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": 2.4801,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0,
+      "loss": 2.4284,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 2.2651,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 2.411,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0,
+      "loss": 2.8299,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0,
+      "loss": 2.2188,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3082402064781276e-06,
+      "loss": 1.345,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9623603097171917e-06,
+      "loss": 0.5695,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.262883767531511e-06,
+      "loss": 0.8812,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5555756797431724e-06,
+      "loss": 0.8725,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7786547836457785e-06,
+      "loss": 0.4851,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9589528137043157e-06,
+      "loss": 0.8354,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.110267503805303e-06,
+      "loss": 0.4873,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.2406394020168525e-06,
+      "loss": 0.6584,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3551671365864186e-06,
+      "loss": 0.529,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.4572878450621517e-06,
+      "loss": 0.5812,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.5494288615482305e-06,
+      "loss": 0.5163,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.6333682331099297e-06,
+      "loss": 0.6595,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.710447450306277e-06,
+      "loss": 0.4775,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.7817036126729157e-06,
+      "loss": 0.682,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.84795507876713e-06,
+      "loss": 0.7048,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.909858960648549e-06,
+      "loss": 0.9478,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.9679508875196075e-06,
+      "loss": 0.5932,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.022673220704539e-06,
+      "loss": 0.5733,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.074395524884577e-06,
+      "loss": 0.5817,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.123429713794031e-06,
+      "loss": 0.6372,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.170041450985754e-06,
+      "loss": 0.5108,
+      "step": 108
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.214458864668026e-06,
+      "loss": 0.6262,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.256879301905398e-06,
+      "loss": 0.8594,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.297474628787183e-06,
+      "loss": 0.7241,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.336395436735046e-06,
+      "loss": 0.6471,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.373774415149143e-06,
+      "loss": 0.6144,
+      "step": 128
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.409729081127459e-06,
+      "loss": 0.672,
+      "step": 132
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.444364007946065e-06,
+      "loss": 0.6802,
+      "step": 136
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.4777726588457195e-06,
+      "loss": 0.514,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.510038907149524e-06,
+      "loss": 0.715,
+      "step": 144
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.541238304971202e-06,
+      "loss": 0.5189,
+      "step": 148
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.5714391488166745e-06,
+      "loss": 0.7188,
+      "step": 152
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.600703379889684e-06,
+      "loss": 0.5829,
+      "step": 156
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.629087348946707e-06,
+      "loss": 0.5551,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.656642469442713e-06,
+      "loss": 0.7016,
+      "step": 164
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.683415777991895e-06,
+      "loss": 0.5357,
+      "step": 168
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.709450417491796e-06,
+      "loss": 0.6232,
+      "step": 172
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.734786055373451e-06,
+      "loss": 0.7218,
+      "step": 176
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.759459247158257e-06,
+      "loss": 0.51,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.783503753685794e-06,
+      "loss": 0.8871,
+      "step": 184
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.806950818921448e-06,
+      "loss": 0.8177,
+      "step": 188
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8298294140798465e-06,
+      "loss": 0.5602,
+      "step": 192
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.852166452849314e-06,
+      "loss": 0.7395,
+      "step": 196
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8739869817278244e-06,
+      "loss": 0.7008,
+      "step": 200
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.89531434884623e-06,
+      "loss": 0.7096,
+      "step": 204
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.916170354132174e-06,
+      "loss": 0.7617,
+      "step": 208
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.936575383236021e-06,
+      "loss": 0.637,
+      "step": 212
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.956548527281403e-06,
+      "loss": 0.7149,
+      "step": 216
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.976107690203556e-06,
+      "loss": 0.585,
+      "step": 220
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.995269685187989e-06,
+      "loss": 0.4153,
+      "step": 224
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.998404594767071e-06,
+      "loss": 0.8215,
+      "step": 228
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.995213784301213e-06,
+      "loss": 0.6051,
+      "step": 232
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.992022973835355e-06,
+      "loss": 0.6785,
+      "step": 236
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.988832163369496e-06,
+      "loss": 0.6774,
+      "step": 240
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.985641352903638e-06,
+      "loss": 0.5039,
+      "step": 244
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.982450542437779e-06,
+      "loss": 0.5724,
+      "step": 248
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.979259731971921e-06,
+      "loss": 0.8469,
+      "step": 252
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.976068921506063e-06,
+      "loss": 0.9735,
+      "step": 256
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.972878111040205e-06,
+      "loss": 0.8594,
+      "step": 260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.969687300574346e-06,
+      "loss": 0.6523,
+      "step": 264
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9664964901084875e-06,
+      "loss": 0.6641,
+      "step": 268
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.96330567964263e-06,
+      "loss": 0.6586,
+      "step": 272
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.960114869176771e-06,
+      "loss": 0.6019,
+      "step": 276
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.956924058710913e-06,
+      "loss": 0.7117,
+      "step": 280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.953733248245054e-06,
+      "loss": 0.4632,
+      "step": 284
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.950542437779196e-06,
+      "loss": 0.733,
+      "step": 288
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947351627313338e-06,
+      "loss": 0.7205,
+      "step": 292
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9441608168474795e-06,
+      "loss": 0.6121,
+      "step": 296
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.940970006381621e-06,
+      "loss": 0.7859,
+      "step": 300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9377791959157625e-06,
+      "loss": 0.8041,
+      "step": 304
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.934588385449905e-06,
+      "loss": 0.7834,
+      "step": 308
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.931397574984046e-06,
+      "loss": 0.6704,
+      "step": 312
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.928206764518188e-06,
+      "loss": 0.8402,
+      "step": 316
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.925015954052329e-06,
+      "loss": 0.7851,
+      "step": 320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9218251435864715e-06,
+      "loss": 0.501,
+      "step": 324
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.918634333120613e-06,
+      "loss": 0.6039,
+      "step": 328
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9154435226547544e-06,
+      "loss": 0.64,
+      "step": 332
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.912252712188896e-06,
+      "loss": 0.5726,
+      "step": 336
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.909061901723038e-06,
+      "loss": 0.6605,
+      "step": 340
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.90587109125718e-06,
+      "loss": 0.8105,
+      "step": 344
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.902680280791321e-06,
+      "loss": 0.8422,
+      "step": 348
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.8994894703254635e-06,
+      "loss": 0.5242,
+      "step": 352
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.896298659859605e-06,
+      "loss": 0.6062,
+      "step": 356
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.8931078493937464e-06,
+      "loss": 0.7289,
+      "step": 360
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.889917038927888e-06,
+      "loss": 0.6916,
+      "step": 364
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.88672622846203e-06,
+      "loss": 0.8526,
+      "step": 368
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.883535417996172e-06,
+      "loss": 1.0668,
+      "step": 372
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.880344607530313e-06,
+      "loss": 0.6912,
+      "step": 376
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.877153797064455e-06,
+      "loss": 0.7383,
+      "step": 380
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.873962986598597e-06,
+      "loss": 0.77,
+      "step": 384
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.870772176132738e-06,
+      "loss": 0.8328,
+      "step": 388
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.86758136566688e-06,
+      "loss": 0.7135,
+      "step": 392
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.864390555201021e-06,
+      "loss": 0.7976,
+      "step": 396
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.861199744735164e-06,
+      "loss": 0.5799,
+      "step": 400
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.858008934269305e-06,
+      "loss": 0.5246,
+      "step": 404
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.854818123803447e-06,
+      "loss": 0.5895,
+      "step": 408
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.851627313337588e-06,
+      "loss": 0.7751,
+      "step": 412
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.84843650287173e-06,
+      "loss": 0.7469,
+      "step": 416
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.845245692405872e-06,
+      "loss": 0.5013,
+      "step": 420
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.842054881940013e-06,
+      "loss": 0.5398,
+      "step": 424
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.838864071474155e-06,
+      "loss": 0.4547,
+      "step": 428
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.835673261008296e-06,
+      "loss": 0.8732,
+      "step": 432
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.832482450542439e-06,
+      "loss": 0.7671,
+      "step": 436
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.82929164007658e-06,
+      "loss": 0.6574,
+      "step": 440
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8261008296107215e-06,
+      "loss": 0.7173,
+      "step": 444
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.822910019144863e-06,
+      "loss": 0.4371,
+      "step": 448
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.819719208679005e-06,
+      "loss": 0.6992,
+      "step": 452
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.816528398213147e-06,
+      "loss": 0.6827,
+      "step": 456
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.813337587747288e-06,
+      "loss": 0.4919,
+      "step": 460
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.81014677728143e-06,
+      "loss": 0.9571,
+      "step": 464
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.806955966815571e-06,
+      "loss": 0.5202,
+      "step": 468
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8037651563497135e-06,
+      "loss": 0.7919,
+      "step": 472
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.800574345883855e-06,
+      "loss": 0.5517,
+      "step": 476
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.7973835354179965e-06,
+      "loss": 0.3889,
+      "step": 480
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.794192724952138e-06,
+      "loss": 0.5933,
+      "step": 484
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.79100191448628e-06,
+      "loss": 0.9298,
+      "step": 488
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.787811104020422e-06,
+      "loss": 0.4758,
+      "step": 492
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.784620293554563e-06,
+      "loss": 0.5162,
+      "step": 496
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.781429483088705e-06,
+      "loss": 0.6675,
+      "step": 500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.778238672622846e-06,
+      "loss": 0.8493,
+      "step": 504
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.7750478621569885e-06,
+      "loss": 0.6583,
+      "step": 508
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.77185705169113e-06,
+      "loss": 0.4897,
+      "step": 512
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.768666241225271e-06,
+      "loss": 0.6633,
+      "step": 516
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.765475430759413e-06,
+      "loss": 0.782,
+      "step": 520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.762284620293555e-06,
+      "loss": 0.815,
+      "step": 524
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.759093809827697e-06,
+      "loss": 0.4498,
+      "step": 528
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.755902999361838e-06,
+      "loss": 0.6006,
+      "step": 532
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.75271218889598e-06,
+      "loss": 0.9473,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.749521378430121e-06,
+      "loss": 0.4036,
+      "step": 540
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.746330567964263e-06,
+      "loss": 0.555,
+      "step": 544
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.743139757498405e-06,
+      "loss": 0.7843,
+      "step": 548
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.739948947032546e-06,
+      "loss": 0.8376,
+      "step": 552
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.736758136566688e-06,
+      "loss": 0.5423,
+      "step": 556
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.73356732610083e-06,
+      "loss": 0.5533,
+      "step": 560
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7303765156349716e-06,
