diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4819 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.37280395172188824,
+  "eval_steps": 500,
+  "global_step": 4000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 3.3689,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 3.3976,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-05,
+      "loss": 3.4177,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 3.3808,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5e-05,
+      "loss": 3.372,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 3.3435,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5e-05,
+      "loss": 3.2703,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4e-05,
+      "loss": 3.226,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.5e-05,
+      "loss": 3.1441,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 3.084,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 2.9698,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 2.8584,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 2.8436,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7e-05,
+      "loss": 2.7909,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 2.7415,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8e-05,
+      "loss": 2.6519,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.5e-05,
+      "loss": 2.6261,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9e-05,
+      "loss": 2.6152,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.5e-05,
+      "loss": 2.5768,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 2.5184,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999994539972005e-05,
+      "loss": 2.5154,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999978159899939e-05,
+      "loss": 2.4047,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999950859819579e-05,
+      "loss": 2.4332,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999912639790549e-05,
+      "loss": 2.4058,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99986349989632e-05,
+      "loss": 2.3825,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999803440244217e-05,
+      "loss": 2.3568,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.99973246096541e-05,
+      "loss": 2.3518,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999650562214917e-05,
+      "loss": 2.3121,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999557744171606e-05,
+      "loss": 2.3241,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999454007038194e-05,
+      "loss": 2.3026,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999339351041244e-05,
+      "loss": 2.2761,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.999213776431165e-05,
+      "loss": 2.2644,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.999077283482214e-05,
+      "loss": 2.282,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.998929872492492e-05,
+      "loss": 2.2837,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.998771543783947e-05,
+      "loss": 2.226,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.998602297702371e-05,
+      "loss": 2.2303,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.9984221346174e-05,
+      "loss": 2.2321,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.998231054922511e-05,
+      "loss": 2.2387,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.998029059035026e-05,
+      "loss": 2.2234,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.997816147396102e-05,
+      "loss": 2.2029,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.997592320470746e-05,
+      "loss": 2.1887,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.997357578747792e-05,
+      "loss": 2.1855,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.997111922739926e-05,
+      "loss": 2.1253,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.996855352983658e-05,
+      "loss": 2.1688,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.996587870039341e-05,
+      "loss": 2.2177,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.996309474491161e-05,
+      "loss": 2.1997,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.996020166947136e-05,
+      "loss": 2.1497,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.99571994803912e-05,
+      "loss": 2.1817,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.995408818422789e-05,
+      "loss": 2.1096,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.995086778777658e-05,
+      "loss": 2.1553,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994753829807064e-05,
+      "loss": 2.1459,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994409972238172e-05,
+      "loss": 2.1393,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9.994055206821968e-05,
+      "loss": 2.1548,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.993689534333266e-05,
+      "loss": 2.1226,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.993312955570699e-05,
+      "loss": 2.1166,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.99292547135672e-05,
+      "loss": 2.1187,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.992527082537595e-05,
+      "loss": 2.119,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.992117789983413e-05,
+      "loss": 2.0724,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.991697594588073e-05,
+      "loss": 2.0518,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.991266497269286e-05,
+      "loss": 2.0458,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.990824498968573e-05,
+      "loss": 2.1092,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.990371600651263e-05,
+      "loss": 2.0913,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.989907803306493e-05,
+      "loss": 2.0883,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.9894331079472e-05,
+      "loss": 2.054,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.988947515610124e-05,
+      "loss": 2.1106,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.988451027355805e-05,
+      "loss": 2.048,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.987943644268578e-05,
+      "loss": 2.0647,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.987425367456573e-05,
+      "loss": 2.0585,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.986896198051712e-05,
+      "loss": 2.0544,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.98635613720971e-05,
+      "loss": 2.0515,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985805186110064e-05,
+      "loss": 2.0492,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.985243345956054e-05,
+      "loss": 2.0497,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984670617974752e-05,
+      "loss": 2.0459,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.984087003416997e-05,
+      "loss": 2.0444,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.983492503557412e-05,
+      "loss": 1.9953,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.982887119694391e-05,
+      "loss": 2.0654,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.982270853150097e-05,
+      "loss": 2.0119,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.981643705270466e-05,
+      "loss": 2.0142,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.981005677425196e-05,
+      "loss": 1.9854,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.980356771007745e-05,
+      "loss": 2.0105,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.979696987435332e-05,
+      "loss": 2.0252,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.979026328148933e-05,
+      "loss": 2.0253,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.978344794613277e-05,
+      "loss": 2.0274,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.977652388316836e-05,
+      "loss": 2.0209,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.976949110771838e-05,
+      "loss": 2.0166,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.976234963514244e-05,
+      "loss": 1.9947,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.975509948103765e-05,
+      "loss": 1.97,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.97477406612384e-05,
+      "loss": 1.9561,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.974027319181642e-05,
+      "loss": 2.0248,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.973269708908078e-05,
+      "loss": 1.9457,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.972501236957776e-05,
+      "loss": 1.9587,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.971721905009086e-05,
+      "loss": 1.9487,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970931714764081e-05,
+      "loss": 1.9365,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.970130667948541e-05,
+      "loss": 1.9687,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.969318766311963e-05,
+      "loss": 1.9459,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.968496011627551e-05,
+      "loss": 1.9277,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.967662405692209e-05,
+      "loss": 1.9468,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.966817950326541e-05,
+      "loss": 2.005,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.965962647374848e-05,
+      "loss": 1.9656,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.965096498705121e-05,
+      "loss": 1.9499,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.964219506209039e-05,
+      "loss": 1.9517,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.963331671801961e-05,
+      "loss": 1.9082,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.962432997422932e-05,
+      "loss": 1.9464,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.961523485034662e-05,
+      "loss": 1.9419,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.960603136623538e-05,
+      "loss": 1.9711,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.959671954199612e-05,
+      "loss": 1.9587,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.958729939796597e-05,
+      "loss": 1.9386,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.957777095471862e-05,
+      "loss": 1.9042,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.95681342330643e-05,
+      "loss": 1.9234,
+      "step": 545
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.955838925404971e-05,
+      "loss": 1.9313,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.954853603895802e-05,
+      "loss": 1.9266,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.953857460930872e-05,
+      "loss": 1.9387,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.952850498685772e-05,
+      "loss": 1.9255,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.951832719359717e-05,
+      "loss": 1.9483,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.950804125175551e-05,
+      "loss": 1.8852,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.94976471837973e-05,
+      "loss": 1.9282,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.948714501242335e-05,
