "step": 1732
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7954690491384816e-06,
+      "loss": 0.7169,
+      "step": 1736
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.792278238672623e-06,
+      "loss": 0.6066,
+      "step": 1740
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7890874282067645e-06,
+      "loss": 0.7224,
+      "step": 1744
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.785896617740907e-06,
+      "loss": 0.5065,
+      "step": 1748
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7827058072750483e-06,
+      "loss": 0.3499,
+      "step": 1752
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.77951499680919e-06,
+      "loss": 0.6114,
+      "step": 1756
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7763241863433313e-06,
+      "loss": 0.612,
+      "step": 1760
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.773133375877473e-06,
+      "loss": 0.9457,
+      "step": 1764
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.769942565411615e-06,
+      "loss": 0.5096,
+      "step": 1768
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7667517549457565e-06,
+      "loss": 0.5297,
+      "step": 1772
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.763560944479898e-06,
+      "loss": 0.3232,
+      "step": 1776
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.76037013401404e-06,
+      "loss": 0.4437,
+      "step": 1780
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7571793235481818e-06,
+      "loss": 0.5564,
+      "step": 1784
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7539885130823233e-06,
+      "loss": 0.598,
+      "step": 1788
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7507977026164647e-06,
+      "loss": 0.4804,
+      "step": 1792
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7476068921506066e-06,
+      "loss": 0.7052,
+      "step": 1796
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.744416081684748e-06,
+      "loss": 0.4539,
+      "step": 1800
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.74122527121889e-06,
+      "loss": 0.6006,
+      "step": 1804
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.738034460753032e-06,
+      "loss": 0.5061,
+      "step": 1808
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7348436502871733e-06,
+      "loss": 0.7403,
+      "step": 1812
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.731652839821315e-06,
+      "loss": 0.8152,
+      "step": 1816
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7284620293554563e-06,
+      "loss": 0.5466,
+      "step": 1820
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7252712188895986e-06,
+      "loss": 0.4677,
+      "step": 1824
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.72208040842374e-06,
+      "loss": 0.468,
+      "step": 1828
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7188895979578815e-06,
+      "loss": 0.759,
+      "step": 1832
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.715698787492023e-06,
+      "loss": 0.6796,
+      "step": 1836
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7125079770261653e-06,
+      "loss": 0.7896,
+      "step": 1840
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.709317166560307e-06,
+      "loss": 0.683,
+      "step": 1844
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7061263560944483e-06,
+      "loss": 0.4372,
+      "step": 1848
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7029355456285897e-06,
+      "loss": 0.7383,
+      "step": 1852
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6997447351627312e-06,
+      "loss": 0.5463,
+      "step": 1856
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6965539246968735e-06,
+      "loss": 0.6441,
+      "step": 1860
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.693363114231015e-06,
+      "loss": 0.5837,
+      "step": 1864
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6901723037651565e-06,
+      "loss": 0.6439,
+      "step": 1868
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.686981493299298e-06,
+      "loss": 0.5672,
+      "step": 1872
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6837906828334403e-06,
+      "loss": 0.4494,
+      "step": 1876
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6805998723675817e-06,
+      "loss": 0.8502,
+      "step": 1880
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.677409061901723e-06,
+      "loss": 0.4624,
+      "step": 1884
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6742182514358647e-06,
+      "loss": 0.4339,
+      "step": 1888
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6710274409700066e-06,
+      "loss": 0.9334,
+      "step": 1892
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6678366305041485e-06,
+      "loss": 0.9478,
+      "step": 1896
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.66464582003829e-06,
+      "loss": 0.3005,
+      "step": 1900
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6614550095724314e-06,
+      "loss": 0.5926,
+      "step": 1904
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6582641991065733e-06,
+      "loss": 0.4696,
+      "step": 1908
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.655073388640715e-06,
+      "loss": 0.4898,
+      "step": 1912
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6518825781748567e-06,
+      "loss": 0.7388,
+      "step": 1916
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6486917677089985e-06,
+      "loss": 0.3915,
+      "step": 1920
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.64550095724314e-06,
+      "loss": 0.3001,
+      "step": 1924
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6423101467772815e-06,
+      "loss": 0.5252,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6391193363114234e-06,
+      "loss": 0.718,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6359285258455653e-06,
+      "loss": 0.7228,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6327377153797067e-06,
+      "loss": 0.769,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6295469049138482e-06,
+      "loss": 0.8362,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.62635609444799e-06,
+      "loss": 0.6777,
+      "step": 1948
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.623165283982132e-06,
+      "loss": 0.8677,
+      "step": 1952
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6199744735162735e-06,
+      "loss": 0.7517,
+      "step": 1956
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.616783663050415e-06,
+      "loss": 0.3785,
+      "step": 1960
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6135928525845564e-06,
+      "loss": 0.5111,
+      "step": 1964
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6104020421186987e-06,
+      "loss": 0.4939,
+      "step": 1968
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.60721123165284e-06,
+      "loss": 0.5917,
+      "step": 1972
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6040204211869817e-06,
+      "loss": 0.7104,
+      "step": 1976
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.600829610721123e-06,
+      "loss": 0.7933,
+      "step": 1980
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5976388002552655e-06,
+      "loss": 0.4683,
+      "step": 1984
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.594447989789407e-06,
+      "loss": 0.596,
+      "step": 1988
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5912571793235484e-06,
+      "loss": 0.6115,
+      "step": 1992
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.58806636885769e-06,
+      "loss": 0.3607,
+      "step": 1996
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5848755583918313e-06,
+      "loss": 0.4731,
+      "step": 2000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5816847479259737e-06,
+      "loss": 0.6709,
+      "step": 2004
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.578493937460115e-06,
+      "loss": 0.6253,
+      "step": 2008
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5753031269942566e-06,
+      "loss": 0.6493,
+      "step": 2012
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5721123165283985e-06,
+      "loss": 0.6203,
+      "step": 2016
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5689215060625404e-06,
+      "loss": 0.5535,
+      "step": 2020
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.565730695596682e-06,
+      "loss": 0.6694,
+      "step": 2024
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5625398851308233e-06,
+      "loss": 0.7581,
+      "step": 2028
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5593490746649652e-06,
+      "loss": 0.6517,
+      "step": 2032
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5561582641991067e-06,
+      "loss": 0.588,
+      "step": 2036
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5529674537332486e-06,
+      "loss": 0.7555,
+      "step": 2040
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.54977664326739e-06,
+      "loss": 0.6893,
+      "step": 2044
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.546585832801532e-06,
+      "loss": 0.5,
+      "step": 2048
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5433950223356734e-06,
+      "loss": 0.3433,
+      "step": 2052
