diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6246 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5184,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002999993113941634,
+      "loss": 1.412,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00029999724558297603,
+      "loss": 1.2243,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002999946013584945,
+      "loss": 1.6929,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00029999005663396985,
+      "loss": 2.0793,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00029998413478906613,
+      "loss": 2.4653,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029997683587815427,
+      "loss": 2.1023,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029996815996824846,
+      "loss": 1.4915,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000299958107139006,
+      "loss": 1.6685,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029995370049014505,
+      "loss": 2.2307,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999490735552899,
+      "loss": 1.7744,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999365425123301,
+      "loss": 2.0726,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029992263484032635,
+      "loss": 1.5018,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999073506669708,
+      "loss": 1.887,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029989069013259374,
+      "loss": 1.4459,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029987265339016236,
+      "loss": 1.906,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002998532406052794,
+      "loss": 1.8526,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029983245195618156,
+      "loss": 2.3401,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000299810287633738,
+      "loss": 1.7854,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029978674784144856,
+      "loss": 2.0329,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029976183279544166,
+      "loss": 1.5686,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002997355427244727,
+      "loss": 1.3467,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029970787786992147,
+      "loss": 1.7594,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002996788384857905,
+      "loss": 1.1309,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029964842483870226,
+      "loss": 1.845,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029961663720789696,
+      "loss": 1.7158,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029958347588522984,
+      "loss": 1.6382,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002995489411751688,
+      "loss": 1.2599,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029951303339479106,
+      "loss": 1.4989,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002994757528737808,
+      "loss": 2.1127,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029943709995442576,
+      "loss": 1.2286,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002993970749916143,
+      "loss": 1.719,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029935567835283203,
+      "loss": 1.3937,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029931291041815865,
+      "loss": 1.3153,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000299268771580264,
+      "loss": 1.673,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029922326224440495,
+      "loss": 1.2415,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002991763828284214,
+      "loss": 1.9943,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002991281337627326,
+      "loss": 1.383,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000299078515490333,
+      "loss": 1.5981,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002990275284667884,
+      "loss": 1.1489,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002989751731602316,
+      "loss": 1.6789,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002989214500513582,
+      "loss": 1.9486,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029886635963342217,
+      "loss": 1.4526,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002988099024122313,
+      "loss": 1.7287,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002987520789061425,
+      "loss": 1.1696,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002986928896460572,
+      "loss": 1.4302,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029863233517541635,
+      "loss": 2.153,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029857041605019534,
+      "loss": 1.6862,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002985071328388992,
+      "loss": 1.3822,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002984424861225571,
+      "loss": 1.858,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029837647649471715,
+      "loss": 1.6337,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002983091045614408,
+      "loss": 2.4304,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002982403709412975,
+      "loss": 1.7602,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029817027626535895,
+      "loss": 1.3822,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029809882117719314,
+      "loss": 1.3085,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002980260063328586,
+      "loss": 1.6656,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029795183240089834,
+      "loss": 1.3397,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029787630006233374,
+      "loss": 1.8768,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002977994100106583,
+      "loss": 1.3506,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002977211629518312,
+      "loss": 1.976,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029764155960427076,
+      "loss": 1.5258,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029756060069884813,
+      "loss": 1.1524,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002974782869788803,
+      "loss": 1.7161,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029739461920012334,
+      "loss": 2.1702,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002973095981307655,
+      "loss": 2.1052,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002972232245514203,
+      "loss": 1.429,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000297135499255119,
+      "loss": 1.3791,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002970464230473036,
+      "loss": 1.6739,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002969559967458194,
+      "loss": 1.7643,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029686422118090745,
+      "loss": 1.7344,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029677109719519697,
+      "loss": 1.6696,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029667662564369755,
+      "loss": 1.7149,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002965808073937914,
+      "loss": 0.9939,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002964836433252251,
+      "loss": 1.7887,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029638513433010214,
+      "loss": 1.5627,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029628528131287396,
+      "loss": 1.4555,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002961840851903322,
+      "loss": 1.8208,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002960815468916,
+      "loss": 1.4685,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029597766735812373,
+      "loss": 1.3059,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029587244754366416,
+      "loss": 1.9683,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029576588841428753,
+      "loss": 1.6673,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029565799094835707,
+      "loss": 1.9922,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029554875613652375,
+      "loss": 1.4377,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029543818498171736,
+      "loss": 2.0199,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002953262784991371,
+      "loss": 1.3798,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002952130377162424,
+      "loss": 1.466,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029509846367274336,
+      "loss": 1.548,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029498255742059157,
+      "loss": 1.7984,
+      "step": 435
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002948653200239698,
+      "loss": 1.4208,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002947467525592829,
+      "loss": 1.6134,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029462685611514745,
+      "loss": 1.4112,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000294505631792382,
+      "loss": 2.0775,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000294383080703997,
+      "loss": 1.5725,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002942592039751843,
+      "loss": 1.5205,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029413400274330703,
+      "loss": 2.3163,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002940074781578893,
+      "loss": 1.8447,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002938796313806052,
+      "loss": 1.903,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002937504635852687,
+      "loss": 1.4324,
+      "step": 485
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002936199759578224,
+      "loss": 1.2384,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002934881696963268,
+      "loss": 1.9072,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029335504601094947,
+      "loss": 1.5876,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029322060612395356,
+      "loss": 2.3766,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002930848512696871,
+      "loss": 1.5821,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002929477826945711,
+      "loss": 1.9367,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002928094016570886,
+      "loss": 2.2331,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002926697094277728,
+      "loss": 1.7781,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002925287072891954,
+      "loss": 1.7931,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029238639653595506,
+      "loss": 1.7712,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029224277847466533,
+      "loss": 2.0733,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002920978544239427,
+      "loss": 1.6243,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002919516257143944,
+      "loss": 1.831,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029180409368860635,
+      "loss": 1.9509,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002916552597011308,
+      "loss": 1.1314,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029150512511847375,
+      "loss": 1.4511,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029135369131908255,
+      "loss": 1.6977,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002912009596933331,
+      "loss": 1.3039,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029104693164351726,
+      "loss": 1.2062,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029089160858382997,
+      "loss": 1.5104,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002907349919403561,
+      "loss": 2.2833,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002905770831510573,
+      "loss": 1.5332,
+      "step": 595
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029041788366575933,
+      "loss": 1.9756,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002902573949461379,
+      "loss": 1.6587,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029009561846570604,
+      "loss": 1.5348,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002899325557098001,
+      "loss": 1.9013,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00028976820817556623,
+      "loss": 1.3699,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002896025773719469,
+      "loss": 1.9326,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002894356648196664,
+      "loss": 1.6738,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00028926747205121757,
+      "loss": 1.9836,
+      "step": 635
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002890980006108475,
+      "loss": 2.0351,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00028892725205454295,
+      "loss": 2.0746,
+      "step": 645
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002887552279500168,
+      "loss": 1.9237,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000288581929876693,
+      "loss": 1.6076,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002884073594256924,
+      "loss": 1.5154,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00028823151819981814,
+      "loss": 2.1754,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00028805440781354074,
+      "loss": 1.6397,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00028787602989298355,
+      "loss": 1.6161,
+      "step": 675
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00028769638607590756,
+      "loss": 1.8723,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002875154780116965,
+      "loss": 2.3238,
+      "step": 685
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00028733330736134175,
+      "loss": 1.5394,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002871498757974269,
+      "loss": 1.5529,
+      "step": 695
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028696518500411254,
+      "loss": 1.3396,
+      "step": 700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028677923667712085,
+      "loss": 1.4043,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002865920325237198,
+      "loss": 2.2666,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028640357426270777,
+      "loss": 1.4279,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002862138636243974,
+      "loss": 1.2806,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028602290235060025,
+      "loss": 1.6162,
+      "step": 725
