diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,6056 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9719117504130625,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.000224999999151214,
+      "loss": 0.6016,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00022499999570302102,
+      "loss": 0.584,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00022499998960237195,
+      "loss": 0.5738,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00022499998084926694,
+      "loss": 0.5683,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00022499996944370622,
+      "loss": 0.5583,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499995538569008,
+      "loss": 0.5523,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499993867521888,
+      "loss": 0.5533,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499991931229302,
+      "loss": 0.5478,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499989729691294,
+      "loss": 0.542,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499987262907924,
+      "loss": 0.537,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002249998453087925,
+      "loss": 0.545,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499981533605342,
+      "loss": 0.5385,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002249997827108627,
+      "loss": 0.5295,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00022499974743322114,
+      "loss": 0.5302,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002249997095031297,
+      "loss": 0.5341,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002249996689205892,
+      "loss": 0.5324,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022499963454479418,
+      "loss": 0.585,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022499958918784804,
+      "loss": 0.53,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002249995411784559,
+      "loss": 0.5247,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022499949051661884,
+      "loss": 0.5205,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002249994372023382,
+      "loss": 0.5262,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022499938123561527,
+      "loss": 0.5229,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002249993226164514,
+      "loss": 0.5213,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002249992613448481,
+      "loss": 0.5204,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022499919742080687,
+      "loss": 0.5208,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499913084432925,
+      "loss": 0.5201,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499906161541688,
+      "loss": 0.5203,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002249989897340715,
+      "loss": 0.5217,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499891520029482,
+      "loss": 0.5224,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002249988380140888,
+      "loss": 0.5235,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499875817545524,
+      "loss": 0.5158,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499867568439615,
+      "loss": 0.5186,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499859054091355,
+      "loss": 0.5166,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499850274500954,
+      "loss": 0.5165,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00022499841229668634,
+      "loss": 0.5241,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002249983191959461,
+      "loss": 0.5134,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002249982428056152,
+      "loss": 0.5743,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002249981449305302,
+      "loss": 0.5115,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00022499804440303481,
+      "loss": 0.5178,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00022499794122313147,
+      "loss": 0.5104,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00022499783539082276,
+      "loss": 0.5112,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002249977269061113,
+      "loss": 0.5076,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00022499761576899973,
+      "loss": 0.5136,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002249975019794908,
+      "loss": 0.5144,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00022499738553758734,
+      "loss": 0.5126,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002249972664432922,
+      "loss": 0.5077,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499714469660837,
+      "loss": 0.509,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002249970202975388,
+      "loss": 0.5099,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499689324608654,
+      "loss": 0.5107,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002249967635422548,
+      "loss": 0.504,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499663118604672,
+      "loss": 0.5088,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499649617746555,
+      "loss": 0.507,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499635851651472,
+      "loss": 0.5103,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499621820319752,
+      "loss": 0.5101,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499607523751748,
+      "loss": 0.5053,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00022499592961947807,
+      "loss": 0.5121,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499578134908292,
+      "loss": 0.5132,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499563042633575,
+      "loss": 0.508,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499547685124013,
+      "loss": 0.5034,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499532062379997,
+      "loss": 0.5014,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002249951617440191,
+      "loss": 0.5055,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499500021190143,
+      "loss": 0.5057,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499483602745094,
+      "loss": 0.5052,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00022499466919067172,
+      "loss": 0.5086,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002249944997015678,
+      "loss": 0.5058,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002249943275601435,
+      "loss": 0.5092,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002249941527664029,
+      "loss": 0.5019,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499397532035043,
+      "loss": 0.4976,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002249937952219904,
+      "loss": 0.5052,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499361247132736,
+      "loss": 0.5051,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002249934270683657,
+      "loss": 0.5002,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499323901311005,
+      "loss": 0.504,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499304830556505,
+      "loss": 0.4937,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499285494573539,
+      "loss": 0.5066,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499265893362587,
+      "loss": 0.5054,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00022499246026924126,
+      "loss": 0.5081,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002249922589525866,
+      "loss": 0.5025,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499205498366666,
+      "loss": 0.499,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499184836248664,
+      "loss": 0.5034,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499163908905155,
+      "loss": 0.4987,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002249914271633666,
+      "loss": 0.4984,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499121258543698,
+      "loss": 0.4968,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499099535526801,
+      "loss": 0.4976,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499077547286507,
+      "loss": 0.5024,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499055293823353,
+      "loss": 0.4972,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002249903277513789,
+      "loss": 0.5025,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00022499009991230675,
+      "loss": 0.4984,
+      "step": 435
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498986942102273,
+      "loss": 0.5012,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498963627753244,
+      "loss": 0.4983,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498940048184174,
+      "loss": 0.5045,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498916203395639,
+      "loss": 0.4942,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498892093388224,
+      "loss": 0.4962,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002249886771816253,
+      "loss": 0.4984,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498843077719158,
+      "loss": 0.5033,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002249881817205871,
+      "loss": 0.5009,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498793001181808,
+      "loss": 0.501,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00022498767565089068,
+      "loss": 0.4931,
+      "step": 485
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498741863781122,
+      "loss": 0.4937,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000224987158972586,
+      "loss": 0.5013,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498689665522144,
+      "loss": 0.4987,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000224986631685724,
+      "loss": 0.5003,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498636406410025,
+      "loss": 0.4955,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002249860937903568,
+      "loss": 0.4928,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498582086450026,
+      "loss": 0.5013,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002249855452865374,
+      "loss": 0.4974,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498526705647505,
+      "loss": 0.5018,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498498617431998,
+      "loss": 0.4971,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00022498470264007922,
+      "loss": 0.5024,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498441645375976,
+      "loss": 0.4977,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498412761536862,
+      "loss": 0.4974,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498383612491287,
+      "loss": 0.4964,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498354198239985,
+      "loss": 0.495,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002249832451878367,
+      "loss": 0.495,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498294574123078,
+      "loss": 0.5012,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002249826436425895,
+      "loss": 0.4973,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002249823388919203,
+      "loss": 0.4913,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002249820314892307,
+      "loss": 0.4886,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00022498172143452824,
+      "loss": 0.4887,
+      "step": 590
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022498140872782062,
+      "loss": 0.4916,
+      "step": 595
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022498109336911555,
+      "loss": 0.4936,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002249807753584208,
+      "loss": 0.4893,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022498045469574422,
+      "loss": 0.49,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022498013138109373,
+      "loss": 0.4932,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002249798054144773,
+      "loss": 0.4942,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000224979476795903,
+      "loss": 0.4933,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022497914552537885,
+      "loss": 0.4922,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022497881160291313,
+      "loss": 0.4902,
+      "step": 635
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00022497847502851406,
+      "loss": 0.4892,
+      "step": 640
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497813580218986,
+      "loss": 0.4867,
+      "step": 645
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000224977793923949,
+      "loss": 0.4894,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497744939379985,
+      "loss": 0.4977,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497710221175098,
+      "loss": 0.4925,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497675237781087,
+      "loss": 0.4931,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002249763998919882,
+      "loss": 0.4955,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497604475429169,
+      "loss": 0.4956,
+      "step": 675
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00022497568696473004,
+      "loss": 0.4974,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002249753265233121,
+      "loss": 0.4931,
+      "step": 685
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002249749634300468,
+      "loss": 0.495,
+      "step": 690
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002249745976849431,
+      "loss": 0.4972,
+      "step": 695
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497422928800998,
+      "loss": 0.4888,
+      "step": 700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497385823925655,
+      "loss": 0.4908,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497348453869197,