+      "loss": 1.9281,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.947653476057051e-05,
+      "loss": 1.8943,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.946581645141167e-05,
+      "loss": 1.8684,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.945499010835575e-05,
+      "loss": 1.9072,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.944405575504761e-05,
+      "loss": 1.8783,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.943301341536799e-05,
+      "loss": 1.8541,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.942186311343349e-05,
+      "loss": 1.8921,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.94106048735965e-05,
+      "loss": 1.9469,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.939923872044512e-05,
+      "loss": 1.8303,
+      "step": 625
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.938776467880316e-05,
+      "loss": 1.9101,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.937618277373009e-05,
+      "loss": 1.9002,
+      "step": 635
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.936449303052086e-05,
+      "loss": 1.8879,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.935269547470607e-05,
+      "loss": 1.8567,
+      "step": 645
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.934079013205166e-05,
+      "loss": 1.8308,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.932877702855904e-05,
+      "loss": 1.8308,
+      "step": 655
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.931665619046501e-05,
+      "loss": 1.8671,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.930442764424155e-05,
+      "loss": 1.8329,
+      "step": 665
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.929209141659598e-05,
+      "loss": 1.8379,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.927964753447074e-05,
+      "loss": 1.8479,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.926709602504344e-05,
+      "loss": 1.8711,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.925443691572668e-05,
+      "loss": 1.8552,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.924167023416813e-05,
+      "loss": 1.8451,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.922879600825033e-05,
+      "loss": 1.8401,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.921581426609074e-05,
+      "loss": 1.8684,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.920272503604167e-05,
+      "loss": 1.8408,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.91895283466901e-05,
+      "loss": 1.8794,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.917622422685776e-05,
+      "loss": 1.8686,
+      "step": 715
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.916281270560101e-05,
+      "loss": 1.8495,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.914929381221074e-05,
+      "loss": 1.8613,
+      "step": 725
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.91356675762124e-05,
+      "loss": 1.8185,
+      "step": 730
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.91219340273658e-05,
+      "loss": 1.8896,
+      "step": 735
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.91080931956652e-05,
+      "loss": 1.8572,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.909414511133909e-05,
+      "loss": 1.8716,
+      "step": 745
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.90800898048503e-05,
+      "loss": 1.8332,
+      "step": 750
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.906592730689573e-05,
+      "loss": 1.8083,
+      "step": 755
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.905165764840645e-05,
+      "loss": 1.7887,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.903728086054755e-05,
+      "loss": 1.8642,
+      "step": 765
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.90227969747181e-05,
+      "loss": 1.8146,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.900820602255108e-05,
+      "loss": 1.8422,
+      "step": 775
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.899350803591326e-05,
+      "loss": 1.8225,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.897870304690523e-05,
+      "loss": 1.8178,
+      "step": 785
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.896379108786126e-05,
+      "loss": 1.829,
+      "step": 790
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.894877219134921e-05,
+      "loss": 1.8032,
+      "step": 795
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.893364639017055e-05,
+      "loss": 1.8336,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.891841371736017e-05,
+      "loss": 1.8154,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.890307420618641e-05,
+      "loss": 1.8013,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.888762789015095e-05,
+      "loss": 1.7777,
+      "step": 815
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.88720748029887e-05,
+      "loss": 1.7992,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.885641497866777e-05,
+      "loss": 1.7858,
+      "step": 825
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.88406484513894e-05,
+      "loss": 1.801,
+      "step": 830
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.882477525558786e-05,
+      "loss": 1.799,
+      "step": 835
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.880879542593038e-05,
+      "loss": 1.7973,
+      "step": 840
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.879270899731711e-05,
+      "loss": 1.7619,
+      "step": 845
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.877651600488099e-05,
+      "loss": 1.7945,
+      "step": 850
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.876021648398766e-05,
+      "loss": 1.7808,
+      "step": 855
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.874381047023549e-05,
+      "loss": 1.7764,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.87272979994554e-05,
+      "loss": 1.7606,
+      "step": 865
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.871067910771076e-05,
+      "loss": 1.8235,
+      "step": 870
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.869395383129749e-05,
+      "loss": 1.7805,
+      "step": 875
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.867712220674374e-05,
+      "loss": 1.7579,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.866018427080996e-05,
+      "loss": 1.7907,
+      "step": 885
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.86431400604888e-05,
+      "loss": 1.774,
+      "step": 890
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.862598961300503e-05,
+      "loss": 1.7479,
+      "step": 895
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.860873296581539e-05,
+      "loss": 1.8033,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.85913701566086e-05,
+      "loss": 1.7726,
+      "step": 905
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.857390122330522e-05,
+      "loss": 1.7737,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.855632620405762e-05,
+      "loss": 1.8069,
+      "step": 915
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.853864513724981e-05,
+      "loss": 1.8007,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.852085806149746e-05,
+      "loss": 1.7751,
+      "step": 925
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.850296501564773e-05,
+      "loss": 1.8256,
+      "step": 930
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.848496603877925e-05,
+      "loss": 1.7958,
+      "step": 935
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.846686117020196e-05,
+      "loss": 1.7686,
+      "step": 940
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.844865044945713e-05,
+      "loss": 1.7612,
+      "step": 945
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.843033391631713e-05,
+      "loss": 1.8147,
+      "step": 950
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.84119116107855e-05,
+      "loss": 1.7497,
+      "step": 955
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.839338357309679e-05,
+      "loss": 1.77,
+      "step": 960
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.837474984371642e-05,
+      "loss": 1.7907,
+      "step": 965
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.835601046334065e-05,
+      "loss": 1.7852,
+      "step": 970
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.833716547289648e-05,
+      "loss": 1.7559,
+      "step": 975
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.831821491354162e-05,
+      "loss": 1.7679,
+      "step": 980
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.829915882666429e-05,
+      "loss": 1.746,
+      "step": 985
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.827999725388322e-05,
+      "loss": 1.7945,
+      "step": 990
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.826073023704745e-05,
+      "loss": 1.7727,
+      "step": 995
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.82413578182364e-05,
+      "loss": 1.7675,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.822188003975962e-05,
+      "loss": 1.7334,
+      "step": 1005
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.820229694415681e-05,
+      "loss": 1.7625,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.81826085741977e-05,
+      "loss": 1.7423,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.816281497288186e-05,
+      "loss": 1.7642,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.814291618343876e-05,
+      "loss": 1.7183,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.81229122493276e-05,
+      "loss": 1.7055,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.810280321423715e-05,
+      "loss": 1.7162,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.80825891220858e-05,
+      "loss": 1.776,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.806227001702135e-05,
+      "loss": 1.7554,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.804184594342096e-05,
+      "loss": 1.7621,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.8021316945891e-05,
+      "loss": 1.7487,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.800068306926708e-05,
+      "loss": 1.738,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.797994435861379e-05,
+      "loss": 1.7301,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.79591008592247e-05,
+      "loss": 1.7027,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.793815261662226e-05,
+      "loss": 1.7678,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.791709967655766e-05,
+      "loss": 1.6961,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.789594208501076e-05,
+      "loss": 1.7152,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.787467988818997e-05,
+      "loss": 1.7228,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.785331313253219e-05,
+      "loss": 1.7126,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.783184186470262e-05,
+      "loss": 1.6957,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.781026613159476e-05,