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5402042118698153e-06,
+      "loss": 0.6995,
+      "step": 2056
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.537013401403957e-06,
+      "loss": 0.7064,
+      "step": 2060
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5338225909380987e-06,
+      "loss": 0.4641,
+      "step": 2064
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.53063178047224e-06,
+      "loss": 0.5876,
+      "step": 2068
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5274409700063816e-06,
+      "loss": 0.6002,
+      "step": 2072
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.524250159540524e-06,
+      "loss": 0.6374,
+      "step": 2076
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5210593490746654e-06,
+      "loss": 0.5119,
+      "step": 2080
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.517868538608807e-06,
+      "loss": 0.7321,
+      "step": 2084
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5146777281429483e-06,
+      "loss": 0.9369,
+      "step": 2088
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5114869176770907e-06,
+      "loss": 0.6868,
+      "step": 2092
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.508296107211232e-06,
+      "loss": 0.3827,
+      "step": 2096
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5051052967453736e-06,
+      "loss": 0.4267,
+      "step": 2100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.501914486279515e-06,
+      "loss": 0.5196,
+      "step": 2104
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4987236758136565e-06,
+      "loss": 0.533,
+      "step": 2108
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.495532865347799e-06,
+      "loss": 0.5341,
+      "step": 2112
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4923420548819403e-06,
+      "loss": 0.6026,
+      "step": 2116
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.489151244416082e-06,
+      "loss": 0.5235,
+      "step": 2120
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4859604339502233e-06,
+      "loss": 0.6593,
+      "step": 2124
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4827696234843656e-06,
+      "loss": 0.5826,
+      "step": 2128
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.479578813018507e-06,
+      "loss": 0.4661,
+      "step": 2132
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4763880025526485e-06,
+      "loss": 0.913,
+      "step": 2136
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.47319719208679e-06,
+      "loss": 0.5038,
+      "step": 2140
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.470006381620932e-06,
+      "loss": 0.4766,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4668155711550738e-06,
+      "loss": 0.512,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4636247606892152e-06,
+      "loss": 0.5954,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.460433950223357e-06,
+      "loss": 0.6475,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4572431397574986e-06,
+      "loss": 0.5139,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4540523292916405e-06,
+      "loss": 0.5509,
+      "step": 2164
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.450861518825782e-06,
+      "loss": 0.6483,
+      "step": 2168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.447670708359924e-06,
+      "loss": 0.7284,
+      "step": 2172
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4444798978940653e-06,
+      "loss": 0.4801,
+      "step": 2176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.441289087428207e-06,
+      "loss": 0.6135,
+      "step": 2180
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4380982769623487e-06,
+      "loss": 0.3792,
+      "step": 2184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4349074664964906e-06,
+      "loss": 0.4799,
+      "step": 2188
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.431716656030632e-06,
+      "loss": 0.5068,
+      "step": 2192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4285258455647735e-06,
+      "loss": 0.661,
+      "step": 2196
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4253350350989154e-06,
+      "loss": 0.592,
+      "step": 2200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4221442246330573e-06,
+      "loss": 0.6508,
+      "step": 2204
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.418953414167199e-06,
+      "loss": 0.5552,
+      "step": 2208
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4157626037013403e-06,
+      "loss": 0.5729,
+      "step": 2212
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4125717932354817e-06,
+      "loss": 0.6907,
+      "step": 2216
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.409380982769624e-06,
+      "loss": 0.4163,
+      "step": 2220
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4061901723037655e-06,
+      "loss": 0.5248,
+      "step": 2224
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.402999361837907e-06,
+      "loss": 0.4138,
+      "step": 2228
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3998085513720485e-06,
+      "loss": 0.4799,
+      "step": 2232
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3966177409061908e-06,
+      "loss": 0.4414,
+      "step": 2236
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3934269304403322e-06,
+      "loss": 0.5975,
+      "step": 2240
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3902361199744737e-06,
+      "loss": 0.5745,
+      "step": 2244
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.387045309508615e-06,
+      "loss": 0.4061,
+      "step": 2248
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3838544990427567e-06,
+      "loss": 0.6807,
+      "step": 2252
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.380663688576899e-06,
+      "loss": 0.4239,
+      "step": 2256
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3774728781110404e-06,
+      "loss": 0.5731,
+      "step": 2260
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.374282067645182e-06,
+      "loss": 0.4826,
+      "step": 2264
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.371091257179324e-06,
+      "loss": 0.5014,
+      "step": 2268
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3679004467134657e-06,
+      "loss": 0.4492,
+      "step": 2272
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.364709636247607e-06,
+      "loss": 0.5584,
+      "step": 2276
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3615188257817486e-06,
+      "loss": 0.5931,
+      "step": 2280
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3583280153158905e-06,
+      "loss": 0.5623,
+      "step": 2284
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.355137204850032e-06,
+      "loss": 0.7667,
+      "step": 2288
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.351946394384174e-06,
+      "loss": 0.4875,
+      "step": 2292
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3487555839183154e-06,
+      "loss": 0.3926,
+      "step": 2296
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3455647734524573e-06,
+      "loss": 0.5208,
+      "step": 2300
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3423739629865987e-06,
+      "loss": 0.5546,
+      "step": 2304
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3391831525207406e-06,
+      "loss": 0.5181,
+      "step": 2308
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3359923420548825e-06,
+      "loss": 0.5089,
+      "step": 2312
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.332801531589024e-06,
+      "loss": 0.6503,
+      "step": 2316
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3296107211231655e-06,
+      "loss": 0.5613,
+      "step": 2320
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.326419910657307e-06,
+      "loss": 0.5959,
+      "step": 2324
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3232291001914492e-06,
+      "loss": 0.6191,
+      "step": 2328
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3200382897255907e-06,
+      "loss": 0.5365,
+      "step": 2332
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.316847479259732e-06,
+      "loss": 0.362,
+      "step": 2336
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3136566687938737e-06,
+      "loss": 0.6065,
+      "step": 2340
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.310465858328016e-06,
+      "loss": 0.6887,
+      "step": 2344
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3072750478621574e-06,
+      "loss": 0.8036,
+      "step": 2348
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.304084237396299e-06,
+      "loss": 0.5733,
+      "step": 2352
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3008934269304404e-06,
+      "loss": 0.4841,
+      "step": 2356
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.297702616464582e-06,
+      "loss": 0.3678,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.294511805998724e-06,
+      "loss": 0.3585,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2913209955328656e-06,
+      "loss": 0.5592,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.288130185067007e-06,
+      "loss": 0.5168,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2849393746011486e-06,
+      "loss": 0.744,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.281748564135291e-06,