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028583069219461016,
+      "loss": 1.4642,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00028563723492118764,
+      "loss": 1.639,
+      "step": 735
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002854425323065435,
+      "loss": 1.9078,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002852465861383224,
+      "loss": 1.5942,
+      "step": 745
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002850493982155867,
+      "loss": 1.6403,
+      "step": 750
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002848509703487999,
+      "loss": 1.7215,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002846513043598097,
+      "loss": 0.7587,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028445040208183164,
+      "loss": 1.7533,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028424826535943226,
+      "loss": 1.5318,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002840448960485118,
+      "loss": 1.5995,
+      "step": 775
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002838402960162876,
+      "loss": 1.7403,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002836344671412766,
+      "loss": 1.3032,
+      "step": 785
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002834274113132784,
+      "loss": 1.2807,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002832191304333576,
+      "loss": 1.5225,
+      "step": 795
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002830096264138267,
+      "loss": 1.3622,
+      "step": 800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002827989011782282,
+      "loss": 1.4411,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002825869566613171,
+      "loss": 1.5546,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028237379480904317,
+      "loss": 1.7998,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028215941757853304,
+      "loss": 1.5628,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028194382693807216,
+      "loss": 1.8731,
+      "step": 825
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002817270248670869,
+      "loss": 1.6162,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028150901335612615,
+      "loss": 1.2305,
+      "step": 835
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028128979440684326,
+      "loss": 1.1375,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002810693700319775,
+      "loss": 1.5766,
+      "step": 845
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002808477422553358,
+      "loss": 1.7187,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002806249131117738,
+      "loss": 1.471,
+      "step": 855
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002804008846471775,
+      "loss": 1.5664,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028017565891844435,
+      "loss": 1.1007,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00027994923799346433,
+      "loss": 1.8754,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002797216239511011,
+      "loss": 1.4075,
+      "step": 875
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002794928188811727,
+      "loss": 1.9204,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002792628248844327,
+      "loss": 1.5619,
+      "step": 885
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002790316440725504,
+      "loss": 1.0522,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00027879927856809197,
+      "loss": 1.0902,
+      "step": 895
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002785657305045005,
+      "loss": 1.1806,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000278331002026077,
+      "loss": 1.4444,
+      "step": 905
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00027809509528796,
+      "loss": 1.9458,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002778580124561063,
+      "loss": 1.3607,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00027761975570727085,
+      "loss": 1.4185,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00027738032722898683,
+      "loss": 1.6605,
+      "step": 925
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002771397292195455,
+      "loss": 1.6314,
+      "step": 930
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002768979638879761,
+      "loss": 1.5514,
+      "step": 935
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002766550334540257,
+      "loss": 1.9447,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00027641094014813826,
+      "loss": 2.1207,
+      "step": 945
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00027616568621143493,
+      "loss": 1.5794,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000275919273895693,
+      "loss": 1.5328,
+      "step": 955
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027567170546332503,
+      "loss": 1.6665,
+      "step": 960
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002754229831873588,
+      "loss": 2.2289,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027517310935141565,
+      "loss": 1.4299,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002749220862496901,
+      "loss": 1.7164,
+      "step": 975
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027466991618692836,
+      "loss": 1.1377,
+      "step": 980
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027441660147840754,
+      "loss": 1.8367,
+      "step": 985
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027416214444991416,
+      "loss": 1.7409,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027390654743772273,
+      "loss": 1.5072,
+      "step": 995
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002736498127885745,
+      "loss": 1.8343,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002733919428596559,
+      "loss": 1.6706,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027313294001857685,
+      "loss": 2.049,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027287280664334875,
+      "loss": 1.992,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002726115451223632,
+      "loss": 1.3719,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002723491578543695,
+      "loss": 1.1513,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002720856472484533,
+      "loss": 1.0665,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002718210157240135,
+      "loss": 1.9594,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002715552657107411,
+      "loss": 1.6081,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002712883996485963,
+      "loss": 1.9058,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000271020419987786,
+      "loss": 1.5741,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027075132918874176,
+      "loss": 1.0194,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002704811297220967,
+      "loss": 1.4918,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027020982406866334,
+      "loss": 1.7175,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00026993741471941044,
+      "loss": 1.6855,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002696639041754402,
+      "loss": 1.6377,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00026938929494796534,
+      "loss": 1.7234,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002691135895582861,
+      "loss": 1.9078,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002688367905377671,
+      "loss": 1.5927,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00026855890042781387,
+      "loss": 1.8701,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002682799217798498,
+      "loss": 1.8161,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002679998571552925,
+      "loss": 1.5415,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00026771870912553055,
+      "loss": 1.3043,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002674364802718995,
+      "loss": 1.2023,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002671531731856586,
+      "loss": 1.648,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002668687904679666,
+      "loss": 1.5379,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002665833347298582,
+      "loss": 2.1148,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00026629680859221983,
+      "loss": 1.5271,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002660092146857659,
+      "loss": 1.6819,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002657205556510142,
+      "loss": 2.1589,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00026543083413826203,
+      "loss": 1.2536,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00026514005280756174,
+      "loss": 1.2755,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002648482143286963,
+      "loss": 2.0277,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00026455532138115474,
+      "loss": 2.1261,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026426137665410765,
+      "loss": 1.479,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002639663828463824,
+      "loss": 1.6757,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026367034266643844,
+      "loss": 1.6926,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002633732588323424,
+      "loss": 1.7553,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002630751340717431,
+      "loss": 2.3215,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002627759711218466,
+      "loss": 1.505,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002624757727293908,
+      "loss": 1.5585,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026217454165062073,
+      "loss": 2.3911,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026187228065126266,
+      "loss": 1.3217,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026156899250649913,
+      "loss": 1.0692,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026126468000094334,
+      "loss": 1.784,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002609593459286135,
+      "loss": 1.3396,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002606529930929072,
+      "loss": 1.8426,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002603456243065759,
+      "loss": 1.7322,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026003724239169874,
+      "loss": 1.5849,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00025972785017965683,
+      "loss": 1.853,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002594174505111074,
+      "loss": 1.3056,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002591060462359573,
+      "loss": 1.7144,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00025879364021333726,
+      "loss": 1.5749,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00025848023531157546,
+      "loss": 1.575,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002581658344081711,
+      "loss": 1.4125,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002578504403897682,
+      "loss": 1.7587,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002575340561521289,
+      "loss": 1.6839,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00025721668460010696,
+      "loss": 1.3159,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00025689832864762103,
+      "loss": 1.4817,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00025657899121762814,
+      "loss": 1.4595,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002562586752420964,
+      "loss": 1.5914,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00025593738366197857,
+      "loss": 1.3689,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00025561511942718467,
+      "loss": 1.3941,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002552918854965552,
+      "loss": 1.3645,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002549676848378338,
+      "loss": 1.8918,
+      "step": 1320
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002546425204276399,
+      "loss": 1.5738,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00025431639525144175,
+      "loss": 1.4829,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002539893123035286,
+      "loss": 1.9539,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002536612745869836,
+      "loss": 1.619,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00025333228511365577,
+      "loss": 2.1926,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002530023469041328,
+      "loss": 1.0672,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000252671462987713,
+      "loss": 1.3778,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002523396364023777,
+      "loss": 2.2418,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00025200687019476313,
+      "loss": 1.3776,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002516731674201328,
+      "loss": 1.7229,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00025133853114234905,
+      "loss": 1.8493,
+      "step": 1375
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000251002964433845,
+      "loss": 1.7188,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002506664703755967,
+      "loss": 1.5356,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002503290520570944,
+      "loss": 2.0891,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002499907125763144,
+      "loss": 1.3363,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00024965145503969065,
+      "loss": 0.9084,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00024931128256208603,
+      "loss": 1.4176,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00024897019826676395,
+      "loss": 2.092,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00024862820528535954,
+      "loss": 1.302,
+      "step": 1415
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00024828530675785094,
+      "loss": 1.4547,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002479415058325306,
+      "loss": 1.571,
+      "step": 1425
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00024759680566597604,
+      "loss": 1.6497,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002472512094230214,
+      "loss": 2.3622,