+      "loss": 0.5212,
+      "step": 564
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.727185705169113e-06,
+      "loss": 0.8054,
+      "step": 568
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7239948947032545e-06,
+      "loss": 0.438,
+      "step": 572
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.720804084237397e-06,
+      "loss": 0.6025,
+      "step": 576
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.717613273771538e-06,
+      "loss": 0.8118,
+      "step": 580
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.71442246330568e-06,
+      "loss": 0.6911,
+      "step": 584
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.711231652839821e-06,
+      "loss": 0.7022,
+      "step": 588
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7080408423739636e-06,
+      "loss": 0.5918,
+      "step": 592
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.704850031908105e-06,
+      "loss": 0.6012,
+      "step": 596
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7016592214422465e-06,
+      "loss": 0.8031,
+      "step": 600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.698468410976389e-06,
+      "loss": 0.7864,
+      "step": 604
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.69527760051053e-06,
+      "loss": 0.6361,
+      "step": 608
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.692086790044672e-06,
+      "loss": 0.6619,
+      "step": 612
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.688895979578813e-06,
+      "loss": 0.5132,
+      "step": 616
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6857051691129555e-06,
+      "loss": 0.6111,
+      "step": 620
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.682514358647097e-06,
+      "loss": 0.7884,
+      "step": 624
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.6793235481812385e-06,
+      "loss": 0.4355,
+      "step": 628
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.67613273771538e-06,
+      "loss": 0.7325,
+      "step": 632
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.672941927249522e-06,
+      "loss": 0.5633,
+      "step": 636
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.669751116783664e-06,
+      "loss": 0.6415,
+      "step": 640
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.666560306317805e-06,
+      "loss": 0.6508,
+      "step": 644
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.663369495851947e-06,
+      "loss": 0.5909,
+      "step": 648
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.660178685386089e-06,
+      "loss": 0.5651,
+      "step": 652
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.6569878749202305e-06,
+      "loss": 0.6729,
+      "step": 656
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.653797064454372e-06,
+      "loss": 0.842,
+      "step": 660
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.650606253988513e-06,
+      "loss": 0.5844,
+      "step": 664
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.647415443522656e-06,
+      "loss": 0.7394,
+      "step": 668
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.644224633056797e-06,
+      "loss": 0.6725,
+      "step": 672
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.641033822590939e-06,
+      "loss": 0.6416,
+      "step": 676
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.63784301212508e-06,
+      "loss": 0.7926,
+      "step": 680
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.634652201659222e-06,
+      "loss": 0.5941,
+      "step": 684
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.631461391193364e-06,
+      "loss": 0.9582,
+      "step": 688
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.628270580727505e-06,
+      "loss": 0.4289,
+      "step": 692
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.625079770261647e-06,
+      "loss": 0.6518,
+      "step": 696
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.621888959795788e-06,
+      "loss": 0.8722,
+      "step": 700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.618698149329931e-06,
+      "loss": 0.5419,
+      "step": 704
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.615507338864072e-06,
+      "loss": 0.6891,
+      "step": 708
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.612316528398214e-06,
+      "loss": 0.5157,
+      "step": 712
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.609125717932355e-06,
+      "loss": 0.7015,
+      "step": 716
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6059349074664965e-06,
+      "loss": 0.546,
+      "step": 720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.602744097000639e-06,
+      "loss": 0.6735,
+      "step": 724
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.59955328653478e-06,
+      "loss": 0.5564,
+      "step": 728
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.596362476068922e-06,
+      "loss": 0.5182,
+      "step": 732
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.593171665603063e-06,
+      "loss": 0.4053,
+      "step": 736
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.5899808551372056e-06,
+      "loss": 0.4039,
+      "step": 740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.586790044671347e-06,
+      "loss": 0.6502,
+      "step": 744
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5835992342054885e-06,
+      "loss": 0.8062,
+      "step": 748
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.58040842373963e-06,
+      "loss": 0.4143,
+      "step": 752
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.5772176132737715e-06,
+      "loss": 0.5539,
+      "step": 756
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.574026802807914e-06,
+      "loss": 0.5926,
+      "step": 760
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.570835992342055e-06,
+      "loss": 0.751,
+      "step": 764
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.567645181876197e-06,
+      "loss": 0.5886,
+      "step": 768
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.564454371410338e-06,
+      "loss": 0.677,
+      "step": 772
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5612635609444805e-06,
+      "loss": 0.7097,
+      "step": 776
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.558072750478622e-06,
+      "loss": 0.56,
+      "step": 780
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.5548819400127634e-06,
+      "loss": 0.4481,
+      "step": 784
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.551691129546905e-06,
+      "loss": 0.4959,
+      "step": 788
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.548500319081046e-06,
+      "loss": 0.8399,
+      "step": 792
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.545309508615189e-06,
+      "loss": 0.6904,
+      "step": 796
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.54211869814933e-06,
+      "loss": 0.8689,
+      "step": 800
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.538927887683472e-06,
+      "loss": 0.6232,
+      "step": 804
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.535737077217613e-06,
+      "loss": 0.6428,
+      "step": 808
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.5325462667517554e-06,
+      "loss": 0.7462,
+      "step": 812
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.529355456285897e-06,
+      "loss": 0.529,
+      "step": 816
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.526164645820038e-06,
+      "loss": 0.4875,
+      "step": 820
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.52297383535418e-06,
+      "loss": 0.6747,
+      "step": 824
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.519783024888322e-06,
+      "loss": 0.7061,
+      "step": 828
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.516592214422464e-06,
+      "loss": 0.7865,
+      "step": 832
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.513401403956605e-06,
+      "loss": 0.5122,
+      "step": 836
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5102105934907466e-06,
+      "loss": 0.4014,
+      "step": 840
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.507019783024889e-06,
+      "loss": 0.7509,
+      "step": 844
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.50382897255903e-06,
+      "loss": 0.8073,
+      "step": 848
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.500638162093172e-06,
+      "loss": 0.3459,
+      "step": 852
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.497447351627314e-06,
+      "loss": 0.6814,
+      "step": 856
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.494256541161456e-06,
+      "loss": 1.1027,
+      "step": 860
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.491065730695597e-06,
+      "loss": 0.5254,
+      "step": 864
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.4878749202297385e-06,
+      "loss": 0.7436,
+      "step": 868
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.484684109763881e-06,
+      "loss": 0.4877,
+      "step": 872
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.481493299298022e-06,
+      "loss": 0.657,
+      "step": 876
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.478302488832164e-06,
+      "loss": 0.7193,
+      "step": 880
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.475111678366305e-06,
+      "loss": 0.5461,
+      "step": 884
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.471920867900448e-06,
+      "loss": 0.5707,
+      "step": 888
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.468730057434589e-06,
+      "loss": 0.9755,
+      "step": 892
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4655392469687305e-06,
+      "loss": 0.551,
+      "step": 896
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.462348436502872e-06,
+      "loss": 0.499,
+      "step": 900
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.459157626037014e-06,
+      "loss": 0.4268,
+      "step": 904
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.455966815571156e-06,
+      "loss": 0.6658,
+      "step": 908
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.452776005105297e-06,
+      "loss": 0.5642,
+      "step": 912
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.449585194639439e-06,
+      "loss": 0.6943,
+      "step": 916
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.446394384173581e-06,
+      "loss": 0.5404,
+      "step": 920
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4432035737077225e-06,
+      "loss": 0.7934,
+      "step": 924
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.440012763241864e-06,
+      "loss": 0.7138,
+      "step": 928
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4368219527760055e-06,
+      "loss": 0.5249,
+      "step": 932
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.433631142310147e-06,
+      "loss": 0.9614,
+      "step": 936
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.430440331844289e-06,
+      "loss": 0.5915,
+      "step": 940
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.427249521378431e-06,
+      "loss": 0.6766,
+      "step": 944
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.424058710912572e-06,
+      "loss": 0.6641,
+      "step": 948
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.420867900446714e-06,
+      "loss": 0.4849,
+      "step": 952
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.417677089980856e-06,
+      "loss": 0.7182,
+      "step": 956
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4144862795149974e-06,
+      "loss": 0.6782,
+      "step": 960
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.411295469049139e-06,
+      "loss": 0.4837,
+      "step": 964
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.40810465858328e-06,
+      "loss": 0.7323,
+      "step": 968
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.404913848117422e-06,
+      "loss": 0.5807,
+      "step": 972
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.401723037651564e-06,
+      "loss": 0.373,
+      "step": 976
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.398532227185706e-06,
+      "loss": 0.5072,
+      "step": 980
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.395341416719847e-06,
+      "loss": 0.5952,
+      "step": 984
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.392150606253989e-06,
+      "loss": 0.549,
+      "step": 988
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.388959795788131e-06,
+      "loss": 0.5918,
+      "step": 992
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.385768985322272e-06,
+      "loss": 0.4411,
+      "step": 996
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.382578174856414e-06,
+      "loss": 0.7001,
+      "step": 1000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.379387364390555e-06,
+      "loss": 0.744,
+      "step": 1004
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.376196553924697e-06,
+      "loss": 0.4091,
+      "step": 1008
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.373005743458839e-06,
+      "loss": 0.7464,
+      "step": 1012
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3698149329929806e-06,
+      "loss": 0.6164,
+      "step": 1016
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.366624122527122e-06,
+      "loss": 0.6213,
+      "step": 1020
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3634333120612635e-06,
+      "loss": 0.6991,
+      "step": 1024
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.360242501595406e-06,
+      "loss": 0.5268,
+      "step": 1028
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.357051691129547e-06,
+      "loss": 0.7768,
+      "step": 1032
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.353860880663689e-06,
+      "loss": 0.9204,
+      "step": 1036
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.35067007019783e-06,
+      "loss": 0.5844,
+      "step": 1040
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.347479259731972e-06,
+      "loss": 0.5198,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.344288449266114e-06,
+      "loss": 0.3069,
+      "step": 1048
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3410976388002555e-06,
+      "loss": 0.5465,
+      "step": 1052
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.337906828334397e-06,
+      "loss": 0.4729,
+      "step": 1056
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3347160178685384e-06,
+      "loss": 0.6514,
+      "step": 1060
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.331525207402681e-06,
+      "loss": 0.8142,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.328334396936822e-06,
+      "loss": 0.6477,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.325143586470964e-06,
+      "loss": 0.4601,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.321952776005105e-06,
+      "loss": 0.6687,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.3187619655392475e-06,
+      "loss": 0.4565,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.315571155073389e-06,
+      "loss": 0.646,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.31238034460753e-06,
+      "loss": 0.6145,
+      "step": 1088
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.309189534141672e-06,
+      "loss": 0.3854,
+      "step": 1092
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.305998723675814e-06,
+      "loss": 0.6016,
+      "step": 1096
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.302807913209956e-06,
+      "loss": 0.5223,
