diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4809 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.999053926206244,
+  "global_step": 3960,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.289308176100629e-06,
+      "loss": 5.7433,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2578616352201259e-05,
+      "loss": 5.2911,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 4.8586,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5157232704402517e-05,
+      "loss": 4.5404,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.144654088050314e-05,
+      "loss": 4.2673,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.7735849056603776e-05,
+      "loss": 4.0794,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.402515723270441e-05,
+      "loss": 3.9428,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.0314465408805034e-05,
+      "loss": 3.8166,
+      "step": 40
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.660377358490566e-05,
+      "loss": 3.7152,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.289308176100629e-05,
+      "loss": 3.6501,
+      "step": 50
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 6.918238993710691e-05,
+      "loss": 3.583,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.547169811320755e-05,
+      "loss": 3.5229,
+      "step": 60
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.176100628930818e-05,
+      "loss": 3.4693,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.805031446540882e-05,
+      "loss": 3.4358,
+      "step": 70
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.433962264150944e-05,
+      "loss": 3.3779,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010062893081761007,
+      "loss": 3.3463,
+      "step": 80
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001069182389937107,
+      "loss": 3.3215,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011320754716981132,
+      "loss": 3.2891,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00011949685534591195,
+      "loss": 3.2593,
+      "step": 95
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012578616352201257,
+      "loss": 3.2389,
+      "step": 100
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001320754716981132,
+      "loss": 3.208,
+      "step": 105
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00013836477987421382,
+      "loss": 3.1893,
+      "step": 110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00014465408805031446,
+      "loss": 3.1749,
+      "step": 115
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001509433962264151,
+      "loss": 3.1352,
+      "step": 120
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00015723270440251574,
+      "loss": 3.1069,
+      "step": 125
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00016352201257861635,
+      "loss": 3.0995,
+      "step": 130
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000169811320754717,
+      "loss": 3.0828,
+      "step": 135
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00017610062893081763,
+      "loss": 3.0757,
+      "step": 140
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018238993710691825,
+      "loss": 3.0406,
+      "step": 145
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00018867924528301889,
+      "loss": 3.0378,
+      "step": 150
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001949685534591195,
+      "loss": 3.0082,
+      "step": 155
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019999996584345097,
+      "loss": 3.0023,
+      "step": 160
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019999877036668484,
+      "loss": 2.9836,
+      "step": 165
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001999958670858004,
+      "loss": 2.9812,
+      "step": 170
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019999125605038058,
+      "loss": 2.9464,
+      "step": 175
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019998493733917384,
+      "loss": 2.9401,
+      "step": 180
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019997691106009272,
+      "loss": 2.924,
+      "step": 185
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000199967177350212,
+      "loss": 2.9024,
+      "step": 190
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019995573637576647,
+      "loss": 2.8813,
+      "step": 195
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001999425883321479,
+      "loss": 2.8854,
+      "step": 200
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019992773344390194,
+      "loss": 2.8847,
+      "step": 205
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019991117196472405,
+      "loss": 2.8749,
+      "step": 210
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019989290417745542,
+      "loss": 2.8411,
+      "step": 215
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001998729303940778,
+      "loss": 2.8315,
+      "step": 220
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019985125095570855,
+      "loss": 2.8267,
+      "step": 225
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019982786623259452,
+      "loss": 2.8072,
+      "step": 230
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001998027766241059,
+      "loss": 2.805,
+      "step": 235
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001997759825587294,
+      "loss": 2.7963,
+      "step": 240
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019974748449406069,
+      "loss": 2.7773,
+      "step": 245
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001997172829167969,
+      "loss": 2.7808,
+      "step": 250
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001996853783427282,
+      "loss": 2.7796,
+      "step": 255
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001996517713167289,
+      "loss": 2.7296,
+      "step": 260
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001996164624127482,
+      "loss": 2.7506,
+      "step": 265
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019957945223380045,
+      "loss": 2.7212,
+      "step": 270
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001995407414119548,
+      "loss": 2.7339,
+      "step": 275
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001995003306083244,
+      "loss": 2.7255,
+      "step": 280
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019945822051305507,
+      "loss": 2.6909,
+      "step": 285
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001994144118453136,
+      "loss": 2.6944,
+      "step": 290
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019936890535327543,
+      "loss": 2.6877,
+      "step": 295
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019932170181411184,
+      "loss": 2.6761,
+      "step": 300
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001992728020339768,
+      "loss": 2.6946,
+      "step": 305
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019922220684799293,
+      "loss": 2.6618,
+      "step": 310
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001991699171202376,
+      "loss": 2.6439,
+      "step": 315
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019911593374372788,