+      "loss": 0.3844,
+      "step": 2380
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2785577536694324e-06,
+      "loss": 0.7852,
+      "step": 2384
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.275366943203574e-06,
+      "loss": 0.5413,
+      "step": 2388
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2721761327377153e-06,
+      "loss": 0.3793,
+      "step": 2392
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.268985322271857e-06,
+      "loss": 0.699,
+      "step": 2396
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.265794511805999e-06,
+      "loss": 0.5154,
+      "step": 2400
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2626037013401406e-06,
+      "loss": 0.5389,
+      "step": 2404
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2594128908742825e-06,
+      "loss": 0.5196,
+      "step": 2408
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.256222080408424e-06,
+      "loss": 0.5163,
+      "step": 2412
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.253031269942566e-06,
+      "loss": 0.4499,
+      "step": 2416
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2498404594767073e-06,
+      "loss": 0.398,
+      "step": 2420
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.246649649010849e-06,
+      "loss": 0.6095,
+      "step": 2424
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2434588385449907e-06,
+      "loss": 0.5462,
+      "step": 2428
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.240268028079132e-06,
+      "loss": 0.6632,
+      "step": 2432
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.237077217613274e-06,
+      "loss": 0.489,
+      "step": 2436
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.233886407147416e-06,
+      "loss": 0.6678,
+      "step": 2440
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2306955966815574e-06,
+      "loss": 0.3573,
+      "step": 2444
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.227504786215699e-06,
+      "loss": 0.4844,
+      "step": 2448
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2243139757498408e-06,
+      "loss": 0.6646,
+      "step": 2452
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2211231652839826e-06,
+      "loss": 0.7763,
+      "step": 2456
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.217932354818124e-06,
+      "loss": 0.4772,
+      "step": 2460
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2147415443522656e-06,
+      "loss": 0.3963,
+      "step": 2464
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.211550733886407e-06,
+      "loss": 0.579,
+      "step": 2468
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.2083599234205494e-06,
+      "loss": 0.4923,
+      "step": 2472
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.205169112954691e-06,
+      "loss": 0.3719,
+      "step": 2476
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.2019783024888323e-06,
+      "loss": 0.3468,
+      "step": 2480
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.1987874920229738e-06,
+      "loss": 0.4931,
+      "step": 2484
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.195596681557116e-06,
+      "loss": 0.5714,
+      "step": 2488
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1924058710912576e-06,
+      "loss": 0.4509,
+      "step": 2492
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.189215060625399e-06,
+      "loss": 0.3998,
+      "step": 2496
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1860242501595405e-06,
+      "loss": 0.425,
+      "step": 2500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.182833439693682e-06,
+      "loss": 0.438,
+      "step": 2504
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1796426292278243e-06,
+      "loss": 0.5492,
+      "step": 2508
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1764518187619658e-06,
+      "loss": 0.6652,
+      "step": 2512
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1732610082961072e-06,
+      "loss": 0.5869,
+      "step": 2516
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.170070197830249e-06,
+      "loss": 0.2636,
+      "step": 2520
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.166879387364391e-06,
+      "loss": 0.5958,
+      "step": 2524
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1636885768985325e-06,
+      "loss": 0.8346,
+      "step": 2528
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.160497766432674e-06,
+      "loss": 0.4926,
+      "step": 2532
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.157306955966816e-06,
+      "loss": 0.4706,
+      "step": 2536
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1541161455009573e-06,
+      "loss": 0.5029,
+      "step": 2540
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1509253350350992e-06,
+      "loss": 0.5374,
+      "step": 2544
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1477345245692407e-06,
+      "loss": 0.6324,
+      "step": 2548
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1445437141033826e-06,
+      "loss": 0.6013,
+      "step": 2552
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.141352903637524e-06,
+      "loss": 0.5026,
+      "step": 2556
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.138162093171666e-06,
+      "loss": 0.2409,
+      "step": 2560
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.134971282705808e-06,
+      "loss": 0.4474,
+      "step": 2564
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1317804722399493e-06,
+      "loss": 0.5061,
+      "step": 2568
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1285896617740908e-06,
+      "loss": 0.8622,
+      "step": 2572
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1253988513082323e-06,
+      "loss": 0.673,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1222080408423746e-06,
+      "loss": 0.2795,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.119017230376516e-06,
+      "loss": 0.396,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1158264199106575e-06,
+      "loss": 0.5924,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.112635609444799e-06,
+      "loss": 0.5165,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1094447989789413e-06,
+      "loss": 0.4082,
+      "step": 2596
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1062539885130828e-06,
+      "loss": 0.4698,
+      "step": 2600
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1030631780472242e-06,
+      "loss": 0.2842,
+      "step": 2604
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0998723675813657e-06,
+      "loss": 0.4788,
+      "step": 2608
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.096681557115507e-06,
+      "loss": 0.3991,
+      "step": 2612
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.0934907466496495e-06,
+      "loss": 0.6134,
+      "step": 2616
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.090299936183791e-06,
+      "loss": 0.9227,
+      "step": 2620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0871091257179324e-06,
+      "loss": 0.5832,
+      "step": 2624
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.083918315252074e-06,
+      "loss": 0.4432,
+      "step": 2628
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0807275047862162e-06,
+      "loss": 0.3732,
+      "step": 2632
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0775366943203577e-06,
+      "loss": 0.6914,
+      "step": 2636
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.074345883854499e-06,
+      "loss": 0.5396,
+      "step": 2640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0711550733886406e-06,
+      "loss": 0.597,
+      "step": 2644
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0679642629227825e-06,
+      "loss": 0.6673,
+      "step": 2648
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0647734524569244e-06,
+      "loss": 0.4989,
+      "step": 2652
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.061582641991066e-06,
+      "loss": 0.4416,
+      "step": 2656
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0583918315252078e-06,
+      "loss": 0.3489,
+      "step": 2660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0552010210593493e-06,
+      "loss": 0.3579,
+      "step": 2664
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.052010210593491e-06,
+      "loss": 0.4483,
+      "step": 2668
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0488194001276326e-06,
+      "loss": 0.498,
+      "step": 2672
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0456285896617745e-06,
+      "loss": 0.5452,
+      "step": 2676
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.042437779195916e-06,
+      "loss": 0.2506,
+      "step": 2680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0392469687300575e-06,
+      "loss": 0.2542,
+      "step": 2684
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0360561582641993e-06,
+      "loss": 0.4921,
+      "step": 2688
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0328653477983412e-06,
+      "loss": 0.4847,
+      "step": 2692
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0296745373324827e-06,
+      "loss": 0.6631,
+      "step": 2696
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.026483726866624e-06,
+      "loss": 0.6046,