+      "step": 1100
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.299617102744097e-06,
+      "loss": 0.6356,
+      "step": 1104
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.2964262922782395e-06,
+      "loss": 0.4599,
+      "step": 1108
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.293235481812381e-06,
+      "loss": 0.6452,
+      "step": 1112
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.290044671346522e-06,
+      "loss": 0.386,
+      "step": 1116
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.286853860880664e-06,
+      "loss": 0.6384,
+      "step": 1120
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.283663050414806e-06,
+      "loss": 0.7654,
+      "step": 1124
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.280472239948948e-06,
+      "loss": 0.6019,
+      "step": 1128
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.277281429483089e-06,
+      "loss": 0.6078,
+      "step": 1132
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.274090619017231e-06,
+      "loss": 0.5181,
+      "step": 1136
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.270899808551373e-06,
+      "loss": 0.6731,
+      "step": 1140
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.267708998085514e-06,
+      "loss": 0.4956,
+      "step": 1144
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.264518187619656e-06,
+      "loss": 0.6115,
+      "step": 1148
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.261327377153797e-06,
+      "loss": 0.7712,
+      "step": 1152
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.25813656668794e-06,
+      "loss": 0.5086,
+      "step": 1156
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.254945756222081e-06,
+      "loss": 0.7241,
+      "step": 1160
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.251754945756223e-06,
+      "loss": 0.5275,
+      "step": 1164
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.248564135290364e-06,
+      "loss": 0.7552,
+      "step": 1168
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.245373324824506e-06,
+      "loss": 0.4292,
+      "step": 1172
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.242182514358648e-06,
+      "loss": 0.7575,
+      "step": 1176
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.238991703892789e-06,
+      "loss": 0.5653,
+      "step": 1180
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.235800893426931e-06,
+      "loss": 0.6882,
+      "step": 1184
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.232610082961072e-06,
+      "loss": 0.6488,
+      "step": 1188
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.2294192724952146e-06,
+      "loss": 0.5522,
+      "step": 1192
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.226228462029356e-06,
+      "loss": 0.578,
+      "step": 1196
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.2230376515634975e-06,
+      "loss": 0.7412,
+      "step": 1200
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.219846841097639e-06,
+      "loss": 0.5138,
+      "step": 1204
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.216656030631781e-06,
+      "loss": 0.6943,
+      "step": 1208
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.213465220165923e-06,
+      "loss": 0.4599,
+      "step": 1212
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.210274409700064e-06,
+      "loss": 0.8815,
+      "step": 1216
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.207083599234206e-06,
+      "loss": 0.6245,
+      "step": 1220
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.203892788768347e-06,
+      "loss": 0.5513,
+      "step": 1224
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2007019783024895e-06,
+      "loss": 0.4635,
+      "step": 1228
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.197511167836631e-06,
+      "loss": 0.5711,
+      "step": 1232
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1943203573707724e-06,
+      "loss": 0.5078,
+      "step": 1236
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.191129546904914e-06,
+      "loss": 0.4304,
+      "step": 1240
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.187938736439056e-06,
+      "loss": 0.715,
+      "step": 1244
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.184747925973198e-06,
+      "loss": 0.6305,
+      "step": 1248
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.181557115507339e-06,
+      "loss": 0.6243,
+      "step": 1252
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.178366305041481e-06,
+      "loss": 0.6439,
+      "step": 1256
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.175175494575622e-06,
+      "loss": 0.4782,
+      "step": 1260
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.171984684109764e-06,
+      "loss": 0.4523,
+      "step": 1264
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.168793873643906e-06,
+      "loss": 0.4884,
+      "step": 1268
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.165603063178047e-06,
+      "loss": 0.3461,
+      "step": 1272
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.162412252712189e-06,
+      "loss": 0.2459,
+      "step": 1276
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.159221442246331e-06,
+      "loss": 0.8138,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.156030631780473e-06,
+      "loss": 0.6026,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.152839821314614e-06,
+      "loss": 0.5463,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.1496490108487556e-06,
+      "loss": 0.4317,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.146458200382897e-06,
+      "loss": 0.6244,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.143267389917039e-06,
+      "loss": 0.554,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.140076579451181e-06,
+      "loss": 0.6441,
+      "step": 1304
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.136885768985322e-06,
+      "loss": 0.6233,
+      "step": 1308
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.133694958519464e-06,
+      "loss": 0.5561,
+      "step": 1312
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.130504148053606e-06,
+      "loss": 0.7524,
+      "step": 1316
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.1273133375877475e-06,
+      "loss": 0.4338,
+      "step": 1320
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.124122527121889e-06,
+      "loss": 0.4495,
+      "step": 1324
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.1209317166560305e-06,
+      "loss": 0.5139,
+      "step": 1328
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.117740906190173e-06,
+      "loss": 0.6545,
+      "step": 1332
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.114550095724314e-06,
+      "loss": 0.5588,
+      "step": 1336
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.111359285258456e-06,
+      "loss": 0.609,
+      "step": 1340
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.108168474792597e-06,
+      "loss": 0.553,
+      "step": 1344
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.1049776643267395e-06,
+      "loss": 0.5844,
+      "step": 1348
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.101786853860881e-06,
+      "loss": 0.5779,
+      "step": 1352
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.0985960433950225e-06,
+      "loss": 0.4207,
+      "step": 1356
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.095405232929165e-06,
+      "loss": 0.4617,
+      "step": 1360
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.092214422463306e-06,
+      "loss": 0.6092,
+      "step": 1364
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.089023611997448e-06,
+      "loss": 0.4607,
+      "step": 1368
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.085832801531589e-06,
+      "loss": 0.4239,
+      "step": 1372
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0826419910657315e-06,
+      "loss": 0.5438,
+      "step": 1376
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.079451180599873e-06,
+      "loss": 0.5006,
+      "step": 1380
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.0762603701340144e-06,
+      "loss": 0.6889,
+      "step": 1384
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.073069559668156e-06,
+      "loss": 0.5742,
+      "step": 1388
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.069878749202298e-06,
+      "loss": 0.8366,
+      "step": 1392
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.06668793873644e-06,
+      "loss": 0.5182,
+      "step": 1396
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.063497128270581e-06,
+      "loss": 0.4807,
+      "step": 1400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.060306317804723e-06,
+      "loss": 0.3995,
+      "step": 1404
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.057115507338865e-06,
+      "loss": 0.5958,
+      "step": 1408
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.0539246968730064e-06,
+      "loss": 0.4855,
+      "step": 1412
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.050733886407148e-06,
+      "loss": 0.5908,
+      "step": 1416
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.047543075941289e-06,
+      "loss": 0.7867,
+      "step": 1420
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.044352265475432e-06,
+      "loss": 0.7617,
+      "step": 1424
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.041161455009573e-06,
+      "loss": 0.4752,
+      "step": 1428
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.037970644543715e-06,
+      "loss": 0.4732,
+      "step": 1432
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.034779834077856e-06,
+      "loss": 0.635,
+      "step": 1436
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.0315890236119976e-06,
+      "loss": 0.4924,
+      "step": 1440
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.02839821314614e-06,
+      "loss": 0.4416,
+      "step": 1444
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.025207402680281e-06,
+      "loss": 0.4448,
+      "step": 1448
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.022016592214423e-06,
+      "loss": 0.7631,
+      "step": 1452
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.018825781748564e-06,
+      "loss": 0.5035,
+      "step": 1456
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.015634971282707e-06,
+      "loss": 0.3779,
+      "step": 1460
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.012444160816848e-06,
+      "loss": 0.4924,
+      "step": 1464
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0092533503509896e-06,
+      "loss": 0.3932,
+      "step": 1468
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.006062539885131e-06,
+      "loss": 0.6974,
+      "step": 1472
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.0028717294192725e-06,
+      "loss": 0.7347,
+      "step": 1476
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.999680918953415e-06,
+      "loss": 0.5564,
+      "step": 1480
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.996490108487556e-06,
+      "loss": 0.4424,
+      "step": 1484
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.993299298021698e-06,
+      "loss": 0.5323,
+      "step": 1488
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.990108487555839e-06,
+      "loss": 0.6138,
+      "step": 1492
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 3.9869176770899815e-06,
+      "loss": 0.5156,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.983726866624123e-06,
+      "loss": 0.282,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.9805360561582645e-06,
+      "loss": 0.5392,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.977345245692406e-06,
+      "loss": 0.5721,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.974154435226547e-06,
+      "loss": 0.6967,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 3.97096362476069e-06,
+      "loss": 0.5348,
+      "step": 1516
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.967772814294831e-06,
+      "loss": 0.6884,
+      "step": 1520
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.964582003828973e-06,
+      "loss": 0.5065,
+      "step": 1524
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.961391193363114e-06,
+      "loss": 0.4505,
+      "step": 1528
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9582003828972565e-06,
+      "loss": 0.6881,
+      "step": 1532
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.955009572431398e-06,
+      "loss": 0.5952,
+      "step": 1536
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.951818761965539e-06,
+      "loss": 0.5656,
+      "step": 1540
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.948627951499681e-06,
+      "loss": 0.6437,
+      "step": 1544
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.945437141033822e-06,
+      "loss": 0.5179,
+      "step": 1548
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.942246330567965e-06,
+      "loss": 0.5278,
+      "step": 1552
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.939055520102106e-06,
+      "loss": 0.6951,
+      "step": 1556
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.935864709636248e-06,
+      "loss": 0.5468,
+      "step": 1560
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.932673899170389e-06,
+      "loss": 0.5132,
+      "step": 1564
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.929483088704531e-06,
+      "loss": 0.6297,
+      "step": 1568
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.926292278238673e-06,
+      "loss": 0.5472,
+      "step": 1572
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.923101467772814e-06,
+      "loss": 0.6623,
+      "step": 1576
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.919910657306956e-06,
+      "loss": 0.6216,
+      "step": 1580
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.916719846841098e-06,
+      "loss": 0.5332,
+      "step": 1584
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.91352903637524e-06,
+      "loss": 0.4792,
+      "step": 1588
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.910338225909381e-06,
+      "loss": 0.4573,
+      "step": 1592
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.9071474154435225e-06,
+      "loss": 0.5135,
+      "step": 1596
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.903956604977665e-06,
+      "loss": 0.7619,
+      "step": 1600
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.900765794511806e-06,
+      "loss": 0.6681,
+      "step": 1604
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.897574984045948e-06,
+      "loss": 0.7789,
+      "step": 1608
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.89438417358009e-06,
+      "loss": 0.6078,
+      "step": 1612
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8911933631142316e-06,
+      "loss": 0.4812,
+      "step": 1616
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.888002552648373e-06,
+      "loss": 0.5893,
+      "step": 1620
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.8848117421825145e-06,
+      "loss": 0.4775,
+      "step": 1624
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.881620931716657e-06,
+      "loss": 0.5012,
+      "step": 1628
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.878430121250798e-06,
+      "loss": 0.4752,
+      "step": 1632
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.87523931078494e-06,
+      "loss": 0.4365,