+      "loss": 0.4923,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497310818632543,
+      "loss": 0.4957,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002249727291821663,
+      "loss": 0.493,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002249723475262238,
+      "loss": 0.4949,
+      "step": 725
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497196321850745,
+      "loss": 0.4899,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497157625902671,
+      "loss": 0.49,
+      "step": 735
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00022497118664779113,
+      "loss": 0.4941,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002249707943848103,
+      "loss": 0.4902,
+      "step": 745
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002249703994700939,
+      "loss": 0.489,
+      "step": 750
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002249700019036517,
+      "loss": 0.4925,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002249696016854935,
+      "loss": 0.4924,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00022496919881562916,
+      "loss": 0.4868,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00022496879329406865,
+      "loss": 0.4935,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00022496838512082193,
+      "loss": 0.4956,
+      "step": 775
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00022496797429589913,
+      "loss": 0.4892,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002249675608193103,
+      "loss": 0.4921,
+      "step": 785
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00022496714469106576,
+      "loss": 0.4915,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002249667259111757,
+      "loss": 0.4881,
+      "step": 795
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496630447965045,
+      "loss": 0.4856,
+      "step": 800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496588039650042,
+      "loss": 0.4896,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496545366173612,
+      "loss": 0.4858,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000224965024275368,
+      "loss": 0.4856,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002249645922374067,
+      "loss": 0.4887,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496415754786288,
+      "loss": 0.4897,
+      "step": 825
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496372020674727,
+      "loss": 0.4885,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496328021407065,
+      "loss": 0.4882,
+      "step": 835
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496283756984385,
+      "loss": 0.485,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00022496239227407785,
+      "loss": 0.4848,
+      "step": 845
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002249619443267836,
+      "loss": 0.4904,
+      "step": 850
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022496149372797214,
+      "loss": 0.4882,
+      "step": 855
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022496104047765462,
+      "loss": 0.4926,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002249605845758422,
+      "loss": 0.4949,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022496012602254617,
+      "loss": 0.4882,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022495966481777779,
+      "loss": 0.4892,
+      "step": 875
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022495920096154848,
+      "loss": 0.491,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00022495873445386964,
+      "loss": 0.4881,
+      "step": 885
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002249582652947528,
+      "loss": 0.4927,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002249577934842096,
+      "loss": 0.4921,
+      "step": 895
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002249573190222516,
+      "loss": 0.4855,
+      "step": 900
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495684190889054,
+      "loss": 0.4812,
+      "step": 905
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495636214413817,
+      "loss": 0.4864,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495587972800633,
+      "loss": 0.487,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495539466050696,
+      "loss": 0.491,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495490694165202,
+      "loss": 0.4901,
+      "step": 925
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495441657145352,
+      "loss": 0.4866,
+      "step": 930
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495392354992356,
+      "loss": 0.4817,
+      "step": 935
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002249534278770743,
+      "loss": 0.4895,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495292955291798,
+      "loss": 0.4848,
+      "step": 945
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00022495242857746694,
+      "loss": 0.4885,
+      "step": 950
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022495192495073345,
+      "loss": 0.49,
+      "step": 955
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022495141867273004,
+      "loss": 0.4827,
+      "step": 960
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022495090974346914,
+      "loss": 0.4891,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022495039816296329,
+      "loss": 0.487,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022494988393122512,
+      "loss": 0.4817,
+      "step": 975
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022494936704826735,
+      "loss": 0.4882,
+      "step": 980
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022494884751410278,
+      "loss": 0.4902,
+      "step": 985
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002249483253287441,
+      "loss": 0.4895,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002249478004922043,
+      "loss": 0.4875,
+      "step": 995
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00022494727300449626,
+      "loss": 0.4884,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.4691707491874695,
+      "eval_runtime": 43.4799,
+      "eval_samples_per_second": 114.996,
+      "eval_steps_per_second": 1.219,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494674286563306,
+      "loss": 0.4873,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494621007562775,
+      "loss": 0.4817,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494567463449347,
+      "loss": 0.4837,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494513654224344,
+      "loss": 0.4838,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494459579889092,
+      "loss": 0.4811,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494405240444928,
+      "loss": 0.4861,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002249435063589319,
+      "loss": 0.4899,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494295766235228,
+      "loss": 0.4885,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494240631472395,
+      "loss": 0.4931,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00022494185231606052,
+      "loss": 0.4927,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002249412956663756,
+      "loss": 0.4902,
+      "step": 1055
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022494073636568303,
+      "loss": 0.4835,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002249401744139965,
+      "loss": 0.4852,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493960981132997,
+      "loss": 0.4803,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493904255769734,
+      "loss": 0.4822,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493847265311258,
+      "loss": 0.4811,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493790009758978,
+      "loss": 0.4871,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493732489114305,
+      "loss": 0.4793,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493674703378657,
+      "loss": 0.4906,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493616652553462,
+      "loss": 0.484,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00022493558336640152,
+      "loss": 0.484,
+      "step": 1105
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493499755640167,
+      "loss": 0.4884,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002249344090955495,
+      "loss": 0.488,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493381798385957,
+      "loss": 0.4821,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493322422134643,
+      "loss": 0.4865,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493262780802473,
+      "loss": 0.4835,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002249320287439092,
+      "loss": 0.4817,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493142702901463,
+      "loss": 0.4863,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493082266335582,
+      "loss": 0.4834,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022493021564694776,
+      "loss": 0.4832,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00022492960597980533,
+      "loss": 0.4839,
+      "step": 1155
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492899366194366,
+      "loss": 0.482,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492837869337782,
+      "loss": 0.4868,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492776107412298,
+      "loss": 0.4908,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492714080419438,
+      "loss": 0.4797,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492651788360736,
+      "loss": 0.4793,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492589231237723,
+      "loss": 0.4825,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492526409051945,
+      "loss": 0.4866,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492463321804957,
+      "loss": 0.4808,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492399969498307,
+      "loss": 0.4825,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00022492336352133565,
+      "loss": 0.48,
+      "step": 1205
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000224922724697123,
+      "loss": 0.4804,
+      "step": 1210
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022492208322236085,
+      "loss": 0.483,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000224921439097065,
+      "loss": 0.4833,
+      "step": 1220
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022492079232125142,
+      "loss": 0.4829,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022492014289493604,
+      "loss": 0.4849,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022491949081813485,
+      "loss": 0.4843,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000224918836090864,
+      "loss": 0.4795,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022491817871313961,
+      "loss": 0.4814,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022491751868497788,
+      "loss": 0.4825,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00022491685600639513,
+      "loss": 0.4819,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002249161906774077,
+      "loss": 0.4822,
+      "step": 1260
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000224915522698032,
+      "loss": 0.4798,
+      "step": 1265
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002249148520682845,
+      "loss": 0.4846,
+      "step": 1270
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002249141787881818,
+      "loss": 0.4793,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022491350285774045,
+      "loss": 0.4789,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022491282427697718,
+      "loss": 0.4823,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022491214304590866,
+      "loss": 0.4847,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002249114591645518,
+      "loss": 0.4852,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002249107726329234,
+      "loss": 0.4807,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022491008345104042,
+      "loss": 0.477,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00022490939161891983,
+      "loss": 0.4827,
+      "step": 1310
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490869713657878,
+      "loss": 0.4836,
+      "step": 1315
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490800000403435,
+      "loss": 0.4895,
+      "step": 1320
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490730022130373,
+      "loss": 0.4808,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490659778840423,
+      "loss": 0.4782,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490589270535315,
+      "loss": 0.488,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490518497216792,
+      "loss": 0.4777,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490447458886593,
+      "loss": 0.4817,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002249037615554648,
+      "loss": 0.4813,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002249030458719821,
+      "loss": 0.4794,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00022490232753843542,
+      "loss": 0.4872,
+      "step": 1360
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022490160655484255,
+      "loss": 0.4721,
+      "step": 1365
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022490088292122126,
+      "loss": 0.4778,
+      "step": 1370
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022490015663758942,
+      "loss": 0.474,
+      "step": 1375
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022489942770396494,