+      "loss": 1.7471,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.778858598033028e-05,
+      "loss": 1.7292,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.776680145825883e-05,
+      "loss": 1.7031,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.774491261295807e-05,
+      "loss": 1.7118,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.772291949223349e-05,
+      "loss": 1.7266,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.77008221441183e-05,
+      "loss": 1.7371,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.767862061687337e-05,
+      "loss": 1.7422,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.765631495898709e-05,
+      "loss": 1.6968,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.763390521917523e-05,
+      "loss": 1.7195,
+      "step": 1145
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.761139144638094e-05,
+      "loss": 1.7593,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.758877368977455e-05,
+      "loss": 1.7263,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.756605199875348e-05,
+      "loss": 1.7469,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.754322642294217e-05,
+      "loss": 1.7119,
+      "step": 1165
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.752029701219196e-05,
+      "loss": 1.6991,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.749726381658089e-05,
+      "loss": 1.7405,
+      "step": 1175
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.747412688641372e-05,
+      "loss": 1.7007,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.745088627222181e-05,
+      "loss": 1.7548,
+      "step": 1185
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.74275420247629e-05,
+      "loss": 1.6719,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.740409419502105e-05,
+      "loss": 1.7285,
+      "step": 1195
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.738054283420665e-05,
+      "loss": 1.7057,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.735688799375609e-05,
+      "loss": 1.6361,
+      "step": 1205
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.733312972533182e-05,
+      "loss": 1.6987,
+      "step": 1210
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.730926808082215e-05,
+      "loss": 1.7064,
+      "step": 1215
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.72853031123412e-05,
+      "loss": 1.6926,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.726123487222871e-05,
+      "loss": 1.6368,
+      "step": 1225
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.723706341305002e-05,
+      "loss": 1.7183,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.721278878759582e-05,
+      "loss": 1.7138,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.71884110488822e-05,
+      "loss": 1.6776,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.716393025015041e-05,
+      "loss": 1.7207,
+      "step": 1245
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.713934644486675e-05,
+      "loss": 1.6893,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.711465968672257e-05,
+      "loss": 1.6584,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.708987002963404e-05,
+      "loss": 1.6757,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.7064977527742e-05,
+      "loss": 1.6782,
+      "step": 1265
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.703998223541196e-05,
+      "loss": 1.6478,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.701488420723395e-05,
+      "loss": 1.6956,
+      "step": 1275
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.698968349802233e-05,
+      "loss": 1.6849,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.696438016281573e-05,
+      "loss": 1.6495,
+      "step": 1285
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.69389742568769e-05,
+      "loss": 1.6761,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.691346583569265e-05,
+      "loss": 1.6705,
+      "step": 1295
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.688785495497366e-05,
+      "loss": 1.669,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.686214167065434e-05,
+      "loss": 1.6822,
+      "step": 1305
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.683632603889283e-05,
+      "loss": 1.7048,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.681040811607075e-05,
+      "loss": 1.681,
+      "step": 1315
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.678438795879312e-05,
+      "loss": 1.6281,
+      "step": 1320
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.675826562388828e-05,
+      "loss": 1.6707,
+      "step": 1325
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.673204116840768e-05,
+      "loss": 1.6443,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.670571464962583e-05,
+      "loss": 1.6993,
+      "step": 1335
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.667928612504015e-05,
+      "loss": 1.6534,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.665275565237082e-05,
+      "loss": 1.6662,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.66261232895607e-05,
+      "loss": 1.6289,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.659938909477517e-05,
+      "loss": 1.6691,
+      "step": 1355
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.6572553126402e-05,
+      "loss": 1.6578,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.654561544305125e-05,
+      "loss": 1.6734,
+      "step": 1365
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.651857610355512e-05,
+      "loss": 1.6563,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.649143516696784e-05,
+      "loss": 1.6246,
+      "step": 1375
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.646419269256552e-05,
+      "loss": 1.709,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.643684873984602e-05,
+      "loss": 1.6707,
+      "step": 1385
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.640940336852884e-05,
+      "loss": 1.6361,
+      "step": 1390
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.638185663855497e-05,
+      "loss": 1.5929,
+      "step": 1395
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.63542086100868e-05,
+      "loss": 1.6479,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.632645934350792e-05,
+      "loss": 1.6897,
+      "step": 1405
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.629860889942302e-05,
+      "loss": 1.6457,
+      "step": 1410
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.627065733865781e-05,
+      "loss": 1.6222,
+      "step": 1415
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.624260472225882e-05,
+      "loss": 1.659,
+      "step": 1420
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.621445111149324e-05,
+      "loss": 1.6597,
+      "step": 1425
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.61861965678489e-05,
+      "loss": 1.6329,
+      "step": 1430
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.615784115303403e-05,
+      "loss": 1.6409,
+      "step": 1435
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.612938492897719e-05,
+      "loss": 1.6227,
+      "step": 1440
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.610082795782707e-05,
+      "loss": 1.6708,
+      "step": 1445
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.607217030195242e-05,
+      "loss": 1.6286,
+      "step": 1450
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.604341202394188e-05,
+      "loss": 1.6288,
+      "step": 1455
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.601455318660386e-05,
+      "loss": 1.6439,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.598559385296636e-05,
+      "loss": 1.6161,
+      "step": 1465
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.595653408627694e-05,
+      "loss": 1.681,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.59273739500024e-05,
+      "loss": 1.646,
+      "step": 1475
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.589811350782884e-05,
+      "loss": 1.6611,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.586875282366139e-05,
+      "loss": 1.6333,
+      "step": 1485
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.583929196162408e-05,
+      "loss": 1.6411,
+      "step": 1490
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.58097309860598e-05,
+      "loss": 1.6579,
+      "step": 1495
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.578006996153004e-05,
+      "loss": 1.6161,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.57503089528148e-05,
+      "loss": 1.6403,
+      "step": 1505
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.572044802491247e-05,
+      "loss": 1.627,
+      "step": 1510
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.569048724303964e-05,
+      "loss": 1.6097,
+      "step": 1515
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.566042667263101e-05,
+      "loss": 1.6172,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.563026637933918e-05,
+      "loss": 1.6401,
+      "step": 1525
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.56000064290346e-05,
+      "loss": 1.5866,
+      "step": 1530
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.556964688780529e-05,
+      "loss": 1.6104,
+      "step": 1535
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.553918782195688e-05,
+      "loss": 1.6477,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.550862929801228e-05,
+      "loss": 1.6379,
+      "step": 1545
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.547797138271166e-05,
+      "loss": 1.6234,
+      "step": 1550
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.544721414301225e-05,
+      "loss": 1.6281,
+      "step": 1555
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.54163576460882e-05,
+      "loss": 1.5896,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.538540195933044e-05,
+      "loss": 1.655,
+      "step": 1565
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.535434715034656e-05,
+      "loss": 1.6159,
+      "step": 1570
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.532319328696058e-05,
+      "loss": 1.5794,
+      "step": 1575
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.529194043721292e-05,
+      "loss": 1.5948,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.526058866936013e-05,
+      "loss": 1.6332,
+      "step": 1585
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.522913805187483e-05,
+      "loss": 1.6252,
+      "step": 1590
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.51975886534455e-05,
+      "loss": 1.627,
+      "step": 1595
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.516594054297642e-05,
+      "loss": 1.6466,
+      "step": 1600
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.513419378958739e-05,
+      "loss": 1.6431,
+      "step": 1605
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.510234846261368e-05,