+      "step": 1435
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00024690472027672774,
+      "loss": 1.6807,
+      "step": 1440
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00024655734140835445,
+      "loss": 1.3086,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002462090760073296,
+      "loss": 1.4113,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002458599272712212,
+      "loss": 1.9112,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002455098984057073,
+      "loss": 1.6202,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002451589926245468,
+      "loss": 1.955,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00024480721314955,
+      "loss": 0.958,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00024445456321054885,
+      "loss": 1.2851,
+      "step": 1475
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002441010460453674,
+      "loss": 1.5792,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002437466648997922,
+      "loss": 1.6548,
+      "step": 1485
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024339142302754214,
+      "loss": 1.9109,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024303532369023898,
+      "loss": 1.5052,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024267837015737716,
+      "loss": 1.5422,
+      "step": 1500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002423205657062938,
+      "loss": 1.2181,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024196191362213862,
+      "loss": 1.3582,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024160241719784398,
+      "loss": 1.6467,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002412420797340943,
+      "loss": 1.1192,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024088090453929596,
+      "loss": 1.5344,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000240518894929547,
+      "loss": 1.3969,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002401560542286064,
+      "loss": 1.8009,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023979238576786387,
+      "loss": 1.2987,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002394278928863091,
+      "loss": 2.0219,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023906257893050104,
+      "loss": 1.8617,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023869644725453735,
+      "loss": 1.5987,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023832950122002345,
+      "loss": 2.2982,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002379617441960418,
+      "loss": 1.9818,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023759317955912083,
+      "loss": 1.1432,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00023722381069320398,
+      "loss": 1.7589,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002368536409896187,
+      "loss": 1.4148,
+      "step": 1580
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023648267384704526,
+      "loss": 1.278,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023611091267148547,
+      "loss": 1.9946,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023573836087623155,
+      "loss": 2.0949,
+      "step": 1595
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023536502188183472,
+      "loss": 1.7023,
+      "step": 1600
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023499089911607377,
+      "loss": 2.1282,
+      "step": 1605
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023461599601392368,
+      "loss": 0.9914,
+      "step": 1610
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002342403160175239,
+      "loss": 1.5996,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023386386257614705,
+      "loss": 1.6788,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002334866391461668,
+      "loss": 0.9417,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00023310864919102667,
+      "loss": 1.8813,
+      "step": 1630
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023272989618120787,
+      "loss": 1.2746,
+      "step": 1635
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023235038359419743,
+      "loss": 1.8115,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002319701149144565,
+      "loss": 1.8113,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023158909363338816,
+      "loss": 1.7773,
+      "step": 1650
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023120732324930543,
+      "loss": 1.5997,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023082480726739922,
+      "loss": 1.5149,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023044154919970596,
+      "loss": 1.732,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023005755256507563,
+      "loss": 1.3192,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002296728208891392,
+      "loss": 1.529,
+      "step": 1675
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022928735770427633,
+      "loss": 1.9453,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000228901166549583,
+      "loss": 1.6557,
+      "step": 1685
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022851425097083906,
+      "loss": 1.5921,
+      "step": 1690
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022812661452047546,
+      "loss": 1.3226,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022773826075754181,
+      "loss": 1.8023,
+      "step": 1700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002273491932476737,
+      "loss": 1.6419,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022695941556305998,
+      "loss": 0.9914,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002265689312824097,
+      "loss": 1.3565,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002261777439909197,
+      "loss": 1.6539,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002257858572802413,
+      "loss": 1.5568,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022539327474844752,
+      "loss": 1.2279,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000225,
+      "loss": 1.8043,
+      "step": 1735
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022460603664571586,
+      "loss": 1.633,
+      "step": 1740
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002242113883027347,
+      "loss": 1.3194,
+      "step": 1745
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022381605859448514,
+      "loss": 2.1916,
+      "step": 1750
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022342005115065186,
+      "loss": 1.4717,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022302336960714202,
+      "loss": 1.6223,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002226260176060519,
+      "loss": 1.1155,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022222799879563357,
+      "loss": 1.9552,
+      "step": 1770
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022182931683026148,
+      "loss": 1.6145,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002214299753703987,
+      "loss": 1.4544,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022102997808256318,
+      "loss": 1.4521,
+      "step": 1785
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022062932863929464,
+      "loss": 1.895,
+      "step": 1790
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022022803071912027,
+      "loss": 1.5163,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021982608800652132,
+      "loss": 1.9087,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021942350419189912,
+      "loss": 1.3489,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002190202829715411,
+      "loss": 1.4871,
+      "step": 1810
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002186164280475871,
+      "loss": 1.0417,
+      "step": 1815
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021821194312799527,
+      "loss": 1.1155,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021780683192650796,
+      "loss": 1.8977,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002174010981626176,
+      "loss": 1.6845,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021699474556153276,
+      "loss": 1.7154,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00021658777785414375,
+      "loss": 1.8151,
+      "step": 1840
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021618019877698842,
+      "loss": 1.7573,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002157720120722178,
+      "loss": 1.4932,
+      "step": 1850
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002153632214875619,
+      "loss": 1.4163,
+      "step": 1855
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021495383077629513,
+      "loss": 1.4138,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021454384369720195,
+      "loss": 1.7746,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002141332640145423,
+      "loss": 1.9383,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021372209549801706,
+      "loss": 1.58,
+      "step": 1875
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021331034192273343,
+      "loss": 1.7546,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00021289800706917025,
+      "loss": 1.5329,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002124850947231434,
+      "loss": 1.1873,
+      "step": 1890
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021207160867577087,
+      "loss": 1.7837,
+      "step": 1895
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021165755272343806,
+      "loss": 1.4512,
+      "step": 1900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002112429306677629,
+      "loss": 1.7442,
+      "step": 1905
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021082774631556102,
+      "loss": 1.5365,
+      "step": 1910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021041200347881057,
+      "loss": 1.8142,
+      "step": 1915
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002099957059746175,
+      "loss": 1.9143,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00020957885762518043,
+      "loss": 2.0534,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002091614622577554,
+      "loss": 1.7375,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00020874352370462085,
+      "loss": 1.6084,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00020832504580304247,
+      "loss": 1.1369,
+      "step": 1940
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000207906032395238,
+      "loss": 1.473,
+      "step": 1945
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020748648732834182,
+      "loss": 1.5715,
+      "step": 1950
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002070664144543696,
+      "loss": 1.5102,
+      "step": 1955
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020664581763018324,
+      "loss": 1.2595,
+      "step": 1960
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020622470071745498,
+      "loss": 1.7648,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002058030675826324,
+      "loss": 1.5256,
+      "step": 1970
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020538092209690265,
+      "loss": 1.5349,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020495826813615706,
+      "loss": 1.1114,
+      "step": 1980
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002045351095809553,
+      "loss": 1.5913,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002041114503164902,
+      "loss": 1.7158,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00020368729423255155,
+      "loss": 1.6397,
+      "step": 1995
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002032626452234908,
+      "loss": 1.7161,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020283750718818501,
+      "loss": 1.4809,
+      "step": 2005
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020241188403000138,
+      "loss": 1.8065,
+      "step": 2010
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002019857796567611,
+      "loss": 1.5828,
+      "step": 2015
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020155919798070366,
+      "loss": 1.5952,
+      "step": 2020
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020113214291845067,
+      "loss": 1.3225,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020070461839097038,
+      "loss": 1.2879,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020027662832354122,
+      "loss": 1.4086,
+      "step": 2035
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019984817664571596,
+      "loss": 2.7161,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019941926729128565,
+      "loss": 1.661,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019898990419824333,
+      "loss": 1.1885,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019856009130874816,
+      "loss": 1.3895,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019812983256908893,
+      "loss": 1.6416,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001976991319296482,
+      "loss": 1.2385,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019726799334486545,
+      "loss": 1.658,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019683642077320134,
+      "loss": 1.2902,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019640441817710117,
+      "loss": 1.8334,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001959719895229583,
+      "loss": 1.1194,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000195539138781078,
+      "loss": 1.3623,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019510586992564093,
+      "loss": 1.6465,
+      "step": 2095
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019467218693466665,
+      "loss": 1.2476,
+      "step": 2100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019423809378997696,
+      "loss": 1.6968,
+      "step": 2105
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019380359447715944,
+      "loss": 1.379,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000193368692985531,
+      "loss": 1.4896,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000192933393308101,
+      "loss": 1.5245,
+      "step": 2120