+      "loss": 0.4731,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002248986961203658,
+      "loss": 0.4766,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022489796188681006,
+      "loss": 0.4825,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022489722500331584,
+      "loss": 0.479,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022489648546990127,
+      "loss": 0.479,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002248957432865847,
+      "loss": 0.486,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00022489499845338435,
+      "loss": 0.4809,
+      "step": 1410
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022489425097031862,
+      "loss": 0.4761,
+      "step": 1415
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000224893500837406,
+      "loss": 0.4784,
+      "step": 1420
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022489274805466495,
+      "loss": 0.4867,
+      "step": 1425
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022489199262211407,
+      "loss": 0.4772,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022489123453977195,
+      "loss": 0.4805,
+      "step": 1435
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022489047380765736,
+      "loss": 0.4788,
+      "step": 1440
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022488971042578903,
+      "loss": 0.4791,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022488894439418584,
+      "loss": 0.481,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022488817571286664,
+      "loss": 0.4842,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00022488740438185044,
+      "loss": 0.4816,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002248866304011562,
+      "loss": 0.4803,
+      "step": 1465
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488585377080308,
+      "loss": 0.4757,
+      "step": 1470
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488507449081025,
+      "loss": 0.479,
+      "step": 1475
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002248842925611969,
+      "loss": 0.4782,
+      "step": 1480
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488350798198232,
+      "loss": 0.4799,
+      "step": 1485
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488272075318593,
+      "loss": 0.4789,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488193087482708,
+      "loss": 0.4823,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488113834692532,
+      "loss": 0.4788,
+      "step": 1500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022488034316950014,
+      "loss": 0.4815,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022487954534257122,
+      "loss": 0.4781,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00022487874486615823,
+      "loss": 0.4779,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002248779417402809,
+      "loss": 0.479,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487713596495908,
+      "loss": 0.4799,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487632754021258,
+      "loss": 0.48,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487551646606144,
+      "loss": 0.4808,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487470274252565,
+      "loss": 0.4825,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002248738863696252,
+      "loss": 0.4795,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487306734738033,
+      "loss": 0.4773,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487224567581125,
+      "loss": 0.4715,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00022487142135493819,
+      "loss": 0.4734,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002248705943847815,
+      "loss": 0.4774,
+      "step": 1565
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002248697647653616,
+      "loss": 0.4782,
+      "step": 1570
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486893249669892,
+      "loss": 0.4788,
+      "step": 1575
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486809757881406,
+      "loss": 0.4738,
+      "step": 1580
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486726001172755,
+      "loss": 0.4805,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002248664197954601,
+      "loss": 0.4811,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486557693003246,
+      "loss": 0.4786,
+      "step": 1595
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486473141546538,
+      "loss": 0.4785,
+      "step": 1600
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486388325177972,
+      "loss": 0.4785,
+      "step": 1605
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486303243899645,
+      "loss": 0.4778,
+      "step": 1610
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486217897713654,
+      "loss": 0.4785,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00022486132286622106,
+      "loss": 0.4751,
+      "step": 1620
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002248604641062711,
+      "loss": 0.4792,
+      "step": 1625
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485960269730788,
+      "loss": 0.4751,
+      "step": 1630
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485873863935268,
+      "loss": 0.4853,
+      "step": 1635
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485787193242675,
+      "loss": 0.4792,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485700257655154,
+      "loss": 0.4768,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485613057174843,
+      "loss": 0.4733,
+      "step": 1650
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000224855255918039,
+      "loss": 0.4727,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485437861544482,
+      "loss": 0.4817,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00022485349866398753,
+      "loss": 0.4818,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002248526160636888,
+      "loss": 0.4746,
+      "step": 1670
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002248517308145705,
+      "loss": 0.4756,
+      "step": 1675
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022485084291665438,
+      "loss": 0.4751,
+      "step": 1680
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002248499523699624,
+      "loss": 0.4763,
+      "step": 1685
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484905917451653,
+      "loss": 0.4771,
+      "step": 1690
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484816333033877,
+      "loss": 0.475,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484726483745128,
+      "loss": 0.4812,
+      "step": 1700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484636369587619,
+      "loss": 0.4715,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484545990563573,
+      "loss": 0.481,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484455346675222,
+      "loss": 0.4792,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00022484364437924803,
+      "loss": 0.4849,
+      "step": 1720
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022484273264314561,
+      "loss": 0.4804,
+      "step": 1725
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002248418182584674,
+      "loss": 0.4775,
+      "step": 1730
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022484090122523598,
+      "loss": 0.4778,
+      "step": 1735
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000224839981543474,
+      "loss": 0.4803,
+      "step": 1740
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022483905921320417,
+      "loss": 0.4801,
+      "step": 1745
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022483813423444918,
+      "loss": 0.4799,
+      "step": 1750
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022483720660723192,
+      "loss": 0.4773,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002248362763315752,
+      "loss": 0.4812,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022483534340750206,
+      "loss": 0.4768,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022483440783503547,
+      "loss": 0.4745,
+      "step": 1770
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022483346961419854,
+      "loss": 0.4705,
+      "step": 1775
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022483252874501438,
+      "loss": 0.4762,
+      "step": 1780
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022483158522750626,
+      "loss": 0.4771,
+      "step": 1785
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022483063906169741,
+      "loss": 0.47,
+      "step": 1790
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002248296902476112,
+      "loss": 0.4733,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022482873878527105,
+      "loss": 0.4716,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002248277846747004,
+      "loss": 0.4763,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022482682791592282,
+      "loss": 0.4751,
+      "step": 1810
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022482586850896194,
+      "loss": 0.4752,
+      "step": 1815
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002248249064538414,
+      "loss": 0.4731,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002248239417505849,
+      "loss": 0.4811,
+      "step": 1825
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022482297439921633,
+      "loss": 0.4737,
+      "step": 1830
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002248220043997595,
+      "loss": 0.4774,
+      "step": 1835
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022482103175223837,
+      "loss": 0.4758,
+      "step": 1840
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022482005645667688,
+      "loss": 0.4709,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481907851309917,
+      "loss": 0.4714,
+      "step": 1850
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481809792152937,
+      "loss": 0.4789,
+      "step": 1855
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481711468199162,
+      "loss": 0.4766,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481612879451018,
+      "loss": 0.4748,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481514025910946,
+      "loss": 0.4796,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022481414907581375,
+      "loss": 0.4758,
+      "step": 1875
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002248131552446476,
+      "loss": 0.4765,
+      "step": 1880
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022481215876563545,
+      "loss": 0.4752,
+      "step": 1885
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022481115963880193,
+      "loss": 0.4763,
+      "step": 1890
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022481015786417164,
+      "loss": 0.4744,
+      "step": 1895
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480915344176938,
+      "loss": 0.4736,
+      "step": 1900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480814637161988,
+      "loss": 0.4751,
+      "step": 1905
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480713665374801,
+      "loss": 0.4768,
+      "step": 1910
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480612428817868,
+      "loss": 0.4707,
+      "step": 1915
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480510927493685,
+      "loss": 0.4737,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022480409161404756,
+      "loss": 0.475,
+      "step": 1925
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022480307130553595,
+      "loss": 0.4709,
+      "step": 1930
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022480204834942724,
+      "loss": 0.4781,
+      "step": 1935
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022480102274574653,
+      "loss": 0.4756,
+      "step": 1940
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479999449451926,
+      "loss": 0.47,
+      "step": 1945
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479896359577073,
+      "loss": 0.4735,
+      "step": 1950
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479793004952643,
+      "loss": 0.4714,
+      "step": 1955
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002247968938558118,
+      "loss": 0.4783,
+      "step": 1960
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479585501465244,
+      "loss": 0.4793,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479481352607396,
+      "loss": 0.4805,
+      "step": 1970
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002247937693901021,
+      "loss": 0.4735,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00022479272260676262,
+      "loss": 0.4737,
+      "step": 1980
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00022479167317608127,
+      "loss": 0.4779,
+      "step": 1985
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00022479062109808404,
+      "loss": 0.4768,
+      "step": 1990
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00022478956637279684,
+      "loss": 0.475,
+      "step": 1995
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002247885090002457,
+      "loss": 0.4711,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.4555172920227051,
+      "eval_runtime": 41.4031,
+      "eval_samples_per_second": 120.764,
+      "eval_steps_per_second": 1.28,
+      "step": 2000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002247874489804567,
+      "loss": 0.4748,
+      "step": 2005
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00022478638631345606,
+      "loss": 0.4744,
+      "step": 2010
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002247853209992699,
