diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8461 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9981341625944022,
+  "global_step": 4218,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5294117647058825e-06,
+      "loss": 3.8713,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.058823529411765e-06,
+      "loss": 3.4478,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 3.2108,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.411764705882353e-05,
+      "loss": 2.9405,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 2.6643,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1176470588235296e-05,
+      "loss": 2.4552,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4705882352941178e-05,
+      "loss": 2.4638,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.823529411764706e-05,
+      "loss": 2.182,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.176470588235294e-05,
+      "loss": 2.0387,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 1.9353,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882352941176471e-05,
+      "loss": 1.9432,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 1.8205,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.588235294117647e-05,
+      "loss": 1.7445,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9411764705882355e-05,
+      "loss": 1.638,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.294117647058824e-05,
+      "loss": 1.6144,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.647058823529412e-05,
+      "loss": 1.597,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6e-05,
+      "loss": 1.5933,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.352941176470588e-05,
+      "loss": 1.6111,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.705882352941176e-05,
+      "loss": 1.5282,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.058823529411765e-05,
+      "loss": 1.499,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.411764705882354e-05,
+      "loss": 1.4422,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 7.764705882352942e-05,
+      "loss": 1.4665,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.11764705882353e-05,
+      "loss": 1.4113,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.470588235294118e-05,
+      "loss": 1.3772,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 8.823529411764706e-05,
+      "loss": 1.3554,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.176470588235295e-05,
+      "loss": 1.3363,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.529411764705883e-05,
+      "loss": 1.3221,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.882352941176471e-05,
+      "loss": 1.231,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.999994222115565e-05,
+      "loss": 1.2604,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.999963888258794e-05,
+      "loss": 1.1821,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.999907554116106e-05,
+      "loss": 1.2191,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.999825219980447e-05,
+      "loss": 1.1625,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.999716886279962e-05,
+      "loss": 1.1489,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.999582553577995e-05,
+      "loss": 1.1297,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.999422222573088e-05,
+      "loss": 1.1592,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.999235894098979e-05,
+      "loss": 1.1427,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.999023569124595e-05,
+      "loss": 1.0463,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.998785248754044e-05,
+      "loss": 1.0977,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.998520934226615e-05,
+      "loss": 1.1207,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.99823062691677e-05,
+      "loss": 1.0715,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.99791432833413e-05,
+      "loss": 1.0612,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.997572040123483e-05,
+      "loss": 0.9976,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.997203764064758e-05,
+      "loss": 0.9853,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.996809502073023e-05,
+      "loss": 0.9514,
+      "step": 132
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.996389256198482e-05,
+      "loss": 0.9948,
+      "step": 135
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.995943028626451e-05,
+      "loss": 0.9322,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.995470821677356e-05,
+      "loss": 0.9208,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.994972637806718e-05,
+      "loss": 0.893,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.994448479605138e-05,
+      "loss": 0.8938,
+      "step": 147
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.993898349798294e-05,
+      "loss": 0.869,
+      "step": 150
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.993322251246907e-05,
+      "loss": 0.8657,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.992720186946747e-05,
+      "loss": 0.9,
+      "step": 156
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.992092160028605e-05,
+      "loss": 0.8594,
+      "step": 159
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.991438173758282e-05,
+      "loss": 0.9043,
+      "step": 162
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.990758231536565e-05,
+      "loss": 0.8273,
+      "step": 165
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.990052336899223e-05,
+      "loss": 0.8295,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.989320493516969e-05,
+      "loss": 0.818,
+      "step": 171
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.988562705195463e-05,
+      "loss": 0.8361,
+      "step": 174
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.987778975875273e-05,
+      "loss": 0.7643,
+      "step": 177
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.986969309631866e-05,
+      "loss": 0.8122,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.986133710675585e-05,
+      "loss": 0.7909,
+      "step": 183
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.985272183351621e-05,
+      "loss": 0.7521,
+      "step": 186
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.984384732140001e-05,
+      "loss": 0.7736,
+      "step": 189
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.983471361655554e-05,
+      "loss": 0.757,
+      "step": 192
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.982532076647893e-05,
+      "loss": 0.7525,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.981566882001388e-05,
+      "loss": 0.7251,
+      "step": 198
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.980575782735142e-05,
+      "loss": 0.6862,
+      "step": 201
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.979558784002968e-05,
+      "loss": 0.7208,
+      "step": 204
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.978515891093355e-05,
+      "loss": 0.7133,
+      "step": 207
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.977447109429443e-05,
+      "loss": 0.7081,
+      "step": 210
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.976352444568999e-05,
+      "loss": 0.6588,
+      "step": 213
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.975231902204385e-05,
+      "loss": 0.6957,
+      "step": 216
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.974085488162525e-05,
+      "loss": 0.7015,
+      "step": 219
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.972913208404884e-05,
+      "loss": 0.6278,
+      "step": 222
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.971715069027426e-05,
+      "loss": 0.7023,
+      "step": 225
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.970491076260587e-05,
+      "loss": 0.6808,
+      "step": 228
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.969241236469249e-05,
+      "loss": 0.6916,
+      "step": 231
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.967965556152696e-05,
+      "loss": 0.6941,
+      "step": 234
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.966664041944586e-05,
+      "loss": 0.6286,
+      "step": 237
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.965336700612917e-05,
+      "loss": 0.7053,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.963983539059989e-05,
+      "loss": 0.7111,
+      "step": 243
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.962604564322371e-05,
+      "loss": 0.6585,
+      "step": 246
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.96119978357086e-05,
+      "loss": 0.6242,
+      "step": 249
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.959769204110453e-05,
+      "loss": 0.6834,
+      "step": 252
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.958312833380296e-05,
+      "loss": 0.6638,
+      "step": 255
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.956830678953654e-05,
+      "loss": 0.6184,
+      "step": 258
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.955322748537875e-05,
+      "loss": 0.5992,
+      "step": 261
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.953789049974334e-05,
+      "loss": 0.5889,
+      "step": 264
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.952229591238411e-05,
+      "loss": 0.5633,
+      "step": 267
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.950644380439442e-05,
+      "loss": 0.6044,
+      "step": 270
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.949033425820671e-05,
+      "loss": 0.577,
+      "step": 273
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.947396735759213e-05,
+      "loss": 0.5375,
+      "step": 276
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.945734318766014e-05,
+      "loss": 0.5692,
+      "step": 279
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.9440461834858e-05,
+      "loss": 0.5442,
+      "step": 282
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.942332338697032e-05,
+      "loss": 0.5688,
+      "step": 285
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.940592793311866e-05,
+      "loss": 0.5424,
+      "step": 288
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.938827556376106e-05,
+      "loss": 0.5837,
+      "step": 291
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.937036637069147e-05,
+      "loss": 0.5909,
+      "step": 294
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.935220044703941e-05,
+      "loss": 0.5029,
+      "step": 297
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.93337778872694e-05,
+      "loss": 0.4939,
+      "step": 300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.931509878718051e-05,
+      "loss": 0.5687,
+      "step": 303
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.929616324390581e-05,
+      "loss": 0.5074,
+      "step": 306
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.927697135591194e-05,
+      "loss": 0.556,
+      "step": 309
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.92575232229985e-05,
+      "loss": 0.4956,
+      "step": 312
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.923781894629767e-05,
+      "loss": 0.5471,
+      "step": 315
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.921785862827353e-05,
+      "loss": 0.5511,
+      "step": 318
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.919764237272162e-05,
+      "loss": 0.5238,
+      "step": 321
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.917717028476839e-05,
+      "loss": 0.4945,
+      "step": 324
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.915644247087064e-05,
+      "loss": 0.5521,
+      "step": 327
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.913545903881497e-05,
+      "loss": 0.5046,
+      "step": 330
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.91142200977172e-05,
+      "loss": 0.4685,
+      "step": 333
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.909272575802187e-05,
+      "loss": 0.517,
+      "step": 336
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.907097613150153e-05,
+      "loss": 0.4779,
+      "step": 339
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.904897133125633e-05,
+      "loss": 0.4686,
+      "step": 342
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.902671147171331e-05,
+      "loss": 0.471,
+      "step": 345
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.900419666862587e-05,
+      "loss": 0.4871,
+      "step": 348
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.89814270390731e-05,
+      "loss": 0.4345,
+      "step": 351
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.895840270145921e-05,
+      "loss": 0.4674,
+      "step": 354
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.893512377551298e-05,
+      "loss": 0.4669,
+      "step": 357
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.891159038228699e-05,
+      "loss": 0.4817,
+      "step": 360
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.888780264415713e-05,
+      "loss": 0.4559,
+      "step": 363
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.88637606848219e-05,
+      "loss": 0.4319,
+      "step": 366
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.883946462930177e-05,
+      "loss": 0.4799,
+      "step": 369
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.881491460393853e-05,
+      "loss": 0.4257,
+      "step": 372
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.879011073639459e-05,
+      "loss": 0.5319,
+      "step": 375
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.876505315565248e-05,
+      "loss": 0.4366,
+      "step": 378
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.873974199201396e-05,
+      "loss": 0.468,
+      "step": 381
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.871417737709949e-05,
+      "loss": 0.4477,
+      "step": 384
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.868835944384746e-05,
+      "loss": 0.4231,
+      "step": 387
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.866228832651361e-05,
+      "loss": 0.4449,
+      "step": 390
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.86359641606702e-05,
+      "loss": 0.406,
+      "step": 393
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.860938708320539e-05,
+      "loss": 0.399,
+      "step": 396
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.858255723232249e-05,
+      "loss": 0.4787,
+      "step": 399
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.855547474753928e-05,
+      "loss": 0.4314,
+      "step": 402
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.852813976968723e-05,
+      "loss": 0.3915,
+      "step": 405
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.850055244091084e-05,
+      "loss": 0.3818,
+      "step": 408
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.847271290466682e-05,
+      "loss": 0.4128,
+      "step": 411
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.844462130572339e-05,
+      "loss": 0.4149,
+      "step": 414
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.841627779015953e-05,
+      "loss": 0.4368,
+      "step": 417
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.838768250536421e-05,
+      "loss": 0.3788,
+      "step": 420
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.835883560003561e-05,
+      "loss": 0.4315,
+      "step": 423
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.83297372241804e-05,
+      "loss": 0.4171,
+      "step": 426
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.830038752911283e-05,
+      "loss": 0.4019,
+      "step": 429
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.827078666745418e-05,
+      "loss": 0.4521,
+      "step": 432
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.824093479313165e-05,
+      "loss": 0.343,
+      "step": 435
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.821083206137788e-05,
+      "loss": 0.4224,
+      "step": 438
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.81804786287299e-05,
+      "loss": 0.4192,
+      "step": 441
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.814987465302843e-05,
+      "loss": 0.4035,
+      "step": 444
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.811902029341704e-05,
+      "loss": 0.3777,
+      "step": 447
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.808791571034135e-05,
+      "loss": 0.4148,
+      "step": 450
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.80565610655481e-05,
+      "loss": 0.3854,
+      "step": 453
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.802495652208445e-05,
+      "loss": 0.3892,
+      "step": 456
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.799310224429701e-05,
+      "loss": 0.4206,
+      "step": 459
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.796099839783103e-05,
+      "loss": 0.4363,
+      "step": 462
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.792864514962956e-05,
+      "loss": 0.3719,
+      "step": 465
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.789604266793257e-05,
+      "loss": 0.3671,
+      "step": 468
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.786319112227603e-05,
+      "loss": 0.4127,
+      "step": 471
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.783009068349109e-05,
+      "loss": 0.4179,
+      "step": 474
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.77967415237032e-05,
+      "loss": 0.371,
+      "step": 477
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.776314381633111e-05,
+      "loss": 0.3564,
+      "step": 480
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.77292977360861e-05,
+      "loss": 0.3611,
+      "step": 483
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.769520345897099e-05,
+      "loss": 0.3295,
+      "step": 486
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.766086116227926e-05,
+      "loss": 0.3418,
+      "step": 489
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.76262710245941e-05,
+      "loss": 0.3541,
+      "step": 492
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.759143322578751e-05,
+      "loss": 0.3386,
+      "step": 495
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.755634794701933e-05,
+      "loss": 0.3182,
+      "step": 498
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.752101537073634e-05,
+      "loss": 0.3904,
+      "step": 501
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.74854356806713e-05,
+      "loss": 0.3687,
+      "step": 504
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.744960906184194e-05,
+      "loss": 0.3435,
+      "step": 507
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.741353570055009e-05,
+      "loss": 0.3185,
+      "step": 510
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.737721578438064e-05,
+      "loss": 0.3605,
+      "step": 513
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.734064950220061e-05,
+      "loss": 0.3285,
+      "step": 516
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.730383704415811e-05,
+      "loss": 0.3139,
+      "step": 519
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.726677860168142e-05,
+      "loss": 0.3675,
+      "step": 522
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.722947436747796e-05,
+      "loss": 0.3471,
+      "step": 525
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.719192453553325e-05,
+      "loss": 0.3423,
+      "step": 528
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.715412930111002e-05,
+      "loss": 0.3339,
+      "step": 531
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.711608886074703e-05,
+      "loss": 0.351,
+      "step": 534
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.707780341225821e-05,
+      "loss": 0.2691,
+      "step": 537
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.703927315473149e-05,
+      "loss": 0.3315,
+      "step": 540
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.700049828852786e-05,
+      "loss": 0.3261,
+      "step": 543
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.696147901528032e-05,
+      "loss": 0.2997,
+      "step": 546
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.692221553789277e-05,
+      "loss": 0.3094,
+      "step": 549
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.6882708060539e-05,
+      "loss": 0.3018,
+      "step": 552
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.684295678866169e-05,
+      "loss": 0.2936,