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001924976994415348,
+      "loss": 1.6641,
+      "step": 2125
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019206161538611692,
+      "loss": 1.4637,
+      "step": 2130
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019162514514571433,
+      "loss": 1.4069,
+      "step": 2135
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019118829272773985,
+      "loss": 1.3833,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019075106214311503,
+      "loss": 1.519,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019031345740623372,
+      "loss": 1.5278,
+      "step": 2150
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001898754825349249,
+      "loss": 1.4175,
+      "step": 2155
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018943714155041593,
+      "loss": 2.3051,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001889984384772957,
+      "loss": 1.9164,
+      "step": 2165
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018855937734347736,
+      "loss": 1.6033,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018811996218016176,
+      "loss": 1.5784,
+      "step": 2175
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001876801970218002,
+      "loss": 1.6539,
+      "step": 2180
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018724008590605742,
+      "loss": 1.3472,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018679963287377452,
+      "loss": 1.577,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018635884196893188,
+      "loss": 1.4079,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00018591771723861197,
+      "loss": 1.2238,
+      "step": 2200
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018547626273296238,
+      "loss": 2.0988,
+      "step": 2205
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018503448250515843,
+      "loss": 1.7247,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018459238061136602,
+      "loss": 1.7499,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001841499611107044,
+      "loss": 1.3963,
+      "step": 2220
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001837072280652089,
+      "loss": 1.68,
+      "step": 2225
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018326418553979367,
+      "loss": 1.5986,
+      "step": 2230
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001828208376022143,
+      "loss": 1.817,
+      "step": 2235
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018237718832303043,
+      "loss": 2.1561,
+      "step": 2240
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018193324177556854,
+      "loss": 1.5795,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018148900203588444,
+      "loss": 1.3028,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018104447318272583,
+      "loss": 1.516,
+      "step": 2255
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001805996592974948,
+      "loss": 1.6465,
+      "step": 2260
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018015456446421059,
+      "loss": 1.3915,
+      "step": 2265
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001797091927694718,
+      "loss": 1.2381,
+      "step": 2270
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017926354830241924,
+      "loss": 1.4619,
+      "step": 2275
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017881763515469783,
+      "loss": 2.0805,
+      "step": 2280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017837145742041955,
+      "loss": 1.2531,
+      "step": 2285
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001779250191961256,
+      "loss": 1.3414,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017747832458074898,
+      "loss": 1.9254,
+      "step": 2295
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017703137767557652,
+      "loss": 1.2747,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001765841825842116,
+      "loss": 1.7596,
+      "step": 2305
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017613674341253606,
+      "loss": 1.7383,
+      "step": 2310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017568906426867312,
+      "loss": 2.1931,
+      "step": 2315
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017524114926294887,
+      "loss": 0.9475,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001747930025078553,
+      "loss": 1.419,
+      "step": 2325
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001743446281180119,
+      "loss": 1.4287,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001738960302101284,
+      "loss": 1.8408,
+      "step": 2335
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017344721290296655,
+      "loss": 1.7793,
+      "step": 2340
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001729981803173026,
+      "loss": 1.3913,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017254893657588934,
+      "loss": 1.6219,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001720994858034183,
+      "loss": 0.8477,
+      "step": 2355
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017164983212648177,
+      "loss": 2.008,
+      "step": 2360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017119997967353514,
+      "loss": 1.2596,
+      "step": 2365
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001707499325748586,
+      "loss": 2.0161,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017029969496251966,
+      "loss": 1.4243,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016984927097033495,
+      "loss": 1.6643,
+      "step": 2380
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016939866473383237,
+      "loss": 1.4719,
+      "step": 2385
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016894788039021304,
+      "loss": 1.4121,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001684969220783133,
+      "loss": 2.2479,
+      "step": 2395
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016804579393856688,
+      "loss": 1.4241,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016759450011296674,
+      "loss": 1.6946,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00016714304474502696,
+      "loss": 1.6951,
+      "step": 2410
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016669143197974495,
+      "loss": 1.6273,
+      "step": 2415
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016623966596356315,
+      "loss": 1.8207,
+      "step": 2420
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016578775084433116,
+      "loss": 1.781,
+      "step": 2425
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016533569077126749,
+      "loss": 1.2038,
+      "step": 2430
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016488348989492145,
+      "loss": 1.4287,
+      "step": 2435
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016443115236713528,
+      "loss": 1.4447,
+      "step": 2440
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016397868234100584,
+      "loss": 1.1601,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001635260839708464,
+      "loss": 1.4291,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016307336141214873,
+      "loss": 1.4518,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016262051882154477,
+      "loss": 1.6841,
+      "step": 2460
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016216756035676845,
+      "loss": 1.4064,
+      "step": 2465
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001617144901766178,
+      "loss": 0.8423,
+      "step": 2470
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016126131244091643,
+      "loss": 1.6949,
+      "step": 2475
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016080803131047534,
+      "loss": 1.3659,
+      "step": 2480
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016035465094705502,
+      "loss": 1.547,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015990117551332702,
+      "loss": 1.4655,
+      "step": 2490
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001594476091728357,
+      "loss": 1.5663,
+      "step": 2495
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015899395608996015,
+      "loss": 1.1099,
+      "step": 2500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001585402204298758,
+      "loss": 1.6187,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00015808640635851638,
+      "loss": 1.8124,
+      "step": 2510
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015763251804253538,
+      "loss": 1.1152,
+      "step": 2515
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015717855964926812,
+      "loss": 2.091,
+      "step": 2520
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015672453534669318,
+      "loss": 1.3521,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015627044930339442,
+      "loss": 1.3461,
+      "step": 2530
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001558163056885225,
+      "loss": 1.7911,
+      "step": 2535
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015536210867175666,
+      "loss": 1.4027,
+      "step": 2540
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015490786242326643,
+      "loss": 1.1494,
+      "step": 2545
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015445357111367337,
+      "loss": 2.7537,
+      "step": 2550
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015399923891401286,
+      "loss": 1.2787,
+      "step": 2555
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015354486999569558,
+      "loss": 1.7394,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00015309046853046937,
+      "loss": 1.5708,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015263603869038086,
+      "loss": 1.9391,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001521815846477374,
+      "loss": 1.6462,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015172711057506827,
+      "loss": 1.1655,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015127262064508688,
+      "loss": 1.6431,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015081811903065205,
+      "loss": 1.4204,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015036360990473008,
+      "loss": 1.3463,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000149909097440356,
+      "loss": 1.807,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001494545858105958,
+      "loss": 1.1753,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001490000791885075,
+      "loss": 1.5372,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014854558174710322,
+      "loss": 1.9133,
+      "step": 2615
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014809109765931097,
+      "loss": 1.6995,
+      "step": 2620
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014763663109793596,
+      "loss": 2.2809,
+      "step": 2625
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014718218623562256,
+      "loss": 1.2495,
+      "step": 2630
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014672776724481584,
+      "loss": 1.4304,
+      "step": 2635
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014627337829772346,
+      "loss": 1.4201,
+      "step": 2640
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014581902356627718,
+      "loss": 1.6049,
+      "step": 2645
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014536470722209449,
+      "loss": 1.6997,
+      "step": 2650
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014491043343644065,
+      "loss": 1.5557,
+      "step": 2655
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014445620638019004,
+      "loss": 1.5758,
+      "step": 2660
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.000144002030223788,
+      "loss": 0.9978,
+      "step": 2665
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001435479091372126,
+      "loss": 1.3099,
+      "step": 2670
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001430938472899362,
+      "loss": 1.6165,
+      "step": 2675
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001426398488508873,
+      "loss": 1.7321,
+      "step": 2680
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014218591798841228,
+      "loss": 1.5079,
+      "step": 2685
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014173205887023702,
+      "loss": 1.5661,
+      "step": 2690
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014127827566342863,
+      "loss": 1.4923,
+      "step": 2695
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014082457253435739,
+      "loss": 1.2609,
+      "step": 2700
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001403709536486582,
+      "loss": 1.7925,
+      "step": 2705
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013991742317119257,
+      "loss": 1.6333,
+      "step": 2710
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013946398526601034,
+      "loss": 1.649,
+      "step": 2715
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013901064409631132,
+      "loss": 1.6406,
+      "step": 2720
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013855740382440714,
+      "loss": 1.6813,
+      "step": 2725
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013810426861168314,
+      "loss": 2.0483,
+      "step": 2730
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013765124261856,
+      "loss": 1.2954,
+      "step": 2735
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013719833000445564,
+      "loss": 1.8643,
+      "step": 2740
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013674553492774697,
+      "loss": 1.0872,
+      "step": 2745
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013629286154573174,
+      "loss": 1.8002,
+      "step": 2750
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013584031401459045,
+      "loss": 1.2758,
+      "step": 2755
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013538789648934795,
+      "loss": 1.6906,
+      "step": 2760
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013493561312383565,