+      "loss": 2.656,
+      "step": 320
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001990602576404055,
+      "loss": 2.6382,
+      "step": 325
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019900288976112084,
+      "loss": 2.6359,
+      "step": 330
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019894383108561697,
+      "loss": 2.6341,
+      "step": 335
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019888308262251285,
+      "loss": 2.617,
+      "step": 340
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000198820645409286,
+      "loss": 2.6151,
+      "step": 345
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001987565205122547,
+      "loss": 2.6267,
+      "step": 350
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019869070902656018,
+      "loss": 2.6007,
+      "step": 355
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019862321207614748,
+      "loss": 2.6052,
+      "step": 360
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019855403081374648,
+      "loss": 2.5964,
+      "step": 365
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019848316642085215,
+      "loss": 2.5614,
+      "step": 370
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019841062010770435,
+      "loss": 2.5562,
+      "step": 375
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001983363931132673,
+      "loss": 2.5723,
+      "step": 380
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019826048670520824,
+      "loss": 2.5707,
+      "step": 385
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019818290217987587,
+      "loss": 2.5718,
+      "step": 390
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019810364086227819,
+      "loss": 2.5659,
+      "step": 395
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019802270410605987,
+      "loss": 2.5516,
+      "step": 400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019794009329347924,
+      "loss": 2.5571,
+      "step": 405
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019785580983538444,
+      "loss": 2.534,
+      "step": 410
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019776985517118957,
+      "loss": 2.5348,
+      "step": 415
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019768223076884996,
+      "loss": 2.5275,
+      "step": 420
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019759293812483713,
+      "loss": 2.5146,
+      "step": 425
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019750197876411334,
+      "loss": 2.5079,
+      "step": 430
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019740935424010538,
+      "loss": 2.5012,
+      "step": 435
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019731506613467816,
+      "loss": 2.5222,
+      "step": 440
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019721911605810757,
+      "loss": 2.4999,
+      "step": 445
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019712150564905317,
+      "loss": 2.5052,
+      "step": 450
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019702223657453002,
+      "loss": 2.4901,
+      "step": 455
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019692131052988034,
+      "loss": 2.4705,
+      "step": 460
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001968187292387444,
+      "loss": 2.484,
+      "step": 465
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019671449445303135,
+      "loss": 2.4761,
+      "step": 470
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019660860795288896,
+      "loss": 2.4538,
+      "step": 475
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019650107154667353,
+      "loss": 2.4634,
+      "step": 480
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001963918870709188,
+      "loss": 2.4823,
+      "step": 485
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019628105639030467,
+      "loss": 2.4622,
+      "step": 490
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019616858139762534,
+      "loss": 2.4447,
+      "step": 495
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000196054464013757,
+      "loss": 2.459,
+      "step": 500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019593870618762497,
+      "loss": 2.4341,
+      "step": 505
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019582130989617051,
+      "loss": 2.4287,
+      "step": 510
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019570227714431698,
+      "loss": 2.4302,
+      "step": 515
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019558160996493556,
+      "loss": 2.4237,
+      "step": 520
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001954593104188107,
+      "loss": 2.4327,
+      "step": 525
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019533538059460475,
+      "loss": 2.419,
+      "step": 530
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019520982260882238,
+      "loss": 2.4153,
+      "step": 535
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00019508263860577442,
+      "loss": 2.414,
+      "step": 540
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001949538307575412,
+      "loss": 2.4247,
+      "step": 545
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019482340126393552,
+      "loss": 2.3946,
+      "step": 550
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019469135235246503,
+      "loss": 2.4084,
+      "step": 555
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019455768627829422,
+      "loss": 2.3901,
+      "step": 560
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019442240532420584,
+      "loss": 2.3726,
+      "step": 565
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00019428551180056204,
+      "loss": 2.3769,
+      "step": 570
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001941470080452648,
+      "loss": 2.3772,
+      "step": 575
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019400689642371606,
+      "loss": 2.3811,
+      "step": 580
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019386517932877724,
+      "loss": 2.3667,
+      "step": 585
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001937218591807285,
+      "loss": 2.3726,
+      "step": 590
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00019357693842722736,
+      "loss": 2.3663,
+      "step": 595
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001934304195432668,
+      "loss": 2.3675,
+      "step": 600
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019328230503113316,
+      "loss": 2.3655,
+      "step": 605
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00019313259742036328,
+      "loss": 2.368,
+      "step": 610
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00019298129926770133,
+      "loss": 2.3418,
+      "step": 615
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019282841315705515,
+      "loss": 2.3395,
+      "step": 620
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019267394169945217,
+      "loss": 2.3583,
+      "step": 625
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001925178875329947,
+      "loss": 2.3253,