+      "loss": 1.6338,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.507040463160583e-05,
+      "loss": 1.6619,
+      "step": 1615
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.503836236632955e-05,
+      "loss": 1.6408,
+      "step": 1620
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.50062217367655e-05,
+      "loss": 1.6272,
+      "step": 1625
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.497398281310914e-05,
+      "loss": 1.6198,
+      "step": 1630
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.49416456657707e-05,
+      "loss": 1.605,
+      "step": 1635
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.490921036537481e-05,
+      "loss": 1.6779,
+      "step": 1640
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.487667698276057e-05,
+      "loss": 1.6273,
+      "step": 1645
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.484404558898125e-05,
+      "loss": 1.655,
+      "step": 1650
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.481131625530417e-05,
+      "loss": 1.5789,
+      "step": 1655
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.477848905321055e-05,
+      "loss": 1.6236,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.474556405439538e-05,
+      "loss": 1.6144,
+      "step": 1665
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.471254133076724e-05,
+      "loss": 1.5772,
+      "step": 1670
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.467942095444809e-05,
+      "loss": 1.6018,
+      "step": 1675
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.464620299777324e-05,
+      "loss": 1.6139,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.461288753329106e-05,
+      "loss": 1.629,
+      "step": 1685
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.45794746337629e-05,
+      "loss": 1.5827,
+      "step": 1690
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.454596437216292e-05,
+      "loss": 1.6154,
+      "step": 1695
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.45123568216779e-05,
+      "loss": 1.6202,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.447865205570708e-05,
+      "loss": 1.5832,
+      "step": 1705
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.444485014786208e-05,
+      "loss": 1.5866,
+      "step": 1710
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.441095117196663e-05,
+      "loss": 1.6026,
+      "step": 1715
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.43769552020565e-05,
+      "loss": 1.6326,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.434286231237922e-05,
+      "loss": 1.6026,
+      "step": 1725
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.430867257739408e-05,
+      "loss": 1.6248,
+      "step": 1730
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.427438607177181e-05,
+      "loss": 1.6132,
+      "step": 1735
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.424000287039459e-05,
+      "loss": 1.6116,
+      "step": 1740
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.420552304835564e-05,
+      "loss": 1.6226,
+      "step": 1745
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.417094668095932e-05,
+      "loss": 1.5531,
+      "step": 1750
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.413627384372078e-05,
+      "loss": 1.5848,
+      "step": 1755
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.410150461236592e-05,
+      "loss": 1.6077,
+      "step": 1760
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.406663906283109e-05,
+      "loss": 1.6075,
+      "step": 1765
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.403167727126306e-05,
+      "loss": 1.6338,
+      "step": 1770
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.399661931401877e-05,
+      "loss": 1.5491,
+      "step": 1775
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.396146526766522e-05,
+      "loss": 1.6025,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.39262152089792e-05,
+      "loss": 1.6172,
+      "step": 1785
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.389086921494724e-05,
+      "loss": 1.5842,
+      "step": 1790
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.38554273627654e-05,
+      "loss": 1.5858,
+      "step": 1795
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.381988972983908e-05,
+      "loss": 1.5482,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.378425639378288e-05,
+      "loss": 1.592,
+      "step": 1805
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.374852743242037e-05,
+      "loss": 1.5815,
+      "step": 1810
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.371270292378404e-05,
+      "loss": 1.5603,
+      "step": 1815
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.3676782946115e-05,
+      "loss": 1.5997,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.364076757786286e-05,
+      "loss": 1.5678,
+      "step": 1825
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.360465689768563e-05,
+      "loss": 1.6015,
+      "step": 1830
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.356845098444942e-05,
+      "loss": 1.5783,
+      "step": 1835
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.353214991722834e-05,
+      "loss": 1.6035,
+      "step": 1840
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.349575377530435e-05,
+      "loss": 1.6347,
+      "step": 1845
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.345926263816702e-05,
+      "loss": 1.5895,
+      "step": 1850
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.34226765855134e-05,
+      "loss": 1.6016,
+      "step": 1855
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.338599569724785e-05,
+      "loss": 1.5756,
+      "step": 1860
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.334922005348183e-05,
+      "loss": 1.5572,
+      "step": 1865
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.331234973453375e-05,
+      "loss": 1.5404,
+      "step": 1870
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.327538482092882e-05,
+      "loss": 1.586,
+      "step": 1875
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.323832539339881e-05,
+      "loss": 1.5408,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.320117153288192e-05,
+      "loss": 1.5674,
+      "step": 1885
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.316392332052262e-05,
+      "loss": 1.5696,
+      "step": 1890
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.31265808376714e-05,
+      "loss": 1.5651,
+      "step": 1895
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.308914416588468e-05,
+      "loss": 1.5824,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.305161338692455e-05,
+      "loss": 1.5555,
+      "step": 1905
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.301398858275868e-05,
+      "loss": 1.5853,
+      "step": 1910
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.297626983556002e-05,
+      "loss": 1.557,
+      "step": 1915
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.293845722770678e-05,
+      "loss": 1.5696,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.29005508417821e-05,
+      "loss": 1.5863,
+      "step": 1925
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.286255076057395e-05,
+      "loss": 1.5396,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.282445706707493e-05,
+      "loss": 1.52,
+      "step": 1935
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.278626984448212e-05,
+      "loss": 1.5863,
+      "step": 1940
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.27479891761968e-05,
+      "loss": 1.5633,
+      "step": 1945
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.270961514582438e-05,
+      "loss": 1.5943,
+      "step": 1950
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.267114783717423e-05,
+      "loss": 1.5701,
+      "step": 1955
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.263258733425933e-05,
+      "loss": 1.5403,
+      "step": 1960
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.259393372129626e-05,
+      "loss": 1.5903,
+      "step": 1965
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.255518708270496e-05,
+      "loss": 1.5887,
+      "step": 1970
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.251634750310849e-05,
+      "loss": 1.5641,
+      "step": 1975
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.247741506733296e-05,
+      "loss": 1.5339,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.243838986040722e-05,
+      "loss": 1.5486,
+      "step": 1985
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.239927196756279e-05,
+      "loss": 1.526,
+      "step": 1990
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.236006147423356e-05,
+      "loss": 1.5766,
+      "step": 1995
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.23207584660557e-05,
+      "loss": 1.5736,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.22813630288674e-05,
+      "loss": 1.5243,
+      "step": 2005
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.224187524870876e-05,
+      "loss": 1.5458,
+      "step": 2010
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.220229521182152e-05,
+      "loss": 1.5692,
+      "step": 2015
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.216262300464893e-05,
+      "loss": 1.556,
+      "step": 2020
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.212285871383554e-05,
+      "loss": 1.5774,
+      "step": 2025
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.208300242622698e-05,
+      "loss": 1.5171,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.204305422886987e-05,
+      "loss": 1.5907,
+      "step": 2035
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.200301420901148e-05,
+      "loss": 1.5895,
+      "step": 2040
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.19628824540997e-05,
+      "loss": 1.5464,
+      "step": 2045
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.192265905178269e-05,
+      "loss": 1.5381,
+      "step": 2050
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.188234408990884e-05,
+      "loss": 1.5594,
+      "step": 2055
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.184193765652648e-05,
+      "loss": 1.5542,
+      "step": 2060
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.18014398398837e-05,
+      "loss": 1.5674,
+      "step": 2065
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.176085072842818e-05,
+      "loss": 1.5551,
+      "step": 2070
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.172017041080702e-05,
+      "loss": 1.5251,
+      "step": 2075
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.167939897586647e-05,
+      "loss": 1.5218,
+      "step": 2080
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.163853651265179e-05,
+      "loss": 1.5978,
+      "step": 2085
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.159758311040707e-05,
+      "loss": 1.5264,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.1556538858575e-05,