+      "loss": 1.6229,
+      "step": 2765
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000134483468070653,
+      "loss": 1.5937,
+      "step": 2770
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013403146548112975,
+      "loss": 1.6342,
+      "step": 2775
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013357960950528743,
+      "loss": 1.5929,
+      "step": 2780
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013312790429180154,
+      "loss": 1.9885,
+      "step": 2785
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001326763539879634,
+      "loss": 1.8865,
+      "step": 2790
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013222496273964204,
+      "loss": 1.1842,
+      "step": 2795
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000131773734691246,
+      "loss": 1.088,
+      "step": 2800
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013132267398568555,
+      "loss": 1.5499,
+      "step": 2805
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013087178476433456,
+      "loss": 1.6725,
+      "step": 2810
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013042107116699228,
+      "loss": 1.3022,
+      "step": 2815
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00012997053733184558,
+      "loss": 1.1577,
+      "step": 2820
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001295201873954308,
+      "loss": 1.751,
+      "step": 2825
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001290700254925959,
+      "loss": 1.4562,
+      "step": 2830
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012862005575646246,
+      "loss": 1.5687,
+      "step": 2835
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012817028231838757,
+      "loss": 1.9392,
+      "step": 2840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012772070930792612,
+      "loss": 1.5673,
+      "step": 2845
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001272713408527927,
+      "loss": 1.2749,
+      "step": 2850
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012682218107882393,
+      "loss": 1.5739,
+      "step": 2855
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012637323410994033,
+      "loss": 1.4778,
+      "step": 2860
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012592450406810862,
+      "loss": 1.5713,
+      "step": 2865
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012547599507330379,
+      "loss": 1.2776,
+      "step": 2870
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00012502771124347137,
+      "loss": 1.8842,
+      "step": 2875
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001245796566944895,
+      "loss": 2.065,
+      "step": 2880
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012413183554013112,
+      "loss": 1.5904,
+      "step": 2885
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012368425189202654,
+      "loss": 1.8088,
+      "step": 2890
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012323690985962515,
+      "loss": 1.8503,
+      "step": 2895
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012278981355015811,
+      "loss": 1.211,
+      "step": 2900
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001223429670686005,
+      "loss": 1.4097,
+      "step": 2905
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001218963745176335,
+      "loss": 1.3921,
+      "step": 2910
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012145003999760698,
+      "loss": 1.6914,
+      "step": 2915
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012100396760650171,
+      "loss": 2.0817,
+      "step": 2920
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001205581614398917,
+      "loss": 1.4066,
+      "step": 2925
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012011262559090664,
+      "loss": 1.9319,
+      "step": 2930
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011966736415019432,
+      "loss": 1.3314,
+      "step": 2935
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011922238120588319,
+      "loss": 1.2883,
+      "step": 2940
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011877768084354448,
+      "loss": 1.9158,
+      "step": 2945
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011833326714615522,
+      "loss": 1.7002,
+      "step": 2950
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001178891441940602,
+      "loss": 1.1873,
+      "step": 2955
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011744531606493482,
+      "loss": 1.4304,
+      "step": 2960
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011700178683374769,
+      "loss": 1.5897,
+      "step": 2965
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.000116558560572723,
+      "loss": 1.7278,
+      "step": 2970
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011611564135130329,
+      "loss": 1.8035,
+      "step": 2975
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00011567303323611203,
+      "loss": 1.7011,
+      "step": 2980
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011523074029091633,
+      "loss": 1.3019,
+      "step": 2985
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011478876657658953,
+      "loss": 1.5944,
+      "step": 2990
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011434711615107404,
+      "loss": 1.1372,
+      "step": 2995
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011390579306934402,
+      "loss": 1.4772,
+      "step": 3000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011346480138336811,
+      "loss": 1.2776,
+      "step": 3005
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011302414514207233,
+      "loss": 1.9572,
+      "step": 3010
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011258382839130281,
+      "loss": 1.4264,
+      "step": 3015
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011214385517378864,
+      "loss": 2.216,
+      "step": 3020
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011170422952910482,
+      "loss": 1.4154,
+      "step": 3025
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001112649554936352,
+      "loss": 1.5587,
+      "step": 3030
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011082603710053528,
+      "loss": 1.9853,
+      "step": 3035
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011038747837969526,
+      "loss": 1.2621,
+      "step": 3040
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010994928335770312,
+      "loss": 1.5037,
+      "step": 3045
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010951145605780751,
+      "loss": 1.6375,
+      "step": 3050
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010907400049988088,
+      "loss": 1.3973,
+      "step": 3055
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001086369207003825,
+      "loss": 1.5211,
+      "step": 3060
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010820022067232186,
+      "loss": 1.1739,
+      "step": 3065
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010776390442522137,
+      "loss": 1.7858,
+      "step": 3070
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010732797596507992,
+      "loss": 1.4643,
+      "step": 3075
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00010689243929433593,
+      "loss": 1.5519,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010645729841183066,
+      "loss": 1.2495,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010602255731277142,
+      "loss": 1.3118,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010558821998869503,
+      "loss": 0.9586,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010515429042743104,
+      "loss": 1.6216,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010472077261306509,
+      "loss": 1.2725,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010428767052590258,
+      "loss": 1.3954,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010385498814243176,
+      "loss": 1.3674,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010342272943528756,
+      "loss": 1.7398,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010299089837321489,
+      "loss": 1.4755,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010255949892103225,
+      "loss": 1.7468,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010212853503959535,
+      "loss": 1.2841,
+      "step": 3135
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010169801068576088,
+      "loss": 1.3969,
+      "step": 3140
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010126792981234985,
+      "loss": 1.4495,
+      "step": 3145
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010083829636811161,
+      "loss": 1.7657,
+      "step": 3150
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010040911429768748,
+      "loss": 1.6876,
+      "step": 3155
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.998038754157452e-05,
+      "loss": 1.1386,
+      "step": 3160
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.95521200360893e-05,
+      "loss": 1.2276,
+      "step": 3165
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.91243157133319e-05,
+      "loss": 1.4063,
+      "step": 3170
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.869697850114969e-05,
+      "loss": 1.1884,
+      "step": 3175
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.827011232310132e-05,
+      "loss": 1.0424,
+      "step": 3180
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.784372109842071e-05,
+      "loss": 1.5932,
+      "step": 3185
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.7417808741981e-05,
+      "loss": 1.5054,
+      "step": 3190
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.699237916425861e-05,
+      "loss": 1.648,
+      "step": 3195
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.656743627129745e-05,
+      "loss": 1.5159,
+      "step": 3200
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.614298396467291e-05,
+      "loss": 1.6202,
+      "step": 3205
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.571902614145616e-05,
+      "loss": 1.4917,
+      "step": 3210
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.529556669417823e-05,
+      "loss": 1.4311,
+      "step": 3215
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.487260951079448e-05,
+      "loss": 1.4103,
+      "step": 3220
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.445015847464868e-05,
+      "loss": 1.4298,
+      "step": 3225
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.402821746443744e-05,
+      "loss": 1.3865,
+      "step": 3230
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.360679035417479e-05,
+      "loss": 1.4313,
+      "step": 3235
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.318588101315616e-05,
+      "loss": 1.5334,
+      "step": 3240
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.276549330592334e-05,
+      "loss": 1.0339,
+      "step": 3245
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.234563109222874e-05,
+      "loss": 1.4073,
+      "step": 3250
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.192629822699993e-05,
+      "loss": 1.2445,
+      "step": 3255
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.150749856030429e-05,
+      "loss": 1.3676,
+      "step": 3260
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.10892359373139e-05,
+      "loss": 0.9688,
+      "step": 3265
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.067151419826976e-05,
+      "loss": 1.3644,
+      "step": 3270
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.025433717844695e-05,
+      "loss": 1.648,
+      "step": 3275
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 8.983770870811918e-05,
+      "loss": 1.5491,
+      "step": 3280
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 8.942163261252378e-05,
+      "loss": 1.1729,
+      "step": 3285
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 8.900611271182636e-05,
+      "loss": 1.6664,
+      "step": 3290
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.85911528210861e-05,
+      "loss": 1.4335,
+      "step": 3295
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.817675675022034e-05,
+      "loss": 1.6384,
+      "step": 3300
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.77629283039697e-05,
+      "loss": 1.5178,
+      "step": 3305
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.734967128186338e-05,
+      "loss": 0.9798,
+      "step": 3310
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.693698947818397e-05,
+      "loss": 1.7922,
+      "step": 3315
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.652488668193286e-05,
+      "loss": 1.8276,
+      "step": 3320
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.611336667679511e-05,
+      "loss": 1.3232,
+      "step": 3325
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.570243324110529e-05,
+      "loss": 1.4273,
+      "step": 3330
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.529209014781201e-05,
+      "loss": 1.5293,
+      "step": 3335
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.488234116444404e-05,
+      "loss": 1.2581,
+      "step": 3340
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.447319005307532e-05,
+      "loss": 1.3106,
+      "step": 3345
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.406464057029023e-05,
+      "loss": 0.7905,
+      "step": 3350
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.365669646714983e-05,
+      "loss": 1.51,
+      "step": 3355
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.324936148915641e-05,
+      "loss": 1.6895,
+      "step": 3360
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.284263937622001e-05,
+      "loss": 1.5375,
+      "step": 3365
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.243653386262362e-05,
+      "loss": 2.1622,
+      "step": 3370
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.203104867698876e-05,
+      "loss": 1.4393,
+      "step": 3375
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.162618754224177e-05,