+      "loss": 0.4697,
+      "step": 2015
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002247842530379246,
+      "loss": 0.4751,
+      "step": 2020
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002247831824294464,
+      "loss": 0.4774,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00022478210917386185,
+      "loss": 0.4725,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022478103327119734,
+      "loss": 0.4736,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002247799547214794,
+      "loss": 0.4728,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022477887352473475,
+      "loss": 0.4751,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022477778968098996,
+      "loss": 0.4757,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022477670319027182,
+      "loss": 0.4731,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002247756140526071,
+      "loss": 0.4775,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022477452226802276,
+      "loss": 0.4743,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00022477342783654565,
+      "loss": 0.476,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002247723307582028,
+      "loss": 0.4683,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002247712310330213,
+      "loss": 0.4725,
+      "step": 2080
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022477012866102826,
+      "loss": 0.4714,
+      "step": 2085
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002247690236422509,
+      "loss": 0.4738,
+      "step": 2090
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476791597671647,
+      "loss": 0.474,
+      "step": 2095
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002247668056644523,
+      "loss": 0.4777,
+      "step": 2100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476569270548583,
+      "loss": 0.4772,
+      "step": 2105
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476457709984447,
+      "loss": 0.4697,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476345884755574,
+      "loss": 0.4761,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476233794864724,
+      "loss": 0.4769,
+      "step": 2120
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022476121440314666,
+      "loss": 0.4815,
+      "step": 2125
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002247600882110817,
+      "loss": 0.4795,
+      "step": 2130
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475895937248015,
+      "loss": 0.4735,
+      "step": 2135
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475782788736984,
+      "loss": 0.4735,
+      "step": 2140
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475669375577874,
+      "loss": 0.4731,
+      "step": 2145
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002247555569777348,
+      "loss": 0.4784,
+      "step": 2150
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475441755326608,
+      "loss": 0.4773,
+      "step": 2155
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475327548240065,
+      "loss": 0.4698,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475213076516675,
+      "loss": 0.471,
+      "step": 2165
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022475098340159262,
+      "loss": 0.4673,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002247498333917065,
+      "loss": 0.4745,
+      "step": 2175
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022474868073553685,
+      "loss": 0.4709,
+      "step": 2180
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022474752543311205,
+      "loss": 0.4696,
+      "step": 2185
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474636748446067,
+      "loss": 0.4775,
+      "step": 2190
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474520688961123,
+      "loss": 0.4729,
+      "step": 2195
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474404364859242,
+      "loss": 0.4705,
+      "step": 2200
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474287776143284,
+      "loss": 0.475,
+      "step": 2205
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474170922816134,
+      "loss": 0.4731,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022474053804880673,
+      "loss": 0.4714,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022473936422339792,
+      "loss": 0.4767,
+      "step": 2220
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022473818775196384,
+      "loss": 0.4747,
+      "step": 2225
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022473700863453354,
+      "loss": 0.4765,
+      "step": 2230
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00022473582687113617,
+      "loss": 0.4734,
+      "step": 2235
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022473464246180078,
+      "loss": 0.4744,
+      "step": 2240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002247334554065567,
+      "loss": 0.4719,
+      "step": 2245
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002247322657054331,
+      "loss": 0.4765,
+      "step": 2250
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022473107335845943,
+      "loss": 0.4753,
+      "step": 2255
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022472987836566507,
+      "loss": 0.4685,
+      "step": 2260
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022472868072707957,
+      "loss": 0.4743,
+      "step": 2265
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022472748044273234,
+      "loss": 0.4691,
+      "step": 2270
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022472627751265315,
+      "loss": 0.4676,
+      "step": 2275
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002247250719368716,
+      "loss": 0.4764,
+      "step": 2280
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00022472386371541746,
+      "loss": 0.4705,
+      "step": 2285
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002247226528483205,
+      "loss": 0.4694,
+      "step": 2290
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022472143933561066,
+      "loss": 0.4736,
+      "step": 2295
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022472022317731783,
+      "loss": 0.4739,
+      "step": 2300
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471900437347203,
+      "loss": 0.4723,
+      "step": 2305
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471778292410338,
+      "loss": 0.4747,
+      "step": 2310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471655882924196,
+      "loss": 0.4735,
+      "step": 2315
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471533208891798,
+      "loss": 0.4704,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471410270316172,
+      "loss": 0.4708,
+      "step": 2325
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471287067200356,
+      "loss": 0.4731,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002247116359954738,
+      "loss": 0.4675,
+      "step": 2335
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00022471039867360298,
+      "loss": 0.4735,
+      "step": 2340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470915870642164,
+      "loss": 0.4712,
+      "step": 2345
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002247079160939603,
+      "loss": 0.4771,
+      "step": 2350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470667083624967,
+      "loss": 0.4751,
+      "step": 2355
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470542293332048,
+      "loss": 0.4748,
+      "step": 2360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470417238520352,
+      "loss": 0.4728,
+      "step": 2365
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470291919192965,
+      "loss": 0.4667,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470166335352974,
+      "loss": 0.4655,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022470040487003484,
+      "loss": 0.4671,
+      "step": 2380
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022469914374147594,
+      "loss": 0.4707,
+      "step": 2385
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00022469787996788422,
+      "loss": 0.4718,
+      "step": 2390
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022469661354929086,
+      "loss": 0.4696,
+      "step": 2395
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022469534448572704,
+      "loss": 0.4716,
+      "step": 2400
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022469407277722416,
+      "loss": 0.473,
+      "step": 2405
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002246927984238135,
+      "loss": 0.4707,
+      "step": 2410
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002246915214255266,
+      "loss": 0.4725,
+      "step": 2415
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002246902417823949,
+      "loss": 0.4712,
+      "step": 2420
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022468895949445003,
+      "loss": 0.4735,
+      "step": 2425
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022468767456172356,
+      "loss": 0.4716,
+      "step": 2430
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022468638698424728,
+      "loss": 0.4739,
+      "step": 2435
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002246850967620529,
+      "loss": 0.4704,
+      "step": 2440
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022468380389517224,
+      "loss": 0.4718,
+      "step": 2445
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022468250838363725,
+      "loss": 0.4652,
+      "step": 2450
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022468121022747986,
+      "loss": 0.4696,
+      "step": 2455
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467990942673214,
+      "loss": 0.4691,
+      "step": 2460
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467860598142614,
+      "loss": 0.4662,
+      "step": 2465
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467729989159406,
+      "loss": 0.4664,
+      "step": 2470
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002246759911572681,
+      "loss": 0.4731,
+      "step": 2475
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467467977848056,
+      "loss": 0.4736,
+      "step": 2480
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002246733657552638,
+      "loss": 0.4705,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467204908765024,
+      "loss": 0.4687,
+      "step": 2490
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022467072977567235,
+      "loss": 0.4739,
+      "step": 2495
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002246694078193627,
+      "loss": 0.4754,
+      "step": 2500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022466808321875393,
+      "loss": 0.4689,
+      "step": 2505
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022466675597387868,
+      "loss": 0.4703,
+      "step": 2510
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022466542608476974,
+      "loss": 0.4683,
+      "step": 2515
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002246640935514599,
+      "loss": 0.4675,
+      "step": 2520
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022466275837398203,
+      "loss": 0.473,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002246614205523691,
+      "loss": 0.4741,
+      "step": 2530
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022466008008665409,
+      "loss": 0.4748,
+      "step": 2535
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002246587369768701,
+      "loss": 0.4692,
+      "step": 2540
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022465739122305025,
+      "loss": 0.4747,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002246560428252278,
+      "loss": 0.4691,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022465469178343592,
+      "loss": 0.4674,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000224653338097708,
+      "loss": 0.4698,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022465198176807748,
+      "loss": 0.4664,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022465062279457778,
+      "loss": 0.4645,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022464926117724243,
+      "loss": 0.4662,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022464789691610502,
+      "loss": 0.4695,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022464653001119927,
+      "loss": 0.4725,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022464516046255884,
+      "loss": 0.4674,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00022464378827021756,
+      "loss": 0.4691,
+      "step": 2595
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022464241343420923,
+      "loss": 0.468,
+      "step": 2600
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022464103595456786,
+      "loss": 0.4698,
+      "step": 2605
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463965583132737,
+      "loss": 0.4661,
+      "step": 2610
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463827306452188,
+      "loss": 0.4697,
+      "step": 2615
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463688765418543,
+      "loss": 0.472,
+      "step": 2620
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463549960035227,
+      "loss": 0.4663,
+      "step": 2625
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002246341089030566,
+      "loss": 0.4718,
+      "step": 2630
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463271556233275,
+      "loss": 0.4663,
+      "step": 2635
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00022463131957821509,