+      "step": 630
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00019236025332281507,
+      "loss": 2.3288,
+      "step": 635
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00019220104176102985,
+      "loss": 2.3621,
+      "step": 640
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001920402555666941,
+      "loss": 2.3427,
+      "step": 645
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00019187789748575481,
+      "loss": 2.3394,
+      "step": 650
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00019171397029100416,
+      "loss": 2.3375,
+      "step": 655
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019154847678203184,
+      "loss": 2.3196,
+      "step": 660
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019138141978517764,
+      "loss": 2.3157,
+      "step": 665
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00019121280215348286,
+      "loss": 2.2923,
+      "step": 670
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001910426267666418,
+      "loss": 2.3152,
+      "step": 675
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00019087089653095252,
+      "loss": 2.3209,
+      "step": 680
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.000190697614379267,
+      "loss": 2.31,
+      "step": 685
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00019052278327094145,
+      "loss": 2.3068,
+      "step": 690
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001903464061917854,
+      "loss": 2.2994,
+      "step": 695
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00019016848615401095,
+      "loss": 2.3125,
+      "step": 700
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00018998902619618116,
+      "loss": 2.3146,
+      "step": 705
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00018980802938315826,
+      "loss": 2.2878,
+      "step": 710
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00018962549880605133,
+      "loss": 2.2895,
+      "step": 715
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018944143758216337,
+      "loss": 2.3017,
+      "step": 720
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00018925584885493823,
+      "loss": 2.2918,
+      "step": 725
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00018906873579390681,
+      "loss": 2.2867,
+      "step": 730
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00018888010159463293,
+      "loss": 2.2766,
+      "step": 735
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00018868994947865883,
+      "loss": 2.2821,
+      "step": 740
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00018849828269345016,
+      "loss": 2.2982,
+      "step": 745
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00018830510451234037,
+      "loss": 2.2653,
+      "step": 750
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00018811041823447498,
+      "loss": 2.2641,
+      "step": 755
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00018791422718475516,
+      "loss": 2.2767,
+      "step": 760
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001877165347137809,
+      "loss": 2.2804,
+      "step": 765
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001875173441977939,
+      "loss": 2.2629,
+      "step": 770
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00018731665903861985,
+      "loss": 2.2589,
+      "step": 775
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00018711448266361025,
+      "loss": 2.2489,
+      "step": 780
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00018691081852558407,
+      "loss": 2.2529,
+      "step": 785
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00018670567010276858,
+      "loss": 2.2726,
+      "step": 790
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001864990408987401,
+      "loss": 2.2599,
+      "step": 795
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00018629093444236402,
+      "loss": 2.2437,
+      "step": 800
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00018608135428773478,
+      "loss": 2.2449,
+      "step": 805
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001858703040141148,
+      "loss": 2.2381,
+      "step": 810
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0001856577872258737,
+      "loss": 2.2333,
+      "step": 815
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001854438075524266,
+      "loss": 2.2305,
+      "step": 820
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001852283686481721,
+      "loss": 2.2371,
+      "step": 825
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0001850114741924299,
+      "loss": 2.2247,
+      "step": 830
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.000184793127889378,
+      "loss": 2.235,
+      "step": 835
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001845733334679894,
+      "loss": 2.2326,
+      "step": 840
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00018435209468196847,
+      "loss": 2.2293,
+      "step": 845
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00018412941530968676,
+      "loss": 2.2337,
+      "step": 850
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00018390529915411852,
+      "loss": 2.2283,
+      "step": 855
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00018367975004277573,
+      "loss": 2.2207,
+      "step": 860
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00018345277182764278,
+      "loss": 2.2269,
+      "step": 865
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00018322436838511064,
+      "loss": 2.2116,
+      "step": 870
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00018299454361591066,
+      "loss": 2.2105,
+      "step": 875
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00018276330144504803,
+      "loss": 2.2061,
+      "step": 880
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00018253064582173455,
+      "loss": 2.2157,
+      "step": 885
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00018229658071932146,
+      "loss": 2.2018,
+      "step": 890
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00018206111013523133,
+      "loss": 2.2106,
+      "step": 895
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00018182423809088997,
+      "loss": 2.2141,
+      "step": 900
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00018158596863165763,
+      "loss": 2.2089,
+      "step": 905
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00018134630582675998,
+      "loss": 2.1962,
+      "step": 910
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00018110525376921862,
+      "loss": 2.1802,
+      "step": 915
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00018086281657578112,
+      "loss": 2.1924,
+      "step": 920
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001806189983868508,
+      "loss": 2.1845,
+      "step": 925
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001803738033664159,
+      "loss": 2.198,
+      "step": 930
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00018012723570197858,
+      "loss": 2.1849,
+      "step": 935
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00017987929960448332,