+      "loss": 1.5568,
+      "step": 3380
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.122195417557919e-05,
+      "loss": 1.1689,
+      "step": 3385
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.081835228843367e-05,
+      "loss": 1.4658,
+      "step": 3390
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.041538558644015e-05,
+      "loss": 2.065,
+      "step": 3395
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.001305776940163e-05,
+      "loss": 1.5075,
+      "step": 3400
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.961137253125504e-05,
+      "loss": 1.2665,
+      "step": 3405
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.921033356003797e-05,
+      "loss": 1.4331,
+      "step": 3410
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.880994453785393e-05,
+      "loss": 1.2724,
+      "step": 3415
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.841020914083923e-05,
+      "loss": 1.4093,
+      "step": 3420
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.801113103912882e-05,
+      "loss": 1.1195,
+      "step": 3425
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.76127138968229e-05,
+      "loss": 1.741,
+      "step": 3430
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.721496137195308e-05,
+      "loss": 1.4885,
+      "step": 3435
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.681787711644877e-05,
+      "loss": 1.4523,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 7.64214647761038e-05,
+      "loss": 1.3673,
+      "step": 3445
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.602572799054291e-05,
+      "loss": 1.4642,
+      "step": 3450
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.563067039318815e-05,
+      "loss": 1.6162,
+      "step": 3455
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.523629561122577e-05,
+      "loss": 1.6921,
+      "step": 3460
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.484260726557283e-05,
+      "loss": 1.8285,
+      "step": 3465
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.444960897084388e-05,
+      "loss": 1.6991,
+      "step": 3470
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.405730433531792e-05,
+      "loss": 1.282,
+      "step": 3475
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.366569696090496e-05,
+      "loss": 1.7964,
+      "step": 3480
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.327479044311346e-05,
+      "loss": 1.7076,
+      "step": 3485
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.288458837101675e-05,
+      "loss": 1.4987,
+      "step": 3490
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.249509432722056e-05,
+      "loss": 1.7264,
+      "step": 3495
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.210631188782988e-05,
+      "loss": 1.5205,
+      "step": 3500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.171824462241607e-05,
+      "loss": 1.3453,
+      "step": 3505
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.133089609398425e-05,
+      "loss": 1.6853,
+      "step": 3510
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.094426985894059e-05,
+      "loss": 1.5102,
+      "step": 3515
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.055836946705939e-05,
+      "loss": 1.773,
+      "step": 3520
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.017319846145099e-05,
+      "loss": 1.6526,
+      "step": 3525
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.978876037852862e-05,
+      "loss": 1.4008,
+      "step": 3530
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.940505874797639e-05,
+      "loss": 1.64,
+      "step": 3535
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.902209709271685e-05,
+      "loss": 1.6294,
+      "step": 3540
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.863987892887826e-05,
+      "loss": 1.656,
+      "step": 3545
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 6.825840776576282e-05,
+      "loss": 1.35,
+      "step": 3550
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.787768710581414e-05,
+      "loss": 1.1362,
+      "step": 3555
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.749772044458505e-05,
+      "loss": 1.6395,
+      "step": 3560
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.711851127070579e-05,
+      "loss": 1.6092,
+      "step": 3565
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.67400630658516e-05,
+      "loss": 1.4654,
+      "step": 3570
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.636237930471108e-05,
+      "loss": 1.5386,
+      "step": 3575
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.598546345495417e-05,
+      "loss": 1.5661,
+      "step": 3580
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.56093189772002e-05,
+      "loss": 1.4736,
+      "step": 3585
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.52339493249864e-05,
+      "loss": 1.3499,
+      "step": 3590
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.485935794473571e-05,
+      "loss": 0.8297,
+      "step": 3595
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.448554827572567e-05,
+      "loss": 1.4855,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.411252375005663e-05,
+      "loss": 1.6239,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.374028779262003e-05,
+      "loss": 1.4557,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.336884382106723e-05,
+      "loss": 1.5754,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.299819524577817e-05,
+      "loss": 1.3462,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.262834546982969e-05,
+      "loss": 1.528,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.225929788896473e-05,
+      "loss": 0.67,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.189105589156076e-05,
+      "loss": 1.187,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.152362285859906e-05,
+      "loss": 1.5347,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.115700216363342e-05,
+      "loss": 1.5134,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.079119717275905e-05,
+      "loss": 1.1493,
+      "step": 3650
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.04262112445821e-05,
+      "loss": 0.9713,
+      "step": 3655
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.006204773018831e-05,
+      "loss": 1.4658,
+      "step": 3660
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.96987099731127e-05,
+      "loss": 1.8118,
+      "step": 3665
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.933620130930867e-05,
+      "loss": 1.8907,
+      "step": 3670
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.897452506711719e-05,
+      "loss": 1.2703,
+      "step": 3675
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.86136845672366e-05,
+      "loss": 1.4875,
+      "step": 3680
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.8253683122691944e-05,
+      "loss": 1.6493,
+      "step": 3685
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.7894524038804373e-05,
+      "loss": 1.481,
+      "step": 3690
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.75362106131612e-05,
+      "loss": 1.376,
+      "step": 3695
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.7178746135585276e-05,
+      "loss": 1.4796,
+      "step": 3700
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.682213388810497e-05,
+      "loss": 2.0087,
+      "step": 3705
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6466377144923985e-05,
+      "loss": 1.8715,
+      "step": 3710
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.6111479172391136e-05,
+      "loss": 1.7617,
+      "step": 3715
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.575744322897074e-05,
+      "loss": 1.5077,
+      "step": 3720
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.540427256521226e-05,
+      "loss": 1.6198,
+      "step": 3725
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.505197042372082e-05,
+      "loss": 1.6346,
+      "step": 3730
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.470054003912727e-05,
+      "loss": 1.7283,
+      "step": 3735
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.434998463805837e-05,
+      "loss": 0.9581,
+      "step": 3740
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.4000307439107434e-05,
+      "loss": 1.6349,
+      "step": 3745
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.3651511652804663e-05,
+      "loss": 1.8797,
+      "step": 3750
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.330360048158741e-05,
+      "loss": 1.2728,
+      "step": 3755
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.2956577119771405e-05,
+      "loss": 1.306,
+      "step": 3760
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.261044475352064e-05,
+      "loss": 1.7036,
+      "step": 3765
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.226520656081875e-05,
+      "loss": 1.3375,
+      "step": 3770
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.192086571143956e-05,
+      "loss": 1.1267,
+      "step": 3775
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.1577425366917915e-05,
+      "loss": 1.4579,
+      "step": 3780
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.12348886805209e-05,
+      "loss": 0.9636,
+      "step": 3785
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.089325879721856e-05,
+      "loss": 1.844,
+      "step": 3790
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.05525388536554e-05,
+      "loss": 1.3017,
+      "step": 3795
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.021273197812134e-05,
+      "loss": 1.299,
+      "step": 3800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.987384129052291e-05,
+      "loss": 1.3614,
+      "step": 3805
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.953586990235492e-05,
+      "loss": 1.7793,
+      "step": 3810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.9198820916671634e-05,
+      "loss": 1.561,
+      "step": 3815
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.886269742805836e-05,
+      "loss": 1.7795,
+      "step": 3820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.8527502522603104e-05,
+      "loss": 0.8945,
+      "step": 3825
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.8193239277868004e-05,
+      "loss": 1.5663,
+      "step": 3830
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.7859910762861406e-05,
+      "loss": 1.4586,
+      "step": 3835
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.752752003800951e-05,
+      "loss": 1.4256,
+      "step": 3840
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.719607015512809e-05,
+      "loss": 1.6858,
+      "step": 3845
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.686556415739488e-05,
+      "loss": 1.5589,
+      "step": 3850
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.653600507932137e-05,
+      "loss": 1.6629,
+      "step": 3855
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.620739594672488e-05,
+      "loss": 1.7728,
+      "step": 3860
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.587973977670103e-05,
+      "loss": 1.8571,
+      "step": 3865
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.555303957759569e-05,
+      "loss": 0.9403,
+      "step": 3870
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.522729834897795e-05,
+      "loss": 1.4019,
+      "step": 3875
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.490251908161175e-05,
+      "loss": 1.4581,
+      "step": 3880
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.4578704757429186e-05,
+      "loss": 1.4361,
+      "step": 3885
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.4255858349502724e-05,
+      "loss": 1.587,
+      "step": 3890
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.3933982822017876e-05,
+      "loss": 2.0539,
+      "step": 3895
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.361308113024626e-05,
+      "loss": 1.2896,
+      "step": 3900
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.32931562205183e-05,
+      "loss": 1.1724,
+      "step": 3905
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2974211030196e-05,
+      "loss": 1.6518,
+      "step": 3910
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.265624848764639e-05,
+      "loss": 1.6734,
+      "step": 3915
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.233927151221432e-05,
+      "loss": 1.8174,
+      "step": 3920
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.2023283014195596e-05,
+      "loss": 1.1279,
+      "step": 3925
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.170828589481071e-05,
+      "loss": 1.5869,
+      "step": 3930
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.1394283046177605e-05,
+      "loss": 1.2266,
+      "step": 3935
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.108127735128561e-05,
+      "loss": 1.4214,
+      "step": 3940
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.076927168396865e-05,
+      "loss": 1.0697,
+      "step": 3945
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.0458268908879056e-05,
+      "loss": 1.6432,
+      "step": 3950
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.014827188146122e-05,
+      "loss": 1.6831,
+      "step": 3955
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.983928344792523e-05,
+      "loss": 1.1097,
+      "step": 3960
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 3.953130644522096e-05,
+      "loss": 1.6208,
+      "step": 3965
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.922434370101198e-05,
+      "loss": 1.7125,
+      "step": 3970
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.891839803364934e-05,
+      "loss": 1.145,
+      "step": 3975
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.861347225214603e-05,
+      "loss": 1.1312,
+      "step": 3980
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.830956915615106e-05,
+      "loss": 1.5828,
+      "step": 3985