+      "step": 555
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.680296192897114e-05,
+      "loss": 0.3058,
+      "step": 558
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.676272368944448e-05,
+      "loss": 0.3237,
+      "step": 561
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.672224227932433e-05,
+      "loss": 0.272,
+      "step": 564
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.668151790911786e-05,
+      "loss": 0.2933,
+      "step": 567
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.664055079059566e-05,
+      "loss": 0.3095,
+      "step": 570
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.659934113679064e-05,
+      "loss": 0.2977,
+      "step": 573
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.655788916199688e-05,
+      "loss": 0.294,
+      "step": 576
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.651619508176859e-05,
+      "loss": 0.2843,
+      "step": 579
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.647425911291894e-05,
+      "loss": 0.28,
+      "step": 582
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.643208147351893e-05,
+      "loss": 0.2794,
+      "step": 585
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.638966238289628e-05,
+      "loss": 0.3104,
+      "step": 588
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.63470020616343e-05,
+      "loss": 0.2917,
+      "step": 591
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.630410073157069e-05,
+      "loss": 0.2701,
+      "step": 594
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.626095861579643e-05,
+      "loss": 0.2661,
+      "step": 597
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.621757593865459e-05,
+      "loss": 0.2869,
+      "step": 600
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.617395292573923e-05,
+      "loss": 0.2935,
+      "step": 603
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.613008980389416e-05,
+      "loss": 0.3391,
+      "step": 606
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.608598680121172e-05,
+      "loss": 0.2654,
+      "step": 609
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.604164414703175e-05,
+      "loss": 0.2974,
+      "step": 612
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.599706207194022e-05,
+      "loss": 0.3128,
+      "step": 615
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.595224080776817e-05,
+      "loss": 0.28,
+      "step": 618
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.59071805875904e-05,
+      "loss": 0.2658,
+      "step": 621
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.586188164572435e-05,
+      "loss": 0.244,
+      "step": 624
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.581634421772878e-05,
+      "loss": 0.2884,
+      "step": 627
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.577056854040262e-05,
+      "loss": 0.2768,
+      "step": 630
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.572455485178377e-05,
+      "loss": 0.2649,
+      "step": 633
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.567830339114776e-05,
+      "loss": 0.3047,
+      "step": 636
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.563181439900655e-05,
+      "loss": 0.296,
+      "step": 639
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.558508811710731e-05,
+      "loss": 0.2854,
+      "step": 642
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.553812478843114e-05,
+      "loss": 0.2982,
+      "step": 645
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.549092465719178e-05,
+      "loss": 0.2845,
+      "step": 648
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.544348796883441e-05,
+      "loss": 0.2385,
+      "step": 651
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.539581497003429e-05,
+      "loss": 0.2782,
+      "step": 654
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.534790590869552e-05,
+      "loss": 0.2555,
+      "step": 657
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.529976103394978e-05,
+      "loss": 0.2351,
+      "step": 660
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.525138059615495e-05,
+      "loss": 0.2713,
+      "step": 663
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.520276484689392e-05,
+      "loss": 0.2828,
+      "step": 666
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.515391403897317e-05,
+      "loss": 0.2697,
+      "step": 669
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.510482842642157e-05,
+      "loss": 0.2552,
+      "step": 672
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.50555082644889e-05,
+      "loss": 0.2809,
+      "step": 675
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.50059538096447e-05,
+      "loss": 0.2809,
+      "step": 678
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.495616531957684e-05,
+      "loss": 0.2342,
+      "step": 681
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.490614305319017e-05,
+      "loss": 0.2461,
+      "step": 684
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.485588727060521e-05,
+      "loss": 0.249,
+      "step": 687
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.480539823315682e-05,
+      "loss": 0.2609,
+      "step": 690
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.475467620339275e-05,
+      "loss": 0.2744,
+      "step": 693
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.470372144507237e-05,
+      "loss": 0.2688,
+      "step": 696
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.465253422316524e-05,
+      "loss": 0.2487,
+      "step": 699
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.460111480384979e-05,
+      "loss": 0.2488,
+      "step": 702
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.454946345451183e-05,
+      "loss": 0.2727,
+      "step": 705
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.449758044374331e-05,
+      "loss": 0.2446,
+      "step": 708
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.44454660413408e-05,
+      "loss": 0.2686,
+      "step": 711
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.439312051830409e-05,
+      "loss": 0.2424,
+      "step": 714
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.434054414683489e-05,
+      "loss": 0.2853,
+      "step": 717
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.42877372003353e-05,
+      "loss": 0.2317,
+      "step": 720
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.423469995340648e-05,
+      "loss": 0.245,
+      "step": 723
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.41814326818471e-05,
+      "loss": 0.2661,
+      "step": 726
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.412793566265201e-05,
+      "loss": 0.2433,
+      "step": 729
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.407420917401082e-05,
+      "loss": 0.2316,
+      "step": 732
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.402025349530633e-05,
+      "loss": 0.2172,
+      "step": 735
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.39660689071132e-05,
+      "loss": 0.2379,
+      "step": 738
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.391165569119638e-05,
+      "loss": 0.227,
+      "step": 741
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.38570141305098e-05,
+      "loss": 0.2239,
+      "step": 744
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.380214450919474e-05,
+      "loss": 0.2413,
+      "step": 747
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.374704711257843e-05,
+      "loss": 0.2193,
+      "step": 750
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.369172222717251e-05,
+      "loss": 0.2278,
+      "step": 753
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.363617014067168e-05,
+      "loss": 0.2092,
+      "step": 756
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.358039114195203e-05,
+      "loss": 0.2193,
+      "step": 759
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.352438552106964e-05,
+      "loss": 0.2251,
+      "step": 762
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.346815356925904e-05,
+      "loss": 0.1976,
+      "step": 765
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.341169557893171e-05,
+      "loss": 0.2125,
+      "step": 768
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.335501184367455e-05,
+      "loss": 0.2191,
+      "step": 771
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.329810265824837e-05,
+      "loss": 0.2488,
+      "step": 774
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.324096831858628e-05,
+      "loss": 0.2331,
+      "step": 777
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.318360912179232e-05,
+      "loss": 0.2159,
+      "step": 780
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.312602536613968e-05,
+      "loss": 0.2258,
+      "step": 783
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.30682173510694e-05,
+      "loss": 0.2157,
+      "step": 786
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.301018537718862e-05,
+      "loss": 0.2438,
+      "step": 789
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.29519297462691e-05,
+      "loss": 0.2276,
+      "step": 792
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.289345076124562e-05,
+      "loss": 0.2056,
+      "step": 795
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.283474872621448e-05,
+      "loss": 0.2274,
+      "step": 798
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.277582394643179e-05,
+      "loss": 0.2272,
+      "step": 801
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.271667672831199e-05,
+      "loss": 0.2033,
+      "step": 804
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.265730737942625e-05,
+      "loss": 0.2068,
+      "step": 807
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.259771620850077e-05,
+      "loss": 0.2137,
+      "step": 810
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.253790352541532e-05,
+      "loss": 0.2475,
+      "step": 813
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.24778696412015e-05,
+      "loss": 0.2361,
+      "step": 816
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.241761486804119e-05,
+      "loss": 0.2405,
+      "step": 819
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.235713951926493e-05,
+      "loss": 0.2116,
+      "step": 822
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.229644390935024e-05,
+      "loss": 0.2054,
+      "step": 825
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.223552835392008e-05,
+      "loss": 0.231,
+      "step": 828
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.217439316974108e-05,
+      "loss": 0.2106,
+      "step": 831
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.2113038674722e-05,
+      "loss": 0.2189,
+      "step": 834
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.205146518791204e-05,
+      "loss": 0.2116,
+      "step": 837
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.198967302949912e-05,
+      "loss": 0.2278,
+      "step": 840
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.192766252080837e-05,
+      "loss": 0.2017,
+      "step": 843
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.18654339843003e-05,
+      "loss": 0.2019,
+      "step": 846
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.180298774356917e-05,
+      "loss": 0.1702,
+      "step": 849
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.174032412334135e-05,
+      "loss": 0.2299,
+      "step": 852
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.167744344947363e-05,
+      "loss": 0.1961,
+      "step": 855
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.161434604895146e-05,
+      "loss": 0.1934,
+      "step": 858
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.155103224988728e-05,
+      "loss": 0.1995,
+      "step": 861
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.148750238151887e-05,
+      "loss": 0.2047,
+      "step": 864
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.142375677420756e-05,
+      "loss": 0.2101,
+      "step": 867
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.135979575943653e-05,
+      "loss": 0.2227,
+      "step": 870
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.129561966980913e-05,
+      "loss": 0.1714,
+      "step": 873
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.123122883904711e-05,
+      "loss": 0.2298,
+      "step": 876
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.116662360198888e-05,
+      "loss": 0.2121,
+      "step": 879
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.11018042945878e-05,
+      "loss": 0.2061,
+      "step": 882
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.10367712539104e-05,
+      "loss": 0.1741,
+      "step": 885
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.097152481813466e-05,
+      "loss": 0.2126,
+      "step": 888
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.090606532654823e-05,
+      "loss": 0.1853,
+      "step": 891
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.084039311954667e-05,
+      "loss": 0.2013,
+      "step": 894
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.077450853863168e-05,
+      "loss": 0.1789,
+      "step": 897
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.070841192640931e-05,
+      "loss": 0.2109,
+      "step": 900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.064210362658819e-05,
+      "loss": 0.2062,
+      "step": 903
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.057558398397781e-05,
+      "loss": 0.2128,
+      "step": 906
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.050885334448658e-05,
+      "loss": 0.1825,
+      "step": 909
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.044191205512013e-05,
+      "loss": 0.2,
+      "step": 912
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.037476046397957e-05,
+      "loss": 0.2142,
+      "step": 915
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.030739892025947e-05,
+      "loss": 0.1759,
+      "step": 918
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.023982777424627e-05,
+      "loss": 0.205,
+      "step": 921
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.017204737731632e-05,
+      "loss": 0.1673,
+      "step": 924
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.010405808193412e-05,
+      "loss": 0.1765,
+      "step": 927
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.003586024165044e-05,
+      "loss": 0.1781,
+      "step": 930
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.996745421110051e-05,
+      "loss": 0.1831,
+      "step": 933
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.989884034600217e-05,
+      "loss": 0.1855,
+      "step": 936
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.983001900315402e-05,
+      "loss": 0.1725,
+      "step": 939
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.976099054043357e-05,
+      "loss": 0.1843,
+      "step": 942
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.969175531679533e-05,
+      "loss": 0.1838,
+      "step": 945
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.962231369226906e-05,
+      "loss": 0.1848,
+      "step": 948
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 8.955266602795779e-05,
+      "loss": 0.1707,
+      "step": 951
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 8.948281268603595e-05,
+      "loss": 0.1491,
+      "step": 954
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 8.941275402974754e-05,
+      "loss": 0.1794,
+      "step": 957
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 8.934249042340421e-05,
+      "loss": 0.19,
+      "step": 960
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 8.92720222323834e-05,
+      "loss": 0.1985,
+      "step": 963
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 8.920134982312633e-05,
+      "loss": 0.1989,
+      "step": 966
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 8.913047356313625e-05,
+      "loss": 0.1682,
+      "step": 969
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 8.905939382097643e-05,
+      "loss": 0.1681,
+      "step": 972
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 8.898811096626826e-05,
+      "loss": 0.1606,
+      "step": 975
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.891662536968932e-05,
+      "loss": 0.1593,
+      "step": 978
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.884493740297152e-05,
+      "loss": 0.193,
+      "step": 981
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.877304743889905e-05,
+      "loss": 0.1803,
+      "step": 984
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.870095585130653e-05,
+      "loss": 0.1911,
+      "step": 987
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.862866301507709e-05,
+      "loss": 0.181,
+      "step": 990
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.855616930614027e-05,
+      "loss": 0.1673,
+      "step": 993
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.848347510147026e-05,
+      "loss": 0.1811,
+      "step": 996
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.841058077908379e-05,
+      "loss": 0.1868,
+      "step": 999
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.833748671803826e-05,
+      "loss": 0.1542,
+      "step": 1002
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.826419329842972e-05,
+      "loss": 0.1497,
+      "step": 1005
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.819070090139086e-05,
+      "loss": 0.1908,
+      "step": 1008
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.811700990908914e-05,
+      "loss": 0.182,
+      "step": 1011
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.80431207047247e-05,
+      "loss": 0.1592,
+      "step": 1014
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 8.796903367252837e-05,
+      "loss": 0.1718,
+      "step": 1017
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.789474919775979e-05,
+      "loss": 0.1649,
+      "step": 1020
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.782026766670522e-05,
+      "loss": 0.1837,
+      "step": 1023
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.774558946667569e-05,
+      "loss": 0.1811,
+      "step": 1026
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.767071498600492e-05,
+      "loss": 0.1839,
+      "step": 1029
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 8.759564461404731e-05,
+      "loss": 0.185,
+      "step": 1032
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 8.75203787411759e-05,
+      "loss": 0.1817,
+      "step": 1035
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 8.744491775878036e-05,
+      "loss": 0.1937,
+      "step": 1038
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 8.736926205926495e-05,
+      "loss": 0.1834,
+      "step": 1041
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 8.729341203604646e-05,
+      "loss": 0.1789,
+      "step": 1044
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 8.72173680835522e-05,
+      "loss": 0.1389,
+      "step": 1047
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.714113059721796e-05,
+      "loss": 0.1955,
+      "step": 1050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.706469997348585e-05,
+      "loss": 0.154,
+      "step": 1053
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.698807660980237e-05,
+      "loss": 0.1485,
+      "step": 1056
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.691126090461628e-05,
+      "loss": 0.1723,
+      "step": 1059
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 8.683425325737651e-05,
+      "loss": 0.185,
+      "step": 1062
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.675705406853014e-05,
+      "loss": 0.1656,
+      "step": 1065
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.667966373952023e-05,
+      "loss": 0.1458,
+      "step": 1068
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.660208267278387e-05,
+      "loss": 0.1592,
+      "step": 1071
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 8.652431127174994e-05,
+      "loss": 0.177,
+      "step": 1074
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.644634994083707e-05,
+      "loss": 0.1663,
+      "step": 1077
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.636819908545157e-05,
+      "loss": 0.1685,
+      "step": 1080
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.628985911198533e-05,