+      "loss": 2.1755,
+      "step": 940
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00017962999930824512,
+      "loss": 2.1981,
+      "step": 945
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00017937933907087703,
+      "loss": 2.1859,
+      "step": 950
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00017912732317321752,
+      "loss": 2.1664,
+      "step": 955
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00017887395591925745,
+      "loss": 2.1795,
+      "step": 960
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001786192416360664,
+      "loss": 2.1862,
+      "step": 965
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00017836318467371888,
+      "loss": 2.1519,
+      "step": 970
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00017810578940521995,
+      "loss": 2.1541,
+      "step": 975
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0001778470602264308,
+      "loss": 2.1666,
+      "step": 980
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00017758700155599317,
+      "loss": 2.1724,
+      "step": 985
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00017732561783525446,
+      "loss": 2.1677,
+      "step": 990
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.0888257026672363,
+      "eval_runtime": 106.8696,
+      "eval_samples_per_second": 124.629,
+      "eval_steps_per_second": 15.58,
+      "step": 990
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00017706291352819145,
+      "loss": 2.5238,
+      "step": 995
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00017679889312133439,
+      "loss": 2.096,
+      "step": 1000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00017653356112368998,
+      "loss": 2.0933,
+      "step": 1005
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00017626692206666476,
+      "loss": 2.1039,
+      "step": 1010
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00017599898050398754,
+      "loss": 2.0939,
+      "step": 1015
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00017572974101163165,
+      "loss": 2.0904,
+      "step": 1020
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00017545920818773668,
+      "loss": 2.1099,
+      "step": 1025
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00017518738665253027,
+      "loss": 2.0936,
+      "step": 1030
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001749142810482488,
+      "loss": 2.0855,
+      "step": 1035
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00017463989603905841,
+      "loss": 2.0943,
+      "step": 1040
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00017436423631097517,
+      "loss": 2.0859,
+      "step": 1045
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00017408730657178517,
+      "loss": 2.0732,
+      "step": 1050
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00017380911155096408,
+      "loss": 2.0809,
+      "step": 1055
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00017352965599959634,
+      "loss": 2.0675,
+      "step": 1060
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00017324894469029404,
+      "loss": 2.0893,
+      "step": 1065
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00017296698241711544,
+      "loss": 2.0714,
+      "step": 1070
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00017268377399548302,
+      "loss": 2.0635,
+      "step": 1075
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001723993242621014,
+      "loss": 2.0781,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00017211363807487456,
+      "loss": 2.096,
+      "step": 1085
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00017182672031282296,
+      "loss": 2.0508,
+      "step": 1090
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001715385758760002,
+      "loss": 2.0748,
+      "step": 1095
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00017124920968540936,
+      "loss": 2.0777,
+      "step": 1100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00017095862668291894,
+      "loss": 2.0694,
+      "step": 1105
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001706668318311784,
+      "loss": 2.0762,
+      "step": 1110
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00017037383011353355,
+      "loss": 2.0849,
+      "step": 1115
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00017007962653394133,
+      "loss": 2.0816,
+      "step": 1120
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001697842261168843,
+      "loss": 2.0551,
+      "step": 1125
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0001694876339072851,
+      "loss": 2.0851,
+      "step": 1130
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00016918985497041994,
+      "loss": 2.0696,
+      "step": 1135
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001688908943918322,
+      "loss": 2.0712,
+      "step": 1140
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00016859075727724586,
+      "loss": 2.0522,
+      "step": 1145
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0001682894487524779,
+      "loss": 2.0487,
+      "step": 1150
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.000167986973963351,
+      "loss": 2.053,
+      "step": 1155
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00016768333807560558,
+      "loss": 2.0575,
+      "step": 1160
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0001673785462748116,
+      "loss": 2.062,
+      "step": 1165
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00016707260376628002,
+      "loss": 2.0396,
+      "step": 1170
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00016676551577497382,
+      "loss": 2.0562,
+      "step": 1175
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001664572875454188,
+      "loss": 2.0456,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001661479243416142,
+      "loss": 2.0543,
+      "step": 1185
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00016583743144694241,
+      "loss": 2.0695,
+      "step": 1190
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00016552581416407917,
+      "loss": 2.0613,
+      "step": 1195
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00016521307781490268,
+      "loss": 2.0313,
+      "step": 1200
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001648992277404028,
+      "loss": 2.0416,
+      "step": 1205
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00016458426930059003,
+      "loss": 2.0529,
+      "step": 1210
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00016426820787440364,
+      "loss": 2.0344,
+      "step": 1215
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00016395104885962014,
+      "loss": 2.0335,
+      "step": 1220
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00016363279767276075,
+      "loss": 2.0373,
+      "step": 1225
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00016331345974899923,
+      "loss": 2.0416,
+      "step": 1230
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00016299304054206886,
+      "loss": 2.0357,
+      "step": 1235