+      "loss": 0.4652,
+      "step": 2640
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002246299209507381,
+      "loss": 0.4741,
+      "step": 2645
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462851967993626,
+      "loss": 0.4703,
+      "step": 2650
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462711576584415,
+      "loss": 0.4679,
+      "step": 2655
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462570920849643,
+      "loss": 0.4724,
+      "step": 2660
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462430000792777,
+      "loss": 0.4628,
+      "step": 2665
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462288816417296,
+      "loss": 0.4742,
+      "step": 2670
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462147367726683,
+      "loss": 0.4785,
+      "step": 2675
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022462005654724432,
+      "loss": 0.4648,
+      "step": 2680
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002246186367741403,
+      "loss": 0.4721,
+      "step": 2685
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00022461721435798992,
+      "loss": 0.4702,
+      "step": 2690
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002246157892988282,
+      "loss": 0.4727,
+      "step": 2695
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002246143615966903,
+      "loss": 0.472,
+      "step": 2700
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022461293125161148,
+      "loss": 0.4719,
+      "step": 2705
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022461149826362705,
+      "loss": 0.4638,
+      "step": 2710
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022461006263277232,
+      "loss": 0.4639,
+      "step": 2715
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022460862435908272,
+      "loss": 0.4659,
+      "step": 2720
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022460718344259376,
+      "loss": 0.4693,
+      "step": 2725
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000224605739883341,
+      "loss": 0.4665,
+      "step": 2730
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022460429368135999,
+      "loss": 0.4678,
+      "step": 2735
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022460284483668652,
+      "loss": 0.4673,
+      "step": 2740
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022460139334935626,
+      "loss": 0.471,
+      "step": 2745
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00022459993921940504,
+      "loss": 0.4712,
+      "step": 2750
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459848244686876,
+      "loss": 0.4716,
+      "step": 2755
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459702303178334,
+      "loss": 0.4742,
+      "step": 2760
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459556097418482,
+      "loss": 0.4722,
+      "step": 2765
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459409627410923,
+      "loss": 0.4682,
+      "step": 2770
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459262893159275,
+      "loss": 0.4717,
+      "step": 2775
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022459115894667156,
+      "loss": 0.4668,
+      "step": 2780
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022458968631938194,
+      "loss": 0.4705,
+      "step": 2785
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022458821104976025,
+      "loss": 0.4664,
+      "step": 2790
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022458673313784287,
+      "loss": 0.4683,
+      "step": 2795
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00022458525258366624,
+      "loss": 0.4718,
+      "step": 2800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002245837693872669,
+      "loss": 0.465,
+      "step": 2805
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00022458228354868152,
+      "loss": 0.4714,
+      "step": 2810
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00022458079506794664,
+      "loss": 0.4673,
+      "step": 2815
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002245793039450991,
+      "loss": 0.4681,
+      "step": 2820
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002245778101801756,
+      "loss": 0.4708,
+      "step": 2825
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002245763137732131,
+      "loss": 0.4728,
+      "step": 2830
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002245748147242484,
+      "loss": 0.4659,
+      "step": 2835
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00022457331303331858,
+      "loss": 0.4656,
+      "step": 2840
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00022457180870046064,
+      "loss": 0.4666,
+      "step": 2845
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00022457030172571174,
+      "loss": 0.4652,
+      "step": 2850
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.000224568792109109,
+      "loss": 0.4713,
+      "step": 2855
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022456727985068975,
+      "loss": 0.4665,
+      "step": 2860
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022456576495049124,
+      "loss": 0.4721,
+      "step": 2865
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002245642474085509,
+      "loss": 0.4613,
+      "step": 2870
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022456272722490614,
+      "loss": 0.4648,
+      "step": 2875
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002245612043995945,
+      "loss": 0.4661,
+      "step": 2880
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002245596789326535,
+      "loss": 0.4648,
+      "step": 2885
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002245581508241208,
+      "loss": 0.4638,
+      "step": 2890
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022455662007403413,
+      "loss": 0.468,
+      "step": 2895
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022455508668243126,
+      "loss": 0.4721,
+      "step": 2900
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00022455355064934996,
+      "loss": 0.4696,
+      "step": 2905
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002245520119748282,
+      "loss": 0.466,
+      "step": 2910
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022455047065890392,
+      "loss": 0.4691,
+      "step": 2915
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454892670161522,
+      "loss": 0.4703,
+      "step": 2920
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454738010300004,
+      "loss": 0.4672,
+      "step": 2925
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454583086309668,
+      "loss": 0.4636,
+      "step": 2930
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454427898194328,
+      "loss": 0.4681,
+      "step": 2935
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454272445957823,
+      "loss": 0.4659,
+      "step": 2940
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022454116729603976,
+      "loss": 0.4619,
+      "step": 2945
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002245396074913664,
+      "loss": 0.4647,
+      "step": 2950
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00022453804504559658,
+      "loss": 0.4645,
+      "step": 2955
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022453647995876886,
+      "loss": 0.4693,
+      "step": 2960
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022453491223092185,
+      "loss": 0.4667,
+      "step": 2965
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022453334186209426,
+      "loss": 0.4659,
+      "step": 2970
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002245317688523248,
+      "loss": 0.4666,
+      "step": 2975
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022453019320165232,
+      "loss": 0.4713,
+      "step": 2980
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022452861491011568,
+      "loss": 0.467,
+      "step": 2985
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022452703397775382,
+      "loss": 0.4703,
+      "step": 2990
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022452545040460578,
+      "loss": 0.4686,
+      "step": 2995
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022452386419071057,
+      "loss": 0.4702,
+      "step": 3000
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.44886377453804016,
+      "eval_runtime": 47.4173,
+      "eval_samples_per_second": 105.447,
+      "eval_steps_per_second": 1.118,
+      "step": 3000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00022452227533610735,
+      "loss": 0.4693,
+      "step": 3005
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022452068384083538,
+      "loss": 0.4681,
+      "step": 3010
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022451908970493387,
+      "loss": 0.4631,
+      "step": 3015
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022451749292844213,
+      "loss": 0.4612,
+      "step": 3020
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022451589351139964,
+      "loss": 0.4647,
+      "step": 3025
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022451429145384582,
+      "loss": 0.4658,
+      "step": 3030
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002245126867558202,
+      "loss": 0.4644,
+      "step": 3035
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022451107941736233,
+      "loss": 0.4727,
+      "step": 3040
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022450946943851198,
+      "loss": 0.4692,
+      "step": 3045
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022450785681930876,
+      "loss": 0.466,
+      "step": 3050
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00022450624155979254,
+      "loss": 0.4656,
+      "step": 3055
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002245046236600031,
+      "loss": 0.4725,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022450300311998046,
+      "loss": 0.4689,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022450137993976454,
+      "loss": 0.469,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022449975411939537,
+      "loss": 0.4685,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002244981256589131,
+      "loss": 0.4668,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022449649455835792,
+      "loss": 0.4695,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002244948608177701,
+      "loss": 0.4664,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022449322443718981,
+      "loss": 0.4745,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022449158541665762,
+      "loss": 0.4724,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022448994375621386,
+      "loss": 0.4609,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00022448829945589903,
+      "loss": 0.4666,
+      "step": 3110
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022448665251575376,
+      "loss": 0.4651,
+      "step": 3115
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002244850029358186,
+      "loss": 0.4644,
+      "step": 3120
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022448335071613434,
+      "loss": 0.4686,
+      "step": 3125
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022448169585674173,
+      "loss": 0.4678,
+      "step": 3130
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022448003835768155,
+      "loss": 0.4668,
+      "step": 3135
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022447837821899474,
+      "loss": 0.4664,
+      "step": 3140
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022447671544072227,
+      "loss": 0.4647,
+      "step": 3145
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022447505002290516,
+      "loss": 0.4689,
+      "step": 3150
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022447338196558448,
+      "loss": 0.4697,
+      "step": 3155
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00022447171126880138,
+      "loss": 0.4724,
+      "step": 3160
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00022447003793259714,
+      "loss": 0.4604,
+      "step": 3165
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.000224468361957013,
+      "loss": 0.4605,
+      "step": 3170
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002244666833420903,
+      "loss": 0.4657,
+      "step": 3175
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002244650020878705,
+      "loss": 0.4684,
+      "step": 3180
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002244633181943951,
+      "loss": 0.467,
+      "step": 3185
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002244616316617056,
+      "loss": 0.4622,
+      "step": 3190
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00022445994248984364,
+      "loss": 0.4677,
+      "step": 3195
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00022445825067885086,
+      "loss": 0.4597,
+      "step": 3200
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.000224456556228769,
+      "loss": 0.4648,
+      "step": 3205
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00022445485913963995,
+      "loss": 0.46,
+      "step": 3210
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00022445315941150554,
+      "loss": 0.4626,
+      "step": 3215
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002244514570444077,
+      "loss": 0.4669,
+      "step": 3220
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002244497520383884,
+      "loss": 0.4707,
+      "step": 3225
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022444804439348978,