+      "loss": 1.5572,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.15154038467967e-05,
+      "loss": 1.5216,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.147417816491148e-05,
+      "loss": 1.5454,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.143286190295668e-05,
+      "loss": 1.5599,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.13914551511675e-05,
+      "loss": 1.5479,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.134995799997673e-05,
+      "loss": 1.6014,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.130837054001463e-05,
+      "loss": 1.554,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.126669286210867e-05,
+      "loss": 1.5455,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.122492505728337e-05,
+      "loss": 1.5272,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.118306721676008e-05,
+      "loss": 1.5671,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.11411194319568e-05,
+      "loss": 1.5282,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.109908179448795e-05,
+      "loss": 1.5794,
+      "step": 2150
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.105695439616421e-05,
+      "loss": 1.5587,
+      "step": 2155
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.101473732899229e-05,
+      "loss": 1.5445,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.097243068517472e-05,
+      "loss": 1.5839,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.09300345571097e-05,
+      "loss": 1.5617,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.088754903739085e-05,
+      "loss": 1.5204,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.084497421880701e-05,
+      "loss": 1.5237,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.080231019434207e-05,
+      "loss": 1.5603,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.075955705717474e-05,
+      "loss": 1.556,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.071671490067833e-05,
+      "loss": 1.5355,
+      "step": 2195
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.067378381842063e-05,
+      "loss": 1.5248,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.063076390416354e-05,
+      "loss": 1.5193,
+      "step": 2205
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.05876552518631e-05,
+      "loss": 1.5642,
+      "step": 2210
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.054445795566907e-05,
+      "loss": 1.5506,
+      "step": 2215
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.050117210992479e-05,
+      "loss": 1.5409,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.045779780916708e-05,
+      "loss": 1.5238,
+      "step": 2225
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.041433514812589e-05,
+      "loss": 1.5373,
+      "step": 2230
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.037078422172414e-05,
+      "loss": 1.4849,
+      "step": 2235
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.032714512507755e-05,
+      "loss": 1.5352,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.028341795349441e-05,
+      "loss": 1.5183,
+      "step": 2245
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.023960280247534e-05,
+      "loss": 1.5222,
+      "step": 2250
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.01956997677131e-05,
+      "loss": 1.56,
+      "step": 2255
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.015170894509244e-05,
+      "loss": 1.5435,
+      "step": 2260
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.01076304306898e-05,
+      "loss": 1.5025,
+      "step": 2265
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.006346432077312e-05,
+      "loss": 1.5396,
+      "step": 2270
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.001921071180175e-05,
+      "loss": 1.557,
+      "step": 2275
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.9974869700426e-05,
+      "loss": 1.5021,
+      "step": 2280
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.993044138348716e-05,
+      "loss": 1.5324,
+      "step": 2285
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.988592585801719e-05,
+      "loss": 1.5372,
+      "step": 2290
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.984132322123844e-05,
+      "loss": 1.5042,
+      "step": 2295
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.979663357056363e-05,
+      "loss": 1.5212,
+      "step": 2300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 8.975185700359542e-05,
+      "loss": 1.5306,
+      "step": 2305
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.970699361812637e-05,
+      "loss": 1.527,
+      "step": 2310
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.966204351213858e-05,
+      "loss": 1.5456,
+      "step": 2315
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.961700678380359e-05,
+      "loss": 1.5445,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.957188353148213e-05,
+      "loss": 1.5563,
+      "step": 2325
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.952667385372389e-05,
+      "loss": 1.5304,
+      "step": 2330
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.94813778492673e-05,
+      "loss": 1.5326,
+      "step": 2335
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.943599561703936e-05,
+      "loss": 1.5432,
+      "step": 2340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.939052725615536e-05,
+      "loss": 1.5116,
+      "step": 2345
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.93449728659187e-05,
+      "loss": 1.507,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.929933254582069e-05,
+      "loss": 1.5025,
+      "step": 2355
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.92536063955403e-05,
+      "loss": 1.5386,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.920779451494396e-05,
+      "loss": 1.5108,
+      "step": 2365
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.916189700408533e-05,
+      "loss": 1.5611,
+      "step": 2370
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.911591396320508e-05,
+      "loss": 1.5521,
+      "step": 2375
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.906984549273069e-05,
+      "loss": 1.5403,
+      "step": 2380
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.90236916932762e-05,
+      "loss": 1.5283,
+      "step": 2385
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.897745266564204e-05,
+      "loss": 1.5405,
+      "step": 2390
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.893112851081478e-05,
+      "loss": 1.5354,
+      "step": 2395
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.888471932996687e-05,
+      "loss": 1.5018,
+      "step": 2400
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.883822522445648e-05,
+      "loss": 1.5329,
+      "step": 2405
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 8.879164629582726e-05,
+      "loss": 1.5248,
+      "step": 2410
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.874498264580812e-05,
+      "loss": 1.4923,
+      "step": 2415
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.869823437631299e-05,
+      "loss": 1.5475,
+      "step": 2420
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.865140158944062e-05,
+      "loss": 1.497,
+      "step": 2425
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.860448438747433e-05,
+      "loss": 1.5153,
+      "step": 2430
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.855748287288182e-05,
+      "loss": 1.493,
+      "step": 2435
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.851039714831492e-05,
+      "loss": 1.5199,
+      "step": 2440
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.84632273166094e-05,
+      "loss": 1.5209,
+      "step": 2445
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.841597348078466e-05,
+      "loss": 1.5493,
+      "step": 2450
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.836863574404362e-05,
+      "loss": 1.5292,
+      "step": 2455
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.832121420977246e-05,
+      "loss": 1.5133,
+      "step": 2460
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.827370898154033e-05,
+      "loss": 1.5543,
+      "step": 2465
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.822612016309915e-05,
+      "loss": 1.4902,
+      "step": 2470
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.817844785838344e-05,
+      "loss": 1.5451,
+      "step": 2475
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.813069217151009e-05,
+      "loss": 1.5008,
+      "step": 2480
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.8082853206778e-05,
+      "loss": 1.5011,
+      "step": 2485
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.803493106866805e-05,
+      "loss": 1.5327,
+      "step": 2490
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.79869258618427e-05,
+      "loss": 1.527,
+      "step": 2495
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.793883769114586e-05,
+      "loss": 1.5219,
+      "step": 2500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.789066666160264e-05,
+      "loss": 1.5127,
+      "step": 2505
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.78424128784191e-05,
+      "loss": 1.4652,
+      "step": 2510
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.779407644698205e-05,
+      "loss": 1.5121,
+      "step": 2515
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 8.774565747285881e-05,
+      "loss": 1.4739,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.769715606179693e-05,
+      "loss": 1.548,
+      "step": 2525
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.764857231972408e-05,
+      "loss": 1.4915,
+      "step": 2530
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.759990635274764e-05,
+      "loss": 1.5482,
+      "step": 2535
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.755115826715467e-05,
+      "loss": 1.5057,
+      "step": 2540
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.750232816941153e-05,
+      "loss": 1.509,
+      "step": 2545
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.745341616616368e-05,
+      "loss": 1.5054,
+      "step": 2550
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.740442236423551e-05,
+      "loss": 1.4871,
+      "step": 2555
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.735534687062999e-05,
+      "loss": 1.5222,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.73061897925286e-05,
+      "loss": 1.5012,
+      "step": 2565
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.725695123729091e-05,
+      "loss": 1.4899,
+      "step": 2570
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.72076313124545e-05,
+      "loss": 1.5451,
+      "step": 2575
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.715823012573462e-05,
+      "loss": 1.5408,
+      "step": 2580
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.710874778502403e-05,