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.8006691535923673e-05,
+      "loss": 1.1374,
+      "step": 3990
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7704842172307894e-05,
+      "loss": 1.1668,
+      "step": 3995
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.740402383670674e-05,
+      "loss": 1.6879,
+      "step": 4000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.7104239291057085e-05,
+      "loss": 1.3145,
+      "step": 4005
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.6805491287804144e-05,
+      "loss": 1.4949,
+      "step": 4010
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.650778256987607e-05,
+      "loss": 1.916,
+      "step": 4015
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.62111158706591e-05,
+      "loss": 1.3152,
+      "step": 4020
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.591549391397208e-05,
+      "loss": 1.6343,
+      "step": 4025
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.562091941404179e-05,
+      "loss": 1.4102,
+      "step": 4030
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.532739507547788e-05,
+      "loss": 1.6651,
+      "step": 4035
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.5034923593247844e-05,
+      "loss": 1.7953,
+      "step": 4040
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.474350765265279e-05,
+      "loss": 1.1191,
+      "step": 4045
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.4453149929302145e-05,
+      "loss": 2.0854,
+      "step": 4050
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.416385308908962e-05,
+      "loss": 1.0797,
+      "step": 4055
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.387561978816847e-05,
+      "loss": 2.0707,
+      "step": 4060
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.358845267292706e-05,
+      "loss": 1.244,
+      "step": 4065
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.330235437996478e-05,
+      "loss": 1.3131,
+      "step": 4070
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.301732753606776e-05,
+      "loss": 1.6666,
+      "step": 4075
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.2733374758184513e-05,
+      "loss": 1.7402,
+      "step": 4080
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.245049865340236e-05,
+      "loss": 1.4406,
+      "step": 4085
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.216870181892303e-05,
+      "loss": 1.5123,
+      "step": 4090
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.1887986842039165e-05,
+      "loss": 1.5698,
+      "step": 4095
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.160835630011038e-05,
+      "loss": 1.3733,
+      "step": 4100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.132981276053966e-05,
+      "loss": 1.6571,
+      "step": 4105
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.105235878074981e-05,
+      "loss": 1.7105,
+      "step": 4110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.077599690815981e-05,
+      "loss": 1.6224,
+      "step": 4115
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 3.0500729680161663e-05,
+      "loss": 1.4937,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.0226559624097034e-05,
+      "loss": 1.7792,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.9953489257233844e-05,
+      "loss": 1.3071,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.9681521086743422e-05,
+      "loss": 1.4754,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.9410657609677434e-05,
+      "loss": 1.7724,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.914090131294475e-05,
+      "loss": 1.4711,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.8872254673288904e-05,
+      "loss": 1.4936,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.860472015726506e-05,
+      "loss": 1.7267,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.8338300221217646e-05,
+      "loss": 1.6548,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.807299731125773e-05,
+      "loss": 1.3744,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.780881386324027e-05,
+      "loss": 1.6085,
+      "step": 4170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7545752302742245e-05,
+      "loss": 1.5895,
+      "step": 4175
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.7283815045039885e-05,
+      "loss": 1.3282,
+      "step": 4180
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.702300449508691e-05,
+      "loss": 1.6033,
+      "step": 4185
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.6763323047492258e-05,
+      "loss": 1.6469,
+      "step": 4190
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.650477308649797e-05,
+      "loss": 1.727,
+      "step": 4195
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.624735698595758e-05,
+      "loss": 1.4403,
+      "step": 4200
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5991077109314206e-05,
+      "loss": 1.6997,
+      "step": 4205
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5735935809578656e-05,
+      "loss": 1.7687,
+      "step": 4210
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5481935429308176e-05,
+      "loss": 1.6448,
+      "step": 4215
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5229078300584693e-05,
+      "loss": 1.3343,
+      "step": 4220
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4977366744993432e-05,
+      "loss": 1.6547,
+      "step": 4225
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.472680307360175e-05,
+      "loss": 1.7316,
+      "step": 4230
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.447738958693764e-05,
+      "loss": 1.7601,
+      "step": 4235
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.422912857496894e-05,
+      "loss": 1.7816,
+      "step": 4240
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.3982022317082032e-05,
+      "loss": 1.8284,
+      "step": 4245
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.3736073082061074e-05,
+      "loss": 1.5585,
+      "step": 4250
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.3491283128067174e-05,
+      "loss": 1.5799,
+      "step": 4255
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.3247654702617454e-05,
+      "loss": 1.3991,
+      "step": 4260
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.3005190042564697e-05,
+      "loss": 1.2329,
+      "step": 4265
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.276389137407669e-05,
+      "loss": 1.7101,
+      "step": 4270
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.25237609126156e-05,
+      "loss": 1.6476,
+      "step": 4275
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.2284800862918068e-05,
+      "loss": 1.5978,
+      "step": 4280
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.2047013418974446e-05,
+      "loss": 2.0835,
+      "step": 4285
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.1810400764009072e-05,
+      "loss": 1.1277,
+      "step": 4290
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.1574965070460043e-05,
+      "loss": 1.24,
+      "step": 4295
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.1340708499959197e-05,
+      "loss": 1.5577,
+      "step": 4300
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.1107633203312452e-05,
+      "loss": 1.1969,
+      "step": 4305
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.0875741320479967e-05,
+      "loss": 1.3219,
+      "step": 4310
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.0645034980556408e-05,
+      "loss": 1.548,
+      "step": 4315
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.04155163017516e-05,
+      "loss": 1.3474,
+      "step": 4320
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.0187187391370824e-05,
+      "loss": 1.349,
+      "step": 4325
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.996005034579573e-05,
+      "loss": 1.2657,
+      "step": 4330
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9734107250464926e-05,
+      "loss": 0.997,
+      "step": 4335
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.950936017985485e-05,
+      "loss": 1.8957,
+      "step": 4340
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.928581119746081e-05,
+      "loss": 1.589,
+      "step": 4345
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.9063462355777847e-05,
+      "loss": 1.077,
+      "step": 4350
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8842315696282152e-05,
+      "loss": 1.7227,
+      "step": 4355
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8622373249412137e-05,
+      "loss": 2.0054,
+      "step": 4360
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8403637034549785e-05,
+      "loss": 1.3692,
+      "step": 4365
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8186109060002262e-05,
+      "loss": 1.7883,
+      "step": 4370
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.796979132298335e-05,
+      "loss": 1.2446,
+      "step": 4375
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.775468580959508e-05,
+      "loss": 1.3009,
+      "step": 4380
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.754079449480969e-05,
+      "loss": 1.6341,
+      "step": 4385
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.7328119342451165e-05,
+      "loss": 1.6593,
+      "step": 4390
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.7116662305177654e-05,
+      "loss": 1.9338,
+      "step": 4395
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6906425324463058e-05,
+      "loss": 1.6533,
+      "step": 4400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6697410330579586e-05,
+      "loss": 1.5444,
+      "step": 4405
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6489619242579856e-05,
+      "loss": 1.1762,
+      "step": 4410
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.628305396827922e-05,
+      "loss": 1.4165,
+      "step": 4415
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.607771640423842e-05,
+      "loss": 1.4529,
+      "step": 4420
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.587360843574609e-05,
+      "loss": 1.3064,
+      "step": 4425
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.567073193680134e-05,
+      "loss": 0.9361,
+      "step": 4430
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.546908877009676e-05,
+      "loss": 1.3433,
+      "step": 4435
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.526868078700119e-05,
+      "loss": 1.2926,
+      "step": 4440
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.5069509827542648e-05,
+      "loss": 0.8124,
+      "step": 4445
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4871577720391708e-05,
+      "loss": 1.6786,
+      "step": 4450
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.467488628284434e-05,
+      "loss": 1.9403,
+      "step": 4455
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4479437320805548e-05,
+      "loss": 1.5124,
+      "step": 4460
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4285232628772602e-05,
+      "loss": 1.4309,
+      "step": 4465
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4092273989818626e-05,
+      "loss": 1.5283,
+      "step": 4470
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.3900563175576224e-05,
+      "loss": 1.753,
+      "step": 4475
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.371010194622117e-05,
+      "loss": 1.3044,
+      "step": 4480
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.3520892050456361e-05,
+      "loss": 1.6555,
+      "step": 4485
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.3332935225495656e-05,
+      "loss": 1.189,
+      "step": 4490
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.3146233197047895e-05,
+      "loss": 2.1216,
+      "step": 4495
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2960787679301171e-05,
+      "loss": 1.1811,
+      "step": 4500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.277660037490707e-05,
+      "loss": 1.8617,
+      "step": 4505
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2593672974964936e-05,
+      "loss": 1.828,
+      "step": 4510
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2412007159006465e-05,
+      "loss": 1.8121,
+      "step": 4515
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.2231604594980155e-05,
+      "loss": 1.3953,
+      "step": 4520
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.205246693923616e-05,
+      "loss": 1.4313,
+      "step": 4525
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1874595836510964e-05,
+      "loss": 1.1439,
+      "step": 4530
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.1697992919912247e-05,
+      "loss": 1.5366,
+      "step": 4535
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.152265981090404e-05,
+      "loss": 1.6374,
+      "step": 4540
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.1348598119291646e-05,
+      "loss": 1.5402,
+      "step": 4545
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.117580944320708e-05,
+      "loss": 1.7178,
+      "step": 4550
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.100429536909423e-05,
+      "loss": 1.5968,
+      "step": 4555
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0834057471694269e-05,
+      "loss": 1.0687,
+      "step": 4560
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0665097314031357e-05,
+      "loss": 1.731,
+      "step": 4565
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0497416447398187e-05,
+      "loss": 1.4297,
+      "step": 4570
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0331016411341703e-05,
+      "loss": 1.8428,
+      "step": 4575
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0165898733649113e-05,
+      "loss": 1.9671,
+      "step": 4580
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.0002064930333637e-05,
+      "loss": 1.6723,
+      "step": 4585
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.839516505620815e-06,
+      "loss": 0.9007,
+      "step": 4590
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.678254951934584e-06,