+      "step": 1636
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.872048500319081e-06,
+      "loss": 0.6722,
+      "step": 1640
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.8688576898532236e-06,
+      "loss": 0.6083,
+      "step": 1644
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.865666879387365e-06,
+      "loss": 0.4533,
+      "step": 1648
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.8624760689215065e-06,
+      "loss": 0.5879,
+      "step": 1652
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.859285258455648e-06,
+      "loss": 0.6564,
+      "step": 1656
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.85609444798979e-06,
+      "loss": 0.5475,
+      "step": 1660
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.852903637523932e-06,
+      "loss": 0.5018,
+      "step": 1664
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.849712827058073e-06,
+      "loss": 0.4544,
+      "step": 1668
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.846522016592215e-06,
+      "loss": 0.6603,
+      "step": 1672
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.843331206126357e-06,
+      "loss": 0.6887,
+      "step": 1676
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.8401403956604985e-06,
+      "loss": 0.7819,
+      "step": 1680
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.83694958519464e-06,
+      "loss": 0.5052,
+      "step": 1684
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.833758774728781e-06,
+      "loss": 0.6689,
+      "step": 1688
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.830567964262923e-06,
+      "loss": 0.5564,
+      "step": 1692
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.827377153797065e-06,
+      "loss": 0.3658,
+      "step": 1696
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.824186343331207e-06,
+      "loss": 0.6376,
+      "step": 1700
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.820995532865348e-06,
+      "loss": 0.5681,
+      "step": 1704
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.81780472239949e-06,
+      "loss": 0.5974,
+      "step": 1708
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.814613911933632e-06,
+      "loss": 0.5623,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.8114231014677734e-06,
+      "loss": 0.6437,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.808232291001915e-06,
+      "loss": 0.6442,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.8050414805360563e-06,
+      "loss": 0.4729,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.801850670070198e-06,
+      "loss": 0.3677,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.79865985960434e-06,
+      "loss": 0.4295,
+      "step": 1732
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7954690491384816e-06,
+      "loss": 0.6049,
+      "step": 1736
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.792278238672623e-06,
+      "loss": 0.6363,
+      "step": 1740
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7890874282067645e-06,
+      "loss": 0.5939,
+      "step": 1744
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.785896617740907e-06,
+      "loss": 0.5011,
+      "step": 1748
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7827058072750483e-06,
+      "loss": 0.5177,
+      "step": 1752
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.77951499680919e-06,
+      "loss": 0.7722,
+      "step": 1756
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7763241863433313e-06,
+      "loss": 0.5204,
+      "step": 1760
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.773133375877473e-06,
+      "loss": 0.455,
+      "step": 1764
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.769942565411615e-06,
+      "loss": 0.5397,
+      "step": 1768
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.7667517549457565e-06,
+      "loss": 0.5528,
+      "step": 1772
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.763560944479898e-06,
+      "loss": 0.5286,
+      "step": 1776
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.76037013401404e-06,
+      "loss": 0.5475,
+      "step": 1780
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7571793235481818e-06,
+      "loss": 0.3887,
+      "step": 1784
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7539885130823233e-06,
+      "loss": 0.6288,
+      "step": 1788
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7507977026164647e-06,
+      "loss": 0.5563,
+      "step": 1792
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.7476068921506066e-06,
+      "loss": 0.6103,
+      "step": 1796
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.744416081684748e-06,
+      "loss": 0.4141,
+      "step": 1800
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.74122527121889e-06,
+      "loss": 0.4075,
+      "step": 1804
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.738034460753032e-06,
+      "loss": 0.3594,
+      "step": 1808
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7348436502871733e-06,
+      "loss": 0.5157,
+      "step": 1812
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.731652839821315e-06,
+      "loss": 0.4918,
+      "step": 1816
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 3.7284620293554563e-06,
+      "loss": 0.4456,
+      "step": 1820
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7252712188895986e-06,
+      "loss": 0.7768,
+      "step": 1824
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.72208040842374e-06,
+      "loss": 0.7511,
+      "step": 1828
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7188895979578815e-06,
+      "loss": 0.4604,
+      "step": 1832
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.715698787492023e-06,
+      "loss": 0.6048,
+      "step": 1836
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 3.7125079770261653e-06,
+      "loss": 0.6261,
+      "step": 1840
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.709317166560307e-06,
+      "loss": 0.7588,
+      "step": 1844
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7061263560944483e-06,
+      "loss": 0.6608,
+      "step": 1848
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.7029355456285897e-06,
+      "loss": 0.5453,
+      "step": 1852
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6997447351627312e-06,
+      "loss": 0.4361,
+      "step": 1856
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.6965539246968735e-06,
+      "loss": 0.5557,
+      "step": 1860
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 3.693363114231015e-06,
+      "loss": 0.6371,
+      "step": 1864
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6901723037651565e-06,
+      "loss": 0.4953,
+      "step": 1868
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.686981493299298e-06,
+      "loss": 0.4157,
+      "step": 1872
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6837906828334403e-06,
+      "loss": 0.5469,
+      "step": 1876
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.6805998723675817e-06,
+      "loss": 0.4933,
+      "step": 1880
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.677409061901723e-06,
+      "loss": 0.4994,
+      "step": 1884
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6742182514358647e-06,
+      "loss": 0.3726,
+      "step": 1888
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6710274409700066e-06,
+      "loss": 0.5413,
+      "step": 1892
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6678366305041485e-06,
+      "loss": 0.574,
+      "step": 1896
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.66464582003829e-06,
+      "loss": 0.2569,
+      "step": 1900
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6614550095724314e-06,
+      "loss": 0.5012,
+      "step": 1904
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 3.6582641991065733e-06,
+      "loss": 0.586,
+      "step": 1908
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.655073388640715e-06,
+      "loss": 0.4588,
+      "step": 1912
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6518825781748567e-06,
+      "loss": 0.3745,
+      "step": 1916
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6486917677089985e-06,
+      "loss": 0.5444,
+      "step": 1920
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.64550095724314e-06,
+      "loss": 0.5545,
+      "step": 1924
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.6423101467772815e-06,
+      "loss": 0.6965,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6391193363114234e-06,
+      "loss": 0.4442,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6359285258455653e-06,
+      "loss": 0.4866,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6327377153797067e-06,
+      "loss": 0.5114,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.6295469049138482e-06,
+      "loss": 0.5922,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.62635609444799e-06,
+      "loss": 0.4787,
+      "step": 1948
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.623165283982132e-06,
+      "loss": 0.6709,
+      "step": 1952
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6199744735162735e-06,
+      "loss": 0.5078,
+      "step": 1956
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.616783663050415e-06,
+      "loss": 0.5999,
+      "step": 1960
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6135928525845564e-06,
+      "loss": 0.5051,
+      "step": 1964
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.6111997447351634e-06,
+      "loss": 0.4373,
+      "step": 1968
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.608008934269305e-06,
+      "loss": 0.7497,
+      "step": 1972
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6048181238034463e-06,
+      "loss": 0.458,
+      "step": 1976
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6016273133375878e-06,
+      "loss": 0.3981,
+      "step": 1980
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.59843650287173e-06,
+      "loss": 0.4995,
+      "step": 1984
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5952456924058716e-06,
+      "loss": 0.493,
+      "step": 1988
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.592054881940013e-06,
+      "loss": 0.462,
+      "step": 1992
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5888640714741545e-06,
+      "loss": 0.5239,
+      "step": 1996
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.585673261008296e-06,
+      "loss": 0.4376,
+      "step": 2000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5824824505424383e-06,
+      "loss": 0.6067,
+      "step": 2004
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5792916400765798e-06,
+      "loss": 0.4091,
+      "step": 2008
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5761008296107212e-06,
+      "loss": 0.5261,
+      "step": 2012
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5729100191448627e-06,
+      "loss": 0.5408,
+      "step": 2016
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.569719208679005e-06,
+      "loss": 0.5867,
+      "step": 2020
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5665283982131465e-06,
+      "loss": 0.636,
+      "step": 2024
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.563337587747288e-06,
+      "loss": 0.4329,
+      "step": 2028
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.56014677728143e-06,
+      "loss": 0.7026,
+      "step": 2032
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5569559668155713e-06,
+      "loss": 0.5245,
+      "step": 2036
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.553765156349713e-06,
+      "loss": 0.4929,
+      "step": 2040
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5505743458838547e-06,
+      "loss": 0.4876,
+      "step": 2044
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.5473835354179966e-06,
+      "loss": 0.45,
+      "step": 2048
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.544192724952138e-06,
+      "loss": 0.5068,
+      "step": 2052
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.54100191448628e-06,
+      "loss": 0.5647,
+      "step": 2056
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5378111040204214e-06,
+      "loss": 0.5048,
+      "step": 2060
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5346202935545633e-06,
+      "loss": 0.457,
+      "step": 2064
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5314294830887048e-06,
+      "loss": 0.4089,
+      "step": 2068
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5282386726228462e-06,
+      "loss": 0.3521,
+      "step": 2072
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.5250478621569886e-06,
+      "loss": 0.3477,
+      "step": 2076
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.52185705169113e-06,
+      "loss": 0.6625,
+      "step": 2080
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5186662412252715e-06,
+      "loss": 0.3829,
+      "step": 2084
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.515475430759413e-06,
+      "loss": 0.4733,
+      "step": 2088
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5122846202935553e-06,
+      "loss": 0.4024,
+      "step": 2092
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5090938098276968e-06,
+      "loss": 0.5733,
+      "step": 2096
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.5059029993618382e-06,
+      "loss": 0.5788,
+      "step": 2100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.5027121888959797e-06,
+      "loss": 0.4806,
+      "step": 2104
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.499521378430121e-06,
+      "loss": 0.5091,
+      "step": 2108
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4963305679642635e-06,
+      "loss": 0.6465,
+      "step": 2112
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.493139757498405e-06,
+      "loss": 0.4821,
+      "step": 2116
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.4899489470325464e-06,
+      "loss": 0.3563,
+      "step": 2120
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.486758136566688e-06,
+      "loss": 0.7174,
+      "step": 2124
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4835673261008302e-06,
+      "loss": 0.3833,
+      "step": 2128
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4803765156349717e-06,
+      "loss": 0.6688,
+      "step": 2132
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.477185705169113e-06,
+      "loss": 0.5733,
+      "step": 2136
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4739948947032546e-06,
+      "loss": 0.5743,
+      "step": 2140
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.4708040842373965e-06,
+      "loss": 0.5219,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4676132737715384e-06,
+      "loss": 0.5964,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.46442246330568e-06,
+      "loss": 0.5981,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4612316528398214e-06,
+      "loss": 0.4068,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.4580408423739632e-06,
+      "loss": 0.3966,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.454850031908105e-06,
+      "loss": 0.2291,
+      "step": 2164
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4516592214422466e-06,
+      "loss": 0.4695,
+      "step": 2168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4484684109763885e-06,
+      "loss": 0.5594,
+      "step": 2172
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.44527760051053e-06,