+      "loss": 1.4806,
+      "step": 2585
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.70591843983927e-05,
+      "loss": 1.5222,
+      "step": 2590
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.700954007408763e-05,
+      "loss": 1.5314,
+      "step": 2595
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.695981492053258e-05,
+      "loss": 1.4753,
+      "step": 2600
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.691000904632785e-05,
+      "loss": 1.4661,
+      "step": 2605
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.686012256025003e-05,
+      "loss": 1.4787,
+      "step": 2610
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.681015557125173e-05,
+      "loss": 1.5311,
+      "step": 2615
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.676010818846145e-05,
+      "loss": 1.4888,
+      "step": 2620
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.670998052118322e-05,
+      "loss": 1.5155,
+      "step": 2625
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.665977267889642e-05,
+      "loss": 1.4674,
+      "step": 2630
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.660948477125556e-05,
+      "loss": 1.5261,
+      "step": 2635
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.655911690809e-05,
+      "loss": 1.4668,
+      "step": 2640
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.650866919940367e-05,
+      "loss": 1.4737,
+      "step": 2645
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.645814175537497e-05,
+      "loss": 1.488,
+      "step": 2650
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.640753468635639e-05,
+      "loss": 1.5266,
+      "step": 2655
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.635684810287436e-05,
+      "loss": 1.5426,
+      "step": 2660
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.630608211562891e-05,
+      "loss": 1.461,
+      "step": 2665
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.625523683549354e-05,
+      "loss": 1.4797,
+      "step": 2670
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.62043123735149e-05,
+      "loss": 1.5399,
+      "step": 2675
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.615330884091261e-05,
+      "loss": 1.4962,
+      "step": 2680
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.610222634907894e-05,
+      "loss": 1.4867,
+      "step": 2685
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.605106500957862e-05,
+      "loss": 1.4869,
+      "step": 2690
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.59998249341486e-05,
+      "loss": 1.5112,
+      "step": 2695
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.59485062346978e-05,
+      "loss": 1.5332,
+      "step": 2700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.589710902330676e-05,
+      "loss": 1.5326,
+      "step": 2705
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.584563341222765e-05,
+      "loss": 1.4751,
+      "step": 2710
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.579407951388372e-05,
+      "loss": 1.5094,
+      "step": 2715
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.57424474408693e-05,
+      "loss": 1.4966,
+      "step": 2720
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.569073730594941e-05,
+      "loss": 1.441,
+      "step": 2725
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.563894922205954e-05,
+      "loss": 1.5072,
+      "step": 2730
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.558708330230546e-05,
+      "loss": 1.4876,
+      "step": 2735
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.553513965996293e-05,
+      "loss": 1.4807,
+      "step": 2740
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.548311840847742e-05,
+      "loss": 1.4652,
+      "step": 2745
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.543101966146396e-05,
+      "loss": 1.4466,
+      "step": 2750
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.537884353270677e-05,
+      "loss": 1.4935,
+      "step": 2755
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.532659013615911e-05,
+      "loss": 1.4784,
+      "step": 2760
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.527425958594299e-05,
+      "loss": 1.5074,
+      "step": 2765
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.52218519963489e-05,
+      "loss": 1.4619,
+      "step": 2770
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.516936748183562e-05,
+      "loss": 1.4961,
+      "step": 2775
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.51168061570299e-05,
+      "loss": 1.5012,
+      "step": 2780
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.506416813672627e-05,
+      "loss": 1.5199,
+      "step": 2785
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.501145353588677e-05,
+      "loss": 1.4816,
+      "step": 2790
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.495866246964066e-05,
+      "loss": 1.4851,
+      "step": 2795
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.490579505328424e-05,
+      "loss": 1.487,
+      "step": 2800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.48528514022805e-05,
+      "loss": 1.4882,
+      "step": 2805
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.4799831632259e-05,
+      "loss": 1.4749,
+      "step": 2810
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.474673585901552e-05,
+      "loss": 1.4658,
+      "step": 2815
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.469356419851177e-05,
+      "loss": 1.4962,
+      "step": 2820
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.464031676687533e-05,
+      "loss": 1.5304,
+      "step": 2825
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.458699368039913e-05,
+      "loss": 1.4989,
+      "step": 2830
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.453359505554139e-05,
+      "loss": 1.4973,
+      "step": 2835
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.448012100892534e-05,
+      "loss": 1.5001,
+      "step": 2840
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.442657165733886e-05,
+      "loss": 1.4837,
+      "step": 2845
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.437294711773436e-05,
+      "loss": 1.5113,
+      "step": 2850
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.431924750722842e-05,
+      "loss": 1.4915,
+      "step": 2855
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.426547294310159e-05,
+      "loss": 1.4915,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.421162354279812e-05,
+      "loss": 1.4856,
+      "step": 2865
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.415769942392573e-05,
+      "loss": 1.4676,
+      "step": 2870
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.410370070425526e-05,
+      "loss": 1.5019,
+      "step": 2875
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.404962750172054e-05,
+      "loss": 1.522,
+      "step": 2880
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.399547993441805e-05,
+      "loss": 1.532,
+      "step": 2885
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.394125812060666e-05,
+      "loss": 1.5481,
+      "step": 2890
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.388696217870747e-05,
+      "loss": 1.4848,
+      "step": 2895
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.383259222730338e-05,
+      "loss": 1.4434,
+      "step": 2900
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.377814838513897e-05,
+      "loss": 1.5135,
+      "step": 2905
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.372363077112022e-05,
+      "loss": 1.4842,
+      "step": 2910
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.366903950431422e-05,
+      "loss": 1.498,
+      "step": 2915
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.361437470394889e-05,
+      "loss": 1.4999,
+      "step": 2920
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.355963648941276e-05,
+      "loss": 1.491,
+      "step": 2925
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.350482498025474e-05,
+      "loss": 1.4368,
+      "step": 2930
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.344994029618374e-05,
+      "loss": 1.4816,
+      "step": 2935
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.339498255706853e-05,
+      "loss": 1.4646,
+      "step": 2940
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.333995188293744e-05,
+      "loss": 1.455,
+      "step": 2945
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.328484839397807e-05,
+      "loss": 1.4989,
+      "step": 2950
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.322967221053707e-05,
+      "loss": 1.4918,
+      "step": 2955
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.317442345311983e-05,
+      "loss": 1.4749,
+      "step": 2960
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.311910224239027e-05,
+      "loss": 1.5029,
+      "step": 2965
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.30637086991705e-05,
+      "loss": 1.4926,
+      "step": 2970
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.300824294444069e-05,
+      "loss": 1.4644,
+      "step": 2975
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.295270509933862e-05,
+      "loss": 1.4709,
+      "step": 2980
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.28970952851596e-05,
+      "loss": 1.5117,
+      "step": 2985
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.284141362335607e-05,
+      "loss": 1.4874,
+      "step": 2990
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.27856602355374e-05,
+      "loss": 1.5238,
+      "step": 2995
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.272983524346962e-05,
+      "loss": 1.4571,
+      "step": 3000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.267393876907514e-05,
+      "loss": 1.5056,
+      "step": 3005
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.261797093443249e-05,
+      "loss": 1.4785,
+      "step": 3010
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.256193186177603e-05,
+      "loss": 1.4421,
+      "step": 3015
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.250582167349575e-05,
+      "loss": 1.4744,
+      "step": 3020
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.24496404921369e-05,
+      "loss": 1.4646,
+      "step": 3025
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.239338844039984e-05,
+      "loss": 1.514,
+      "step": 3030
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.233706564113963e-05,
+      "loss": 1.4644,
+      "step": 3035
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.228067221736595e-05,
+      "loss": 1.5086,
+      "step": 3040
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.222420829224266e-05,
+      "loss": 1.4253,
+      "step": 3045
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.216767398908755e-05,
+      "loss": 1.4686,
+      "step": 3050