+      "step": 2700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.023292916400766e-06,
+      "loss": 0.4858,
+      "step": 2704
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.020102105934908e-06,
+      "loss": 0.4847,
+      "step": 2708
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0169112954690494e-06,
+      "loss": 0.5442,
+      "step": 2712
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.013720485003191e-06,
+      "loss": 0.58,
+      "step": 2716
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0105296745373324e-06,
+      "loss": 0.5303,
+      "step": 2720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0073388640714747e-06,
+      "loss": 0.3083,
+      "step": 2724
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.004148053605616e-06,
+      "loss": 0.6125,
+      "step": 2728
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0009572431397576e-06,
+      "loss": 0.5361,
+      "step": 2732
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.997766432673899e-06,
+      "loss": 0.509,
+      "step": 2736
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9945756222080414e-06,
+      "loss": 0.7011,
+      "step": 2740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.991384811742183e-06,
+      "loss": 0.5548,
+      "step": 2744
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9881940012763244e-06,
+      "loss": 0.343,
+      "step": 2748
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.985003190810466e-06,
+      "loss": 0.6061,
+      "step": 2752
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9818123803446077e-06,
+      "loss": 0.6704,
+      "step": 2756
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9786215698787496e-06,
+      "loss": 0.4994,
+      "step": 2760
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.975430759412891e-06,
+      "loss": 0.3216,
+      "step": 2764
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9722399489470326e-06,
+      "loss": 0.5823,
+      "step": 2768
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9690491384811745e-06,
+      "loss": 0.3471,
+      "step": 2772
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9658583280153163e-06,
+      "loss": 0.2349,
+      "step": 2776
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.962667517549458e-06,
+      "loss": 0.4946,
+      "step": 2780
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9594767070835993e-06,
+      "loss": 0.366,
+      "step": 2784
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.956285896617741e-06,
+      "loss": 0.2581,
+      "step": 2788
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9530950861518827e-06,
+      "loss": 0.4435,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9499042756860245e-06,
+      "loss": 0.6195,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.946713465220166e-06,
+      "loss": 0.4006,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.943522654754308e-06,
+      "loss": 0.4902,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9403318442884494e-06,
+      "loss": 0.8388,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9371410338225913e-06,
+      "loss": 0.631,
+      "step": 2812
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.933950223356733e-06,
+      "loss": 0.5347,
+      "step": 2816
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9307594128908746e-06,
+      "loss": 0.6902,
+      "step": 2820
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.927568602425016e-06,
+      "loss": 0.4865,
+      "step": 2824
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9243777919591576e-06,
+      "loss": 0.4538,
+      "step": 2828
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9211869814933e-06,
+      "loss": 0.4446,
+      "step": 2832
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9179961710274414e-06,
+      "loss": 0.6213,
+      "step": 2836
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.914805360561583e-06,
+      "loss": 0.4046,
+      "step": 2840
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9116145500957243e-06,
+      "loss": 0.4247,
+      "step": 2844
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9084237396298666e-06,
+      "loss": 0.4863,
+      "step": 2848
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.905232929164008e-06,
+      "loss": 0.3327,
+      "step": 2852
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9020421186981496e-06,
+      "loss": 0.5872,
+      "step": 2856
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.898851308232291e-06,
+      "loss": 0.5875,
+      "step": 2860
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8956604977664325e-06,
+      "loss": 0.3457,
+      "step": 2864
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.892469687300575e-06,
+      "loss": 0.3893,
+      "step": 2868
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8892788768347163e-06,
+      "loss": 0.6332,
+      "step": 2872
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8860880663688578e-06,
+      "loss": 0.5131,
+      "step": 2876
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8828972559029992e-06,
+      "loss": 0.586,
+      "step": 2880
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8797064454371415e-06,
+      "loss": 0.2341,
+      "step": 2884
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.876515634971283e-06,
+      "loss": 0.3804,
+      "step": 2888
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8733248245054245e-06,
+      "loss": 0.6579,
+      "step": 2892
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.870134014039566e-06,
+      "loss": 0.617,
+      "step": 2896
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.866943203573708e-06,
+      "loss": 0.3744,
+      "step": 2900
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8637523931078497e-06,
+      "loss": 0.6594,
+      "step": 2904
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8605615826419912e-06,
+      "loss": 0.3151,
+      "step": 2908
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.857370772176133e-06,
+      "loss": 0.3211,
+      "step": 2912
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8541799617102746e-06,
+      "loss": 0.6055,
+      "step": 2916
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8509891512444165e-06,
+      "loss": 0.5992,
+      "step": 2920
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.847798340778558e-06,
+      "loss": 0.4327,
+      "step": 2924
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8446075303127e-06,
+      "loss": 0.4729,
+      "step": 2928
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8414167198468413e-06,
+      "loss": 0.2755,
+      "step": 2932
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8382259093809828e-06,
+      "loss": 0.5271,
+      "step": 2936
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8350350989151247e-06,
+      "loss": 0.447,
+      "step": 2940
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8318442884492666e-06,
+      "loss": 0.2964,
+      "step": 2944
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.828653477983408e-06,
+      "loss": 0.5652,
+      "step": 2948
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8254626675175495e-06,
+      "loss": 0.2943,
+      "step": 2952
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8222718570516914e-06,
+      "loss": 0.615,
+      "step": 2956
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8190810465858333e-06,
+      "loss": 0.4737,
+      "step": 2960
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8158902361199748e-06,
+      "loss": 0.5275,
+      "step": 2964
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8126994256541162e-06,
+      "loss": 0.3709,
+      "step": 2968
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8095086151882577e-06,
+      "loss": 0.4134,
+      "step": 2972
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8063178047224e-06,
+      "loss": 0.5188,
+      "step": 2976
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8031269942565415e-06,
+      "loss": 0.3888,
+      "step": 2980
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.799936183790683e-06,
+      "loss": 0.4893,
+      "step": 2984
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7967453733248244e-06,
+      "loss": 0.4538,
+      "step": 2988
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7935545628589667e-06,
+      "loss": 0.5402,
+      "step": 2992
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7903637523931082e-06,
+      "loss": 0.2905,
+      "step": 2996
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7871729419272497e-06,
+      "loss": 0.2932,
+      "step": 3000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.783982131461391e-06,
+      "loss": 0.5082,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.780791320995533e-06,
+      "loss": 0.6075,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.777600510529675e-06,
+      "loss": 0.4388,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7744097000638164e-06,
+      "loss": 0.5735,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.771218889597958e-06,
+      "loss": 0.5469,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7680280791320998e-06,
+      "loss": 0.4436,
+      "step": 3024