+      "loss": 0.1425,
+      "step": 1083
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.62113304278136e-05,
+      "loss": 0.1742,
+      "step": 1086
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 8.6132613441293e-05,
+      "loss": 0.1788,
+      "step": 1089
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.60537085617593e-05,
+      "loss": 0.1601,
+      "step": 1092
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.597461619952534e-05,
+      "loss": 0.1678,
+      "step": 1095
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.589533676587893e-05,
+      "loss": 0.1641,
+      "step": 1098
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.58158706730806e-05,
+      "loss": 0.1818,
+      "step": 1101
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 8.573621833436158e-05,
+      "loss": 0.166,
+      "step": 1104
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.565638016392158e-05,
+      "loss": 0.156,
+      "step": 1107
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.557635657692665e-05,
+      "loss": 0.1745,
+      "step": 1110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.5496147989507e-05,
+      "loss": 0.1614,
+      "step": 1113
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 8.541575481875493e-05,
+      "loss": 0.1691,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.53351774827225e-05,
+      "loss": 0.1511,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.525441640041951e-05,
+      "loss": 0.1781,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.517347199181124e-05,
+      "loss": 0.1433,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.509234467781627e-05,
+      "loss": 0.1607,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 8.501103488030435e-05,
+      "loss": 0.1533,
+      "step": 1131
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.49295430220941e-05,
+      "loss": 0.145,
+      "step": 1134
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.484786952695091e-05,
+      "loss": 0.1526,
+      "step": 1137
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.476601481958471e-05,
+      "loss": 0.1568,
+      "step": 1140
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.46839793256477e-05,
+      "loss": 0.1585,
+      "step": 1143
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 8.460176347173224e-05,
+      "loss": 0.1598,
+      "step": 1146
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.451936768536858e-05,
+      "loss": 0.1586,
+      "step": 1149
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.443679239502258e-05,
+      "loss": 0.1358,
+      "step": 1152
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.435403803009359e-05,
+      "loss": 0.1371,
+      "step": 1155
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 8.427110502091214e-05,
+      "loss": 0.1559,
+      "step": 1158
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.418799379873775e-05,
+      "loss": 0.1457,
+      "step": 1161
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.410470479575663e-05,
+      "loss": 0.1481,
+      "step": 1164
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.40212384450795e-05,
+      "loss": 0.1593,
+      "step": 1167
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.39375951807393e-05,
+      "loss": 0.1423,
+      "step": 1170
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 8.385377543768888e-05,
+      "loss": 0.1408,
+      "step": 1173
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.376977965179891e-05,
+      "loss": 0.1386,
+      "step": 1176
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.36856082598554e-05,
+      "loss": 0.1524,
+      "step": 1179
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.360126169955756e-05,
+      "loss": 0.1497,
+      "step": 1182
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.351674040951551e-05,
+      "loss": 0.1568,
+      "step": 1185
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 8.343204482924796e-05,
+      "loss": 0.1583,
+      "step": 1188
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.334717539917994e-05,
+      "loss": 0.1399,
+      "step": 1191
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.326213256064053e-05,
+      "loss": 0.1484,
+      "step": 1194
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.317691675586057e-05,
+      "loss": 0.1509,
+      "step": 1197
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.30915284279703e-05,
+      "loss": 0.1418,
+      "step": 1200
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.30059680209971e-05,
+      "loss": 0.139,
+      "step": 1203
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.292023597986323e-05,
+      "loss": 0.1756,
+      "step": 1206
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.283433275038339e-05,
+      "loss": 0.136,
+      "step": 1209
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.27482587792625e-05,
+      "loss": 0.1163,
+      "step": 1212
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.266201451409343e-05,
+      "loss": 0.1489,
+      "step": 1215
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.257560040335449e-05,
+      "loss": 0.1536,
+      "step": 1218
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.248901689640724e-05,
+      "loss": 0.1259,
+      "step": 1221
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.240226444349411e-05,
+      "loss": 0.1449,
+      "step": 1224
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.231534349573609e-05,
+      "loss": 0.1492,
+      "step": 1227
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.222825450513035e-05,
+      "loss": 0.139,
+      "step": 1230
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.214099792454787e-05,
+      "loss": 0.14,
+      "step": 1233
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.205357420773115e-05,
+      "loss": 0.1426,
+      "step": 1236
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.19659838092918e-05,
+      "loss": 0.1341,
+      "step": 1239
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.187822718470818e-05,
+      "loss": 0.141,
+      "step": 1242
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.179030479032305e-05,
+      "loss": 0.1552,
+      "step": 1245
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.170221708334122e-05,
+      "loss": 0.1429,
+      "step": 1248
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.161396452182709e-05,
+      "loss": 0.1202,
+      "step": 1251
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.152554756470234e-05,
+      "loss": 0.1411,
+      "step": 1254
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.143696667174352e-05,
+      "loss": 0.1476,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.134822230357969e-05,
+      "loss": 0.1551,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.125931492168992e-05,
+      "loss": 0.1464,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.117024498840108e-05,
+      "loss": 0.1473,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.108101296688523e-05,
+      "loss": 0.1311,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.099161932115733e-05,
+      "loss": 0.1314,
+      "step": 1272
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.090206451607283e-05,
+      "loss": 0.1427,
+      "step": 1275
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.081234901732517e-05,
+      "loss": 0.1288,
+      "step": 1278
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.072247329144346e-05,
+      "loss": 0.1377,
+      "step": 1281
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.063243780579002e-05,
+      "loss": 0.1492,
+      "step": 1284
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.054224302855787e-05,
+      "loss": 0.1519,
+      "step": 1287
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.045188942876838e-05,
+      "loss": 0.1336,
+      "step": 1290
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.036137747626886e-05,
+      "loss": 0.1249,
+      "step": 1293
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.027070764173002e-05,
+      "loss": 0.1313,
+      "step": 1296
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.017988039664361e-05,
+      "loss": 0.146,
+      "step": 1299
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.00888962133199e-05,
+      "loss": 0.131,
+      "step": 1302
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 7.999775556488526e-05,
+      "loss": 0.1507,
+      "step": 1305
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 7.99064589252797e-05,
+      "loss": 0.1328,
+      "step": 1308
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 7.981500676925438e-05,
+      "loss": 0.1255,
+      "step": 1311
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 7.97233995723692e-05,
+      "loss": 0.1426,
+      "step": 1314
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.963163781099026e-05,
+      "loss": 0.1363,
+      "step": 1317
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.953972196228738e-05,
+      "loss": 0.1436,
+      "step": 1320
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.944765250423173e-05,
+      "loss": 0.1356,
+      "step": 1323
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.935542991559319e-05,
+      "loss": 0.1378,
+      "step": 1326
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 7.926305467593793e-05,
+      "loss": 0.1228,
+      "step": 1329
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.917052726562599e-05,
+      "loss": 0.1234,
+      "step": 1332
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.907784816580868e-05,
+      "loss": 0.1369,
+      "step": 1335
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.898501785842605e-05,
+      "loss": 0.1354,
+      "step": 1338
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 7.889203682620452e-05,
+      "loss": 0.1371,
+      "step": 1341
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.879890555265428e-05,
+      "loss": 0.1358,
+      "step": 1344
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.870562452206674e-05,
+      "loss": 0.1268,
+      "step": 1347
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.861219421951211e-05,
+      "loss": 0.1345,
+      "step": 1350
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.851861513083685e-05,
+      "loss": 0.1428,
+      "step": 1353
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.842488774266106e-05,
+      "loss": 0.1241,
+      "step": 1356
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.833101254237601e-05,
+      "loss": 0.1252,
+      "step": 1359
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.823699001814167e-05,
+      "loss": 0.1455,
+      "step": 1362
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.814282065888408e-05,
+      "loss": 0.1153,
+      "step": 1365
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.804850495429281e-05,
+      "loss": 0.136,
+      "step": 1368
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.795404339481847e-05,
+      "loss": 0.1347,
+      "step": 1371
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.78594364716701e-05,
+      "loss": 0.1175,
+      "step": 1374
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.776468467681268e-05,
+      "loss": 0.1349,
+      "step": 1377
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.766978850296452e-05,
+      "loss": 0.1374,
+      "step": 1380
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 7.75747484435947e-05,
+      "loss": 0.1288,
+      "step": 1383
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.747956499292057e-05,
+      "loss": 0.1127,
+      "step": 1386
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.738423864590506e-05,
+      "loss": 0.1297,
+      "step": 1389
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.728876989825418e-05,
+      "loss": 0.1381,
+      "step": 1392
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.719315924641452e-05,
+      "loss": 0.1243,
+      "step": 1395
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.70974071875705e-05,
+      "loss": 0.1247,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.700151421964187e-05,
+      "loss": 0.1423,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.690548084128122e-05,
+      "loss": 0.1316,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.680930755187115e-05,
+      "loss": 0.1207,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.671299485152192e-05,
+      "loss": 0.1035,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.661654324106865e-05,
+      "loss": 0.1116,
+      "step": 1413
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.651995322206888e-05,
+      "loss": 0.1145,
+      "step": 1416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.642322529679988e-05,
+      "loss": 0.1167,
+      "step": 1419
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.632635996825598e-05,
+      "loss": 0.1166,
+      "step": 1422
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.622935774014609e-05,
+      "loss": 0.1098,
+      "step": 1425
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.6132219116891e-05,
+      "loss": 0.0963,
+      "step": 1428
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.603494460362072e-05,
+      "loss": 0.0987,
+      "step": 1431
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.593753470617196e-05,
+      "loss": 0.1057,
+      "step": 1434
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.58399899310854e-05,
+      "loss": 0.1,
+      "step": 1437
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.574231078560314e-05,
+      "loss": 0.1062,
+      "step": 1440
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.564449777766595e-05,
+      "loss": 0.1115,
+      "step": 1443
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.554655141591077e-05,
+      "loss": 0.1204,
+      "step": 1446
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.544847220966795e-05,
+      "loss": 0.1065,
+      "step": 1449
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.535026066895864e-05,
+      "loss": 0.0947,
+      "step": 1452
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.525191730449219e-05,
+      "loss": 0.0989,
+      "step": 1455
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.515344262766338e-05,
+      "loss": 0.1073,
+      "step": 1458
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.505483715054988e-05,
+      "loss": 0.1014,
+      "step": 1461
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.495610138590949e-05,
+      "loss": 0.1014,
+      "step": 1464
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.485723584717757e-05,
+      "loss": 0.1033,
+      "step": 1467
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.475824104846427e-05,
+      "loss": 0.1299,
+      "step": 1470
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.465911750455192e-05,
+      "loss": 0.1021,
+      "step": 1473
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.455986573089234e-05,
+      "loss": 0.1116,
+      "step": 1476
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.446048624360417e-05,
+      "loss": 0.1016,
+      "step": 1479
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 7.436097955947014e-05,
+      "loss": 0.1017,
+      "step": 1482
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 7.426134619593443e-05,
+      "loss": 0.1117,
+      "step": 1485
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 7.416158667110002e-05,
+      "loss": 0.1176,
+      "step": 1488
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 7.406170150372583e-05,
+      "loss": 0.0998,
+      "step": 1491
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 7.396169121322422e-05,
+      "loss": 0.0899,
+      "step": 1494
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 7.386155631965818e-05,
+      "loss": 0.1014,
+      "step": 1497
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.376129734373862e-05,
+      "loss": 0.1114,
+      "step": 1500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.366091480682173e-05,
+      "loss": 0.1049,
+      "step": 1503
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.356040923090624e-05,
+      "loss": 0.1066,
+      "step": 1506
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.345978113863063e-05,
+      "loss": 0.1046,
+      "step": 1509
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 7.33590310532705e-05,
+      "loss": 0.0984,
+      "step": 1512
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 7.325815949873587e-05,
+      "loss": 0.1061,
+      "step": 1515
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 7.315716699956837e-05,
+      "loss": 0.1059,
+      "step": 1518
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 7.305605408093858e-05,
+      "loss": 0.1102,
+      "step": 1521
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 7.295482126864323e-05,
+      "loss": 0.1044,
+      "step": 1524
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 7.285346908910254e-05,
+      "loss": 0.1026,
+      "step": 1527
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 7.275199806935745e-05,
+      "loss": 0.1199,
+      "step": 1530
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 7.265040873706687e-05,
+      "loss": 0.1055,
+      "step": 1533
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 7.254870162050498e-05,
+      "loss": 0.106,
+      "step": 1536
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 7.24468772485584e-05,
+      "loss": 0.1055,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 7.23449361507235e-05,
+      "loss": 0.1,
+      "step": 1542
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 7.224287885710366e-05,
+      "loss": 0.1091,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 7.214070589840648e-05,
+      "loss": 0.1061,
+      "step": 1548
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 7.203841780594101e-05,
+      "loss": 0.0959,
+      "step": 1551
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 7.193601511161505e-05,
+      "loss": 0.1197,
+      "step": 1554
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 7.183349834793229e-05,
+      "loss": 0.0991,
+      "step": 1557
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 7.173086804798961e-05,
+      "loss": 0.1019,
+      "step": 1560
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 7.16281247454743e-05,
+      "loss": 0.1148,
+      "step": 1563
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 7.152526897466126e-05,
+      "loss": 0.1151,
+      "step": 1566
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 7.142230127041025e-05,
+      "loss": 0.1098,
+      "step": 1569
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 7.131922216816311e-05,
+      "loss": 0.1049,
+      "step": 1572
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 7.121603220394089e-05,
+      "loss": 0.1026,
+      "step": 1575
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 7.111273191434122e-05,
+      "loss": 0.1031,
+      "step": 1578
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 7.100932183653537e-05,
+      "loss": 0.0932,
+      "step": 1581
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.090580250826554e-05,
+      "loss": 0.0985,
+      "step": 1584
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.080217446784206e-05,
+      "loss": 0.0979,
+      "step": 1587
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.069843825414056e-05,
+      "loss": 0.1042,
+      "step": 1590
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.059459440659917e-05,
+      "loss": 0.0973,
+      "step": 1593
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 7.049064346521573e-05,
+      "loss": 0.0944,
+      "step": 1596
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.0386585970545e-05,
+      "loss": 0.1044,
+      "step": 1599
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.028242246369577e-05,
+      "loss": 0.1143,
+      "step": 1602
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.01781534863282e-05,
+      "loss": 0.0982,
+      "step": 1605
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 7.007377958065083e-05,
+      "loss": 0.1065,
+      "step": 1608
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.996930128941783e-05,
+      "loss": 0.0932,
+      "step": 1611
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.986471915592625e-05,