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00016267154552416938,
+      "loss": 2.0326,
+      "step": 1240
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 2.0421,
+      "step": 1245
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00016202535003603273,
+      "loss": 2.0278,
+      "step": 1250
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001617006606016845,
+      "loss": 2.0503,
+      "step": 1255
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00016137491742795632,
+      "loss": 2.0314,
+      "step": 1260
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00016104812607797202,
+      "loss": 2.0495,
+      "step": 1265
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00016072029213275626,
+      "loss": 2.035,
+      "step": 1270
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00016039142119113957,
+      "loss": 2.0221,
+      "step": 1275
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00016006151886966248,
+      "loss": 2.0347,
+      "step": 1280
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001597305908024797,
+      "loss": 2.0261,
+      "step": 1285
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00015939864264126393,
+      "loss": 2.0282,
+      "step": 1290
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00015906568005510919,
+      "loss": 2.0107,
+      "step": 1295
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00015873170873043413,
+      "loss": 2.0284,
+      "step": 1300
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00015839673437088498,
+      "loss": 2.015,
+      "step": 1305
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.000158060762697238,
+      "loss": 2.0027,
+      "step": 1310
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001577237994473018,
+      "loss": 2.0183,
+      "step": 1315
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001573858503758194,
+      "loss": 2.0183,
+      "step": 1320
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0001570469212543699,
+      "loss": 2.012,
+      "step": 1325
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00015670701787127005,
+      "loss": 1.9962,
+      "step": 1330
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00015636614603147512,
+      "loss": 2.019,
+      "step": 1335
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00015602431155648007,
+      "loss": 2.0047,
+      "step": 1340
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00015568152028421988,
+      "loss": 2.0012,
+      "step": 1345
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00015533777806897003,
+      "loss": 1.9839,
+      "step": 1350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00015499309078124645,
+      "loss": 1.9936,
+      "step": 1355
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00015464746430770516,
+      "loss": 1.9975,
+      "step": 1360
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001543009045510419,
+      "loss": 2.0006,
+      "step": 1365
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00015395341742989124,
+      "loss": 2.0144,
+      "step": 1370
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00015360500887872557,
+      "loss": 1.9964,
+      "step": 1375
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00015325568484775358,
+      "loss": 1.9902,
+      "step": 1380
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00015290545130281885,
+      "loss": 2.0048,
+      "step": 1385
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001525543142252978,
+      "loss": 2.0061,
+      "step": 1390
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00015220227961199766,
+      "loss": 1.9854,
+      "step": 1395
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.000151849353475054,
+      "loss": 2.0033,
+      "step": 1400
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00015149554184182802,
+      "loss": 1.9844,
+      "step": 1405
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00015114085075480367,
+      "loss": 1.997,
+      "step": 1410
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001507852862714844,
+      "loss": 1.9905,
+      "step": 1415
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001504288544642898,
+      "loss": 2.0043,
+      "step": 1420
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015007156142045176,
+      "loss": 1.9792,
+      "step": 1425
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00014971341324191068,
+      "loss": 2.0062,
+      "step": 1430
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001493544160452111,
+      "loss": 1.9878,
+      "step": 1435
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00014899457596139729,
+      "loss": 1.9749,
+      "step": 1440
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00014863389913590864,
+      "loss": 1.98,
+      "step": 1445
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001482723917284745,
+      "loss": 1.9756,
+      "step": 1450
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00014791005991300924,
+      "loss": 1.9749,
+      "step": 1455
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001475469098775066,
+      "loss": 1.9647,
+      "step": 1460
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00014718294782393407,
+      "loss": 1.9799,
+      "step": 1465
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00014681817996812705,
+      "loss": 1.9697,
+      "step": 1470
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00014645261253968262,
+      "loss": 1.9613,
+      "step": 1475
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00014608625178185312,
+      "loss": 1.9812,
+      "step": 1480
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00014571910395143965,
+      "loss": 1.9796,
+      "step": 1485
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.000145351175318685,
+      "loss": 1.9723,
+      "step": 1490
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014498247216716682,
+      "loss": 1.9648,
+      "step": 1495
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014461300079369013,
+      "loss": 1.9464,
+      "step": 1500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014424276750817986,
+      "loss": 1.9698,
+      "step": 1505
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014387177863357307,
+      "loss": 1.9431,
+      "step": 1510
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014350004050571092,
+      "loss": 1.9648,
+      "step": 1515
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014312755947323052,
+      "loss": 1.9693,
+      "step": 1520
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001427543418974566,
+      "loss": 1.9606,
+      "step": 1525
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001423803941522925,
+      "loss": 1.9557,
+      "step": 1530
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014200572262411192,
+      "loss": 1.9674,
+      "step": 1535
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014163033371164917,
+      "loss": 1.9663,
+      "step": 1540
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014125423382589048,
+      "loss": 1.9341,