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7648372686662417e-06,
+      "loss": 0.5435,
+      "step": 3028
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.761646458200383e-06,
+      "loss": 0.462,
+      "step": 3032
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7584556477345246e-06,
+      "loss": 0.3768,
+      "step": 3036
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7552648372686665e-06,
+      "loss": 0.4469,
+      "step": 3040
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.752074026802808e-06,
+      "loss": 0.5634,
+      "step": 3044
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.74888321633695e-06,
+      "loss": 0.5286,
+      "step": 3048
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7456924058710913e-06,
+      "loss": 0.4438,
+      "step": 3052
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7425015954052332e-06,
+      "loss": 0.5585,
+      "step": 3056
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7393107849393747e-06,
+      "loss": 0.5876,
+      "step": 3060
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7361199744735166e-06,
+      "loss": 0.6556,
+      "step": 3064
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7329291640076585e-06,
+      "loss": 0.4359,
+      "step": 3068
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7297383535418e-06,
+      "loss": 0.6578,
+      "step": 3072
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7265475430759414e-06,
+      "loss": 0.5182,
+      "step": 3076
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.723356732610083e-06,
+      "loss": 0.492,
+      "step": 3080
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7201659221442252e-06,
+      "loss": 0.3898,
+      "step": 3084
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7169751116783667e-06,
+      "loss": 0.4528,
+      "step": 3088
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.713784301212508e-06,
+      "loss": 0.6559,
+      "step": 3092
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7105934907466496e-06,
+      "loss": 0.578,
+      "step": 3096
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.707402680280792e-06,
+      "loss": 0.2924,
+      "step": 3100
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7042118698149334e-06,
+      "loss": 0.663,
+      "step": 3104
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.701021059349075e-06,
+      "loss": 0.3754,
+      "step": 3108
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6978302488832164e-06,
+      "loss": 0.4799,
+      "step": 3112
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.694639438417358e-06,
+      "loss": 0.4219,
+      "step": 3116
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6914486279515e-06,
+      "loss": 0.4431,
+      "step": 3120
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6882578174856416e-06,
+      "loss": 0.4762,
+      "step": 3124
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.685067007019783e-06,
+      "loss": 0.3265,
+      "step": 3128
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6818761965539246e-06,
+      "loss": 0.3126,
+      "step": 3132
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.678685386088067e-06,
+      "loss": 0.5364,
+      "step": 3136
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6754945756222083e-06,
+      "loss": 0.3967,
+      "step": 3140
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.67230376515635e-06,
+      "loss": 0.3551,
+      "step": 3144
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6691129546904913e-06,
+      "loss": 0.4471,
+      "step": 3148
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.665922144224633e-06,
+      "loss": 0.4281,
+      "step": 3152
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.662731333758775e-06,
+      "loss": 0.5498,
+      "step": 3156
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6595405232929165e-06,
+      "loss": 0.4744,
+      "step": 3160
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6563497128270584e-06,
+      "loss": 0.327,
+      "step": 3164
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6531589023612e-06,
+      "loss": 0.3314,
+      "step": 3168
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.649968091895342e-06,
+      "loss": 0.4735,
+      "step": 3172
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6467772814294833e-06,
+      "loss": 0.3421,
+      "step": 3176
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.643586470963625e-06,
+      "loss": 0.5688,
+      "step": 3180
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6403956604977666e-06,
+      "loss": 0.4909,
+      "step": 3184
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.637204850031908e-06,
+      "loss": 0.5088,
+      "step": 3188
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.63401403956605e-06,
+      "loss": 0.4669,
+      "step": 3192
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.630823229100192e-06,
+      "loss": 0.4551,
+      "step": 3196
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6276324186343334e-06,
+      "loss": 0.4754,
+      "step": 3200
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.624441608168475e-06,
+      "loss": 0.5207,
+      "step": 3204
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6212507977026167e-06,
+      "loss": 0.4255,
+      "step": 3208
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6180599872367586e-06,
+      "loss": 0.5081,
+      "step": 3212
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6148691767709e-06,
+      "loss": 0.4932,
+      "step": 3216
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6116783663050416e-06,
+      "loss": 0.5588,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.608487555839183e-06,
+      "loss": 0.5865,
+      "step": 3224
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6052967453733253e-06,
+      "loss": 0.4913,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.602105934907467e-06,
+      "loss": 0.3849,
+      "step": 3232
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5989151244416083e-06,
+      "loss": 0.5976,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5957243139757497e-06,
+      "loss": 0.4934,
+      "step": 3240
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.592533503509892e-06,
+      "loss": 0.388,
+      "step": 3244
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5893426930440335e-06,
+      "loss": 0.396,
+      "step": 3248
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.586151882578175e-06,
+      "loss": 0.5215,
+      "step": 3252
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5829610721123165e-06,
+      "loss": 0.5468,
+      "step": 3256
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5797702616464584e-06,
+      "loss": 0.4921,
+      "step": 3260
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5765794511806003e-06,
+      "loss": 0.5395,
+      "step": 3264
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5733886407147417e-06,
+      "loss": 0.2728,
+      "step": 3268
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.570197830248883e-06,
+      "loss": 0.4477,
+      "step": 3272
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.567007019783025e-06,
+      "loss": 0.4712,
+      "step": 3276
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.563816209317167e-06,
+      "loss": 0.2961,
+      "step": 3280
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5606253988513085e-06,
+      "loss": 0.5793,
+      "step": 3284
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.55743458838545e-06,
+      "loss": 0.3269,
+      "step": 3288
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.554243777919592e-06,
+      "loss": 0.4204,
+      "step": 3292
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5510529674537333e-06,
+      "loss": 0.6517,
+      "step": 3296
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.547862156987875e-06,
+      "loss": 0.3711,
+      "step": 3300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5446713465220167e-06,
+      "loss": 0.4304,
+      "step": 3304
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5414805360561586e-06,
+      "loss": 0.406,
+      "step": 3308
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5382897255903e-06,
+      "loss": 0.4387,
+      "step": 3312
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.535098915124442e-06,
+      "loss": 0.3445,
+      "step": 3316
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.531908104658584e-06,
+      "loss": 0.5092,
+      "step": 3320
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5287172941927253e-06,
+      "loss": 0.4835,
+      "step": 3324
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5255264837268667e-06,
+      "loss": 0.3742,
+      "step": 3328
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5223356732610082e-06,
+      "loss": 0.397,
+      "step": 3332
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5191448627951505e-06,
+      "loss": 0.509,
+      "step": 3336
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.515954052329292e-06,
+      "loss": 0.3613,
+      "step": 3340
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5127632418634335e-06,
+      "loss": 0.5101,
+      "step": 3344