+      "loss": 0.4659,
+      "step": 3230
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022444633410975391,
+      "loss": 0.4607,
+      "step": 3235
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022444462118722305,
+      "loss": 0.4666,
+      "step": 3240
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022444290562593944,
+      "loss": 0.4645,
+      "step": 3245
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022444118742594537,
+      "loss": 0.4641,
+      "step": 3250
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022443946658728332,
+      "loss": 0.4639,
+      "step": 3255
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022443774310999568,
+      "loss": 0.466,
+      "step": 3260
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00022443601699412497,
+      "loss": 0.467,
+      "step": 3265
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022443428823971383,
+      "loss": 0.4708,
+      "step": 3270
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022443255684680488,
+      "loss": 0.463,
+      "step": 3275
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022443082281544085,
+      "loss": 0.4602,
+      "step": 3280
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022442908614566455,
+      "loss": 0.4658,
+      "step": 3285
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002244273468375188,
+      "loss": 0.4621,
+      "step": 3290
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002244256048910465,
+      "loss": 0.4653,
+      "step": 3295
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002244238603062907,
+      "loss": 0.4685,
+      "step": 3300
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022442211308329433,
+      "loss": 0.4697,
+      "step": 3305
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022442036322210059,
+      "loss": 0.4652,
+      "step": 3310
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00022441861072275266,
+      "loss": 0.464,
+      "step": 3315
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002244168555852938,
+      "loss": 0.4579,
+      "step": 3320
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002244150978097672,
+      "loss": 0.4666,
+      "step": 3325
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022441333739621635,
+      "loss": 0.4667,
+      "step": 3330
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002244115743446846,
+      "loss": 0.4635,
+      "step": 3335
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022440980865521553,
+      "loss": 0.4632,
+      "step": 3340
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022440804032785265,
+      "loss": 0.4616,
+      "step": 3345
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022440626936263963,
+      "loss": 0.4636,
+      "step": 3350
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022440449575962014,
+      "loss": 0.4677,
+      "step": 3355
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00022440271951883797,
+      "loss": 0.471,
+      "step": 3360
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002244009406403369,
+      "loss": 0.4554,
+      "step": 3365
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022439915912416086,
+      "loss": 0.4594,
+      "step": 3370
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002243973749703538,
+      "loss": 0.4636,
+      "step": 3375
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022439558817895973,
+      "loss": 0.4651,
+      "step": 3380
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022439379875002275,
+      "loss": 0.4616,
+      "step": 3385
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.000224392006683587,
+      "loss": 0.4553,
+      "step": 3390
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022439021197969675,
+      "loss": 0.4659,
+      "step": 3395
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002243884146383962,
+      "loss": 0.4686,
+      "step": 3400
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022438661465972972,
+      "loss": 0.4694,
+      "step": 3405
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022438481204374176,
+      "loss": 0.4646,
+      "step": 3410
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002243830067904768,
+      "loss": 0.4655,
+      "step": 3415
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00022438119889997934,
+      "loss": 0.4665,
+      "step": 3420
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000224379388372294,
+      "loss": 0.4616,
+      "step": 3425
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022437757520746545,
+      "loss": 0.4645,
+      "step": 3430
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022437575940553847,
+      "loss": 0.4647,
+      "step": 3435
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022437394096655783,
+      "loss": 0.4691,
+      "step": 3440
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002243721198905684,
+      "loss": 0.465,
+      "step": 3445
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002243702961776151,
+      "loss": 0.4707,
+      "step": 3450
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002243684698277429,
+      "loss": 0.4671,
+      "step": 3455
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022436664084099698,
+      "loss": 0.4651,
+      "step": 3460
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022436480921742233,
+      "loss": 0.4623,
+      "step": 3465
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00022436297495706425,
+      "loss": 0.4675,
+      "step": 3470
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022436113805996796,
+      "loss": 0.4667,
+      "step": 3475
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022435929852617875,
+      "loss": 0.4673,
+      "step": 3480
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022435745635574208,
+      "loss": 0.4669,
+      "step": 3485
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002243556115487033,
+      "loss": 0.4632,
+      "step": 3490
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022435376410510801,
+      "loss": 0.4588,
+      "step": 3495
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002243519140250018,
+      "loss": 0.47,
+      "step": 3500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002243500613084303,
+      "loss": 0.466,
+      "step": 3505
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002243482059554392,
+      "loss": 0.4637,
+      "step": 3510
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022434634796607427,
+      "loss": 0.4661,
+      "step": 3515
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00022434448734038142,
+      "loss": 0.4609,
+      "step": 3520
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022434262407840648,
+      "loss": 0.4634,
+      "step": 3525
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002243407581801955,
+      "loss": 0.4621,
+      "step": 3530
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022433888964579448,
+      "loss": 0.462,
+      "step": 3535
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002243370184752495,
+      "loss": 0.4571,
+      "step": 3540
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022433514466860678,
+      "loss": 0.4613,
+      "step": 3545
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022433326822591252,
+      "loss": 0.4654,
+      "step": 3550
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022433138914721302,
+      "loss": 0.4618,
+      "step": 3555
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002243295074325547,
+      "loss": 0.4631,
+      "step": 3560
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002243276230819839,
+      "loss": 0.4677,
+      "step": 3565
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022432573609554716,
+      "loss": 0.4603,
+      "step": 3570
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00022432384647329105,
+      "loss": 0.4624,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022432195421526218,
+      "loss": 0.4667,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022432005932150723,
+      "loss": 0.4641,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022431816179207297,
+      "loss": 0.4631,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022431626162700623,
+      "loss": 0.4612,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022431435882635386,
+      "loss": 0.4624,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022431245339016285,
+      "loss": 0.4666,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022431054531848018,
+      "loss": 0.4584,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.000224308634611353,
+      "loss": 0.4636,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022430672126882832,
+      "loss": 0.4634,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00022430480529095345,
+      "loss": 0.4637,
+      "step": 3625
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022430288667777567,
+      "loss": 0.4618,
+      "step": 3630
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002243009654293423,
+      "loss": 0.4593,
+      "step": 3635
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022429904154570073,
+      "loss": 0.4696,
+      "step": 3640
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022429711502689842,
+      "loss": 0.4701,
+      "step": 3645
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022429518587298295,
+      "loss": 0.4622,
+      "step": 3650
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022429325408400186,
+      "loss": 0.4613,
+      "step": 3655
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022429131966000287,
+      "loss": 0.4677,
+      "step": 3660
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00022428938260103367,
+      "loss": 0.4637,
+      "step": 3665
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002242874429071421,
+      "loss": 0.4608,
+      "step": 3670
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.000224285500578376,
+      "loss": 0.4593,
+      "step": 3675
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022428355561478325,
+      "loss": 0.4678,
+      "step": 3680
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002242816080164119,
+      "loss": 0.4646,
+      "step": 3685
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022427965778331,
+      "loss": 0.4657,
+      "step": 3690
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022427770491552563,
+      "loss": 0.4629,
+      "step": 3695
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022427574941310703,
+      "loss": 0.4635,
+      "step": 3700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022427379127610238,
+      "loss": 0.4629,
+      "step": 3705
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022427183050456007,
+      "loss": 0.4657,
+      "step": 3710
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022426986709852845,
+      "loss": 0.4674,
+      "step": 3715
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022426790105805594,
+      "loss": 0.4667,
+      "step": 3720
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00022426593238319105,
+      "loss": 0.4668,
+      "step": 3725
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242639610739824,
+      "loss": 0.4603,
+      "step": 3730
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242619871304786,
+      "loss": 0.4597,
+      "step": 3735
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242600105527284,
+      "loss": 0.4631,
+      "step": 3740
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00022425803134078044,
+      "loss": 0.4632,
+      "step": 3745
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242560494946837,
+      "loss": 0.4613,
+      "step": 3750
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00022425406501448705,
+      "loss": 0.4597,
+      "step": 3755
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242520779002394,
+      "loss": 0.4584,
+      "step": 3760
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00022425008815198983,
+      "loss": 0.4654,
+      "step": 3765
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242480957697874,
+      "loss": 0.4591,
+      "step": 3770
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00022424610075368127,
+      "loss": 0.4626,
+      "step": 3775
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002242441031037207,
+      "loss": 0.4605,
+      "step": 3780
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022424210281995496,
+      "loss": 0.4665,
+      "step": 3785
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002242400999024334,
+      "loss": 0.4596,
+      "step": 3790
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022423809435120543,
+      "loss": 0.4662,
+      "step": 3795
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002242360861663206,
+      "loss": 0.4664,
+      "step": 3800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022423407534782837,
+      "loss": 0.4646,
+      "step": 3805
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002242320618957784,
+      "loss": 0.4621,
+      "step": 3810
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022423004581022038,
+      "loss": 0.4582,
+      "step": 3815
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022422802709120402,
+      "loss": 0.4619,
+      "step": 3820
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00022422600573877917,
+      "loss": 0.4578,
+      "step": 3825