+      "step": 1545
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00014087742938996418,
+      "loss": 1.9535,
+      "step": 1550
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00014049992683903104,
+      "loss": 1.9382,
+      "step": 1555
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00014012173262017443,
+      "loss": 1.9495,
+      "step": 1560
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013974285319229025,
+      "loss": 1.9587,
+      "step": 1565
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001393632950259765,
+      "loss": 1.9475,
+      "step": 1570
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013898306460342295,
+      "loss": 1.96,
+      "step": 1575
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013860216841830018,
+      "loss": 1.9517,
+      "step": 1580
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013822061297564896,
+      "loss": 1.9415,
+      "step": 1585
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.000137838404791769,
+      "loss": 1.9566,
+      "step": 1590
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013745555039410763,
+      "loss": 1.9566,
+      "step": 1595
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013707205632114842,
+      "loss": 1.942,
+      "step": 1600
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0001366879291222995,
+      "loss": 1.9478,
+      "step": 1605
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013630317535778163,
+      "loss": 1.9668,
+      "step": 1610
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001359178015985163,
+      "loss": 1.932,
+      "step": 1615
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00013553181442601327,
+      "loss": 1.964,
+      "step": 1620
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00013514522043225855,
+      "loss": 1.9471,
+      "step": 1625
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00013475802621960137,
+      "loss": 1.9414,
+      "step": 1630
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001343702384006418,
+      "loss": 1.9402,
+      "step": 1635
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0001339818635981176,
+      "loss": 1.946,
+      "step": 1640
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00013359290844479118,
+      "loss": 1.943,
+      "step": 1645
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0001332033795833364,
+      "loss": 1.9344,
+      "step": 1650
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00013281328366622494,
+      "loss": 1.9268,
+      "step": 1655
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00013242262735561292,
+      "loss": 1.938,
+      "step": 1660
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00013203141732322693,
+      "loss": 1.9305,
+      "step": 1665
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0001316396602502501,
+      "loss": 1.9316,
+      "step": 1670
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00013124736282720822,
+      "loss": 1.9243,
+      "step": 1675
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00013085453175385513,
+      "loss": 1.9441,
+      "step": 1680
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00013046117373905866,
+      "loss": 1.9184,
+      "step": 1685
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0001300672955006857,
+      "loss": 1.9521,
+      "step": 1690
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00012967290376548782,
+      "loss": 1.927,
+      "step": 1695
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00012927800526898604,
+      "loss": 1.9317,
+      "step": 1700
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00012888260675535622,
+      "loss": 1.9251,
+      "step": 1705
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00012848671497731336,
+      "loss": 1.9138,
+      "step": 1710
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001280903366959968,
+      "loss": 1.9152,
+      "step": 1715
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00012769347868085427,
+      "loss": 1.9302,
+      "step": 1720
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0001272961477095267,
+      "loss": 1.9222,
+      "step": 1725
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00012689835056773228,
+      "loss": 1.9256,
+      "step": 1730
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0001265000940491504,
+      "loss": 1.9017,
+      "step": 1735
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00012610138495530599,
+      "loss": 1.9096,
+      "step": 1740
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00012570223009545308,
+      "loss": 1.9122,
+      "step": 1745
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00012530263628645868,
+      "loss": 1.9296,
+      "step": 1750
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00012490261035268612,
+      "loss": 1.8998,
+      "step": 1755
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001245021591258789,
+      "loss": 1.9104,
+      "step": 1760
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00012410128944504359,
+      "loss": 1.9083,
+      "step": 1765
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00012370000815633327,
+      "loss": 1.918,
+      "step": 1770
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00012329832211293069,
+      "loss": 1.8937,
+      "step": 1775
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00012289623817493087,
+      "loss": 1.912,
+      "step": 1780
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00012249376320922436,
+      "loss": 1.9008,
+      "step": 1785
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00012209090408937971,
+      "loss": 1.9204,
+      "step": 1790
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00012168766769552614,
+      "loss": 1.9103,
+      "step": 1795
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00012128406091423614,
+      "loss": 1.8958,
+      "step": 1800
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00012088009063840762,
+      "loss": 1.8986,
+      "step": 1805
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00012047576376714647,
+      "loss": 1.8898,
+      "step": 1810
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00012007108720564853,
+      "loss": 1.9027,
+      "step": 1815
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00011966606786508182,
+      "loss": 1.905,
+      "step": 1820
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00011926071266246826,
+      "loss": 1.9051,
+      "step": 1825
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00011885502852056594,
+      "loss": 1.9067,
+      "step": 1830
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00011844902236775044,
+      "loss": 1.8823,
+      "step": 1835
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0001180427011378969,
+      "loss": 1.8949,
+      "step": 1840
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00011763607177026131,
+      "loss": 1.8945,
+      "step": 1845
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0001172291412093621,
+      "loss": 1.9152,
+      "step": 1850