+      "loss": 0.1032,
+      "step": 1614
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.976003372401307e-05,
+      "loss": 0.0992,
+      "step": 1617
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.965524553805245e-05,
+      "loss": 0.0939,
+      "step": 1620
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.955035514295292e-05,
+      "loss": 0.0915,
+      "step": 1623
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.944536308415444e-05,
+      "loss": 0.1157,
+      "step": 1626
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.934026990762568e-05,
+      "loss": 0.1028,
+      "step": 1629
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.923507615986109e-05,
+      "loss": 0.12,
+      "step": 1632
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.912978238787813e-05,
+      "loss": 0.1102,
+      "step": 1635
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 6.90243891392144e-05,
+      "loss": 0.0964,
+      "step": 1638
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.891889696192481e-05,
+      "loss": 0.0948,
+      "step": 1641
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.881330640457862e-05,
+      "loss": 0.1052,
+      "step": 1644
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.870761801625676e-05,
+      "loss": 0.1002,
+      "step": 1647
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.860183234654887e-05,
+      "loss": 0.1059,
+      "step": 1650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 6.84959499455504e-05,
+      "loss": 0.0844,
+      "step": 1653
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.838997136385996e-05,
+      "loss": 0.0964,
+      "step": 1656
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.828389715257614e-05,
+      "loss": 0.0959,
+      "step": 1659
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.81777278632949e-05,
+      "loss": 0.0976,
+      "step": 1662
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 6.807146404810664e-05,
+      "loss": 0.0992,
+      "step": 1665
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.796510625959321e-05,
+      "loss": 0.097,
+      "step": 1668
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.785865505082524e-05,
+      "loss": 0.1005,
+      "step": 1671
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.775211097535908e-05,
+      "loss": 0.1216,
+      "step": 1674
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.764547458723401e-05,
+      "loss": 0.1074,
+      "step": 1677
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 6.753874644096935e-05,
+      "loss": 0.0992,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.743192709156159e-05,
+      "loss": 0.1034,
+      "step": 1683
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.732501709448144e-05,
+      "loss": 0.0804,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.721801700567103e-05,
+      "loss": 0.1002,
+      "step": 1689
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.711092738154097e-05,
+      "loss": 0.0954,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 6.700374877896743e-05,
+      "loss": 0.0903,
+      "step": 1695
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.689648175528931e-05,
+      "loss": 0.0922,
+      "step": 1698
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.678912686830528e-05,
+      "loss": 0.0943,
+      "step": 1701
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.668168467627094e-05,
+      "loss": 0.0858,
+      "step": 1704
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 6.657415573789588e-05,
+      "loss": 0.1088,
+      "step": 1707
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.646654061234072e-05,
+      "loss": 0.0955,
+      "step": 1710
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.635883985921435e-05,
+      "loss": 0.1152,
+      "step": 1713
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.625105403857089e-05,
+      "loss": 0.0967,
+      "step": 1716
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.614318371090678e-05,
+      "loss": 0.0926,
+      "step": 1719
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 6.603522943715798e-05,
+      "loss": 0.0912,
+      "step": 1722
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.592719177869693e-05,
+      "loss": 0.1041,
+      "step": 1725
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.581907129732971e-05,
+      "loss": 0.1026,
+      "step": 1728
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.571086855529303e-05,
+      "loss": 0.0923,
+      "step": 1731
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.560258411525139e-05,
+      "loss": 0.09,
+      "step": 1734
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.54942185402942e-05,
+      "loss": 0.0954,
+      "step": 1737
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.538577239393269e-05,
+      "loss": 0.0785,
+      "step": 1740
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.527724624009706e-05,
+      "loss": 0.0986,
+      "step": 1743
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.516864064313362e-05,
+      "loss": 0.0932,
+      "step": 1746
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.505995616780176e-05,
+      "loss": 0.0914,
+      "step": 1749
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.495119337927103e-05,
+      "loss": 0.0889,
+      "step": 1752
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.484235284311828e-05,
+      "loss": 0.0922,
+      "step": 1755
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.47334351253246e-05,
+      "loss": 0.0834,
+      "step": 1758
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.462444079227239e-05,
+      "loss": 0.0889,
+      "step": 1761
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 6.451537041074255e-05,
+      "loss": 0.0919,
+      "step": 1764
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.440622454791139e-05,
+      "loss": 0.0823,
+      "step": 1767
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.429700377134773e-05,
+      "loss": 0.0932,
+      "step": 1770
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.418770864900997e-05,
+      "loss": 0.0819,
+      "step": 1773
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.40783397492431e-05,
+      "loss": 0.0768,
+      "step": 1776
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 6.396889764077575e-05,
+      "loss": 0.0883,
+      "step": 1779
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 6.385938289271725e-05,
+      "loss": 0.1014,
+      "step": 1782
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 6.374979607455465e-05,
+      "loss": 0.0918,
+      "step": 1785
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 6.364013775614984e-05,
+      "loss": 0.0711,
+      "step": 1788
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 6.353040850773644e-05,
+      "loss": 0.0918,
+      "step": 1791
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.342060889991691e-05,
+      "loss": 0.093,
+      "step": 1794
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.331073950365963e-05,
+      "loss": 0.1003,
+      "step": 1797
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.320080089029589e-05,
+      "loss": 0.0888,
+      "step": 1800
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.309079363151685e-05,
+      "loss": 0.0968,
+      "step": 1803
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 6.298071829937072e-05,
+      "loss": 0.0952,
+      "step": 1806
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.287057546625965e-05,
+      "loss": 0.0822,
+      "step": 1809
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.27603657049368e-05,
+      "loss": 0.0797,
+      "step": 1812
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.265008958850336e-05,
+      "loss": 0.0934,
+      "step": 1815
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.253974769040558e-05,
+      "loss": 0.091,
+      "step": 1818
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 6.242934058443181e-05,
+      "loss": 0.1042,
+      "step": 1821
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.231886884470944e-05,
+      "loss": 0.0864,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.2208333045702e-05,
+      "loss": 0.0813,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.209773376220612e-05,
+      "loss": 0.0834,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 6.198707156934851e-05,
+      "loss": 0.0844,
+      "step": 1833
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.187634704258307e-05,
+      "loss": 0.0975,
+      "step": 1836
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.176556075768789e-05,
+      "loss": 0.0865,
+      "step": 1839
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.165471329076208e-05,
+      "loss": 0.0873,
+      "step": 1842
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.154380521822298e-05,
+      "loss": 0.0867,
+      "step": 1845
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 6.14328371168031e-05,
+      "loss": 0.0894,
+      "step": 1848
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.132180956354705e-05,
+      "loss": 0.1123,
+      "step": 1851
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.121072313580866e-05,
+      "loss": 0.0859,
+      "step": 1854
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.109957841124786e-05,
+      "loss": 0.0843,
+      "step": 1857
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.098837596782776e-05,
+      "loss": 0.0851,
+      "step": 1860
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 6.087711638381157e-05,
+      "loss": 0.0797,
+      "step": 1863
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.0765800237759695e-05,
+      "loss": 0.1038,
+      "step": 1866
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.0654428108526615e-05,
+      "loss": 0.0832,
+      "step": 1869
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.0543000575257994e-05,
+      "loss": 0.0878,
+      "step": 1872
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.043151821738753e-05,
+      "loss": 0.0967,
+      "step": 1875
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 6.0319981614634056e-05,
+      "loss": 0.1,
+      "step": 1878
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.020839134699847e-05,
+      "loss": 0.1083,
+      "step": 1881
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.009674799476073e-05,
+      "loss": 0.1017,
+      "step": 1884
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.998505213847686e-05,
+      "loss": 0.0896,
+      "step": 1887
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.987330435897589e-05,
+      "loss": 0.1017,
+      "step": 1890
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.976150523735686e-05,
+      "loss": 0.0933,
+      "step": 1893
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.964965535498579e-05,
+      "loss": 0.0818,
+      "step": 1896
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.953775529349267e-05,
+      "loss": 0.0822,
+      "step": 1899
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.942580563476839e-05,
+      "loss": 0.0895,
+      "step": 1902
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.931380696096184e-05,
+      "loss": 0.0741,
+      "step": 1905
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.920175985447669e-05,
+      "loss": 0.0825,
+      "step": 1908
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.908966489796854e-05,
+      "loss": 0.0982,
+      "step": 1911
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.8977522674341756e-05,
+      "loss": 0.0868,
+      "step": 1914
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.8865333766746534e-05,
+      "loss": 0.1042,
+      "step": 1917
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.8753098758575805e-05,
+      "loss": 0.0807,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.864081823346229e-05,
+      "loss": 0.0764,
+      "step": 1923
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.852849277527534e-05,
+      "loss": 0.0831,
+      "step": 1926
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.8416122968117945e-05,
+      "loss": 0.0689,
+      "step": 1929
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.830370939632378e-05,
+      "loss": 0.083,
+      "step": 1932
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.8191252644454056e-05,
+      "loss": 0.0738,
+      "step": 1935
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.807875329729454e-05,
+      "loss": 0.073,
+      "step": 1938
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.796621193985251e-05,
+      "loss": 0.0803,
+      "step": 1941
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.785362915735368e-05,
+      "loss": 0.0753,
+      "step": 1944
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.774100553523918e-05,
+      "loss": 0.0974,
+      "step": 1947
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.76283416591625e-05,
+      "loss": 0.0806,
+      "step": 1950
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.7515638114986484e-05,
+      "loss": 0.074,
+      "step": 1953
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.740289548878024e-05,
+      "loss": 0.0956,
+      "step": 1956
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.72901143668161e-05,
+      "loss": 0.0919,
+      "step": 1959
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.7177295335566615e-05,
+      "loss": 0.0884,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.7064438981701393e-05,
+      "loss": 0.0785,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.695154589208417e-05,
+      "loss": 0.0801,
+      "step": 1968
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.683861665376973e-05,
+      "loss": 0.0784,
+      "step": 1971
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.6725651854000793e-05,
+      "loss": 0.0827,
+      "step": 1974
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.661265208020505e-05,
+      "loss": 0.0795,
+      "step": 1977
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.6499617919991985e-05,
+      "loss": 0.0899,
+      "step": 1980
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.638654996114997e-05,
+      "loss": 0.0842,
+      "step": 1983
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.627344879164309e-05,
+      "loss": 0.0855,
+      "step": 1986
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.6160314999608155e-05,
+      "loss": 0.0864,
+      "step": 1989
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.604714917335159e-05,
+      "loss": 0.0817,
+      "step": 1992
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5933951901346424e-05,
+      "loss": 0.0811,
+      "step": 1995
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5820723772229186e-05,
+      "loss": 0.0931,
+      "step": 1998
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.570746537479687e-05,
+      "loss": 0.0866,
+      "step": 2001
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.559417729800389e-05,
+      "loss": 0.0698,
+      "step": 2004
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.548086013095897e-05,
+      "loss": 0.0886,
+      "step": 2007
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.53675144629221e-05,
+      "loss": 0.0817,
+      "step": 2010
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.525414088330152e-05,
+      "loss": 0.0779,
+      "step": 2013
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.514073998165057e-05,
+      "loss": 0.0874,
+      "step": 2016
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.502731234766466e-05,
+      "loss": 0.0869,
+      "step": 2019
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.4913858571178265e-05,
+      "loss": 0.0904,
+      "step": 2022
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.480037924216177e-05,
+      "loss": 0.0717,
+      "step": 2025
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.4686874950718435e-05,
+      "loss": 0.0777,
+      "step": 2028
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.457334628708132e-05,
+      "loss": 0.077,
+      "step": 2031
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.4459793841610206e-05,
+      "loss": 0.0953,
+      "step": 2034
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.434621820478863e-05,
+      "loss": 0.0807,
+      "step": 2037
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.42326199672206e-05,
+      "loss": 0.0764,
+      "step": 2040
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.411899971962776e-05,
+      "loss": 0.072,
+      "step": 2043
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.400535805284614e-05,
+      "loss": 0.0885,
+      "step": 2046
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.389169555782319e-05,
+      "loss": 0.091,
+      "step": 2049
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.377801282561461e-05,
+      "loss": 0.0772,
+      "step": 2052
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.3664310447381425e-05,
+      "loss": 0.0767,
+      "step": 2055
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.3550589014386765e-05,
+      "loss": 0.0851,
+      "step": 2058
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.343684911799285e-05,
+      "loss": 0.0778,
+      "step": 2061
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.332309134965792e-05,
+      "loss": 0.0865,
+      "step": 2064
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.320931630093312e-05,
+      "loss": 0.0766,
+      "step": 2067
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.3095524563459495e-05,
+      "loss": 0.076,
+      "step": 2070
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.2981716728964894e-05,
+      "loss": 0.0844,
+      "step": 2073
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.286789338926081e-05,
+      "loss": 0.0756,
+      "step": 2076
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.275405513623941e-05,
+      "loss": 0.0732,
+      "step": 2079
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.2640202561870364e-05,
+      "loss": 0.0724,
+      "step": 2082
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.252633625819786e-05,
+      "loss": 0.0747,
+      "step": 2085
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.241245681733748e-05,
+      "loss": 0.0862,
+      "step": 2088
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.22985648314731e-05,
+      "loss": 0.0814,
+      "step": 2091
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.2184660892853824e-05,
+      "loss": 0.0741,
+      "step": 2094
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.207074559379094e-05,
+      "loss": 0.075,
+      "step": 2097
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.195681952665478e-05,
+      "loss": 0.0815,
+      "step": 2100
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.1842883283871666e-05,
+      "loss": 0.0763,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.172893745792091e-05,
+      "loss": 0.0915,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.161498264133157e-05,
+      "loss": 0.0767,
+      "step": 2109
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1501019426679483e-05,
+      "loss": 0.0777,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.1387048406584196e-05,
+      "loss": 0.083,
+      "step": 2115
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.127307017370575e-05,
+      "loss": 0.0833,
+      "step": 2118
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.115908532074181e-05,
+      "loss": 0.0807,
+      "step": 2121
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.1045094440424377e-05,
+      "loss": 0.0873,
+      "step": 2124
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.093109812551685e-05,
+      "loss": 0.0754,
+      "step": 2127
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.0817096968810875e-05,
+      "loss": 0.0729,
+      "step": 2130
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.070309156312324e-05,
+      "loss": 0.0815,
+      "step": 2133
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.058908250129285e-05,
+      "loss": 0.0868,
+      "step": 2136
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.0475070376177644e-05,
+      "loss": 0.0708,
+      "step": 2139
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.036105578065149e-05,