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.211106943137226e-05,
+      "loss": 1.4589,
+      "step": 3055
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.20543947427217e-05,
+      "loss": 1.5158,
+      "step": 3060
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.199765004691408e-05,
+      "loss": 1.4828,
+      "step": 3065
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.19408354678804e-05,
+      "loss": 1.458,
+      "step": 3070
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.188395112970438e-05,
+      "loss": 1.4592,
+      "step": 3075
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.182699715662203e-05,
+      "loss": 1.4531,
+      "step": 3080
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.176997367302148e-05,
+      "loss": 1.498,
+      "step": 3085
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.171288080344264e-05,
+      "loss": 1.4695,
+      "step": 3090
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.165571867257698e-05,
+      "loss": 1.4889,
+      "step": 3095
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.159848740526723e-05,
+      "loss": 1.5151,
+      "step": 3100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.154118712650713e-05,
+      "loss": 1.4543,
+      "step": 3105
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.148381796144114e-05,
+      "loss": 1.4428,
+      "step": 3110
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.142638003536413e-05,
+      "loss": 1.4678,
+      "step": 3115
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.136887347372119e-05,
+      "loss": 1.4748,
+      "step": 3120
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.131129840210731e-05,
+      "loss": 1.4404,
+      "step": 3125
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.125365494626706e-05,
+      "loss": 1.4371,
+      "step": 3130
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.119594323209441e-05,
+      "loss": 1.4588,
+      "step": 3135
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.11381633856324e-05,
+      "loss": 1.4749,
+      "step": 3140
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.108031553307283e-05,
+      "loss": 1.4211,
+      "step": 3145
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.10223998007561e-05,
+      "loss": 1.4568,
+      "step": 3150
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.096441631517078e-05,
+      "loss": 1.4694,
+      "step": 3155
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.090636520295348e-05,
+      "loss": 1.4609,
+      "step": 3160
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.084824659088845e-05,
+      "loss": 1.4851,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.079006060590741e-05,
+      "loss": 1.5124,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.073180737508921e-05,
+      "loss": 1.457,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.067348702565954e-05,
+      "loss": 1.4647,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.06150996849907e-05,
+      "loss": 1.4237,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.055664548060131e-05,
+      "loss": 1.4707,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.0498124540156e-05,
+      "loss": 1.4926,
+      "step": 3195
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.043953699146515e-05,
+      "loss": 1.4641,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.038088296248462e-05,
+      "loss": 1.4705,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.032216258131548e-05,
+      "loss": 1.4334,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.026337597620369e-05,
+      "loss": 1.4649,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.020452327553986e-05,
+      "loss": 1.4306,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.014560460785896e-05,
+      "loss": 1.475,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.008662010184e-05,
+      "loss": 1.4586,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.00275698863058e-05,
+      "loss": 1.4228,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.996845409022272e-05,
+      "loss": 1.4395,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.990927284270027e-05,
+      "loss": 1.4959,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.9850026272991e-05,
+      "loss": 1.4583,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.979071451049007e-05,
+      "loss": 1.4522,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.973133768473503e-05,
+      "loss": 1.4524,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.967189592540554e-05,
+      "loss": 1.4588,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 7.961238936232307e-05,
+      "loss": 1.4764,
+      "step": 3270
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.95528181254506e-05,
+      "loss": 1.4758,
+      "step": 3275
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.94931823448924e-05,
+      "loss": 1.4747,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.943348215089367e-05,
+      "loss": 1.4478,
+      "step": 3285
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.93737176738403e-05,
+      "loss": 1.4492,
+      "step": 3290
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.93138890442586e-05,
+      "loss": 1.4654,
+      "step": 3295
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.925399639281494e-05,
+      "loss": 1.4578,
+      "step": 3300
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.919403985031557e-05,
+      "loss": 1.4463,
+      "step": 3305
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.913401954770622e-05,
+      "loss": 1.4906,
+      "step": 3310
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.90739356160719e-05,
+      "loss": 1.4543,
+      "step": 3315
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.901378818663663e-05,
+      "loss": 1.4494,
+      "step": 3320
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.895357739076304e-05,
+      "loss": 1.4467,
+      "step": 3325
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.889330335995218e-05,
+      "loss": 1.5219,
+      "step": 3330
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.883296622584323e-05,
+      "loss": 1.435,
+      "step": 3335
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.877256612021312e-05,
+      "loss": 1.4612,
+      "step": 3340
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.87121031749764e-05,
+      "loss": 1.4404,
+      "step": 3345
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.865157752218482e-05,
+      "loss": 1.4348,
+      "step": 3350
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.859098929402709e-05,
+      "loss": 1.5159,
+      "step": 3355
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.853033862282853e-05,
+      "loss": 1.463,
+      "step": 3360
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.846962564105091e-05,
+      "loss": 1.48,
+      "step": 3365
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.84088504812921e-05,
+      "loss": 1.4455,
+      "step": 3370
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 7.834801327628567e-05,
+      "loss": 1.4657,
+      "step": 3375
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.82871141589008e-05,
+      "loss": 1.4799,
+      "step": 3380
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.822615326214183e-05,
+      "loss": 1.4992,
+      "step": 3385
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.816513071914804e-05,
+      "loss": 1.4521,
+      "step": 3390
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.810404666319333e-05,
+      "loss": 1.4061,
+      "step": 3395
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.804290122768601e-05,
+      "loss": 1.4543,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.798169454616835e-05,
+      "loss": 1.4484,
+      "step": 3405
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.792042675231643e-05,
+      "loss": 1.4448,
+      "step": 3410
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.785909797993984e-05,
+      "loss": 1.4099,
+      "step": 3415
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.779770836298124e-05,
+      "loss": 1.4587,
+      "step": 3420
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.773625803551631e-05,
+      "loss": 1.5019,
+      "step": 3425
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.767474713175321e-05,
+      "loss": 1.4506,
+      "step": 3430
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.761317578603245e-05,
+      "loss": 1.4101,
+      "step": 3435
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.755154413282654e-05,
+      "loss": 1.4347,
+      "step": 3440
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.748985230673971e-05,
+      "loss": 1.4568,
+      "step": 3445
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.742810044250761e-05,
+      "loss": 1.4503,
+      "step": 3450
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.736628867499696e-05,
+      "loss": 1.4565,
+      "step": 3455
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.73044171392054e-05,
+      "loss": 1.4709,
+      "step": 3460
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.724248597026103e-05,
+      "loss": 1.4857,
+      "step": 3465
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.71804953034222e-05,
+      "loss": 1.476,
+      "step": 3470
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.711844527407724e-05,
+      "loss": 1.4542,
+      "step": 3475
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.705633601774414e-05,
+      "loss": 1.4696,
+      "step": 3480
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 7.699416767007017e-05,
+      "loss": 1.4303,
+      "step": 3485
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.69319403668317e-05,
+      "loss": 1.475,
+      "step": 3490
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.686965424393389e-05,
+      "loss": 1.4128,
+      "step": 3495
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.680730943741028e-05,
+      "loss": 1.4559,
+      "step": 3500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.674490608342266e-05,
+      "loss": 1.4504,
+      "step": 3505
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.668244431826066e-05,
+      "loss": 1.4778,
+      "step": 3510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.661992427834146e-05,
+      "loss": 1.4665,
+      "step": 3515
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.655734610020952e-05,
+      "loss": 1.4535,
+      "step": 3520
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.649470992053629e-05,
+      "loss": 1.4826,
+      "step": 3525
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.643201587611988e-05,
+      "loss": 1.4314,
+      "step": 3530
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.636926410388481e-05,
+      "loss": 1.4525,
+      "step": 3535
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.630645474088163e-05,