+      "loss": 0.5603,
+      "step": 2176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4420867900446714e-06,
+      "loss": 0.4934,
+      "step": 2180
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4388959795788133e-06,
+      "loss": 0.6316,
+      "step": 2184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.4357051691129552e-06,
+      "loss": 0.3424,
+      "step": 2188
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4325143586470967e-06,
+      "loss": 0.566,
+      "step": 2192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.429323548181238e-06,
+      "loss": 0.3565,
+      "step": 2196
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.42613273771538e-06,
+      "loss": 0.5191,
+      "step": 2200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.422941927249522e-06,
+      "loss": 0.3848,
+      "step": 2204
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.4197511167836634e-06,
+      "loss": 0.6962,
+      "step": 2208
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.416560306317805e-06,
+      "loss": 0.3646,
+      "step": 2212
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4133694958519464e-06,
+      "loss": 0.3756,
+      "step": 2216
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4101786853860887e-06,
+      "loss": 0.2853,
+      "step": 2220
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.40698787492023e-06,
+      "loss": 0.2925,
+      "step": 2224
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.4037970644543716e-06,
+      "loss": 0.3838,
+      "step": 2228
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.400606253988513e-06,
+      "loss": 0.4479,
+      "step": 2232
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3974154435226554e-06,
+      "loss": 0.5207,
+      "step": 2236
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.394224633056797e-06,
+      "loss": 0.3813,
+      "step": 2240
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3910338225909384e-06,
+      "loss": 0.4028,
+      "step": 2244
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.38784301212508e-06,
+      "loss": 0.3406,
+      "step": 2248
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.3846522016592213e-06,
+      "loss": 0.4495,
+      "step": 2252
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3814613911933636e-06,
+      "loss": 0.5411,
+      "step": 2256
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.378270580727505e-06,
+      "loss": 0.3533,
+      "step": 2260
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3750797702616465e-06,
+      "loss": 0.5577,
+      "step": 2264
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.371888959795788e-06,
+      "loss": 0.4198,
+      "step": 2268
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.3686981493299303e-06,
+      "loss": 0.2956,
+      "step": 2272
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.365507338864072e-06,
+      "loss": 0.5714,
+      "step": 2276
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3623165283982133e-06,
+      "loss": 0.3393,
+      "step": 2280
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.359125717932355e-06,
+      "loss": 0.3448,
+      "step": 2284
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3559349074664966e-06,
+      "loss": 0.4956,
+      "step": 2288
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.3527440970006385e-06,
+      "loss": 0.4609,
+      "step": 2292
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.34955328653478e-06,
+      "loss": 0.4499,
+      "step": 2296
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.346362476068922e-06,
+      "loss": 0.3638,
+      "step": 2300
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3431716656030634e-06,
+      "loss": 0.6062,
+      "step": 2304
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3399808551372053e-06,
+      "loss": 0.319,
+      "step": 2308
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3367900446713467e-06,
+      "loss": 0.3106,
+      "step": 2312
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.3335992342054886e-06,
+      "loss": 0.6715,
+      "step": 2316
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.33040842373963e-06,
+      "loss": 0.4007,
+      "step": 2320
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3272176132737716e-06,
+      "loss": 0.5854,
+      "step": 2324
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.324026802807914e-06,
+      "loss": 0.4384,
+      "step": 2328
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.3208359923420554e-06,
+      "loss": 0.5186,
+      "step": 2332
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.317645181876197e-06,
+      "loss": 0.2793,
+      "step": 2336
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3144543714103383e-06,
+      "loss": 0.3945,
+      "step": 2340
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3112635609444806e-06,
+      "loss": 0.4488,
+      "step": 2344
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.308072750478622e-06,
+      "loss": 0.2692,
+      "step": 2348
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.3048819400127635e-06,
+      "loss": 0.4689,
+      "step": 2352
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.301691129546905e-06,
+      "loss": 0.2162,
+      "step": 2356
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.2985003190810465e-06,
+      "loss": 0.3339,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.295309508615189e-06,
+      "loss": 0.5855,
+      "step": 2364
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2921186981493303e-06,
+      "loss": 0.4823,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2889278876834717e-06,
+      "loss": 0.3587,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2857370772176132e-06,
+      "loss": 0.3903,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.2825462667517555e-06,
+      "loss": 0.5669,
+      "step": 2380
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.279355456285897e-06,
+      "loss": 0.413,
+      "step": 2384
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2761646458200385e-06,
+      "loss": 0.3735,
+      "step": 2388
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.27297383535418e-06,
+      "loss": 0.5467,
+      "step": 2392
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.269783024888322e-06,
+      "loss": 0.3738,
+      "step": 2396
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.2665922144224637e-06,
+      "loss": 0.4619,
+      "step": 2400
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.263401403956605e-06,
+      "loss": 0.3739,
+      "step": 2404
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2602105934907467e-06,
+      "loss": 0.4065,
+      "step": 2408
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2570197830248886e-06,
+      "loss": 0.3406,
+      "step": 2412
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.2538289725590305e-06,
+      "loss": 0.4554,
+      "step": 2416
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.250638162093172e-06,
+      "loss": 0.799,
+      "step": 2420
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.247447351627314e-06,
+      "loss": 0.4552,
+      "step": 2424
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2442565411614553e-06,
+      "loss": 0.3708,
+      "step": 2428
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2410657306955968e-06,
+      "loss": 0.3309,
+      "step": 2432
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2378749202297387e-06,
+      "loss": 0.4065,
+      "step": 2436
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.2346841097638806e-06,
+      "loss": 0.627,
+      "step": 2440
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.231493299298022e-06,
+      "loss": 0.3551,
+      "step": 2444
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2283024888321635e-06,
+      "loss": 0.2465,
+      "step": 2448
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2251116783663054e-06,
+      "loss": 0.5104,
+      "step": 2452
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2219208679004473e-06,
+      "loss": 0.5923,
+      "step": 2456
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2187300574345887e-06,
+      "loss": 0.3455,
+      "step": 2460
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2155392469687302e-06,
+      "loss": 0.3767,
+      "step": 2464
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.2123484365028717e-06,
+      "loss": 0.4728,
+      "step": 2468
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.209157626037014e-06,
+      "loss": 0.4427,
+      "step": 2472
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.2059668155711555e-06,
+      "loss": 0.3805,
+      "step": 2476
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.202776005105297e-06,
+      "loss": 0.2417,
+      "step": 2480
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.1995851946394384e-06,
+      "loss": 0.4459,
+      "step": 2484
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.1963943841735807e-06,
+      "loss": 0.5951,
+      "step": 2488
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.193203573707722e-06,
+      "loss": 0.4512,
+      "step": 2492
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1900127632418637e-06,
+      "loss": 0.4038,
+      "step": 2496
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.186821952776005e-06,
+      "loss": 0.5716,
+      "step": 2500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.1836311423101466e-06,
+      "loss": 0.3122,
+      "step": 2504
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.180440331844289e-06,
+      "loss": 0.6523,
+      "step": 2508
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1772495213784304e-06,
+      "loss": 0.4129,
+      "step": 2512
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.174058710912572e-06,
+      "loss": 0.7674,
+      "step": 2516
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1708679004467138e-06,
+      "loss": 0.5227,
+      "step": 2520
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1676770899808557e-06,
+      "loss": 0.457,
+      "step": 2524
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.164486279514997e-06,
+      "loss": 0.3279,
+      "step": 2528
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.1612954690491386e-06,
+      "loss": 0.4809,
+      "step": 2532
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1581046585832805e-06,
+      "loss": 0.3513,
+      "step": 2536
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.154913848117422e-06,
+      "loss": 0.5097,
+      "step": 2540
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.151723037651564e-06,
+      "loss": 0.4727,
+      "step": 2544
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1485322271857053e-06,
+      "loss": 0.3848,
+      "step": 2548
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.1453414167198472e-06,
+      "loss": 0.49,
+      "step": 2552
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1421506062539887e-06,
+      "loss": 0.4166,
+      "step": 2556
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1389597957881306e-06,
+      "loss": 0.4397,
+      "step": 2560
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.135768985322272e-06,
+      "loss": 0.4295,
+      "step": 2564
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.132578174856414e-06,
+      "loss": 0.3807,
+      "step": 2568
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.1293873643905554e-06,
+      "loss": 0.5155,
+      "step": 2572
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.126196553924697e-06,
+      "loss": 0.4183,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.123005743458839e-06,
+      "loss": 0.4173,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1198149329929807e-06,
+      "loss": 0.5842,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.116624122527122e-06,
+      "loss": 0.6673,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.1134333120612636e-06,
+      "loss": 0.4492,
+      "step": 2592
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.110242501595406e-06,
+      "loss": 0.4747,
+      "step": 2596
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1070516911295474e-06,
+      "loss": 0.5746,
+      "step": 2600
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.103860880663689e-06,
+      "loss": 0.5708,
+      "step": 2604
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.1006700701978303e-06,
+      "loss": 0.601,
+      "step": 2608
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.097479259731972e-06,
+      "loss": 0.7205,
+      "step": 2612
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.094288449266114e-06,
+      "loss": 0.4023,
+      "step": 2616
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0910976388002556e-06,
+      "loss": 0.5012,
+      "step": 2620
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.087906828334397e-06,
+      "loss": 0.4779,
+      "step": 2624
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0847160178685385e-06,
+      "loss": 0.5062,
+      "step": 2628
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.081525207402681e-06,
+      "loss": 0.2888,
+      "step": 2632
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.0783343969368223e-06,
+      "loss": 0.4121,
+      "step": 2636
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.075143586470964e-06,
+      "loss": 0.4721,
+      "step": 2640
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0719527760051053e-06,
+      "loss": 0.3422,
+      "step": 2644
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.068761965539247e-06,
+      "loss": 0.5018,
+      "step": 2648
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.065571155073389e-06,
+      "loss": 0.5165,
+      "step": 2652
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.0623803446075305e-06,
+      "loss": 0.3574,
+      "step": 2656
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.059189534141672e-06,
+      "loss": 0.579,
+      "step": 2660
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.055998723675814e-06,
+      "loss": 0.4961,
+      "step": 2664
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0528079132099558e-06,
+      "loss": 0.3949,
+      "step": 2668
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0496171027440973e-06,
+      "loss": 0.3999,
+      "step": 2672
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.046426292278239e-06,
+      "loss": 0.524,
+      "step": 2676
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.0432354818123806e-06,
+      "loss": 0.3688,
+      "step": 2680
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.040044671346522e-06,
+      "loss": 0.403,
+      "step": 2684
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.036853860880664e-06,
+      "loss": 0.3388,
+      "step": 2688
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.033663050414806e-06,
+      "loss": 0.4706,
+      "step": 2692
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0304722399489473e-06,
+      "loss": 0.6817,
+      "step": 2696
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.027281429483089e-06,
+      "loss": 0.3896,
+      "step": 2700
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.0240906190172307e-06,
+      "loss": 0.358,
+      "step": 2704
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0208998085513726e-06,
+      "loss": 0.3115,