+      "loss": 0.0761,
+      "step": 2142
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.024703930760106e-05,
+      "loss": 0.0923,
+      "step": 2145
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.013302154992285e-05,
+      "loss": 0.0694,
+      "step": 2148
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.0019003100519946e-05,
+      "loss": 0.0816,
+      "step": 2151
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.990498455229913e-05,
+      "loss": 0.0815,
+      "step": 2154
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.979096649816766e-05,
+      "loss": 0.0721,
+      "step": 2157
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.9676949531030175e-05,
+      "loss": 0.0711,
+      "step": 2160
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.9562934243785765e-05,
+      "loss": 0.0874,
+      "step": 2163
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.944892122932466e-05,
+      "loss": 0.0818,
+      "step": 2166
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.9334911080525394e-05,
+      "loss": 0.0722,
+      "step": 2169
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.9220904390251524e-05,
+      "loss": 0.0642,
+      "step": 2172
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.910690175134862e-05,
+      "loss": 0.0793,
+      "step": 2175
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.899290375664124e-05,
+      "loss": 0.0894,
+      "step": 2178
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.887891099892972e-05,
+      "loss": 0.0807,
+      "step": 2181
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.8764924070987226e-05,
+      "loss": 0.074,
+      "step": 2184
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.865094356555657e-05,
+      "loss": 0.0698,
+      "step": 2187
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8536970075347176e-05,
+      "loss": 0.0706,
+      "step": 2190
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.842300419303202e-05,
+      "loss": 0.0882,
+      "step": 2193
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.8309046511244435e-05,
+      "loss": 0.0876,
+      "step": 2196
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.819509762257523e-05,
+      "loss": 0.0851,
+      "step": 2199
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.808115811956935e-05,
+      "loss": 0.0804,
+      "step": 2202
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.796722859472305e-05,
+      "loss": 0.0854,
+      "step": 2205
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.7853309640480645e-05,
+      "loss": 0.0927,
+      "step": 2208
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.773940184923148e-05,
+      "loss": 0.073,
+      "step": 2211
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.762550581330688e-05,
+      "loss": 0.0812,
+      "step": 2214
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7511622124976995e-05,
+      "loss": 0.0766,
+      "step": 2217
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7397751376447816e-05,
+      "loss": 0.0779,
+      "step": 2220
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.728389415985803e-05,
+      "loss": 0.0728,
+      "step": 2223
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.7170051067275914e-05,
+      "loss": 0.0634,
+      "step": 2226
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.705622269069637e-05,
+      "loss": 0.0787,
+      "step": 2229
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.694240962203771e-05,
+      "loss": 0.0746,
+      "step": 2232
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.682861245313871e-05,
+      "loss": 0.0768,
+      "step": 2235
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.671483177575541e-05,
+      "loss": 0.092,
+      "step": 2238
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.6601068181558106e-05,
+      "loss": 0.0776,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.648732226212828e-05,
+      "loss": 0.0801,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.637359460895547e-05,
+      "loss": 0.0731,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.625988581343428e-05,
+      "loss": 0.0752,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.614619646686117e-05,
+      "loss": 0.0779,
+      "step": 2253
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.6032527160431554e-05,
+      "loss": 0.0833,
+      "step": 2256
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.591887848523659e-05,
+      "loss": 0.072,
+      "step": 2259
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.580525103226013e-05,
+      "loss": 0.078,
+      "step": 2262
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.569164539237574e-05,
+      "loss": 0.0832,
+      "step": 2265
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.557806215634345e-05,
+      "loss": 0.0768,
+      "step": 2268
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.5464501914806894e-05,
+      "loss": 0.0741,
+      "step": 2271
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.535096525829009e-05,
+      "loss": 0.0772,
+      "step": 2274
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.5237452777194366e-05,
+      "loss": 0.0717,
+      "step": 2277
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.5123965061795425e-05,
+      "loss": 0.0788,
+      "step": 2280
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.5010502702240096e-05,
+      "loss": 0.0773,
+      "step": 2283
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.4897066288543435e-05,
+      "loss": 0.0784,
+      "step": 2286
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.4783656410585515e-05,
+      "loss": 0.071,
+      "step": 2289
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.4670273658108473e-05,
+      "loss": 0.0726,
+      "step": 2292
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.455691862071335e-05,
+      "loss": 0.0921,
+      "step": 2295
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.444359188785705e-05,
+      "loss": 0.0801,
+      "step": 2298
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.433029404884936e-05,
+      "loss": 0.0691,
+      "step": 2301
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.4217025692849765e-05,
+      "loss": 0.0839,
+      "step": 2304
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.4103787408864444e-05,
+      "loss": 0.0685,
+      "step": 2307
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.3990579785743216e-05,
+      "loss": 0.0752,
+      "step": 2310
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.387740341217641e-05,
+      "loss": 0.0672,
+      "step": 2313
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.376425887669194e-05,
+      "loss": 0.0805,
+      "step": 2316
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.365114676765206e-05,
+      "loss": 0.0809,
+      "step": 2319
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.353806767325049e-05,
+      "loss": 0.0723,
+      "step": 2322
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.342502218150923e-05,
+      "loss": 0.0709,
+      "step": 2325
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.331201088027552e-05,
+      "loss": 0.0737,
+      "step": 2328
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.319903435721888e-05,
+      "loss": 0.0633,
+      "step": 2331
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.308609319982787e-05,
+      "loss": 0.0686,
+      "step": 2334
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.297318799540726e-05,
+      "loss": 0.0725,
+      "step": 2337
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.2860319331074786e-05,
+      "loss": 0.0745,
+      "step": 2340
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.274748779375822e-05,
+      "loss": 0.0718,
+      "step": 2343
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.263469397019224e-05,
+      "loss": 0.0594,
+      "step": 2346
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.252193844691537e-05,
+      "loss": 0.0794,
+      "step": 2349
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.240922181026707e-05,
+      "loss": 0.074,
+      "step": 2352
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.229654464638449e-05,
+      "loss": 0.0731,
+      "step": 2355
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.218390754119958e-05,
+      "loss": 0.0689,
+      "step": 2358
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.207131108043596e-05,
+      "loss": 0.0764,
+      "step": 2361
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.1958755849605866e-05,
+      "loss": 0.0688,
+      "step": 2364
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.184624243400721e-05,
+      "loss": 0.0722,
+      "step": 2367
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.1733771418720354e-05,
+      "loss": 0.0723,
+      "step": 2370
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.162134338860527e-05,
+      "loss": 0.0775,
+      "step": 2373
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.1508958928298376e-05,
+      "loss": 0.0752,
+      "step": 2376
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.139661862220947e-05,
+      "loss": 0.0641,
+      "step": 2379
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.128432305451884e-05,
+      "loss": 0.072,
+      "step": 2382
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.117207280917402e-05,
+      "loss": 0.0611,
+      "step": 2385
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.1059868469886956e-05,
+      "loss": 0.0787,
+      "step": 2388
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.094771062013083e-05,
+      "loss": 0.0689,
+      "step": 2391
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.083559984313712e-05,
+      "loss": 0.072,
+      "step": 2394
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.072353672189245e-05,
+      "loss": 0.0723,
+      "step": 2397
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.0611521839135666e-05,
+      "loss": 0.0651,
+      "step": 2400
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.049955577735479e-05,
+      "loss": 0.0654,
+      "step": 2403
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.038763911878393e-05,
+      "loss": 0.0749,
+      "step": 2406
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.0275772445400344e-05,
+      "loss": 0.0699,
+      "step": 2409
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.016395633892132e-05,
+      "loss": 0.0801,
+      "step": 2412
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.0052191380801176e-05,
+      "loss": 0.0713,
+      "step": 2415
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.9940478152228315e-05,
+      "loss": 0.0704,
+      "step": 2418
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.982881723412211e-05,
+      "loss": 0.0699,
+      "step": 2421
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.971720920712989e-05,
+      "loss": 0.0759,
+      "step": 2424
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.9605654651624005e-05,
+      "loss": 0.0691,
+      "step": 2427
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.9494154147698666e-05,
+      "loss": 0.0674,
+      "step": 2430
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.93827082751671e-05,
+      "loss": 0.0632,
+      "step": 2433
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.927131761355837e-05,
+      "loss": 0.0917,
+      "step": 2436
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.915998274211449e-05,
+      "loss": 0.0634,
+      "step": 2439
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.9048704239787345e-05,
+      "loss": 0.0686,
+      "step": 2442
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.8937482685235657e-05,
+      "loss": 0.0693,
+      "step": 2445
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.8826318656822084e-05,
+      "loss": 0.0737,
+      "step": 2448
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.8715212732610076e-05,
+      "loss": 0.0655,
+      "step": 2451
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.860416549036098e-05,
+      "loss": 0.0714,
+      "step": 2454
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.849317750753098e-05,
+      "loss": 0.0705,
+      "step": 2457
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.83822493612681e-05,
+      "loss": 0.0624,
+      "step": 2460
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.8271381628409206e-05,
+      "loss": 0.0787,
+      "step": 2463
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.8160574885477e-05,
+      "loss": 0.0835,
+      "step": 2466
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.804982970867706e-05,
+      "loss": 0.069,
+      "step": 2469
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.79391466738948e-05,
+      "loss": 0.0749,
+      "step": 2472
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.7828526356692476e-05,
+      "loss": 0.0711,
+      "step": 2475
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.7717969332306233e-05,
+      "loss": 0.0636,
+      "step": 2478
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.760747617564304e-05,
+      "loss": 0.0573,
+      "step": 2481
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.7497047461277825e-05,
+      "loss": 0.0694,
+      "step": 2484
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.738668376345032e-05,
+      "loss": 0.0661,
+      "step": 2487
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.727638565606223e-05,
+      "loss": 0.086,
+      "step": 2490
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.716615371267417e-05,
+      "loss": 0.0742,
+      "step": 2493
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.705598850650267e-05,
+      "loss": 0.0769,
+      "step": 2496
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.694589061041727e-05,
+      "loss": 0.075,
+      "step": 2499
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.683586059693742e-05,
+      "loss": 0.075,
+      "step": 2502
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.672589903822964e-05,
+      "loss": 0.0771,
+      "step": 2505
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.661600650610444e-05,
+      "loss": 0.0664,
+      "step": 2508
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.650618357201343e-05,
+      "loss": 0.0656,
+      "step": 2511
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.6396430807046255e-05,
+      "loss": 0.0615,
+      "step": 2514
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.6286748781927654e-05,
+      "loss": 0.0756,
+      "step": 2517
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.617713806701459e-05,
+      "loss": 0.0569,
+      "step": 2520
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.6067599232293145e-05,
+      "loss": 0.0812,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.595813284737567e-05,
+      "loss": 0.0647,
+      "step": 2526
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.5848739481497716e-05,
+      "loss": 0.0693,
+      "step": 2529
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.573941970351512e-05,
+      "loss": 0.0758,
+      "step": 2532
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.5630174081901125e-05,
+      "loss": 0.0738,
+      "step": 2535
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.552100318474331e-05,
+      "loss": 0.0667,
+      "step": 2538
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.5411907579740653e-05,
+      "loss": 0.0589,
+      "step": 2541
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.5302887834200684e-05,
+      "loss": 0.0604,
+      "step": 2544
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.5193944515036346e-05,
+      "loss": 0.0773,
+      "step": 2547
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.5085078188763285e-05,
+      "loss": 0.0773,
+      "step": 2550
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.4976289421496655e-05,
+      "loss": 0.0652,
+      "step": 2553
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.486757877894838e-05,
+      "loss": 0.0676,
+      "step": 2556
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.47589468264241e-05,
+      "loss": 0.066,
+      "step": 2559
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.465039412882029e-05,
+      "loss": 0.0632,
+      "step": 2562
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.454192125062121e-05,
+      "loss": 0.0564,
+      "step": 2565
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.443352875589611e-05,
+      "loss": 0.0745,
+      "step": 2568
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.432521720829626e-05,
+      "loss": 0.0489,
+      "step": 2571
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.421698717105194e-05,
+      "loss": 0.0684,
+      "step": 2574
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.410883920696963e-05,
+      "loss": 0.0657,
+      "step": 2577
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.400077387842897e-05,
+      "loss": 0.0581,
+      "step": 2580
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.389279174737989e-05,
+      "loss": 0.0675,
+      "step": 2583
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.378489337533973e-05,
+      "loss": 0.0794,
+      "step": 2586
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.367707932339024e-05,
+      "loss": 0.0695,
+      "step": 2589
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3569350152174694e-05,
+      "loss": 0.0609,
+      "step": 2592
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3461706421895003e-05,
+      "loss": 0.0652,
+      "step": 2595
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3354148692308736e-05,
+      "loss": 0.0675,
+      "step": 2598
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3246677522726324e-05,
+      "loss": 0.0701,
+      "step": 2601
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.313929347200797e-05,
+      "loss": 0.0693,
+      "step": 2604
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.303199709856096e-05,
+      "loss": 0.0735,
+      "step": 2607
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.2924788960336575e-05,
+      "loss": 0.0652,
+      "step": 2610
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.281766961482732e-05,
+      "loss": 0.0617,
+      "step": 2613
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.271063961906392e-05,
+      "loss": 0.0555,
+      "step": 2616
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.260369952961248e-05,
+      "loss": 0.0572,
+      "step": 2619
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.24968499025716e-05,
+      "loss": 0.0719,
+      "step": 2622
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.239009129356948e-05,
+      "loss": 0.0646,
+      "step": 2625
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.2283424257761005e-05,
+      "loss": 0.0678,
+      "step": 2628
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.2176849349824865e-05,
+      "loss": 0.0711,
+      "step": 2631
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.207036712396065e-05,
+      "loss": 0.069,
+      "step": 2634
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.196397813388605e-05,
+      "loss": 0.0635,
+      "step": 2637
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.1857682932833864e-05,
+      "loss": 0.0713,
+      "step": 2640
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.175148207354927e-05,
+      "loss": 0.0757,
+      "step": 2643
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.164537610828674e-05,
+      "loss": 0.063,
+      "step": 2646
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.1539365588807346e-05,
+      "loss": 0.07,
+      "step": 2649
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.1433451066375856e-05,
+      "loss": 0.0652,
+      "step": 2652
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.132763309175782e-05,
+      "loss": 0.0785,
+      "step": 2655
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.122191221521672e-05,
+      "loss": 0.0688,
+      "step": 2658
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.111628898651113e-05,
+      "loss": 0.0569,
+      "step": 2661
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.101076395489187e-05,
+      "loss": 0.0497,
+      "step": 2664