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002242239817529957,
+      "loss": 0.4596,
+      "step": 3830
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002242219551339035,
+      "loss": 0.4582,
+      "step": 3835
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002242199258815526,
+      "loss": 0.4606,
+      "step": 3840
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022421789399599312,
+      "loss": 0.4705,
+      "step": 3845
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022421585947727515,
+      "loss": 0.4656,
+      "step": 3850
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002242138223254489,
+      "loss": 0.459,
+      "step": 3855
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022421178254056461,
+      "loss": 0.46,
+      "step": 3860
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022420974012267263,
+      "loss": 0.464,
+      "step": 3865
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022420769507182337,
+      "loss": 0.4581,
+      "step": 3870
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022420564738806725,
+      "loss": 0.4569,
+      "step": 3875
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00022420359707145483,
+      "loss": 0.4623,
+      "step": 3880
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00022420154412203667,
+      "loss": 0.4632,
+      "step": 3885
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241994885398635,
+      "loss": 0.4601,
+      "step": 3890
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241974303249859,
+      "loss": 0.4629,
+      "step": 3895
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241953694774548,
+      "loss": 0.462,
+      "step": 3900
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00022419330599732096,
+      "loss": 0.4644,
+      "step": 3905
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241912398846353,
+      "loss": 0.4584,
+      "step": 3910
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00022418917113944882,
+      "loss": 0.4628,
+      "step": 3915
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241870997618126,
+      "loss": 0.4635,
+      "step": 3920
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002241850257517777,
+      "loss": 0.4587,
+      "step": 3925
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00022418294910939529,
+      "loss": 0.462,
+      "step": 3930
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00022418086983471664,
+      "loss": 0.4596,
+      "step": 3935
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022417878792779303,
+      "loss": 0.4602,
+      "step": 3940
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022417670338867584,
+      "loss": 0.4633,
+      "step": 3945
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002241746162174165,
+      "loss": 0.4608,
+      "step": 3950
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002241725264140665,
+      "loss": 0.462,
+      "step": 3955
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022417043397867745,
+      "loss": 0.462,
+      "step": 3960
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022416833891130092,
+      "loss": 0.4627,
+      "step": 3965
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022416624121198862,
+      "loss": 0.4635,
+      "step": 3970
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002241641408807923,
+      "loss": 0.4597,
+      "step": 3975
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00022416203791776384,
+      "loss": 0.4623,
+      "step": 3980
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002241599323229551,
+      "loss": 0.4634,
+      "step": 3985
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.000224157824096418,
+      "loss": 0.4606,
+      "step": 3990
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022415571323820455,
+      "loss": 0.4616,
+      "step": 3995
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002241535997483669,
+      "loss": 0.4583,
+      "step": 4000
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.44417059421539307,
+      "eval_runtime": 48.8396,
+      "eval_samples_per_second": 102.376,
+      "eval_steps_per_second": 1.085,
+      "step": 4000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022415148362695716,
+      "loss": 0.4598,
+      "step": 4005
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022414936487402752,
+      "loss": 0.4636,
+      "step": 4010
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022414724348963032,
+      "loss": 0.4621,
+      "step": 4015
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022414511947381783,
+      "loss": 0.4552,
+      "step": 4020
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002241429928266425,
+      "loss": 0.4637,
+      "step": 4025
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00022414086354815682,
+      "loss": 0.4636,
+      "step": 4030
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002241387316384133,
+      "loss": 0.463,
+      "step": 4035
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022413659709746454,
+      "loss": 0.4637,
+      "step": 4040
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002241344599253632,
+      "loss": 0.4629,
+      "step": 4045
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022413232012216206,
+      "loss": 0.4617,
+      "step": 4050
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022413017768791385,
+      "loss": 0.4641,
+      "step": 4055
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022412803262267147,
+      "loss": 0.4618,
+      "step": 4060
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022412588492648787,
+      "loss": 0.4594,
+      "step": 4065
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.000224123734599416,
+      "loss": 0.4686,
+      "step": 4070
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022412158164150894,
+      "loss": 0.4659,
+      "step": 4075
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022411942605281984,
+      "loss": 0.4659,
+      "step": 4080
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00022411726783340183,
+      "loss": 0.4615,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022411510698330816,
+      "loss": 0.4619,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002241129435025922,
+      "loss": 0.4563,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002241107773913073,
+      "loss": 0.4674,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002241086086495069,
+      "loss": 0.4649,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022410643727724454,
+      "loss": 0.4608,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022410426327457378,
+      "loss": 0.4561,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022410208664154824,
+      "loss": 0.4608,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022409990737822167,
+      "loss": 0.4669,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022409772548464779,
+      "loss": 0.4648,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022409554096088048,
+      "loss": 0.4671,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00022409335380697362,
+      "loss": 0.4629,
+      "step": 4140
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002240911640229812,
+      "loss": 0.4623,
+      "step": 4145
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022408897160895722,
+      "loss": 0.4638,
+      "step": 4150
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022408677656495578,
+      "loss": 0.4653,
+      "step": 4155
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022408457889103105,
+      "loss": 0.4625,
+      "step": 4160
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022408237858723726,
+      "loss": 0.4631,
+      "step": 4165
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022408017565362867,
+      "loss": 0.4665,
+      "step": 4170
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022407797009025968,
+      "loss": 0.458,
+      "step": 4175
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022407576189718468,
+      "loss": 0.4641,
+      "step": 4180
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00022407355107445818,
+      "loss": 0.4614,
+      "step": 4185
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002240713376221347,
+      "loss": 0.4614,
+      "step": 4190
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022406912154026887,
+      "loss": 0.4621,
+      "step": 4195
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022406690282891537,
+      "loss": 0.4664,
+      "step": 4200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022406468148812892,
+      "loss": 0.4556,
+      "step": 4205
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022406245751796437,
+      "loss": 0.4592,
+      "step": 4210
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002240602309184766,
+      "loss": 0.4645,
+      "step": 4215
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022405800168972048,
+      "loss": 0.4602,
+      "step": 4220
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002240557698317511,
+      "loss": 0.4608,
+      "step": 4225
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022405353534462347,
+      "loss": 0.4646,
+      "step": 4230
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022405129822839273,
+      "loss": 0.4577,
+      "step": 4235
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00022404905848311414,
+      "loss": 0.4591,
+      "step": 4240
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022404681610884287,
+      "loss": 0.4572,
+      "step": 4245
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022404457110563429,
+      "loss": 0.4611,
+      "step": 4250
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002240423234735438,
+      "loss": 0.4572,
+      "step": 4255
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022404007321262685,
+      "loss": 0.4562,
+      "step": 4260
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022403782032293898,
+      "loss": 0.4588,
+      "step": 4265
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022403556480453577,
+      "loss": 0.4583,
+      "step": 4270
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022403330665747285,
+      "loss": 0.4571,
+      "step": 4275
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.000224031045881806,
+      "loss": 0.4614,
+      "step": 4280
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002240287824775909,
+      "loss": 0.4587,
+      "step": 4285
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022402651644488347,
+      "loss": 0.4613,
+      "step": 4290
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00022402424778373962,
+      "loss": 0.4595,
+      "step": 4295
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002240219764942153,
+      "loss": 0.4541,
+      "step": 4300
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022401970257636658,
+      "loss": 0.4578,
+      "step": 4305
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022401742603024954,
+      "loss": 0.4593,
+      "step": 4310
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022401514685592037,
+      "loss": 0.4631,
+      "step": 4315
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002240128650534353,
+      "loss": 0.4599,
+      "step": 4320
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022401058062285063,
+      "loss": 0.4625,
+      "step": 4325
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022400829356422272,
+      "loss": 0.4627,
+      "step": 4330
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.000224006003877608,
+      "loss": 0.4688,
+      "step": 4335
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.000224003711563063,
+      "loss": 0.461,
+      "step": 4340
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00022400141662064422,
+      "loss": 0.4625,
+      "step": 4345
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022399911905040835,
+      "loss": 0.4632,
+      "step": 4350
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022399681885241203,
+      "loss": 0.4628,
+      "step": 4355
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022399451602671207,
+      "loss": 0.4606,
+      "step": 4360
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002239922105733652,
+      "loss": 0.4596,
+      "step": 4365
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002239899024924284,
+      "loss": 0.4589,
+      "step": 4370
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022398759178395857,
+      "loss": 0.4623,
+      "step": 4375
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022398527844801275,
+      "loss": 0.4593,
+      "step": 4380
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022398296248464798,
+      "loss": 0.4641,
+      "step": 4385
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022398064389392142,
+      "loss": 0.4563,
+      "step": 4390
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00022397832267589032,
+      "loss": 0.4615,
+      "step": 4395
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022397599883061193,
+      "loss": 0.4594,
+      "step": 4400
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022397367235814353,
+      "loss": 0.4607,
+      "step": 4405
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002239713432585426,
+      "loss": 0.4617,
+      "step": 4410
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022396901153186655,
+      "loss": 0.455,
+      "step": 4415
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022396667717817297,