+      "step": 2708
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.017708998085514e-06,
+      "loss": 0.5322,
+      "step": 2712
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0145181876196555e-06,
+      "loss": 0.4613,
+      "step": 2716
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.011327377153797e-06,
+      "loss": 0.4374,
+      "step": 2720
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.0081365666879393e-06,
+      "loss": 0.4775,
+      "step": 2724
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.004945756222081e-06,
+      "loss": 0.349,
+      "step": 2728
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0017549457562223e-06,
+      "loss": 0.5114,
+      "step": 2732
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9985641352903637e-06,
+      "loss": 0.3901,
+      "step": 2736
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.995373324824506e-06,
+      "loss": 0.4756,
+      "step": 2740
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.9921825143586475e-06,
+      "loss": 0.4669,
+      "step": 2744
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.988991703892789e-06,
+      "loss": 0.5554,
+      "step": 2748
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9858008934269305e-06,
+      "loss": 0.3345,
+      "step": 2752
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.982610082961072e-06,
+      "loss": 0.3653,
+      "step": 2756
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9794192724952143e-06,
+      "loss": 0.4543,
+      "step": 2760
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.9762284620293557e-06,
+      "loss": 0.382,
+      "step": 2764
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.973037651563497e-06,
+      "loss": 0.2821,
+      "step": 2768
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.969846841097639e-06,
+      "loss": 0.4392,
+      "step": 2772
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.966656030631781e-06,
+      "loss": 0.3785,
+      "step": 2776
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9634652201659224e-06,
+      "loss": 0.4799,
+      "step": 2780
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.960274409700064e-06,
+      "loss": 0.4004,
+      "step": 2784
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.957083599234206e-06,
+      "loss": 0.4598,
+      "step": 2788
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9538927887683473e-06,
+      "loss": 0.6889,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.950701978302489e-06,
+      "loss": 0.3401,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9475111678366306e-06,
+      "loss": 0.5162,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.9443203573707725e-06,
+      "loss": 0.3811,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.941129546904914e-06,
+      "loss": 0.3048,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.937938736439056e-06,
+      "loss": 0.5528,
+      "step": 2812
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9347479259731974e-06,
+      "loss": 0.3721,
+      "step": 2816
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9315571155073393e-06,
+      "loss": 0.4877,
+      "step": 2820
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9283663050414807e-06,
+      "loss": 0.3101,
+      "step": 2824
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.925175494575622e-06,
+      "loss": 0.3458,
+      "step": 2828
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9219846841097645e-06,
+      "loss": 0.3741,
+      "step": 2832
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.918793873643906e-06,
+      "loss": 0.7428,
+      "step": 2836
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9156030631780475e-06,
+      "loss": 0.3487,
+      "step": 2840
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.912412252712189e-06,
+      "loss": 0.3184,
+      "step": 2844
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9092214422463313e-06,
+      "loss": 0.3778,
+      "step": 2848
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9060306317804727e-06,
+      "loss": 0.4085,
+      "step": 2852
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.902839821314614e-06,
+      "loss": 0.4859,
+      "step": 2856
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8996490108487557e-06,
+      "loss": 0.5783,
+      "step": 2860
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.896458200382897e-06,
+      "loss": 0.2396,
+      "step": 2864
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8932673899170395e-06,
+      "loss": 0.6563,
+      "step": 2868
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.890076579451181e-06,
+      "loss": 0.3048,
+      "step": 2872
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8868857689853224e-06,
+      "loss": 0.3935,
+      "step": 2876
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.883694958519464e-06,
+      "loss": 0.5063,
+      "step": 2880
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.880504148053606e-06,
+      "loss": 0.5056,
+      "step": 2884
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8773133375877476e-06,
+      "loss": 0.2203,
+      "step": 2888
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.874122527121889e-06,
+      "loss": 0.3044,
+      "step": 2892
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8709317166560306e-06,
+      "loss": 0.4444,
+      "step": 2896
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8677409061901725e-06,
+      "loss": 0.3645,
+      "step": 2900
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8645500957243144e-06,
+      "loss": 0.4594,
+      "step": 2904
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.861359285258456e-06,
+      "loss": 0.4897,
+      "step": 2908
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8581684747925973e-06,
+      "loss": 0.4772,
+      "step": 2912
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.854977664326739e-06,
+      "loss": 0.388,
+      "step": 2916
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.851786853860881e-06,
+      "loss": 0.3869,
+      "step": 2920
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8485960433950226e-06,
+      "loss": 0.4853,
+      "step": 2924
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8454052329291645e-06,
+      "loss": 0.4467,
+      "step": 2928
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.842214422463306e-06,
+      "loss": 0.2356,
+      "step": 2932
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8390236119974474e-06,
+      "loss": 0.4614,
+      "step": 2936
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8358328015315893e-06,
+      "loss": 0.3212,
+      "step": 2940
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.832641991065731e-06,
+      "loss": 0.5037,
+      "step": 2944
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8294511805998727e-06,
+      "loss": 0.4957,
+      "step": 2948
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.826260370134014e-06,
+      "loss": 0.418,
+      "step": 2952
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.823069559668156e-06,
+      "loss": 0.2996,
+      "step": 2956
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.819878749202298e-06,
+      "loss": 0.5421,
+      "step": 2960
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8166879387364394e-06,
+      "loss": 0.5049,
+      "step": 2964
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.813497128270581e-06,
+      "loss": 0.3929,
+      "step": 2968
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8103063178047223e-06,
+      "loss": 0.4045,
+      "step": 2972
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8071155073388646e-06,
+      "loss": 0.3494,
+      "step": 2976
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.803924696873006e-06,
+      "loss": 0.3782,
+      "step": 2980
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8007338864071476e-06,
+      "loss": 0.2768,
+      "step": 2984
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.797543075941289e-06,
+      "loss": 0.531,
+      "step": 2988
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7943522654754314e-06,
+      "loss": 0.4958,
+      "step": 2992
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.791161455009573e-06,
+      "loss": 0.6183,
+      "step": 2996
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7879706445437143e-06,
+      "loss": 0.3521,
+      "step": 3000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7847798340778558e-06,
+      "loss": 0.4406,
+      "step": 3004
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7815890236119973e-06,
+      "loss": 0.4131,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7783982131461396e-06,
+      "loss": 0.5107,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.775207402680281e-06,
+      "loss": 0.2735,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7720165922144225e-06,
+      "loss": 0.3788,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7696234843650286e-06,
+      "loss": 0.5599,
+      "step": 3024
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7664326738991705e-06,
+      "loss": 0.2355,
+      "step": 3028
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7632418634333124e-06,
+      "loss": 0.5358,
+      "step": 3032
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.760051052967454e-06,
+      "loss": 0.3283,
+      "step": 3036
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7568602425015958e-06,
+      "loss": 0.4093,
+      "step": 3040
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7536694320357372e-06,
+      "loss": 0.287,
+      "step": 3044
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.750478621569879e-06,
+      "loss": 0.5271,
+      "step": 3048
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7472878111040206e-06,
+      "loss": 0.3372,
+      "step": 3052
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7440970006381625e-06,
+      "loss": 0.5649,
+      "step": 3056
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.740906190172304e-06,
+      "loss": 0.5017,
+      "step": 3060
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7377153797064454e-06,
+      "loss": 0.6057,
+      "step": 3064
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7345245692405873e-06,
+      "loss": 0.4184,
+      "step": 3068
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7313337587747292e-06,
+      "loss": 0.2892,
+      "step": 3072
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7281429483088707e-06,
+      "loss": 0.5914,
+      "step": 3076
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.724952137843012e-06,
+      "loss": 0.472,
+      "step": 3080
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.721761327377154e-06,
+      "loss": 0.3773,
+      "step": 3084
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.718570516911296e-06,
+      "loss": 0.2942,
+      "step": 3088
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.7153797064454374e-06,
+      "loss": 0.3445,
+      "step": 3092
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.712188895979579e-06,
+      "loss": 0.2773,
+      "step": 3096
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7089980855137204e-06,
+      "loss": 0.4007,
+      "step": 3100
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.7058072750478627e-06,
+      "loss": 0.3083,
+      "step": 3104
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.702616464582004e-06,
+      "loss": 0.4782,
+      "step": 3108
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6994256541161456e-06,
+      "loss": 0.5419,
+      "step": 3112
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.696234843650287e-06,
+      "loss": 0.5713,
+      "step": 3116
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6930440331844294e-06,
+      "loss": 0.3722,
+      "step": 3120
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.689853222718571e-06,
+      "loss": 0.4663,
+      "step": 3124
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6866624122527123e-06,
+      "loss": 0.3208,
+      "step": 3128
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.683471601786854e-06,
+      "loss": 0.351,
+      "step": 3132
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.6802807913209957e-06,
+      "loss": 0.513,
+      "step": 3136
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6770899808551376e-06,
+      "loss": 0.4409,
+      "step": 3140
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.673899170389279e-06,
+      "loss": 0.3335,
+      "step": 3144
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6707083599234205e-06,
+      "loss": 0.3706,
+      "step": 3148
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6675175494575624e-06,
+      "loss": 0.4404,
+      "step": 3152
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6643267389917043e-06,
+      "loss": 0.4186,
+      "step": 3156
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.661135928525846e-06,
+      "loss": 0.3666,
+      "step": 3160
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6579451180599873e-06,
+      "loss": 0.3849,
+      "step": 3164
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.654754307594129e-06,
+      "loss": 0.4564,
+      "step": 3168
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6515634971282706e-06,
+      "loss": 0.3534,
+      "step": 3172
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.6483726866624125e-06,
+      "loss": 0.3735,
+      "step": 3176
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.645181876196554e-06,
+      "loss": 0.4449,
+      "step": 3180
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.641991065730696e-06,
+      "loss": 0.5032,
+      "step": 3184
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6388002552648374e-06,
+      "loss": 0.3677,
+      "step": 3188
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6356094447989793e-06,
+      "loss": 0.5004,
+      "step": 3192
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.632418634333121e-06,
+      "loss": 0.1972,
+      "step": 3196
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.6292278238672626e-06,
+      "loss": 0.4606,
+      "step": 3200
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.626037013401404e-06,
+      "loss": 0.3533,
+      "step": 3204
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6228462029355456e-06,
+      "loss": 0.3607,
+      "step": 3208
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.619655392469688e-06,
+      "loss": 0.5767,
+      "step": 3212
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.6164645820038293e-06,
+      "loss": 0.5316,
+      "step": 3216
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.613273771537971e-06,
+      "loss": 0.2474,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6100829610721123e-06,
+      "loss": 0.3168,
+      "step": 3224
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6068921506062546e-06,
+      "loss": 0.4029,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.603701340140396e-06,
+      "loss": 0.2693,
+      "step": 3232
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.6005105296745375e-06,
+      "loss": 0.3756,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.597319719208679e-06,
+      "loss": 0.3712,
+      "step": 3240
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.5941289087428205e-06,
+      "loss": 0.366,