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.509572431397575e-06,
+      "loss": 0.3241,
+      "step": 3348
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5063816209317173e-06,
+      "loss": 0.3792,
+      "step": 3352
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5031908104658587e-06,
+      "loss": 0.3263,
+      "step": 3356
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5227,
+      "step": 3360
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4968091895341417e-06,
+      "loss": 0.5232,
+      "step": 3364
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4936183790682836e-06,
+      "loss": 0.4621,
+      "step": 3368
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.490427568602425e-06,
+      "loss": 0.3039,
+      "step": 3372
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.487236758136567e-06,
+      "loss": 0.3463,
+      "step": 3376
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4840459476707084e-06,
+      "loss": 0.6041,
+      "step": 3380
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4808551372048503e-06,
+      "loss": 0.4663,
+      "step": 3384
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4776643267389918e-06,
+      "loss": 0.3348,
+      "step": 3388
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4744735162731337e-06,
+      "loss": 0.3892,
+      "step": 3392
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.471282705807275e-06,
+      "loss": 0.6636,
+      "step": 3396
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4680918953414166e-06,
+      "loss": 0.5367,
+      "step": 3400
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4649010848755585e-06,
+      "loss": 0.4974,
+      "step": 3404
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4617102744097e-06,
+      "loss": 0.2935,
+      "step": 3408
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.458519463943842e-06,
+      "loss": 0.4858,
+      "step": 3412
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4553286534779838e-06,
+      "loss": 0.4655,
+      "step": 3416
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4521378430121252e-06,
+      "loss": 0.2403,
+      "step": 3420
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.448947032546267e-06,
+      "loss": 0.3554,
+      "step": 3424
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4457562220804086e-06,
+      "loss": 0.5229,
+      "step": 3428
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4425654116145505e-06,
+      "loss": 0.4178,
+      "step": 3432
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.439374601148692e-06,
+      "loss": 0.4432,
+      "step": 3436
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.436183790682834e-06,
+      "loss": 0.352,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4329929802169753e-06,
+      "loss": 0.6136,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.429802169751117e-06,
+      "loss": 0.4073,
+      "step": 3448
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4266113592852587e-06,
+      "loss": 0.3307,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4234205488194006e-06,
+      "loss": 0.3501,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.420229738353542e-06,
+      "loss": 0.4445,
+      "step": 3460
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.417038927887684e-06,
+      "loss": 0.5194,
+      "step": 3464
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4138481174218254e-06,
+      "loss": 0.5119,
+      "step": 3468
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.410657306955967e-06,
+      "loss": 0.6337,
+      "step": 3472
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4074664964901088e-06,
+      "loss": 0.2655,
+      "step": 3476
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4042756860242502e-06,
+      "loss": 0.5107,
+      "step": 3480
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.401084875558392e-06,
+      "loss": 0.5027,
+      "step": 3484
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3978940650925336e-06,
+      "loss": 0.6517,
+      "step": 3488
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3947032546266755e-06,
+      "loss": 0.3089,
+      "step": 3492
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.391512444160817e-06,
+      "loss": 0.3434,
+      "step": 3496
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.388321633694959e-06,
+      "loss": 0.3684,
+      "step": 3500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3851308232291003e-06,
+      "loss": 0.4419,
+      "step": 3504
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.381940012763242e-06,
+      "loss": 0.4431,
+      "step": 3508
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3787492022973837e-06,
+      "loss": 0.4426,
+      "step": 3512
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.375558391831525e-06,
+      "loss": 0.4228,
+      "step": 3516
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.372367581365667e-06,
+      "loss": 0.2987,
+      "step": 3520
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3691767708998085e-06,
+      "loss": 0.664,
+      "step": 3524
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3659859604339504e-06,
+      "loss": 0.573,
+      "step": 3528
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.362795149968092e-06,
+      "loss": 0.4935,
+      "step": 3532
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3596043395022338e-06,
+      "loss": 0.2314,
+      "step": 3536
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3564135290363753e-06,
+      "loss": 0.4465,
+      "step": 3540
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.353222718570517e-06,
+      "loss": 0.4357,
+      "step": 3544
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3500319081046586e-06,
+      "loss": 0.4599,
+      "step": 3548
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3468410976388005e-06,
+      "loss": 0.5576,
+      "step": 3552
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.343650287172942e-06,
+      "loss": 0.8771,
+      "step": 3556
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.340459476707084e-06,
+      "loss": 0.5068,
+      "step": 3560
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3372686662412258e-06,
+      "loss": 0.5533,
+      "step": 3564
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3340778557753672e-06,
+      "loss": 0.3103,
+      "step": 3568
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.330887045309509e-06,
+      "loss": 0.4112,
+      "step": 3572
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3276962348436506e-06,
+      "loss": 0.3791,
+      "step": 3576
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.324505424377792e-06,
+      "loss": 0.489,
+      "step": 3580
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.321314613911934e-06,
+      "loss": 0.4167,
+      "step": 3584
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3181238034460754e-06,
+      "loss": 0.435,
+      "step": 3588
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3149329929802173e-06,
+      "loss": 0.4934,
+      "step": 3592
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.311742182514359e-06,
+      "loss": 0.3645,
+      "step": 3596
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3085513720485007e-06,
+      "loss": 0.4467,
+      "step": 3600
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.306158264199107e-06,
+      "loss": 0.3215,
+      "step": 3604
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3029674537332483e-06,
+      "loss": 0.3827,
+      "step": 3608
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.29977664326739e-06,
+      "loss": 0.3655,
+      "step": 3612
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2965858328015316e-06,
+      "loss": 0.343,
+      "step": 3616
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2933950223356735e-06,
+      "loss": 0.4173,
+      "step": 3620
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.290204211869815e-06,
+      "loss": 0.3316,
+      "step": 3624
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.287013401403957e-06,
+      "loss": 0.4303,
+      "step": 3628
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2838225909380984e-06,
+      "loss": 0.3626,
+      "step": 3632
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2806317804722402e-06,
+      "loss": 0.4426,
+      "step": 3636
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2774409700063817e-06,
+      "loss": 0.4539,
+      "step": 3640
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.274250159540523e-06,
+      "loss": 0.4288,
+      "step": 3644
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.271059349074665e-06,
+      "loss": 0.5354,
+      "step": 3648
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2678685386088066e-06,
+      "loss": 0.4777,
+      "step": 3652
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2646777281429484e-06,
+      "loss": 0.6505,
+      "step": 3656
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.26148691767709e-06,
+      "loss": 0.3299,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.258296107211232e-06,
+      "loss": 0.3947,
+      "step": 3664
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2551052967453733e-06,
+      "loss": 0.4154,