+      "loss": 0.4597,
+      "step": 4420
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022396434019751943,
+      "loss": 0.4645,
+      "step": 4425
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002239620005899636,
+      "loss": 0.4605,
+      "step": 4430
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022395965835556318,
+      "loss": 0.4627,
+      "step": 4435
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022395731349437602,
+      "loss": 0.4639,
+      "step": 4440
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00022395496600645991,
+      "loss": 0.4582,
+      "step": 4445
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002239526158918728,
+      "loss": 0.4626,
+      "step": 4450
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002239502631506727,
+      "loss": 0.4584,
+      "step": 4455
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002239479077829176,
+      "loss": 0.4584,
+      "step": 4460
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022394554978866574,
+      "loss": 0.4598,
+      "step": 4465
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022394318916797514,
+      "loss": 0.4575,
+      "step": 4470
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022394082592090418,
+      "loss": 0.4588,
+      "step": 4475
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022393846004751107,
+      "loss": 0.4603,
+      "step": 4480
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022393609154785425,
+      "loss": 0.4623,
+      "step": 4485
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022393372042199216,
+      "loss": 0.4701,
+      "step": 4490
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022393134666998324,
+      "loss": 0.4745,
+      "step": 4495
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00022392897029188615,
+      "loss": 0.4779,
+      "step": 4500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022392659128775947,
+      "loss": 0.4711,
+      "step": 4505
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022392420965766188,
+      "loss": 0.4729,
+      "step": 4510
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022392182540165225,
+      "loss": 0.4711,
+      "step": 4515
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022391943851978927,
+      "loss": 0.4638,
+      "step": 4520
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002239170490121319,
+      "loss": 0.4675,
+      "step": 4525
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022391465687873912,
+      "loss": 0.4635,
+      "step": 4530
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022391226211966993,
+      "loss": 0.4661,
+      "step": 4535
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002239098647349834,
+      "loss": 0.4676,
+      "step": 4540
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022390746472473872,
+      "loss": 0.4635,
+      "step": 4545
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00022390506208899507,
+      "loss": 0.4681,
+      "step": 4550
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022390265682781178,
+      "loss": 0.4633,
+      "step": 4555
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022390024894124814,
+      "loss": 0.4556,
+      "step": 4560
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022389783842936362,
+      "loss": 0.4647,
+      "step": 4565
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022389542529221767,
+      "loss": 0.4582,
+      "step": 4570
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022389300952986983,
+      "loss": 0.4609,
+      "step": 4575
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022389059114237968,
+      "loss": 0.4613,
+      "step": 4580
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022388817012980694,
+      "loss": 0.4629,
+      "step": 4585
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002238857464922113,
+      "loss": 0.4552,
+      "step": 4590
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022388332022965263,
+      "loss": 0.4627,
+      "step": 4595
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00022388089134219073,
+      "loss": 0.4551,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022387845982988554,
+      "loss": 0.4576,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002238760256927971,
+      "loss": 0.4625,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022387358893098542,
+      "loss": 0.4607,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022387114954451062,
+      "loss": 0.461,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022386870753343292,
+      "loss": 0.4521,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002238662628978126,
+      "loss": 0.4642,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002238638156377099,
+      "loss": 0.4615,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022386136575318526,
+      "loss": 0.462,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022385891324429915,
+      "loss": 0.4597,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022385645811111199,
+      "loss": 0.4586,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00022385400035368445,
+      "loss": 0.4588,
+      "step": 4655
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00022385153997207712,
+      "loss": 0.4607,
+      "step": 4660
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00022384907696635078,
+      "loss": 0.4547,
+      "step": 4665
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002238466113365661,
+      "loss": 0.4616,
+      "step": 4670
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.000223844143082784,
+      "loss": 0.4612,
+      "step": 4675
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002238416722050653,
+      "loss": 0.4623,
+      "step": 4680
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00022383919870347109,
+      "loss": 0.4573,
+      "step": 4685
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002238367225780623,
+      "loss": 0.4578,
+      "step": 4690
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002238342438289,
+      "loss": 0.4663,
+      "step": 4695
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00022383176245604545,
+      "loss": 0.4643,
+      "step": 4700
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00022382927845955984,
+      "loss": 0.4581,
+      "step": 4705
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022382679183950442,
+      "loss": 0.4594,
+      "step": 4710
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002238243025959406,
+      "loss": 0.4584,
+      "step": 4715
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022382181072892973,
+      "loss": 0.4596,
+      "step": 4720
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022381931623853337,
+      "loss": 0.4583,
+      "step": 4725
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022381681912481304,
+      "loss": 0.4562,
+      "step": 4730
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022381431938783036,
+      "loss": 0.4582,
+      "step": 4735
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022381181702764696,
+      "loss": 0.4565,
+      "step": 4740
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022380931204432462,
+      "loss": 0.4614,
+      "step": 4745
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00022380680443792519,
+      "loss": 0.4569,
+      "step": 4750
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002238042942085105,
+      "loss": 0.4567,
+      "step": 4755
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022380178135614245,
+      "loss": 0.4606,
+      "step": 4760
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022379926588088312,
+      "loss": 0.4625,
+      "step": 4765
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022379674778279452,
+      "loss": 0.4611,
+      "step": 4770
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022379422706193884,
+      "loss": 0.4607,
+      "step": 4775
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022379170371837818,
+      "loss": 0.4517,
+      "step": 4780
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002237891777521749,
+      "loss": 0.4608,
+      "step": 4785
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022378664916339127,
+      "loss": 0.4584,
+      "step": 4790
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022378411795208968,
+      "loss": 0.4536,
+      "step": 4795
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022378158411833263,
+      "loss": 0.4582,
+      "step": 4800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002237790476621826,
+      "loss": 0.463,
+      "step": 4805
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00022377650858370216,
+      "loss": 0.4633,
+      "step": 4810
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022377396688295402,
+      "loss": 0.4575,
+      "step": 4815
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022377142256000082,
+      "loss": 0.4569,
+      "step": 4820
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022376887561490537,
+      "loss": 0.4615,
+      "step": 4825
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022376632604773056,
+      "loss": 0.4549,
+      "step": 4830
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022376377385853924,
+      "loss": 0.4571,
+      "step": 4835
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002237612190473944,
+      "loss": 0.4603,
+      "step": 4840
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002237586616143591,
+      "loss": 0.4601,
+      "step": 4845
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022375610155949638,
+      "loss": 0.4576,
+      "step": 4850
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022375353888286946,
+      "loss": 0.466,
+      "step": 4855
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00022375097358454156,
+      "loss": 0.4605,
+      "step": 4860
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022374840566457597,
+      "loss": 0.4584,
+      "step": 4865
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002237458351230361,
+      "loss": 0.4573,
+      "step": 4870
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022374326195998524,
+      "loss": 0.4638,
+      "step": 4875
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022374068617548703,
+      "loss": 0.4583,
+      "step": 4880
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022373810776960497,
+      "loss": 0.4495,
+      "step": 4885
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022373552674240265,
+      "loss": 0.4516,
+      "step": 4890
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002237329430939438,
+      "loss": 0.4582,
+      "step": 4895
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022373035682429215,
+      "loss": 0.4589,
+      "step": 4900
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002237277679335115,
+      "loss": 0.4604,
+      "step": 4905
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00022372517642166572,
+      "loss": 0.4573,
+      "step": 4910
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022372258228881882,
+      "loss": 0.4599,
+      "step": 4915
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022371998553503473,
+      "loss": 0.4591,
+      "step": 4920
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022371738616037756,
+      "loss": 0.4611,
+      "step": 4925
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022371478416491146,
+      "loss": 0.4547,
+      "step": 4930
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002237121795487006,
+      "loss": 0.4596,
+      "step": 4935
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022370957231180924,
+      "loss": 0.4626,
+      "step": 4940
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022370696245430175,
+      "loss": 0.457,
+      "step": 4945
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022370434997624248,
+      "loss": 0.4613,
+      "step": 4950
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022370173487769595,
+      "loss": 0.4559,
+      "step": 4955
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00022369911715872662,
+      "loss": 0.4617,
+      "step": 4960
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022369649681939913,
+      "loss": 0.4636,
+      "step": 4965
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022369387385977812,
+      "loss": 0.4592,
+      "step": 4970
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022369124827992828,
+      "loss": 0.4593,
+      "step": 4975
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022368862007991443,
+      "loss": 0.4607,
+      "step": 4980
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002236859892598014,
+      "loss": 0.458,
+      "step": 4985
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022368335581965414,
+      "loss": 0.4626,
+      "step": 4990
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022368071975953756,
+      "loss": 0.4562,
+      "step": 4995
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022367808107951674,
+      "loss": 0.4576,
+      "step": 5000
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.44060567021369934,
+      "eval_runtime": 41.3498,
+      "eval_samples_per_second": 120.919,
+      "eval_steps_per_second": 1.282,
+      "step": 5000
+    }
+  ],
+  "max_steps": 100000,
+  "num_train_epochs": 20,
+  "total_flos": 5.257519643492352e+20,
+  "trial_name": null,
+  "trial_params": null
+}