+      "step": 3244
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.590938098276963e-06,
+      "loss": 0.3813,
+      "step": 3248
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5877472878111043e-06,
+      "loss": 0.4442,
+      "step": 3252
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5845564773452457e-06,
+      "loss": 0.4061,
+      "step": 3256
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.581365666879387e-06,
+      "loss": 0.3679,
+      "step": 3260
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.5781748564135295e-06,
+      "loss": 0.2641,
+      "step": 3264
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.574984045947671e-06,
+      "loss": 0.5656,
+      "step": 3268
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5717932354818125e-06,
+      "loss": 0.3672,
+      "step": 3272
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.568602425015954e-06,
+      "loss": 0.3395,
+      "step": 3276
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.565411614550096e-06,
+      "loss": 0.5946,
+      "step": 3280
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.5622208040842377e-06,
+      "loss": 0.3526,
+      "step": 3284
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.559029993618379e-06,
+      "loss": 0.3365,
+      "step": 3288
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.555839183152521e-06,
+      "loss": 0.4003,
+      "step": 3292
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5526483726866626e-06,
+      "loss": 0.3994,
+      "step": 3296
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.5494575622208045e-06,
+      "loss": 0.3623,
+      "step": 3300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.546266751754946e-06,
+      "loss": 0.5994,
+      "step": 3304
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.543075941289088e-06,
+      "loss": 0.3717,
+      "step": 3308
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5398851308232293e-06,
+      "loss": 0.2424,
+      "step": 3312
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5366943203573708e-06,
+      "loss": 0.5083,
+      "step": 3316
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5335035098915127e-06,
+      "loss": 0.2865,
+      "step": 3320
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.5303126994256545e-06,
+      "loss": 0.2184,
+      "step": 3324
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.527121888959796e-06,
+      "loss": 0.5697,
+      "step": 3328
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5239310784939375e-06,
+      "loss": 0.3524,
+      "step": 3332
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5207402680280794e-06,
+      "loss": 0.3922,
+      "step": 3336
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5175494575622213e-06,
+      "loss": 0.3364,
+      "step": 3340
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5143586470963627e-06,
+      "loss": 0.3983,
+      "step": 3344
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5111678366305042e-06,
+      "loss": 0.3812,
+      "step": 3348
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.5079770261646457e-06,
+      "loss": 0.3001,
+      "step": 3352
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.504786215698788e-06,
+      "loss": 0.3159,
+      "step": 3356
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5015954052329295e-06,
+      "loss": 0.316,
+      "step": 3360
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.498404594767071e-06,
+      "loss": 0.2347,
+      "step": 3364
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.495213784301213e-06,
+      "loss": 0.4728,
+      "step": 3368
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.4920229738353543e-06,
+      "loss": 0.4451,
+      "step": 3372
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.488832163369496e-06,
+      "loss": 0.6876,
+      "step": 3376
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4856413529036377e-06,
+      "loss": 0.4799,
+      "step": 3380
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.482450542437779e-06,
+      "loss": 0.3912,
+      "step": 3384
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.479259731971921e-06,
+      "loss": 0.2295,
+      "step": 3388
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.4760689215060625e-06,
+      "loss": 0.2529,
+      "step": 3392
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4728781110402044e-06,
+      "loss": 0.454,
+      "step": 3396
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.469687300574346e-06,
+      "loss": 0.3894,
+      "step": 3400
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4664964901084878e-06,
+      "loss": 0.2908,
+      "step": 3404
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4633056796426292e-06,
+      "loss": 0.499,
+      "step": 3408
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.460114869176771e-06,
+      "loss": 0.3336,
+      "step": 3412
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.4569240587109126e-06,
+      "loss": 0.3016,
+      "step": 3416
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4537332482450545e-06,
+      "loss": 0.4519,
+      "step": 3420
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.450542437779196e-06,
+      "loss": 0.2589,
+      "step": 3424
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.447351627313338e-06,
+      "loss": 0.404,
+      "step": 3428
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4441608168474793e-06,
+      "loss": 0.335,
+      "step": 3432
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4409700063816212e-06,
+      "loss": 0.4312,
+      "step": 3436
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.437779195915763e-06,
+      "loss": 0.2877,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4345883854499046e-06,
+      "loss": 0.3591,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4313975749840465e-06,
+      "loss": 0.3149,
+      "step": 3448
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.428206764518188e-06,
+      "loss": 0.3785,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4250159540523294e-06,
+      "loss": 0.3654,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4218251435864713e-06,
+      "loss": 0.2894,
+      "step": 3460
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4186343331206128e-06,
+      "loss": 0.5198,
+      "step": 3464
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4154435226547547e-06,
+      "loss": 0.4666,
+      "step": 3468
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.412252712188896e-06,
+      "loss": 0.3899,
+      "step": 3472
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.409061901723038e-06,
+      "loss": 0.4248,
+      "step": 3476
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.4058710912571795e-06,
+      "loss": 0.3144,
+      "step": 3480
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.4026802807913214e-06,
+      "loss": 0.3294,
+      "step": 3484
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.399489470325463e-06,
+      "loss": 0.3395,
+      "step": 3488
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3962986598596043e-06,
+      "loss": 0.4384,
+      "step": 3492
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3931078493937462e-06,
+      "loss": 0.3029,
+      "step": 3496
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3899170389278877e-06,
+      "loss": 0.3868,
+      "step": 3500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3867262284620296e-06,
+      "loss": 0.233,
+      "step": 3504
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.383535417996171e-06,
+      "loss": 0.4025,
+      "step": 3508
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.380344607530313e-06,
+      "loss": 0.2714,
+      "step": 3512
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3771537970644544e-06,
+      "loss": 0.4694,
+      "step": 3516
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.3739629865985963e-06,
+      "loss": 0.3092,
+      "step": 3520
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.370772176132738e-06,
+      "loss": 0.3375,
+      "step": 3524
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3675813656668793e-06,
+      "loss": 0.2356,
+      "step": 3528
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.364390555201021e-06,
+      "loss": 0.4403,
+      "step": 3532
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3611997447351626e-06,
+      "loss": 0.4015,
+      "step": 3536
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3580089342693045e-06,
+      "loss": 0.5201,
+      "step": 3540
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.3548181238034464e-06,
+      "loss": 0.4203,
+      "step": 3544
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.351627313337588e-06,
+      "loss": 0.4869,
+      "step": 3548
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3484365028717298e-06,
+      "loss": 0.3923,
+      "step": 3552
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3452456924058712e-06,
+      "loss": 0.6743,
+      "step": 3556
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.342054881940013e-06,
+      "loss": 0.2588,
+      "step": 3560
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3388640714741546e-06,
+      "loss": 0.323,
+      "step": 3564
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.3356732610082965e-06,
+      "loss": 0.2859,
+      "step": 3568
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.332482450542438e-06,
+      "loss": 0.2747,
+      "step": 3572
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.32929164007658e-06,
+      "loss": 0.2221,
+      "step": 3576
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3261008296107213e-06,
+      "loss": 0.3744,
+      "step": 3580
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3229100191448632e-06,
+      "loss": 0.3965,
+      "step": 3584
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.3197192086790047e-06,
+      "loss": 0.4889,
+      "step": 3588
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3165283982131466e-06,
+      "loss": 0.4218,
+      "step": 3592
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.313337587747288e-06,
+      "loss": 0.3016,
+      "step": 3596
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3101467772814295e-06,
+      "loss": 0.3408,
+      "step": 3600
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.3069559668155714e-06,
+      "loss": 0.387,
+      "step": 3604
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.303765156349713e-06,
+      "loss": 0.3845,
+      "step": 3608
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.300574345883855e-06,
+      "loss": 0.2885,
+      "step": 3612
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2973835354179963e-06,
+      "loss": 0.1871,
+      "step": 3616
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.294192724952138e-06,
+      "loss": 0.3516,
+      "step": 3620
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2910019144862796e-06,
+      "loss": 0.4165,
+      "step": 3624
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.2878111040204215e-06,
+      "loss": 0.2891,
+      "step": 3628
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.284620293554563e-06,
+      "loss": 0.3616,
+      "step": 3632
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2814294830887045e-06,
+      "loss": 0.4057,
+      "step": 3636
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2782386726228464e-06,
+      "loss": 0.5166,
+      "step": 3640
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.275047862156988e-06,
+      "loss": 0.3279,
+      "step": 3644
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.2718570516911297e-06,
+      "loss": 0.3537,
+      "step": 3648
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.268666241225271e-06,
+      "loss": 0.3187,
+      "step": 3652
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.265475430759413e-06,
+      "loss": 0.4043,
+      "step": 3656
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2622846202935546e-06,
+      "loss": 0.2799,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.2590938098276964e-06,
+      "loss": 0.3363,
+      "step": 3664
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.255902999361838e-06,
+      "loss": 0.6477,
+      "step": 3668
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.25271218889598e-06,
+      "loss": 0.4967,
+      "step": 3672
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2495213784301213e-06,
+      "loss": 0.4474,
+      "step": 3676
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.246330567964263e-06,
+      "loss": 0.2501,
+      "step": 3680
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2431397574984046e-06,
+      "loss": 0.3448,
+      "step": 3684
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2399489470325465e-06,
+      "loss": 0.3084,
+      "step": 3688
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.2367581365666884e-06,
+      "loss": 0.3165,
+      "step": 3692
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.23356732610083e-06,
+      "loss": 0.405,
+      "step": 3696
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.230376515634972e-06,
+      "loss": 0.3648,
+      "step": 3700
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2271857051691133e-06,
+      "loss": 0.2938,
+      "step": 3704
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2239948947032547e-06,
+      "loss": 0.336,
+      "step": 3708
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.2208040842373966e-06,
+      "loss": 0.4741,
+      "step": 3712
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.217613273771538e-06,
+      "loss": 0.4006,
+      "step": 3716
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.21442246330568e-06,
+      "loss": 0.3443,
+      "step": 3720
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2112316528398215e-06,
+      "loss": 0.2771,
+      "step": 3724
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2080408423739634e-06,
+      "loss": 0.2515,
+      "step": 3728
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.204850031908105e-06,
+      "loss": 0.3897,
+      "step": 3732
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.2016592214422467e-06,
+      "loss": 0.182,
+      "step": 3736
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.198468410976388e-06,
+      "loss": 0.3575,
+      "step": 3740
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1952776005105297e-06,
+      "loss": 0.3662,
+      "step": 3744
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1920867900446716e-06,
+      "loss": 0.4394,
+      "step": 3748
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.188895979578813e-06,
+      "loss": 0.3541,
+      "step": 3752
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.185705169112955e-06,
+      "loss": 0.3837,
+      "step": 3756
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.1825143586470964e-06,
+      "loss": 0.2765,
+      "step": 3760
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1793235481812383e-06,
+      "loss": 0.3349,
+      "step": 3764
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1761327377153797e-06,
+      "loss": 0.3141,
+      "step": 3768
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1729419272495216e-06,
+      "loss": 0.3836,
+      "step": 3772
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.169751116783663e-06,