+      "step": 3668
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.251914486279515e-06,
+      "loss": 0.413,
+      "step": 3672
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.248723675813657e-06,
+      "loss": 0.5292,
+      "step": 3676
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2455328653477985e-06,
+      "loss": 0.2279,
+      "step": 3680
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2423420548819404e-06,
+      "loss": 0.4984,
+      "step": 3684
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.239151244416082e-06,
+      "loss": 0.3523,
+      "step": 3688
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.235960433950224e-06,
+      "loss": 0.2664,
+      "step": 3692
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2327696234843653e-06,
+      "loss": 0.2055,
+      "step": 3696
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.229578813018507e-06,
+      "loss": 0.359,
+      "step": 3700
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2263880025526486e-06,
+      "loss": 0.4491,
+      "step": 3704
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2231971920867905e-06,
+      "loss": 0.4488,
+      "step": 3708
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.220006381620932e-06,
+      "loss": 0.4134,
+      "step": 3712
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2168155711550735e-06,
+      "loss": 0.7982,
+      "step": 3716
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2136247606892154e-06,
+      "loss": 0.3617,
+      "step": 3720
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.210433950223357e-06,
+      "loss": 0.4095,
+      "step": 3724
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2072431397574987e-06,
+      "loss": 0.4942,
+      "step": 3728
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.20405232929164e-06,
+      "loss": 0.3022,
+      "step": 3732
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.200861518825782e-06,
+      "loss": 0.4971,
+      "step": 3736
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.1976707083599236e-06,
+      "loss": 0.4171,
+      "step": 3740
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1944798978940654e-06,
+      "loss": 0.4044,
+      "step": 3744
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.191289087428207e-06,
+      "loss": 0.4036,
+      "step": 3748
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1880982769623484e-06,
+      "loss": 0.2978,
+      "step": 3752
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1849074664964903e-06,
+      "loss": 0.4452,
+      "step": 3756
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1817166560306318e-06,
+      "loss": 0.4334,
+      "step": 3760
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1785258455647736e-06,
+      "loss": 0.3877,
+      "step": 3764
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.175335035098915e-06,
+      "loss": 0.4025,
+      "step": 3768
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.172144224633057e-06,
+      "loss": 0.4121,
+      "step": 3772
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1689534141671985e-06,
+      "loss": 0.4433,
+      "step": 3776
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1657626037013404e-06,
+      "loss": 0.5417,
+      "step": 3780
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.162571793235482e-06,
+      "loss": 0.2991,
+      "step": 3784
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1593809827696237e-06,
+      "loss": 0.5368,
+      "step": 3788
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.156190172303765e-06,
+      "loss": 0.6703,
+      "step": 3792
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.152999361837907e-06,
+      "loss": 0.6356,
+      "step": 3796
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1498085513720486e-06,
+      "loss": 0.2694,
+      "step": 3800
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1466177409061905e-06,
+      "loss": 0.2441,
+      "step": 3804
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.143426930440332e-06,
+      "loss": 0.3316,
+      "step": 3808
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.140236119974474e-06,
+      "loss": 0.3115,
+      "step": 3812
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1370453095086153e-06,
+      "loss": 0.2622,
+      "step": 3816
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.133854499042757e-06,
+      "loss": 0.3383,
+      "step": 3820
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1306636885768987e-06,
+      "loss": 0.3967,
+      "step": 3824
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1274728781110406e-06,
+      "loss": 0.7028,
+      "step": 3828
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.124282067645182e-06,
+      "loss": 0.573,
+      "step": 3832
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.121091257179324e-06,
+      "loss": 0.3881,
+      "step": 3836
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1179004467134654e-06,
+      "loss": 0.4497,
+      "step": 3840
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1147096362476073e-06,
+      "loss": 0.5644,
+      "step": 3844
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1115188257817488e-06,
+      "loss": 0.278,
+      "step": 3848
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1083280153158906e-06,
+      "loss": 0.3726,
+      "step": 3852
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.105137204850032e-06,
+      "loss": 0.4098,
+      "step": 3856
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1019463943841736e-06,
+      "loss": 0.2577,
+      "step": 3860
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.0987555839183155e-06,
+      "loss": 0.5709,
+      "step": 3864
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.095564773452457e-06,
+      "loss": 0.3037,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.092373962986599e-06,
+      "loss": 0.4883,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0891831525207403e-06,
+      "loss": 0.4228,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.085992342054882e-06,
+      "loss": 0.5301,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0828015315890237e-06,
+      "loss": 0.3912,
+      "step": 3884
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0796107211231656e-06,
+      "loss": 0.5301,
+      "step": 3888
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.076419910657307e-06,
+      "loss": 0.4336,
+      "step": 3892
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0732291001914485e-06,
+      "loss": 0.5384,
+      "step": 3896
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0700382897255904e-06,
+      "loss": 0.4925,
+      "step": 3900
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.066847479259732e-06,
+      "loss": 0.392,
+      "step": 3904
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0636566687938738e-06,
+      "loss": 0.3385,
+      "step": 3908
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0604658583280152e-06,
+      "loss": 0.2874,
+      "step": 3912
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.057275047862157e-06,
+      "loss": 0.4633,
+      "step": 3916
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0540842373962986e-06,
+      "loss": 0.3791,
+      "step": 3920
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0508934269304405e-06,
+      "loss": 0.3847,
+      "step": 3924
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0477026164645824e-06,
+      "loss": 0.5012,
+      "step": 3928
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.044511805998724e-06,
+      "loss": 0.3865,
+      "step": 3932
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0413209955328658e-06,
+      "loss": 0.4644,
+      "step": 3936
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0381301850670072e-06,
+      "loss": 0.4693,
+      "step": 3940
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.034939374601149e-06,
+      "loss": 0.3533,
+      "step": 3944
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0317485641352906e-06,
+      "loss": 0.3577,
+      "step": 3948
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0285577536694325e-06,
+      "loss": 0.3633,
+      "step": 3952
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.025366943203574e-06,
+      "loss": 0.3995,
+      "step": 3956
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.022176132737716e-06,
+      "loss": 0.5508,
+      "step": 3960
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0189853222718573e-06,
+      "loss": 0.3738,
+      "step": 3964
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0157945118059988e-06,
+      "loss": 0.4689,
+      "step": 3968
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0126037013401407e-06,
+      "loss": 0.3641,
+      "step": 3972
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.009412890874282e-06,
+      "loss": 0.2933,
+      "step": 3976
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.006222080408424e-06,
+      "loss": 0.3805,
+      "step": 3980
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.0030312699425655e-06,
+      "loss": 0.4096,
+      "step": 3984
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9998404594767074e-06,
+      "loss": 0.395,
+      "step": 3988