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.090533766909908e-05,
+      "loss": 0.0662,
+      "step": 2667
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.080001067735942e-05,
+      "loss": 0.0777,
+      "step": 2670
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.0694783527383255e-05,
+      "loss": 0.0638,
+      "step": 2673
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.058965676636171e-05,
+      "loss": 0.0641,
+      "step": 2676
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.0484630940963932e-05,
+      "loss": 0.0578,
+      "step": 2679
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.0379706597334146e-05,
+      "loss": 0.0627,
+      "step": 2682
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.0274884281088856e-05,
+      "loss": 0.068,
+      "step": 2685
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.0170164537314072e-05,
+      "loss": 0.0632,
+      "step": 2688
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.0065547910562364e-05,
+      "loss": 0.0781,
+      "step": 2691
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.996103494485013e-05,
+      "loss": 0.0632,
+      "step": 2694
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.985662618365467e-05,
+      "loss": 0.0711,
+      "step": 2697
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.975232216991141e-05,
+      "loss": 0.0542,
+      "step": 2700
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.9648123446011146e-05,
+      "loss": 0.0514,
+      "step": 2703
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.9544030553797086e-05,
+      "loss": 0.0641,
+      "step": 2706
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.9440044034562163e-05,
+      "loss": 0.0679,
+      "step": 2709
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.9336164429046098e-05,
+      "loss": 0.0663,
+      "step": 2712
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.9232392277432695e-05,
+      "loss": 0.0761,
+      "step": 2715
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.912872811934698e-05,
+      "loss": 0.07,
+      "step": 2718
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.902517249385236e-05,
+      "loss": 0.0603,
+      "step": 2721
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.8921725939447965e-05,
+      "loss": 0.0604,
+      "step": 2724
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.8818388994065594e-05,
+      "loss": 0.0606,
+      "step": 2727
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.8715162195067214e-05,
+      "loss": 0.064,
+      "step": 2730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.861204607924197e-05,
+      "loss": 0.0714,
+      "step": 2733
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.8509041182803364e-05,
+      "loss": 0.0553,
+      "step": 2736
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.8406148041386698e-05,
+      "loss": 0.0708,
+      "step": 2739
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.830336719004598e-05,
+      "loss": 0.0542,
+      "step": 2742
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.820069916325144e-05,
+      "loss": 0.0566,
+      "step": 2745
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.8098144494886513e-05,
+      "loss": 0.0507,
+      "step": 2748
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.7995703718245157e-05,
+      "loss": 0.0638,
+      "step": 2751
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.78933773660292e-05,
+      "loss": 0.067,
+      "step": 2754
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.7791165970345278e-05,
+      "loss": 0.0731,
+      "step": 2757
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.7689070062702404e-05,
+      "loss": 0.0752,
+      "step": 2760
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.758709017400891e-05,
+      "loss": 0.0673,
+      "step": 2763
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.748522683456992e-05,
+      "loss": 0.0669,
+      "step": 2766
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7383480574084462e-05,
+      "loss": 0.0704,
+      "step": 2769
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7281851921642672e-05,
+      "loss": 0.0586,
+      "step": 2772
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7180341405723235e-05,
+      "loss": 0.0663,
+      "step": 2775
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7078949554190388e-05,
+      "loss": 0.0686,
+      "step": 2778
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.6977676894291414e-05,
+      "loss": 0.0713,
+      "step": 2781
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.6876523952653736e-05,
+      "loss": 0.0606,
+      "step": 2784
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.677549125528219e-05,
+      "loss": 0.0624,
+      "step": 2787
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.6674579327556436e-05,
+      "loss": 0.0594,
+      "step": 2790
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.6573788694227997e-05,
+      "loss": 0.0545,
+      "step": 2793
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.6473119879417755e-05,
+      "loss": 0.0596,
+      "step": 2796
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.63725734066131e-05,
+      "loss": 0.064,
+      "step": 2799
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.6272149798665146e-05,
+      "loss": 0.0585,
+      "step": 2802
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.6171849577786262e-05,
+      "loss": 0.0712,
+      "step": 2805
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.6071673265547003e-05,
+      "loss": 0.0606,
+      "step": 2808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5971621382873774e-05,
+      "loss": 0.0607,
+      "step": 2811
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5871694450045758e-05,
+      "loss": 0.0614,
+      "step": 2814
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5771892986692537e-05,
+      "loss": 0.0579,
+      "step": 2817
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.5672217511791175e-05,
+      "loss": 0.0475,
+      "step": 2820
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5572668543663526e-05,
+      "loss": 0.0618,
+      "step": 2823
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5473246599973722e-05,
+      "loss": 0.0552,
+      "step": 2826
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.5373952197725232e-05,
+      "loss": 0.0527,
+      "step": 2829
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.52747858532584e-05,
+      "loss": 0.051,
+      "step": 2832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.517574808224761e-05,
+      "loss": 0.053,
+      "step": 2835
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.5076839399698605e-05,
+      "loss": 0.0538,
+      "step": 2838
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.497806031994598e-05,
+      "loss": 0.0522,
+      "step": 2841
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.4879411356650223e-05,
+      "loss": 0.0565,
+      "step": 2844
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.478089302279536e-05,
+      "loss": 0.049,
+      "step": 2847
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4682505830686048e-05,
+      "loss": 0.0673,
+      "step": 2850
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.458425029194496e-05,
+      "loss": 0.0579,
+      "step": 2853
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.448612691751027e-05,
+      "loss": 0.0542,
+      "step": 2856
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.4388136217632762e-05,
+      "loss": 0.0554,
+      "step": 2859
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.429027870187342e-05,
+      "loss": 0.0464,
+      "step": 2862
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4192554879100548e-05,
+      "loss": 0.0483,
+      "step": 2865
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.4094965257487286e-05,
+      "loss": 0.049,
+      "step": 2868
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.3997510344508956e-05,
+      "loss": 0.0604,
+      "step": 2871
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.390019064694027e-05,
+      "loss": 0.0524,
+      "step": 2874
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.3803006670852956e-05,
+      "loss": 0.051,
+      "step": 2877
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.3705958921612814e-05,
+      "loss": 0.0665,
+      "step": 2880
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.3609047903877392e-05,
+      "loss": 0.0547,
+      "step": 2883
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.3512274121593163e-05,
+      "loss": 0.0584,
+      "step": 2886
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.3415638077992914e-05,
+      "loss": 0.0466,
+      "step": 2889
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.3319140275593297e-05,
+      "loss": 0.0603,
+      "step": 2892
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.3222781216191948e-05,
+      "loss": 0.0573,
+      "step": 2895
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.3126561400865166e-05,
+      "loss": 0.0534,
+      "step": 2898
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.303048132996511e-05,
+      "loss": 0.0608,
+      "step": 2901
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.2934541503117185e-05,
+      "loss": 0.0459,
+      "step": 2904
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.2838742419217663e-05,
+      "loss": 0.0732,
+      "step": 2907
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.2743084576430775e-05,
+      "loss": 0.0529,
+      "step": 2910
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.264756847218643e-05,
+      "loss": 0.0563,
+      "step": 2913
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.255219460317741e-05,
+      "loss": 0.0591,
+      "step": 2916
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.2456963465356822e-05,
+      "loss": 0.0548,
+      "step": 2919
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.236187555393568e-05,
+      "loss": 0.0498,
+      "step": 2922
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.226693136338005e-05,
+      "loss": 0.0588,
+      "step": 2925
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.2172131387408796e-05,
+      "loss": 0.0617,
+      "step": 2928
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.2077476118990698e-05,
+      "loss": 0.0503,
+      "step": 2931
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.198296605034218e-05,
+      "loss": 0.0523,
+      "step": 2934
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.1888601672924548e-05,
+      "loss": 0.0516,
+      "step": 2937
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.1794383477441453e-05,
+      "loss": 0.0518,
+      "step": 2940
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.170031195383651e-05,
+      "loss": 0.0564,
+      "step": 2943
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.1606387591290476e-05,
+      "loss": 0.0551,
+      "step": 2946
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.1512610878218998e-05,
+      "loss": 0.061,
+      "step": 2949
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.1418982302269864e-05,
+      "loss": 0.052,
+      "step": 2952
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.132550235032048e-05,
+      "loss": 0.0624,
+      "step": 2955
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.1232171508475535e-05,
+      "loss": 0.0541,
+      "step": 2958
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.1138990262064172e-05,
+      "loss": 0.0599,
+      "step": 2961
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.104595909563775e-05,
+      "loss": 0.0465,
+      "step": 2964
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.0953078492967155e-05,
+      "loss": 0.0545,
+      "step": 2967
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.086034893704026e-05,
+      "loss": 0.0507,
+      "step": 2970
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.076777091005962e-05,
+      "loss": 0.0491,
+      "step": 2973
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.0675344893439662e-05,
+      "loss": 0.0578,
+      "step": 2976
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.05830713678045e-05,
+      "loss": 0.0525,
+      "step": 2979
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.049095081298518e-05,
+      "loss": 0.0591,
+      "step": 2982
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.0398983708017316e-05,
+      "loss": 0.0604,
+      "step": 2985
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.0307170531138585e-05,
+      "loss": 0.0565,
+      "step": 2988
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.0215511759786158e-05,
+      "loss": 0.0577,
+      "step": 2991
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.012400787059439e-05,
+      "loss": 0.0541,
+      "step": 2994
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.0032659339392112e-05,
+      "loss": 0.0592,
+      "step": 2997
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.9941466641200386e-05,
+      "loss": 0.0619,
+      "step": 3000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.9850430250229885e-05,
+      "loss": 0.0459,
+      "step": 3003
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.9759550639878422e-05,
+      "loss": 0.0534,
+      "step": 3006
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.9668828282728634e-05,
+      "loss": 0.0554,
+      "step": 3009
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.957826365054532e-05,
+      "loss": 0.0581,
+      "step": 3012
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.9487857214273192e-05,
+      "loss": 0.0548,
+      "step": 3015
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9397609444034288e-05,
+      "loss": 0.0584,
+      "step": 3018
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.93075208091255e-05,
+      "loss": 0.0557,
+      "step": 3021
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9217591778016346e-05,
+      "loss": 0.0617,
+      "step": 3024
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9127822818346235e-05,
+      "loss": 0.0611,
+      "step": 3027
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.9038214396922317e-05,
+      "loss": 0.056,
+      "step": 3030
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8948766979716848e-05,
+      "loss": 0.0518,
+      "step": 3033
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8859481031864874e-05,
+      "loss": 0.0582,
+      "step": 3036
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.87703570176618e-05,
+      "loss": 0.0547,
+      "step": 3039
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8681395400560898e-05,
+      "loss": 0.0552,
+      "step": 3042
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.8592596643171068e-05,
+      "loss": 0.0641,
+      "step": 3045
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8503961207254188e-05,
+      "loss": 0.0654,
+      "step": 3048
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8415489553722988e-05,
+      "loss": 0.0545,
+      "step": 3051
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.832718214263843e-05,
+      "loss": 0.0583,
+      "step": 3054
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.8239039433207384e-05,
+      "loss": 0.0453,
+      "step": 3057
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8151061883780347e-05,
+      "loss": 0.0493,
+      "step": 3060
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.8063249951848844e-05,
+      "loss": 0.0577,
+      "step": 3063
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.7975604094043295e-05,
+      "loss": 0.0542,
+      "step": 3066
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.7888124766130464e-05,
+      "loss": 0.0553,
+      "step": 3069
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.7800812423011087e-05,
+      "loss": 0.0614,
+      "step": 3072
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.7713667518717697e-05,
+      "loss": 0.0588,
+      "step": 3075
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.7626690506411974e-05,
+      "loss": 0.0571,
+      "step": 3078
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.7539881838382684e-05,
+      "loss": 0.0635,
+      "step": 3081
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.74532419660431e-05,
+      "loss": 0.0498,
+      "step": 3084
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.7366771339928774e-05,
+      "loss": 0.0495,
+      "step": 3087
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.728047040969516e-05,
+      "loss": 0.0462,
+      "step": 3090
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.719433962411523e-05,
+      "loss": 0.0467,
+      "step": 3093
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.7108379431077282e-05,
+      "loss": 0.0528,
+      "step": 3096
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.7022590277582436e-05,
+      "loss": 0.0472,
+      "step": 3099
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.6936972609742417e-05,
+      "loss": 0.0558,
+      "step": 3102
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.685152687277724e-05,
+      "loss": 0.0617,
+      "step": 3105
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.6766253511012775e-05,
+      "loss": 0.0541,
+      "step": 3108
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.668115296787865e-05,
+      "loss": 0.0418,
+      "step": 3111
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.6596225685905675e-05,
+      "loss": 0.0566,
+      "step": 3114
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.651147210672383e-05,
+      "loss": 0.0587,
+      "step": 3117
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.642689267105974e-05,
+      "loss": 0.0569,
+      "step": 3120
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.6342487818734443e-05,
+      "loss": 0.0481,
+      "step": 3123
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.625825798866123e-05,
+      "loss": 0.0506,
+      "step": 3126
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.6174203618843132e-05,
+      "loss": 0.0571,
+      "step": 3129
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.6090325146370887e-05,
+      "loss": 0.0506,
+      "step": 3132
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.600662300742048e-05,
+      "loss": 0.0486,
+      "step": 3135
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.5923097637250962e-05,
+      "loss": 0.0493,
+      "step": 3138
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.5839749470202193e-05,
+      "loss": 0.0504,
+      "step": 3141
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.5756578939692477e-05,
+      "loss": 0.05,
+      "step": 3144
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.5673586478216516e-05,
+      "loss": 0.0497,
+      "step": 3147
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.559077251734295e-05,
+      "loss": 0.0615,
+      "step": 3150
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.5508137487712223e-05,
+      "loss": 0.0523,
+      "step": 3153
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.5425681819034326e-05,
+      "loss": 0.0605,
+      "step": 3156
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.534340594008651e-05,
+      "loss": 0.0587,
+      "step": 3159
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.52613102787112e-05,
+      "loss": 0.053,
+      "step": 3162
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.5179395261813544e-05,
+      "loss": 0.0443,
+      "step": 3165
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.5097661315359446e-05,
+      "loss": 0.0536,
+      "step": 3168
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.5016108864373169e-05,
+      "loss": 0.0543,
+      "step": 3171
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4934738332935138e-05,
+      "loss": 0.0538,
+      "step": 3174
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4853550144179884e-05,
+      "loss": 0.0588,
+      "step": 3177
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4772544720293625e-05,
+      "loss": 0.0529,
+      "step": 3180
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4691722482512277e-05,
+      "loss": 0.0465,
+      "step": 3183
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.4611083851119122e-05,
+      "loss": 0.0525,
+      "step": 3186