+      "loss": 0.417,
+      "step": 3776
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.1665603063178046e-06,
+      "loss": 0.339,
+      "step": 3780
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1633694958519465e-06,
+      "loss": 0.4287,
+      "step": 3784
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1601786853860884e-06,
+      "loss": 0.3423,
+      "step": 3788
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.15698787492023e-06,
+      "loss": 0.3367,
+      "step": 3792
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.1537970644543717e-06,
+      "loss": 0.2519,
+      "step": 3796
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.150606253988513e-06,
+      "loss": 0.3884,
+      "step": 3800
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.147415443522655e-06,
+      "loss": 0.2767,
+      "step": 3804
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1442246330567966e-06,
+      "loss": 0.3162,
+      "step": 3808
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1410338225909385e-06,
+      "loss": 0.3722,
+      "step": 3812
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.13784301212508e-06,
+      "loss": 0.462,
+      "step": 3816
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.134652201659222e-06,
+      "loss": 0.4508,
+      "step": 3820
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.1314613911933633e-06,
+      "loss": 0.309,
+      "step": 3824
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.128270580727505e-06,
+      "loss": 0.4566,
+      "step": 3828
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1250797702616467e-06,
+      "loss": 0.3216,
+      "step": 3832
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1218889597957886e-06,
+      "loss": 0.4669,
+      "step": 3836
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.11869814932993e-06,
+      "loss": 0.4764,
+      "step": 3840
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.115507338864072e-06,
+      "loss": 0.3011,
+      "step": 3844
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.1123165283982134e-06,
+      "loss": 0.3308,
+      "step": 3848
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.109125717932355e-06,
+      "loss": 0.4038,
+      "step": 3852
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1059349074664967e-06,
+      "loss": 0.2768,
+      "step": 3856
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.1027440970006382e-06,
+      "loss": 0.374,
+      "step": 3860
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.09955328653478e-06,
+      "loss": 0.3393,
+      "step": 3864
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.0963624760689216e-06,
+      "loss": 0.3846,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0931716656030635e-06,
+      "loss": 0.308,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.089980855137205e-06,
+      "loss": 0.4816,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.086790044671347e-06,
+      "loss": 0.2121,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0835992342054883e-06,
+      "loss": 0.3698,
+      "step": 3884
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.0804084237396298e-06,
+      "loss": 0.3615,
+      "step": 3888
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0772176132737717e-06,
+      "loss": 0.2294,
+      "step": 3892
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.074026802807913e-06,
+      "loss": 0.2515,
+      "step": 3896
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.070835992342055e-06,
+      "loss": 0.3559,
+      "step": 3900
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0676451818761965e-06,
+      "loss": 0.4243,
+      "step": 3904
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.0644543714103384e-06,
+      "loss": 0.3622,
+      "step": 3908
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.06126356094448e-06,
+      "loss": 0.5588,
+      "step": 3912
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0580727504786218e-06,
+      "loss": 0.2169,
+      "step": 3916
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0548819400127632e-06,
+      "loss": 0.4732,
+      "step": 3920
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.051691129546905e-06,
+      "loss": 0.2331,
+      "step": 3924
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0485003190810466e-06,
+      "loss": 0.3388,
+      "step": 3928
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.0453095086151885e-06,
+      "loss": 0.4545,
+      "step": 3932
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.04211869814933e-06,
+      "loss": 0.3886,
+      "step": 3936
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.038927887683472e-06,
+      "loss": 0.2233,
+      "step": 3940
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0357370772176138e-06,
+      "loss": 0.3658,
+      "step": 3944
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0325462667517552e-06,
+      "loss": 0.3229,
+      "step": 3948
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.029355456285897e-06,
+      "loss": 0.1759,
+      "step": 3952
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.0261646458200386e-06,
+      "loss": 0.3737,
+      "step": 3956
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.02297383535418e-06,
+      "loss": 0.3362,
+      "step": 3960
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.019783024888322e-06,
+      "loss": 0.2873,
+      "step": 3964
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0165922144224634e-06,
+      "loss": 0.3454,
+      "step": 3968
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0134014039566053e-06,
+      "loss": 0.3428,
+      "step": 3972
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.0102105934907468e-06,
+      "loss": 0.4089,
+      "step": 3976
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.0070197830248887e-06,
+      "loss": 0.3472,
+      "step": 3980
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.00382897255903e-06,
+      "loss": 0.2868,
+      "step": 3984
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.000638162093172e-06,
+      "loss": 0.3088,
+      "step": 3988
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.9974473516273135e-06,
+      "loss": 0.2471,
+      "step": 3992
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.994256541161455e-06,
+      "loss": 0.2816,
+      "step": 3996
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.991065730695597e-06,
+      "loss": 0.3135,
+      "step": 4000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9878749202297383e-06,
+      "loss": 0.379,
+      "step": 4004
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9846841097638802e-06,
+      "loss": 0.5225,
+      "step": 4008
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9814932992980217e-06,
+      "loss": 0.3229,
+      "step": 4012
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.9783024888321636e-06,
+      "loss": 0.3573,
+      "step": 4016
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.975111678366305e-06,
+      "loss": 0.2219,
+      "step": 4020
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.971920867900447e-06,
+      "loss": 0.2133,
+      "step": 4024
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9687300574345884e-06,
+      "loss": 0.4303,
+      "step": 4028
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.96553924696873e-06,
+      "loss": 0.4735,
+      "step": 4032
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.9631461391193364e-06,
+      "loss": 0.3223,
+      "step": 4036
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.959955328653478e-06,
+      "loss": 0.3124,
+      "step": 4040
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.95676451818762e-06,
+      "loss": 0.4547,
+      "step": 4044
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9535737077217613e-06,
+      "loss": 0.3089,
+      "step": 4048
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.950382897255903e-06,
+      "loss": 0.344,
+      "step": 4052
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.947192086790045e-06,
+      "loss": 0.1488,
+      "step": 4056
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9440012763241865e-06,
+      "loss": 0.4715,
+      "step": 4060
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.9408104658583284e-06,
+      "loss": 0.2866,
+      "step": 4064
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.93761965539247e-06,
+      "loss": 0.3207,
+      "step": 4068
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9344288449266118e-06,
+      "loss": 0.3532,
+      "step": 4072
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9312380344607532e-06,
+      "loss": 0.3416,
+      "step": 4076
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.928047223994895e-06,
+      "loss": 0.6239,
+      "step": 4080
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.9248564135290366e-06,
+      "loss": 0.1806,
+      "step": 4084
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9216656030631785e-06,
+      "loss": 0.3065,
+      "step": 4088
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.91847479259732e-06,
+      "loss": 0.2393,
+      "step": 4092
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9152839821314614e-06,
+      "loss": 0.4581,
+      "step": 4096
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.9120931716656033e-06,
+      "loss": 0.2407,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.908902361199745e-06,
+      "loss": 0.3328,
+      "step": 4104
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.9057115507338867e-06,
+      "loss": 0.2898,
+      "step": 4108
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.9025207402680282e-06,
+      "loss": 0.5888,
+      "step": 4112
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.89932992980217e-06,
+      "loss": 0.3909,
+      "step": 4116
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8961391193363115e-06,
+      "loss": 0.2613,
+      "step": 4120
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.8929483088704534e-06,
+      "loss": 0.2594,
+      "step": 4124
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.889757498404595e-06,
+      "loss": 0.3601,
+      "step": 4128
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8865666879387366e-06,
+      "loss": 0.1791,
+      "step": 4132
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8833758774728783e-06,
+      "loss": 0.3714,
+      "step": 4136
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.88018506700702e-06,
+      "loss": 0.3601,
+      "step": 4140
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8769942565411616e-06,
+      "loss": 0.4697,
+      "step": 4144
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.8738034460753033e-06,
+      "loss": 0.4277,
+      "step": 4148
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.870612635609445e-06,
+      "loss": 0.4183,
+      "step": 4152
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8674218251435867e-06,
+      "loss": 0.2764,
+      "step": 4156
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8642310146777281e-06,
+      "loss": 0.3209,
+      "step": 4160
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.86104020421187e-06,
+      "loss": 0.328,
+      "step": 4164
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.8578493937460115e-06,
+      "loss": 0.3673,
+      "step": 4168
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8546585832801534e-06,
+      "loss": 0.2856,
+      "step": 4172
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8514677728142949e-06,
+      "loss": 0.4248,
+      "step": 4176
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8482769623484368e-06,
+      "loss": 0.419,
+      "step": 4180
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8450861518825782e-06,
+      "loss": 0.3315,
+      "step": 4184
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8418953414167201e-06,
+      "loss": 0.3508,
+      "step": 4188
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.8387045309508616e-06,
+      "loss": 0.2016,
+      "step": 4192
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8355137204850033e-06,
+      "loss": 0.2352,
+      "step": 4196
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.832322910019145e-06,
+      "loss": 0.4638,
+      "step": 4200
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8291320995532866e-06,
+      "loss": 0.4352,
+      "step": 4204
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.8259412890874283e-06,
+      "loss": 0.4832,
+      "step": 4208
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.82275047862157e-06,
+      "loss": 0.295,
+      "step": 4212
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8195596681557117e-06,
+      "loss": 0.3176,
+      "step": 4216
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8163688576898534e-06,
+      "loss": 0.0922,
+      "step": 4220
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.813178047223995e-06,
+      "loss": 0.2375,
+      "step": 4224
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8099872367581367e-06,
+      "loss": 0.3374,
+      "step": 4228
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.8067964262922782e-06,
+      "loss": 0.2551,
+      "step": 4232
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.80360561582642e-06,
+      "loss": 0.3228,
+      "step": 4236
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.8004148053605616e-06,
+      "loss": 0.3102,
+      "step": 4240
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7972239948947035e-06,
+      "loss": 0.2471,
+      "step": 4244
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.794033184428845e-06,
+      "loss": 0.285,
+      "step": 4248
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7908423739629868e-06,
+      "loss": 0.3468,
+      "step": 4252
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.7876515634971283e-06,
+      "loss": 0.2877,
+      "step": 4256
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7844607530312702e-06,
+      "loss": 0.4362,
+      "step": 4260
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7812699425654117e-06,
+      "loss": 0.1789,
+      "step": 4264
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7780791320995533e-06,
+      "loss": 0.3056,
+      "step": 4268
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.774888321633695e-06,
+      "loss": 0.478,
+      "step": 4272
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7716975111678367e-06,
+      "loss": 0.3405,
+      "step": 4276
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7685067007019786e-06,
+      "loss": 0.2038,
+      "step": 4280
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.76531589023612e-06,
+      "loss": 0.2301,
+      "step": 4284
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.762125079770262e-06,
+      "loss": 0.3283,
+      "step": 4288
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7589342693044034e-06,
+      "loss": 0.1711,
+      "step": 4292
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7557434588385453e-06,
+      "loss": 0.241,
+      "step": 4296
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7525526483726868e-06,
+      "loss": 0.2408,
+      "step": 4300
+    }
+  ],
+  "logging_steps": 4,
+  "max_steps": 6468,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "total_flos": 43550404509696.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}