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.996649649010849e-06,
+      "loss": 0.3805,
+      "step": 3992
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9934588385449908e-06,
+      "loss": 0.4663,
+      "step": 3996
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9902680280791322e-06,
+      "loss": 0.3886,
+      "step": 4000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9870772176132737e-06,
+      "loss": 0.3044,
+      "step": 4004
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9838864071474156e-06,
+      "loss": 0.4757,
+      "step": 4008
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.980695596681557e-06,
+      "loss": 0.2762,
+      "step": 4012
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.977504786215699e-06,
+      "loss": 0.431,
+      "step": 4016
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9743139757498404e-06,
+      "loss": 0.3777,
+      "step": 4020
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9711231652839823e-06,
+      "loss": 0.4599,
+      "step": 4024
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.967932354818124e-06,
+      "loss": 0.549,
+      "step": 4028
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9647415443522657e-06,
+      "loss": 0.3346,
+      "step": 4032
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.961550733886407e-06,
+      "loss": 0.2425,
+      "step": 4036
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.958359923420549e-06,
+      "loss": 0.3963,
+      "step": 4040
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9551691129546905e-06,
+      "loss": 0.2525,
+      "step": 4044
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9519783024888324e-06,
+      "loss": 0.3303,
+      "step": 4048
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.948787492022974e-06,
+      "loss": 0.4514,
+      "step": 4052
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9455966815571158e-06,
+      "loss": 0.3102,
+      "step": 4056
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9424058710912573e-06,
+      "loss": 0.3588,
+      "step": 4060
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.939215060625399e-06,
+      "loss": 0.375,
+      "step": 4064
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9360242501595406e-06,
+      "loss": 0.3988,
+      "step": 4068
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9328334396936825e-06,
+      "loss": 0.385,
+      "step": 4072
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.929642629227824e-06,
+      "loss": 0.3828,
+      "step": 4076
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.926451818761966e-06,
+      "loss": 0.4949,
+      "step": 4080
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9232610082961073e-06,
+      "loss": 0.3858,
+      "step": 4084
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9200701978302492e-06,
+      "loss": 0.477,
+      "step": 4088
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9168793873643907e-06,
+      "loss": 0.1664,
+      "step": 4092
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9136885768985326e-06,
+      "loss": 0.3089,
+      "step": 4096
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.910497766432674e-06,
+      "loss": 0.2623,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.907306955966816e-06,
+      "loss": 0.3961,
+      "step": 4104
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.9041161455009574e-06,
+      "loss": 0.2924,
+      "step": 4108
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.900925335035099e-06,
+      "loss": 0.2636,
+      "step": 4112
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8977345245692408e-06,
+      "loss": 0.492,
+      "step": 4116
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8945437141033823e-06,
+      "loss": 0.5656,
+      "step": 4120
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8913529036375242e-06,
+      "loss": 0.3425,
+      "step": 4124
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8881620931716656e-06,
+      "loss": 0.5472,
+      "step": 4128
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8849712827058075e-06,
+      "loss": 0.3806,
+      "step": 4132
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.881780472239949e-06,
+      "loss": 0.2649,
+      "step": 4136
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8785896617740909e-06,
+      "loss": 0.3659,
+      "step": 4140
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8753988513082324e-06,
+      "loss": 0.42,
+      "step": 4144
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.872208040842374e-06,
+      "loss": 0.3603,
+      "step": 4148
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.869017230376516e-06,
+      "loss": 0.5069,
+      "step": 4152
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8658264199106574e-06,
+      "loss": 0.384,
+      "step": 4156
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8626356094447993e-06,
+      "loss": 0.4073,
+      "step": 4160
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8594447989789408e-06,
+      "loss": 0.1713,
+      "step": 4164
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8562539885130827e-06,
+      "loss": 0.3695,
+      "step": 4168
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8530631780472241e-06,
+      "loss": 0.4161,
+      "step": 4172
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8498723675813656e-06,
+      "loss": 0.2599,
+      "step": 4176
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8466815571155075e-06,
+      "loss": 0.4248,
+      "step": 4180
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.843490746649649e-06,
+      "loss": 0.1923,
+      "step": 4184
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8402999361837909e-06,
+      "loss": 0.3949,
+      "step": 4188
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8371091257179323e-06,
+      "loss": 0.3583,
+      "step": 4192
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8339183152520742e-06,
+      "loss": 0.4495,
+      "step": 4196
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8307275047862157e-06,
+      "loss": 0.2304,
+      "step": 4200
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8275366943203576e-06,
+      "loss": 0.3311,
+      "step": 4204
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8243458838544993e-06,
+      "loss": 0.4612,
+      "step": 4208
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8211550733886407e-06,
+      "loss": 0.3376,
+      "step": 4212
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8179642629227826e-06,
+      "loss": 0.5294,
+      "step": 4216
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8147734524569241e-06,
+      "loss": 0.5372,
+      "step": 4220
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.811582641991066e-06,
+      "loss": 0.429,
+      "step": 4224
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8083918315252075e-06,
+      "loss": 0.3858,
+      "step": 4228
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8052010210593494e-06,
+      "loss": 0.3116,
+      "step": 4232
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8020102105934908e-06,
+      "loss": 0.4358,
+      "step": 4236
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7988194001276327e-06,
+      "loss": 0.3932,
+      "step": 4240
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7956285896617742e-06,
+      "loss": 0.4402,
+      "step": 4244
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7924377791959157e-06,
+      "loss": 0.4396,
+      "step": 4248
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7892469687300576e-06,
+      "loss": 0.3562,
+      "step": 4252
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7860561582641992e-06,
+      "loss": 0.4307,
+      "step": 4256
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.782865347798341e-06,
+      "loss": 0.268,
+      "step": 4260
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7796745373324826e-06,
+      "loss": 0.4127,
+      "step": 4264
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7764837268666243e-06,
+      "loss": 0.2596,
+      "step": 4268
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.773292916400766e-06,
+      "loss": 0.1921,
+      "step": 4272
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7701021059349077e-06,
+      "loss": 0.2519,
+      "step": 4276
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7669112954690493e-06,
+      "loss": 0.702,
+      "step": 4280
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7637204850031908e-06,
+      "loss": 0.5359,
+      "step": 4284
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7605296745373327e-06,
+      "loss": 0.4276,
+      "step": 4288
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7573388640714742e-06,
+      "loss": 0.4729,
+      "step": 4292
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.754148053605616e-06,
+      "loss": 0.2199,
+      "step": 4296
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7509572431397575e-06,
+      "loss": 0.3834,
+      "step": 4300
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 6468,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 29035325030400.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}