+      "loss": 1.4489,
+      "step": 3540
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.62435879242867e-05,
+      "loss": 1.494,
+      "step": 3545
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.618066379140186e-05,
+      "loss": 1.4398,
+      "step": 3550
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.611768247965408e-05,
+      "loss": 1.4415,
+      "step": 3555
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.605464412659531e-05,
+      "loss": 1.4829,
+      "step": 3560
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.599154886990199e-05,
+      "loss": 1.449,
+      "step": 3565
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.592839684737485e-05,
+      "loss": 1.443,
+      "step": 3570
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.586518819693864e-05,
+      "loss": 1.446,
+      "step": 3575
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.580192305664174e-05,
+      "loss": 1.4271,
+      "step": 3580
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.573860156465597e-05,
+      "loss": 1.4597,
+      "step": 3585
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.567522385927612e-05,
+      "loss": 1.4253,
+      "step": 3590
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.561179007891983e-05,
+      "loss": 1.45,
+      "step": 3595
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.554830036212721e-05,
+      "loss": 1.4328,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.548475484756047e-05,
+      "loss": 1.4422,
+      "step": 3605
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.542115367400376e-05,
+      "loss": 1.4208,
+      "step": 3610
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.535749698036274e-05,
+      "loss": 1.4153,
+      "step": 3615
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.529378490566435e-05,
+      "loss": 1.4696,
+      "step": 3620
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.523001758905647e-05,
+      "loss": 1.457,
+      "step": 3625
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.516619516980763e-05,
+      "loss": 1.4579,
+      "step": 3630
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.51023177873067e-05,
+      "loss": 1.4604,
+      "step": 3635
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.503838558106263e-05,
+      "loss": 1.4523,
+      "step": 3640
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.497439869070404e-05,
+      "loss": 1.4064,
+      "step": 3645
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.491035725597903e-05,
+      "loss": 1.4579,
+      "step": 3650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.484626141675481e-05,
+      "loss": 1.3878,
+      "step": 3655
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.478211131301743e-05,
+      "loss": 1.4285,
+      "step": 3660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.471790708487139e-05,
+      "loss": 1.4511,
+      "step": 3665
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.465364887253949e-05,
+      "loss": 1.4187,
+      "step": 3670
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.458933681636234e-05,
+      "loss": 1.4455,
+      "step": 3675
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.452497105679824e-05,
+      "loss": 1.465,
+      "step": 3680
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.446055173442268e-05,
+      "loss": 1.4225,
+      "step": 3685
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.439607898992822e-05,
+      "loss": 1.425,
+      "step": 3690
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.433155296412405e-05,
+      "loss": 1.4222,
+      "step": 3695
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.426697379793572e-05,
+      "loss": 1.4147,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.420234163240487e-05,
+      "loss": 1.4277,
+      "step": 3705
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.413765660868885e-05,
+      "loss": 1.4339,
+      "step": 3710
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.407291886806048e-05,
+      "loss": 1.4531,
+      "step": 3715
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.400812855190771e-05,
+      "loss": 1.4692,
+      "step": 3720
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.394328580173333e-05,
+      "loss": 1.3956,
+      "step": 3725
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.387839075915465e-05,
+      "loss": 1.4331,
+      "step": 3730
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.381344356590312e-05,
+      "loss": 1.4476,
+      "step": 3735
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.374844436382418e-05,
+      "loss": 1.4392,
+      "step": 3740
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.368339329487677e-05,
+      "loss": 1.399,
+      "step": 3745
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.36182905011332e-05,
+      "loss": 1.3972,
+      "step": 3750
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.35531361247787e-05,
+      "loss": 1.4426,
+      "step": 3755
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.348793030811111e-05,
+      "loss": 1.4792,
+      "step": 3760
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.342267319354071e-05,
+      "loss": 1.4163,
+      "step": 3765
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.335736492358976e-05,
+      "loss": 1.4636,
+      "step": 3770
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.329200564089224e-05,
+      "loss": 1.4525,
+      "step": 3775
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.322659548819355e-05,
+      "loss": 1.4573,
+      "step": 3780
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.316113460835021e-05,
+      "loss": 1.4468,
+      "step": 3785
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.309562314432952e-05,
+      "loss": 1.4738,
+      "step": 3790
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.303006123920923e-05,
+      "loss": 1.4853,
+      "step": 3795
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.29644490361773e-05,
+      "loss": 1.4292,
+      "step": 3800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.28987866785315e-05,
+      "loss": 1.4719,
+      "step": 3805
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.283307430967916e-05,
+      "loss": 1.4524,
+      "step": 3810
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.276731207313683e-05,
+      "loss": 1.4715,
+      "step": 3815
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.270150011252994e-05,
+      "loss": 1.4379,
+      "step": 3820
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.263563857159262e-05,
+      "loss": 1.3972,
+      "step": 3825
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.256972759416714e-05,
+      "loss": 1.4352,
+      "step": 3830
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.250376732420387e-05,
+      "loss": 1.4548,
+      "step": 3835
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.243775790576073e-05,
+      "loss": 1.4124,
+      "step": 3840
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.237169948300307e-05,
+      "loss": 1.4701,
+      "step": 3845
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.230559220020319e-05,
+      "loss": 1.4365,
+      "step": 3850
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.223943620174014e-05,
+      "loss": 1.4524,
+      "step": 3855
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.217323163209941e-05,
+      "loss": 1.4027,
+      "step": 3860
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.210697863587246e-05,
+      "loss": 1.4173,
+      "step": 3865
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.204067735775662e-05,
+      "loss": 1.4291,
+      "step": 3870
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.197432794255457e-05,
+      "loss": 1.4009,
+      "step": 3875
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.190793053517424e-05,
+      "loss": 1.467,
+      "step": 3880
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.184148528062827e-05,
+      "loss": 1.4117,
+      "step": 3885
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.177499232403384e-05,
+      "loss": 1.4038,
+      "step": 3890
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.170845181061236e-05,
+      "loss": 1.4045,
+      "step": 3895
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.164186388568898e-05,
+      "loss": 1.442,
+      "step": 3900
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.157522869469253e-05,
+      "loss": 1.4667,
+      "step": 3905
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.1508546383155e-05,
+      "loss": 1.4309,
+      "step": 3910
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.144181709671129e-05,
+      "loss": 1.4286,
+      "step": 3915
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.137504098109894e-05,
+      "loss": 1.4431,
+      "step": 3920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.13082181821577e-05,
+      "loss": 1.4327,
+      "step": 3925
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.124134884582933e-05,
+      "loss": 1.4531,
+      "step": 3930
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.117443311815722e-05,
+      "loss": 1.4494,
+      "step": 3935
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.110747114528604e-05,
+      "loss": 1.4133,
+      "step": 3940
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.10404630734615e-05,
+      "loss": 1.3919,
+      "step": 3945
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.097340904903e-05,
+      "loss": 1.4806,
+      "step": 3950
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.090630921843825e-05,
+      "loss": 1.4293,
+      "step": 3955
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.083916372823304e-05,
+      "loss": 1.4641,
+      "step": 3960
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.077197272506089e-05,
+      "loss": 1.4418,
+      "step": 3965
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.070473635566768e-05,
+      "loss": 1.4011,
+      "step": 3970
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.063745476689841e-05,
+      "loss": 1.4254,
+      "step": 3975
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.057012810569681e-05,
+      "loss": 1.4511,
+      "step": 3980
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.050275651910506e-05,
+      "loss": 1.4461,
+      "step": 3985
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.043534015426349e-05,
+      "loss": 1.4318,
+      "step": 3990
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.036787915841016e-05,
+      "loss": 1.4904,
+      "step": 3995
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.030037367888069e-05,
+      "loss": 1.4308,
+      "step": 4000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 10729,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 1.0610259290750976e+19,
+  "trial_name": null,
+  "trial_params": null
+}