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00011682191640486169,
+      "loss": 1.8951,
+      "step": 1855
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0001164144043114475,
+      "loss": 1.9003,
+      "step": 1860
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00011600661188871354,
+      "loss": 1.8892,
+      "step": 1865
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00011559854610104106,
+      "loss": 1.8922,
+      "step": 1870
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00011519021391748023,
+      "loss": 1.9047,
+      "step": 1875
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00011478162231163061,
+      "loss": 1.8902,
+      "step": 1880
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00011437277826152224,
+      "loss": 1.8895,
+      "step": 1885
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00011396368874949652,
+      "loss": 1.8839,
+      "step": 1890
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00011355436076208687,
+      "loss": 1.8859,
+      "step": 1895
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00011314480128989953,
+      "loss": 1.8897,
+      "step": 1900
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00011273501732749398,
+      "loss": 1.882,
+      "step": 1905
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00011232501587326374,
+      "loss": 1.8906,
+      "step": 1910
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00011191480392931655,
+      "loss": 1.8743,
+      "step": 1915
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00011150438850135508,
+      "loss": 1.8777,
+      "step": 1920
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00011109377659855707,
+      "loss": 1.882,
+      "step": 1925
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00011068297523345573,
+      "loss": 1.8653,
+      "step": 1930
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00011027199142181998,
+      "loss": 1.884,
+      "step": 1935
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010986083218253456,
+      "loss": 1.892,
+      "step": 1940
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00010944950453748029,
+      "loss": 1.871,
+      "step": 1945
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010903801551141395,
+      "loss": 1.8827,
+      "step": 1950
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00010862637213184852,
+      "loss": 1.8757,
+      "step": 1955
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00010821458142893301,
+      "loss": 1.8614,
+      "step": 1960
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0001078026504353325,
+      "loss": 1.8783,
+      "step": 1965
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010739058618610798,
+      "loss": 1.8757,
+      "step": 1970
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00010697839571859616,
+      "loss": 1.8664,
+      "step": 1975
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00010656608607228941,
+      "loss": 1.856,
+      "step": 1980
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.82151198387146,
+      "eval_runtime": 106.7172,
+      "eval_samples_per_second": 124.807,
+      "eval_steps_per_second": 15.602,
+      "step": 1980
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00010615366428871542,
+      "loss": 2.1449,
+      "step": 1985
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00010574113741131704,
+      "loss": 1.7986,
+      "step": 1990
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00010532851248533186,
+      "loss": 1.7705,
+      "step": 1995
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00010491579655767203,
+      "loss": 1.7827,
+      "step": 2000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00010450299667680382,
+      "loss": 1.8033,
+      "step": 2005
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00010409011989262725,
+      "loss": 1.7888,
+      "step": 2010
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00010367717325635583,
+      "loss": 1.775,
+      "step": 2015
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00010326416382039588,
+      "loss": 1.7692,
+      "step": 2020
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00010285109863822631,
+      "loss": 1.7874,
+      "step": 2025
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00010243798476427801,
+      "loss": 1.7889,
+      "step": 2030
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00010202482925381358,
+      "loss": 1.7943,
+      "step": 2035
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00010161163916280654,
+      "loss": 1.7865,
+      "step": 2040
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00010119842154782104,
+      "loss": 1.7823,
+      "step": 2045
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00010078518346589134,
+      "loss": 1.7958,
+      "step": 2050
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00010037193197440118,
+      "loss": 1.7666,
+      "step": 2055
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.995867413096333e-05,
+      "loss": 1.7981,
+      "step": 2060
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.954541699329904e-05,
+      "loss": 1.7924,
+      "step": 2065
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 9.913216761911755e-05,
+      "loss": 1.7811,
+      "step": 2070
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.871893306599545e-05,
+      "loss": 1.7969,
+      "step": 2075
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.830572039125623e-05,
+      "loss": 1.7927,
+      "step": 2080
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.789253665184971e-05,
+      "loss": 1.7776,
+      "step": 2085
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.747938890423163e-05,
+      "loss": 1.7733,
+      "step": 2090
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.706628420424298e-05,
+      "loss": 1.7968,
+      "step": 2095
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.665322960698957e-05,
+      "loss": 1.79,
+      "step": 2100
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.624023216672161e-05,
+      "loss": 1.7775,
+      "step": 2105
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.58272989367131e-05,
+      "loss": 1.777,
+      "step": 2110
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.541443696914142e-05,
+      "loss": 1.7659,
+      "step": 2115
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.500165331496705e-05,
+      "loss": 1.792,
+      "step": 2120
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.458895502381284e-05,
+      "loss": 1.7792,
+      "step": 2125
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.41763491438439e-05,
+      "loss": 1.7761,
+      "step": 2130
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.376384272164707e-05,
+      "loss": 1.7733,
+      "step": 2135
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.335144280211066e-05,
+      "loss": 1.7803,
+      "step": 2140
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.293915642830407e-05,
+      "loss": 1.7798,
+      "step": 2145
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.252699064135758e-05,