+      "loss": 1.5605,
+      "step": 4595
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.518281749883528e-06,
+      "loss": 1.8201,
+      "step": 4600
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.359598368247429e-06,
+      "loss": 1.4871,
+      "step": 4605
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.202206263963625e-06,
+      "loss": 1.2976,
+      "step": 4610
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.046106882113751e-06,
+      "loss": 1.5821,
+      "step": 4615
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.89130165591041e-06,
+      "loss": 0.9298,
+      "step": 4620
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.737792006684052e-06,
+      "loss": 1.3903,
+      "step": 4625
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.585579343869881e-06,
+      "loss": 1.7596,
+      "step": 4630
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.434665064994905e-06,
+      "loss": 1.2298,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.28505055566519e-06,
+      "loss": 1.5793,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.136737189553077e-06,
+      "loss": 1.0533,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.989726328384516e-06,
+      "loss": 1.7101,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.844019321926686e-06,
+      "loss": 1.5813,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.699617507975563e-06,
+      "loss": 0.8989,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.5565222123435075e-06,
+      "loss": 1.0196,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.414734748847306e-06,
+      "loss": 1.1217,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.274256419295954e-06,
+      "loss": 1.6232,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 7.135088513478693e-06,
+      "loss": 1.6138,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 6.997232309153389e-06,
+      "loss": 1.4079,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 6.8606890720344585e-06,
+      "loss": 1.36,
+      "step": 4690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.725460055781545e-06,
+      "loss": 1.5277,
+      "step": 4695
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.59154650198781e-06,
+      "loss": 1.4765,
+      "step": 4700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.458949640168675e-06,
+      "loss": 2.1942,
+      "step": 4705
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.327670687750436e-06,
+      "loss": 1.602,
+      "step": 4710
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.197710850059129e-06,
+      "loss": 1.4992,
+      "step": 4715
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 6.0690713203094484e-06,
+      "loss": 1.692,
+      "step": 4720
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.941753279593852e-06,
+      "loss": 1.3642,
+      "step": 4725
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.81575789687157e-06,
+      "loss": 1.6257,
+      "step": 4730
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.691086328958028e-06,
+      "loss": 1.1301,
+      "step": 4735
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5677397205141415e-06,
+      "loss": 1.5586,
+      "step": 4740
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.445719204035875e-06,
+      "loss": 1.4131,
+      "step": 4745
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.325025899843732e-06,
+      "loss": 1.2939,
+      "step": 4750
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.205660916072546e-06,
+      "loss": 1.2161,
+      "step": 4755
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.087625348661339e-06,
+      "loss": 1.6379,
+      "step": 4760
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.9709202813431994e-06,
+      "loss": 1.2947,
+      "step": 4765
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.855546785635283e-06,
+      "loss": 1.5706,
+      "step": 4770
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.741505920829131e-06,
+      "loss": 1.5105,
+      "step": 4775
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.628798733980782e-06,
+      "loss": 1.4684,
+      "step": 4780
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.517426259901291e-06,
+      "loss": 1.6168,
+      "step": 4785
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.407389521147114e-06,
+      "loss": 1.5196,
+      "step": 4790
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.298689528010785e-06,
+      "loss": 1.3683,
+      "step": 4795
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.191327278511653e-06,
+      "loss": 0.9307,
+      "step": 4800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.08530375838666e-06,
+      "loss": 1.8983,
+      "step": 4805
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.980619941081348e-06,
+      "loss": 1.7198,
+      "step": 4810
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.877276787740879e-06,
+      "loss": 0.8416,
+      "step": 4815
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.7752752472012315e-06,
+      "loss": 0.979,
+      "step": 4820
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.674616255980467e-06,
+      "loss": 1.647,
+      "step": 4825
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.5753007382702258e-06,
+      "loss": 1.5859,
+      "step": 4830
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.477329605927048e-06,
+      "loss": 1.505,
+      "step": 4835
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.3807037584642316e-06,
+      "loss": 1.6099,
+      "step": 4840
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.28542408304337e-06,
+      "loss": 1.8445,
+      "step": 4845
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.1914914544663437e-06,
+      "loss": 0.9932,
+      "step": 4850
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.098906735167278e-06,
+      "loss": 1.5865,
+      "step": 4855
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.0076707752044958e-06,
+      "loss": 1.4505,
+      "step": 4860
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9177844122528936e-06,
+      "loss": 1.7117,
+      "step": 4865
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.8292484715961127e-06,
+      "loss": 1.1955,
+      "step": 4870
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.7420637661190124e-06,
+      "loss": 1.1857,
+      "step": 4875
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.656231096300243e-06,
+      "loss": 1.4263,
+      "step": 4880
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.5717512502048342e-06,
+      "loss": 1.2868,
+      "step": 4885
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.488625003476952e-06,
+      "loss": 1.0146,
+      "step": 4890
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.4068531193329022e-06,
+      "loss": 1.3149,
+      "step": 4895
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3264363485539396e-06,
+      "loss": 1.6428,
+      "step": 4900
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.2473754294795032e-06,
+      "loss": 1.2771,
+      "step": 4905
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.169671088000391e-06,
+      "loss": 1.087,
+      "step": 4910
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.0933240375521297e-06,
+      "loss": 1.8595,
+      "step": 4915
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.0183349791083827e-06,
+      "loss": 1.2098,
+      "step": 4920
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.944704601174485e-06,
+      "loss": 1.5938,
+      "step": 4925
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.8724335797812685e-06,
+      "loss": 1.446,
+      "step": 4930
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.801522578478648e-06,
+      "loss": 1.4064,
+      "step": 4935
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.7319722483297106e-06,
+      "loss": 1.1977,
+      "step": 4940
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.663783227904636e-06,
+      "loss": 1.4832,
+      "step": 4945
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.596956143274819e-06,
+      "loss": 1.4754,
+      "step": 4950
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5314916080072404e-06,
+      "loss": 1.8633,
+      "step": 4955
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4673902231586875e-06,
+      "loss": 1.3627,
+      "step": 4960
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4046525772703432e-06,
+      "loss": 1.028,
+      "step": 4965
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3432792463623387e-06,
+      "loss": 1.4373,
+      "step": 4970
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.2832707939284427e-06,
+      "loss": 1.643,
+      "step": 4975
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.2246277709309305e-06,
+      "loss": 1.1771,
+      "step": 4980
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.167350715795523e-06,
+      "loss": 1.3419,
+      "step": 4985
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.11144015440639e-06,
+      "loss": 1.5908,
+      "step": 4990
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0568966001014033e-06,
+      "loss": 1.6198,
+      "step": 4995
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0037205536673588e-06,
+      "loss": 1.6799,
+      "step": 5000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.51912503335378e-07,
+      "loss": 1.7136,
+      "step": 5005
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.01472924776514e-07,
+      "loss": 1.3827,
+      "step": 5010
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.524022810972364e-07,
+      "loss": 1.3498,
+      "step": 5015
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.04701022835319e-07,
+      "loss": 1.1361,
+      "step": 5020
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.583695879556262e-07,
+      "loss": 1.4652,
+      "step": 5025
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.134084018461161e-07,
+      "loss": 1.9276,
+      "step": 5030
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.698178773139107e-07,
+      "loss": 1.3343,
+      "step": 5035
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.275984145815483e-07,
+      "loss": 1.414,
+      "step": 5040
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.867504012832703e-07,
+      "loss": 1.7457,
+      "step": 5045
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.472742124614904e-07,
+      "loss": 1.3583,
+      "step": 5050
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.091702105633145e-07,
+      "loss": 1.5982,
+      "step": 5055
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.724387454372591e-07,
+      "loss": 1.4183,
+      "step": 5060
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.370801543300051e-07,
+      "loss": 1.6305,
+      "step": 5065
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.030947618833158e-07,
+      "loss": 1.4188,
+      "step": 5070
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.7048288013105663e-07,
+      "loss": 1.4879,
+      "step": 5075
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.3924480849633064e-07,
+      "loss": 0.6158,
+      "step": 5080
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.093808337886805e-07,
+      "loss": 1.4235,
+      "step": 5085
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.8089123020157424e-07,
+      "loss": 0.8873,
+      "step": 5090
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.5377625930977363e-07,
+      "loss": 1.2172,
+      "step": 5095
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2803617006695308e-07,
+      "loss": 1.4942,
+      "step": 5100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.0367119880346782e-07,
+      "loss": 1.514,
+      "step": 5105
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8068156922413924e-07,
+      "loss": 1.7758,
+      "step": 5110
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.590674924062063e-07,
+      "loss": 1.7397,
+      "step": 5115
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.3882916679739398e-07,
+      "loss": 1.3743,
+      "step": 5120
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1996677821409784e-07,
+      "loss": 1.4851,
+      "step": 5125
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0248049983965223e-07,
+      "loss": 1.5487,
+      "step": 5130
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.637049222276482e-08,
+      "loss": 0.9175,
+      "step": 5135
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.16369032760511e-08,
+      "loss": 1.3961,
+      "step": 5140
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.827986827463549e-08,
+      "loss": 1.3167,
+      "step": 5145
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.6299509854968994e-08,
+      "loss": 1.7824,
+      "step": 5150
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.5695938013630134e-08,
+      "loss": 1.6007,
+      "step": 5155
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6469250106392338e-08,
+      "loss": 1.2018,
+      "step": 5160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8619530847241436e-08,
+      "loss": 1.6535,
+      "step": 5165
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.2146852307692856e-08,
+      "loss": 1.4676,
+      "step": 5170
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.0512739160755176e-09,
+      "loss": 1.3325,
+      "step": 5175
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3328424569822876e-09,
+      "loss": 1.1266,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0,
+      "step": 5184,
+      "total_flos": 1.830374044626125e+16,
+      "train_loss": 1.552536604029161,
+      "train_runtime": 23883.8454,
+      "train_samples_per_second": 0.217,
+      "train_steps_per_second": 0.217
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 5184,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "total_flos": 1.830374044626125e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}