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4530629245442684e-05,
+      "loss": 0.0469,
+      "step": 3189
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4450359083854553e-05,
+      "loss": 0.0512,
+      "step": 3192
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4370273783767124e-05,
+      "loss": 0.0578,
+      "step": 3195
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.4290373761631593e-05,
+      "loss": 0.047,
+      "step": 3198
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.4210659432935625e-05,
+      "loss": 0.0548,
+      "step": 3201
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.4131131212201287e-05,
+      "loss": 0.0474,
+      "step": 3204
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.4051789512982877e-05,
+      "loss": 0.0496,
+      "step": 3207
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.3972634747864693e-05,
+      "loss": 0.0558,
+      "step": 3210
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.389366732845907e-05,
+      "loss": 0.0558,
+      "step": 3213
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3814887665404047e-05,
+      "loss": 0.0479,
+      "step": 3216
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3736296168361334e-05,
+      "loss": 0.0545,
+      "step": 3219
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3657893246014187e-05,
+      "loss": 0.0522,
+      "step": 3222
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3579679306065173e-05,
+      "loss": 0.049,
+      "step": 3225
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.3501654755234261e-05,
+      "loss": 0.0535,
+      "step": 3228
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.3423819999256437e-05,
+      "loss": 0.0588,
+      "step": 3231
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.334617544287986e-05,
+      "loss": 0.051,
+      "step": 3234
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.3268721489863573e-05,
+      "loss": 0.0434,
+      "step": 3237
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.3191458542975427e-05,
+      "loss": 0.0571,
+      "step": 3240
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.3114387003990125e-05,
+      "loss": 0.0487,
+      "step": 3243
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.303750727368691e-05,
+      "loss": 0.0615,
+      "step": 3246
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2960819751847708e-05,
+      "loss": 0.0435,
+      "step": 3249
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.288432483725489e-05,
+      "loss": 0.0459,
+      "step": 3252
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2808022927689251e-05,
+      "loss": 0.0521,
+      "step": 3255
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.273191441992797e-05,
+      "loss": 0.0521,
+      "step": 3258
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.265599970974245e-05,
+      "loss": 0.0583,
+      "step": 3261
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.2580279191896422e-05,
+      "loss": 0.0551,
+      "step": 3264
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.2504753260143736e-05,
+      "loss": 0.0576,
+      "step": 3267
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.242942230722639e-05,
+      "loss": 0.0434,
+      "step": 3270
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.23542867248725e-05,
+      "loss": 0.0538,
+      "step": 3273
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.2279346903794159e-05,
+      "loss": 0.0537,
+      "step": 3276
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.2204603233685597e-05,
+      "loss": 0.0527,
+      "step": 3279
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.2130056103220926e-05,
+      "loss": 0.0447,
+      "step": 3282
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.2055705900052328e-05,
+      "loss": 0.051,
+      "step": 3285
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1981553010807905e-05,
+      "loss": 0.0587,
+      "step": 3288
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1907597821089656e-05,
+      "loss": 0.0547,
+      "step": 3291
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1833840715471639e-05,
+      "loss": 0.045,
+      "step": 3294
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1760282077497737e-05,
+      "loss": 0.0562,
+      "step": 3297
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1686922289679864e-05,
+      "loss": 0.0505,
+      "step": 3300
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1613761733495865e-05,
+      "loss": 0.0516,
+      "step": 3303
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1540800789387562e-05,
+      "loss": 0.0519,
+      "step": 3306
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1468039836758793e-05,
+      "loss": 0.0554,
+      "step": 3309
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1395479253973351e-05,
+      "loss": 0.0517,
+      "step": 3312
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.13231194183532e-05,
+      "loss": 0.0565,
+      "step": 3315
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1250960706176312e-05,
+      "loss": 0.057,
+      "step": 3318
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1179003492674834e-05,
+      "loss": 0.0513,
+      "step": 3321
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1107248152033106e-05,
+      "loss": 0.047,
+      "step": 3324
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1035695057385653e-05,
+      "loss": 0.0454,
+      "step": 3327
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.096434458081539e-05,
+      "loss": 0.0505,
+      "step": 3330
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0893197093351543e-05,
+      "loss": 0.0476,
+      "step": 3333
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0822252964967782e-05,
+      "loss": 0.0472,
+      "step": 3336
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.0751512564580319e-05,
+      "loss": 0.0493,
+      "step": 3339
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0680976260045889e-05,
+      "loss": 0.0519,
+      "step": 3342
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0610644418160015e-05,
+      "loss": 0.0449,
+      "step": 3345
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0540517404654881e-05,
+      "loss": 0.0592,
+      "step": 3348
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.0470595584197645e-05,
+      "loss": 0.0554,
+      "step": 3351
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.040087932038838e-05,
+      "loss": 0.0555,
+      "step": 3354
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.033136897575827e-05,
+      "loss": 0.0435,
+      "step": 3357
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0262064911767706e-05,
+      "loss": 0.044,
+      "step": 3360
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0192967488804333e-05,
+      "loss": 0.0496,
+      "step": 3363
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0124077066181353e-05,
+      "loss": 0.045,
+      "step": 3366
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.0055394002135466e-05,
+      "loss": 0.0499,
+      "step": 3369
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 9.986918653825117e-06,
+      "loss": 0.0547,
+      "step": 3372
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 9.918651377328619e-06,
+      "loss": 0.0532,
+      "step": 3375
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 9.850592527642232e-06,
+      "loss": 0.0518,
+      "step": 3378
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 9.782742458678473e-06,
+      "loss": 0.0543,
+      "step": 3381
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.715101523264102e-06,
+      "loss": 0.0439,
+      "step": 3384
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.647670073138399e-06,
+      "loss": 0.0526,
+      "step": 3387
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.580448458951303e-06,
+      "loss": 0.0549,
+      "step": 3390
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.51343703026154e-06,
+      "loss": 0.05,
+      "step": 3393
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.446636135534932e-06,
+      "loss": 0.056,
+      "step": 3396
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.380046122142433e-06,
+      "loss": 0.0545,
+      "step": 3399
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.313667336358433e-06,
+      "loss": 0.0603,
+      "step": 3402
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.2475001233589e-06,
+      "loss": 0.0486,
+      "step": 3405
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.181544827219619e-06,
+      "loss": 0.0524,
+      "step": 3408
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.11580179091438e-06,
+      "loss": 0.0466,
+      "step": 3411
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.050271356313172e-06,
+      "loss": 0.057,
+      "step": 3414
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.984953864180495e-06,
+      "loss": 0.0513,
+      "step": 3417
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.91984965417349e-06,
+      "loss": 0.0563,
+      "step": 3420
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.854959064840206e-06,
+      "loss": 0.046,
+      "step": 3423
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.79028243361788e-06,
+      "loss": 0.0494,
+      "step": 3426
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.725820096831077e-06,
+      "loss": 0.0502,
+      "step": 3429
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.661572389690098e-06,
+      "loss": 0.048,
+      "step": 3432
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.597539646289088e-06,
+      "loss": 0.054,
+      "step": 3435
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 8.533722199604388e-06,
+      "loss": 0.0572,
+      "step": 3438
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.470120381492769e-06,
+      "loss": 0.0473,
+      "step": 3441
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.40673452268968e-06,
+      "loss": 0.0485,
+      "step": 3444
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.343564952807626e-06,
+      "loss": 0.0512,
+      "step": 3447
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.280612000334342e-06,
+      "loss": 0.0542,
+      "step": 3450
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.217875992631153e-06,
+      "loss": 0.0622,
+      "step": 3453
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.155357255931234e-06,
+      "loss": 0.0494,
+      "step": 3456
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.093056115337939e-06,
+      "loss": 0.047,
+      "step": 3459
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 8.030972894823119e-06,
+      "loss": 0.0465,
+      "step": 3462
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 7.969107917225365e-06,
+      "loss": 0.0451,
+      "step": 3465
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.907461504248453e-06,
+      "loss": 0.0542,
+      "step": 3468
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.846033976459566e-06,
+      "loss": 0.0528,
+      "step": 3471
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.784825653287664e-06,
+      "loss": 0.0505,
+      "step": 3474
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.723836853021832e-06,
+      "loss": 0.0545,
+      "step": 3477
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 7.663067892809589e-06,
+      "loss": 0.056,
+      "step": 3480
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.602519088655307e-06,
+      "loss": 0.0447,
+      "step": 3483
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.542190755418499e-06,
+      "loss": 0.0488,
+      "step": 3486
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.4820832068122125e-06,
+      "loss": 0.0547,
+      "step": 3489
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.422196755401406e-06,
+      "loss": 0.0498,
+      "step": 3492
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 7.362531712601261e-06,
+      "loss": 0.0405,
+      "step": 3495
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.303088388675694e-06,
+      "loss": 0.0459,
+      "step": 3498
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.243867092735612e-06,
+      "loss": 0.05,
+      "step": 3501
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.1848681327373545e-06,
+      "loss": 0.0503,
+      "step": 3504
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.126091815481112e-06,
+      "loss": 0.0528,
+      "step": 3507
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 7.067538446609295e-06,
+      "loss": 0.0487,
+      "step": 3510
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 7.009208330604971e-06,
+      "loss": 0.0492,
+      "step": 3513
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.9511017707902806e-06,
+      "loss": 0.0525,
+      "step": 3516
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.893219069324824e-06,
+      "loss": 0.0384,
+      "step": 3519
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.835560527204143e-06,
+      "loss": 0.0474,
+      "step": 3522
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.77812644425811e-06,
+      "loss": 0.0514,
+      "step": 3525
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.720917119149417e-06,
+      "loss": 0.0552,
+      "step": 3528
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.66393284937194e-06,
+      "loss": 0.0527,
+      "step": 3531
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.607173931249316e-06,
+      "loss": 0.0505,
+      "step": 3534
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.550640659933277e-06,
+      "loss": 0.0474,
+      "step": 3537
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.4943333294021915e-06,
+      "loss": 0.0466,
+      "step": 3540
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.438252232459535e-06,
+      "loss": 0.0558,
+      "step": 3543
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.382397660732276e-06,
+      "loss": 0.0514,
+      "step": 3546
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.326769904669521e-06,
+      "loss": 0.0425,
+      "step": 3549
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.271369253540865e-06,
+      "loss": 0.0463,
+      "step": 3552
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.216195995434931e-06,
+      "loss": 0.0487,
+      "step": 3555
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.161250417257902e-06,
+      "loss": 0.0459,
+      "step": 3558
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.1065328047319736e-06,
+      "loss": 0.0517,
+      "step": 3561
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.052043442393928e-06,
+      "loss": 0.0542,
+      "step": 3564
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.997782613593606e-06,
+      "loss": 0.0488,
+      "step": 3567
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.943750600492459e-06,
+      "loss": 0.0472,
+      "step": 3570
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.889947684062075e-06,
+      "loss": 0.0559,
+      "step": 3573
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.836374144082718e-06,
+      "loss": 0.0466,
+      "step": 3576
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 5.783030259141881e-06,
+      "loss": 0.0558,
+      "step": 3579
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.729916306632782e-06,
+      "loss": 0.054,
+      "step": 3582
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.677032562753049e-06,
+      "loss": 0.0515,
+      "step": 3585
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.6243793025031354e-06,
+      "loss": 0.0518,
+      "step": 3588
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.571956799684996e-06,
+      "loss": 0.0498,
+      "step": 3591
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 5.519765326900617e-06,
+      "loss": 0.0542,
+      "step": 3594
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.46780515555057e-06,
+      "loss": 0.0532,
+      "step": 3597
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.416076555832706e-06,
+      "loss": 0.0524,
+      "step": 3600
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.3645797967406204e-06,
+      "loss": 0.0427,
+      "step": 3603
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 5.313315146062348e-06,
+      "loss": 0.0436,
+      "step": 3606
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.2622828703789315e-06,
+      "loss": 0.0347,
+      "step": 3609
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.2114832350630365e-06,
+      "loss": 0.0506,
+      "step": 3612
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.16091650427758e-06,
+      "loss": 0.0509,
+      "step": 3615
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.110582940974351e-06,
+      "loss": 0.0543,
+      "step": 3618
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 5.060482806892647e-06,
+      "loss": 0.0509,
+      "step": 3621
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 5.0106163625579124e-06,
+      "loss": 0.0503,
+      "step": 3624
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 4.960983867280384e-06,
+      "loss": 0.0494,
+      "step": 3627
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 4.911585579153732e-06,
+      "loss": 0.0571,
+      "step": 3630
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 4.862421755053747e-06,
+      "loss": 0.0474,
+      "step": 3633
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 4.813492650636964e-06,
+      "loss": 0.0461,
+      "step": 3636
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.764798520339364e-06,
+      "loss": 0.0474,
+      "step": 3639
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.7163396173750375e-06,
+      "loss": 0.0522,
+      "step": 3642
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.668116193734878e-06,
+      "loss": 0.0494,
+      "step": 3645
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.620128500185233e-06,
+      "loss": 0.0554,
+      "step": 3648
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.57237678626668e-06,
+      "loss": 0.0473,
+      "step": 3651
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.5248613002926485e-06,
+      "loss": 0.0449,
+      "step": 3654
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.477582289348164e-06,
+      "loss": 0.0494,
+      "step": 3657
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.430539999288574e-06,
+      "loss": 0.0439,
+      "step": 3660
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.383734674738216e-06,
+      "loss": 0.0499,
+      "step": 3663
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.337166559089239e-06,
+      "loss": 0.05,
+      "step": 3666
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.2908358945002505e-06,
+      "loss": 0.0428,
+      "step": 3669
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.24474292189509e-06,
+      "loss": 0.0549,
+      "step": 3672
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.198887880961583e-06,
+      "loss": 0.0508,
+      "step": 3675
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.153271010150289e-06,
+      "loss": 0.0557,
+      "step": 3678
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 4.107892546673259e-06,
+      "loss": 0.0537,
+      "step": 3681
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 4.0627527265028e-06,
+      "loss": 0.0565,
+      "step": 3684
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 4.017851784370252e-06,
+      "loss": 0.0481,
+      "step": 3687
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.973189953764766e-06,
+      "loss": 0.047,
+      "step": 3690
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 3.928767466932104e-06,
+      "loss": 0.0457,
+      "step": 3693
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.884584554873405e-06,
+      "loss": 0.0396,
+      "step": 3696
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.840641447343968e-06,
+      "loss": 0.05,
+      "step": 3699
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.7969383728521544e-06,
+      "loss": 0.0419,
+      "step": 3702
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 3.7534755586580707e-06,
+      "loss": 0.0404,
+      "step": 3705
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.7102532307724757e-06,
+      "loss": 0.0469,
+      "step": 3708
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.667271613955575e-06,
+      "loss": 0.0473,
+      "step": 3711
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.6245309317158193e-06,
+      "loss": 0.0619,
+      "step": 3714
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.5820314063088388e-06,
+      "loss": 0.0464,
+      "step": 3717
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 3.5397732587361785e-06,
+      "loss": 0.0498,
+      "step": 3720
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.4977567087442153e-06,
+      "loss": 0.0479,
+      "step": 3723
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.4559819748229807e-06,
+      "loss": 0.0451,
+      "step": 3726
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.4144492742050503e-06,
+      "loss": 0.0503,
+      "step": 3729
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.3731588228644086e-06,
+      "loss": 0.051,
+      "step": 3732
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 3.3321108355153063e-06,
+      "loss": 0.054,
+      "step": 3735
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.291305525611177e-06,
+      "loss": 0.0451,
+      "step": 3738
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.25074310534349e-06,
+      "loss": 0.0402,
+      "step": 3741
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.2104237856406703e-06,
+      "loss": 0.0565,
+      "step": 3744
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 3.1703477761670095e-06,
+      "loss": 0.0441,
+      "step": 3747
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.1305152853215504e-06,
+      "loss": 0.0559,
+      "step": 3750
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.0909265202370197e-06,
+      "loss": 0.0509,
+      "step": 3753
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.0515816867787426e-06,
+      "loss": 0.0435,
+      "step": 3756
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.012480989543581e-06,
+      "loss": 0.0566,
+      "step": 3759
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.973624631858879e-06,
+      "loss": 0.0496,
+      "step": 3762
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.9350128157813462e-06,
+      "loss": 0.0436,
+      "step": 3765
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.896645742096116e-06,
+      "loss": 0.0534,
+      "step": 3768
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.858523610315589e-06,
+      "loss": 0.0392,
+      "step": 3771
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.8206466186784743e-06,
+      "loss": 0.043,
+      "step": 3774
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.783014964148706e-06,
+      "loss": 0.0492,
+      "step": 3777
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.7456288424144547e-06,
+      "loss": 0.0528,
+      "step": 3780
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.7084884478870977e-06,
+      "loss": 0.0489,
+      "step": 3783
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.6715939737002095e-06,
+      "loss": 0.0688,
+      "step": 3786
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.6349456117085447e-06,
+      "loss": 0.0478,
+      "step": 3789
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.5985435524870594e-06,
+      "loss": 0.0455,
+      "step": 3792
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.5623879853299115e-06,
+      "loss": 0.0557,
+      "step": 3795
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.5264790982494733e-06,
+      "loss": 0.0462,
+      "step": 3798
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.490817077975366e-06,
+      "loss": 0.0453,
+      "step": 3801
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.455402109953464e-06,
+      "loss": 0.0491,
+      "step": 3804
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.42023437834496e-06,
+      "loss": 0.0526,
+      "step": 3807
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.385314066025396e-06,
+      "loss": 0.0478,
+      "step": 3810
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3506413545837047e-06,
+      "loss": 0.0479,
+      "step": 3813
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3162164243212493e-06,
+      "loss": 0.0605,
+      "step": 3816
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.2820394542509506e-06,
+      "loss": 0.0467,
+      "step": 3819
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.248110622096289e-06,
+      "loss": 0.0542,
+      "step": 3822
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.214430104290399e-06,
+      "loss": 0.0468,
+      "step": 3825
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1809980759751715e-06,
+      "loss": 0.0534,
+      "step": 3828
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.1478147110003122e-06,
+      "loss": 0.0555,
+      "step": 3831
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.1148801819224683e-06,
+      "loss": 0.052,
+      "step": 3834
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.0821946600043163e-06,
+      "loss": 0.0468,
+      "step": 3837
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.0497583152136513e-06,
+      "loss": 0.0529,
+      "step": 3840
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.017571316222555e-06,
+      "loss": 0.0443,
+      "step": 3843
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.985633830406464e-06,
+      "loss": 0.0535,
+      "step": 3846
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.9539460238433283e-06,
+      "loss": 0.0526,
+      "step": 3849
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.92250806131275e-06,
+      "loss": 0.0456,
+      "step": 3852
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.891320106295108e-06,
+      "loss": 0.0521,
+      "step": 3855
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.8603823209707271e-06,
+      "loss": 0.0411,
+      "step": 3858
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.8296948662190226e-06,
+      "loss": 0.0433,
+      "step": 3861
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.7992579016176625e-06,
+      "loss": 0.0475,
+      "step": 3864
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.769071585441745e-06,
+      "loss": 0.0495,
+      "step": 3867
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.7391360746629782e-06,
+      "loss": 0.0536,
+      "step": 3870
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.7094515249488518e-06,
+      "loss": 0.0544,
+      "step": 3873
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.680018090661839e-06,
+      "loss": 0.0567,
+      "step": 3876
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6508359248585848e-06,
+      "loss": 0.0468,
+      "step": 3879
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6219051792891183e-06,
+      "loss": 0.0478,
+      "step": 3882
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.5932260043960535e-06,
+      "loss": 0.0486,
+      "step": 3885
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.564798549313823e-06,
+      "loss": 0.0353,
+      "step": 3888
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5366229618678784e-06,
+      "loss": 0.0516,
+      "step": 3891
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5086993885739476e-06,
+      "loss": 0.0485,
+      "step": 3894
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.4810279746372614e-06,
+      "loss": 0.0442,
+      "step": 3897
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.4536088639517887e-06,
+      "loss": 0.0409,
+      "step": 3900
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.4264421990995036e-06,
+      "loss": 0.0557,
+      "step": 3903
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.3995281213496359e-06,
+      "loss": 0.0489,
+      "step": 3906
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.3728667706579434e-06,
+      "loss": 0.0544,
+      "step": 3909
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.3464582856659747e-06,
+      "loss": 0.0473,
+      "step": 3912
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.3203028037003518e-06,
+      "loss": 0.0514,
+      "step": 3915
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.294400460772055e-06,
+      "loss": 0.0439,
+      "step": 3918
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2687513915757342e-06,
+      "loss": 0.0489,
+      "step": 3921
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2433557294889653e-06,
+      "loss": 0.0537,
+      "step": 3924
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2182136065716054e-06,
+      "loss": 0.049,
+      "step": 3927
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1933251535650726e-06,
+      "loss": 0.0543,
+      "step": 3930
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.168690499891678e-06,
+      "loss": 0.0482,
+      "step": 3933
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1443097736539621e-06,
+      "loss": 0.0511,
+      "step": 3936
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1201831016340036e-06,
+      "loss": 0.048,
+      "step": 3939
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.0963106092927888e-06,
+      "loss": 0.0499,
+      "step": 3942
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.0726924207695387e-06,
+      "loss": 0.0444,
+      "step": 3945
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.0493286588810657e-06,
+      "loss": 0.045,
+      "step": 3948
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.026219445121146e-06,
+      "loss": 0.0487,
+      "step": 3951
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.0033648996598866e-06,
+      "loss": 0.0489,
+      "step": 3954
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 9.807651413430874e-07,
+      "loss": 0.048,
+      "step": 3957
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 9.584202876916304e-07,
+      "loss": 0.0487,
+      "step": 3960
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 9.363304549008689e-07,
+      "loss": 0.0487,
+      "step": 3963
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 9.144957578400281e-07,
+      "loss": 0.0463,
+      "step": 3966
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.929163100516058e-07,
+      "loss": 0.0458,
+      "step": 3969
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.715922237507723e-07,
+      "loss": 0.0544,
+      "step": 3972
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.505236098247993e-07,
+      "loss": 0.0497,
+      "step": 3975
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.297105778324654e-07,
+      "loss": 0.0465,
+      "step": 3978
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.091532360035182e-07,
+      "loss": 0.0538,
+      "step": 3981
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 7.888516912380794e-07,
+      "loss": 0.0511,
+      "step": 3984
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 7.688060491061078e-07,
+      "loss": 0.048,
+      "step": 3987
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 7.490164138468369e-07,
+      "loss": 0.0564,
+      "step": 3990
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 7.294828883682492e-07,
+      "loss": 0.044,
+      "step": 3993
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 7.102055742465252e-07,
+      "loss": 0.044,
+      "step": 3996
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 6.911845717255283e-07,
+      "loss": 0.0476,
+      "step": 3999
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 6.724199797162711e-07,
+      "loss": 0.0581,
+      "step": 4002
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 6.539118957964052e-07,
+      "loss": 0.0562,
+      "step": 4005
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 6.356604162097268e-07,
+      "loss": 0.0509,
+      "step": 4008
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 6.176656358656608e-07,
+      "loss": 0.0568,
+      "step": 4011
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 5.999276483387605e-07,
+      "loss": 0.0457,
+      "step": 4014
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.824465458682538e-07,
+      "loss": 0.0387,
+      "step": 4017
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.65222419357514e-07,
+      "loss": 0.0464,
+      "step": 4020
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.482553583736449e-07,
+      "loss": 0.0582,
+      "step": 4023
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.315454511469698e-07,
+      "loss": 0.0537,
+      "step": 4026
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 5.150927845705811e-07,
+      "loss": 0.0486,
+      "step": 4029
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.988974441999084e-07,
+      "loss": 0.0537,
+      "step": 4032
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.82959514252257e-07,
+      "loss": 0.0456,
+      "step": 4035
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.672790776063696e-07,
+      "loss": 0.0468,
+      "step": 4038
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.518562158020045e-07,
+      "loss": 0.0491,
+      "step": 4041
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.366910090395027e-07,
+      "loss": 0.043,
+      "step": 4044
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 4.2178353617938207e-07,
+      "loss": 0.0456,
+      "step": 4047
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 4.0713387474190514e-07,
+      "loss": 0.0522,
+      "step": 4050
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.927421009067012e-07,
+      "loss": 0.0474,
+      "step": 4053
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.7860828951236103e-07,
+      "loss": 0.0476,
+      "step": 4056
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.6473251405604317e-07,
+      "loss": 0.0492,
+      "step": 4059
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.5111484669309024e-07,
+      "loss": 0.0601,
+      "step": 4062
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.377553582366688e-07,
+      "loss": 0.0495,
+      "step": 4065
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.2465411815738036e-07,
+      "loss": 0.0512,
+      "step": 4068
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.1181119458291164e-07,
+      "loss": 0.0375,
+      "step": 4071
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.992266542976907e-07,
+      "loss": 0.0502,
+      "step": 4074
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.869005627425092e-07,
+      "loss": 0.0515,
+      "step": 4077
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.748329840142172e-07,
+      "loss": 0.0606,
+      "step": 4080
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.6302398086536784e-07,
+      "loss": 0.0459,
+      "step": 4083
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.514736147038954e-07,
+      "loss": 0.0495,
+      "step": 4086
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.401819455927934e-07,
+      "loss": 0.0461,
+      "step": 4089
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.291490322498202e-07,
+      "loss": 0.0386,
+      "step": 4092
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.1837493204716064e-07,
+      "loss": 0.0509,
+      "step": 4095
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.0785970101115938e-07,
+      "loss": 0.0541,
+      "step": 4098
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.976033938220212e-07,
+      "loss": 0.0521,
+      "step": 4101
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.8760606381350575e-07,
+      "loss": 0.0504,
+      "step": 4104
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.7786776297268327e-07,
+      "loss": 0.0496,
+      "step": 4107
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.683885419396347e-07,
+      "loss": 0.0411,
+      "step": 4110
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.591684500072077e-07,
+      "loss": 0.0573,
+      "step": 4113
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.5020753512076104e-07,
+      "loss": 0.0547,
+      "step": 4116
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.4150584387789267e-07,
+      "loss": 0.049,
+      "step": 4119
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3306342152822893e-07,
+      "loss": 0.0468,
+      "step": 4122
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.248803119731634e-07,
+      "loss": 0.0454,
+      "step": 4125
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.1695655776563508e-07,
+      "loss": 0.0436,
+      "step": 4128
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0929220010992281e-07,
+      "loss": 0.0527,
+      "step": 4131
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0188727886141225e-07,
+      "loss": 0.0495,
+      "step": 4134
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.474183252639601e-08,
+      "loss": 0.0489,
+      "step": 4137
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.785589826186825e-08,
+      "loss": 0.0474,
+      "step": 4140
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 8.122951187534144e-08,
+      "loss": 0.0508,
+      "step": 4143
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 7.486270782465222e-08,
+      "loss": 0.048,
+      "step": 4146
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 6.875551921778357e-08,
+      "loss": 0.0614,
+      "step": 4149
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 6.290797781269842e-08,
+      "loss": 0.0474,
+      "step": 4152
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 5.7320114017161976e-08,
+      "loss": 0.0494,
+      "step": 4155
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.199195688860292e-08,
+      "loss": 0.0493,
+      "step": 4158
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 4.692353413393025e-08,
+      "loss": 0.0523,
+      "step": 4161
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 4.2114872109438876e-08,
+      "loss": 0.0512,
+      "step": 4164
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.756599582060982e-08,
+      "loss": 0.0444,
+      "step": 4167
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.3276928922043595e-08,
+      "loss": 0.0579,
+      "step": 4170
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.9247693717288083e-08,
+      "loss": 0.0397,
+      "step": 4173
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.5478311158749768e-08,
+      "loss": 0.043,
+      "step": 4176
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.196880084757158e-08,
+      "loss": 0.038,
+      "step": 4179
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.8719181033549637e-08,
+      "loss": 0.0533,
+      "step": 4182
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.572946861500557e-08,
+      "loss": 0.0576,
+      "step": 4185
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2999679138725462e-08,
+      "loss": 0.0454,
+      "step": 4188
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0529826799882125e-08,
+      "loss": 0.0467,
+      "step": 4191
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.31992444193519e-09,
+      "loss": 0.0484,
+      "step": 4194
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 6.3699835565922406e-09,
+      "loss": 0.0481,
+      "step": 4197
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.680014283725553e-09,
+      "loss": 0.0476,
+      "step": 4200
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.250025411333235e-09,
+      "loss": 0.0444,
+      "step": 4203
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.0800243755003667e-09,
+      "loss": 0.0439,
+      "step": 4206
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.17001726033239e-09,
+      "loss": 0.0498,
+      "step": 4209
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.200087979662093e-10,
+      "loss": 0.0476,
+      "step": 4212
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.3000236849802783e-10,
+      "loss": 0.0439,
+      "step": 4215
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.0418,
+      "step": 4218
+    },
+    {
+      "epoch": 3.0,
+      "step": 4218,
+      "total_flos": 3.2919915274960896e+17,
+      "train_loss": 0.18810525719710525,
+      "train_runtime": 27776.8918,
+      "train_samples_per_second": 19.449,
+      "train_steps_per_second": 0.152
+    }
+  ],
+  "max_steps": 4218,
+  "num_train_epochs": 3,
+  "total_flos": 3.2919915274960896e+17,
+  "trial_name": null,
+  "trial_params": null
+}