+      "loss": 1.7952,
+      "step": 2150
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.211495248034201e-05,
+      "loss": 1.7811,
+      "step": 2155
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.170304898214854e-05,
+      "loss": 1.7879,
+      "step": 2160
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.129128718136867e-05,
+      "loss": 1.7709,
+      "step": 2165
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.087967411017375e-05,
+      "loss": 1.7762,
+      "step": 2170
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.046821679819527e-05,
+      "loss": 1.7718,
+      "step": 2175
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.00569222724045e-05,
+      "loss": 1.7775,
+      "step": 2180
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.964579755699271e-05,
+      "loss": 1.78,
+      "step": 2185
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 8.923484967325104e-05,
+      "loss": 1.7741,
+      "step": 2190
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.882408563945067e-05,
+      "loss": 1.7617,
+      "step": 2195
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.841351247072296e-05,
+      "loss": 1.774,
+      "step": 2200
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.800313717893957e-05,
+      "loss": 1.7651,
+      "step": 2205
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.759296677259291e-05,
+      "loss": 1.7654,
+      "step": 2210
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.718300825667611e-05,
+      "loss": 1.7775,
+      "step": 2215
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 8.677326863256372e-05,
+      "loss": 1.7896,
+      "step": 2220
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.636375489789192e-05,
+      "loss": 1.7585,
+      "step": 2225
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.59544740464392e-05,
+      "loss": 1.7678,
+      "step": 2230
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.554543306800668e-05,
+      "loss": 1.759,
+      "step": 2235
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 8.513663894829898e-05,
+      "loss": 1.7755,
+      "step": 2240
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.472809866880475e-05,
+      "loss": 1.7686,
+      "step": 2245
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.431981920667749e-05,
+      "loss": 1.7535,
+      "step": 2250
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.391180753461645e-05,
+      "loss": 1.7704,
+      "step": 2255
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.350407062074739e-05,
+      "loss": 1.774,
+      "step": 2260
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.309661542850375e-05,
+      "loss": 1.7705,
+      "step": 2265
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 8.268944891650767e-05,
+      "loss": 1.7469,
+      "step": 2270
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.228257803845113e-05,
+      "loss": 1.7598,
+      "step": 2275
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.187600974297714e-05,
+      "loss": 1.7683,
+      "step": 2280
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.146975097356117e-05,
+      "loss": 1.7595,
+      "step": 2285
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 8.10638086683926e-05,
+      "loss": 1.7602,
+      "step": 2290
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.065818976025606e-05,
+      "loss": 1.7545,
+      "step": 2295
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.02529011764131e-05,
+      "loss": 1.7785,
+      "step": 2300
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.984794983848399e-05,
+      "loss": 1.7672,
+      "step": 2305
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 7.944334266232938e-05,
+      "loss": 1.7786,
+      "step": 2310
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 7.903908655793224e-05,
+      "loss": 1.7594,
+      "step": 2315
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 7.863518842927993e-05,
+      "loss": 1.7517,
+      "step": 2320
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 7.823165517424609e-05,
+      "loss": 1.7723,
+      "step": 2325
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 7.782849368447301e-05,
+      "loss": 1.7713,
+      "step": 2330
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 7.74257108452539e-05,
+      "loss": 1.7536,
+      "step": 2335
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 7.702331353541529e-05,
+      "loss": 1.7412,
+      "step": 2340
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 7.662130862719944e-05,
+      "loss": 1.7424,
+      "step": 2345
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 7.621970298614717e-05,
+      "loss": 1.7537,
+      "step": 2350
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 7.581850347098052e-05,
+      "loss": 1.7487,
+      "step": 2355
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 7.541771693348556e-05,
+      "loss": 1.777,
+      "step": 2360
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 7.501735021839549e-05,
+      "loss": 1.7651,
+      "step": 2365
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 7.461741016327359e-05,
+      "loss": 1.7729,
+      "step": 2370
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 7.421790359839657e-05,
+      "loss": 1.7481,
+      "step": 2375
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 7.381883734663797e-05,
+      "loss": 1.7398,
+      "step": 2380
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 7.342021822335143e-05,
+      "loss": 1.7416,
+      "step": 2385
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 7.302205303625444e-05,
+      "loss": 1.7398,
+      "step": 2390
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 7.262434858531208e-05,
+      "loss": 1.7459,
+      "step": 2395
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 7.222711166262089e-05,
+      "loss": 1.7491,
+      "step": 2400
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 7.183034905229279e-05,
+      "loss": 1.7654,
+      "step": 2405
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 7.143406753033933e-05,
+      "loss": 1.7568,
+      "step": 2410
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 7.103827386455586e-05,
+      "loss": 1.7451,
+      "step": 2415
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 7.0642974814406e-05,
+      "loss": 1.7537,
+      "step": 2420
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 7.024817713090628e-05,
+      "loss": 1.7355,
+      "step": 2425
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 6.985388755651069e-05,
+      "loss": 1.7415,
+      "step": 2430
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 6.946011282499558e-05,
+      "loss": 1.73,
+      "step": 2435
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 6.90668596613447e-05,
+      "loss": 1.7344,
+      "step": 2440
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 6.867413478163445e-05,
+      "loss": 1.7402,
+      "step": 2445
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 6.82819448929189e-05,
+      "loss": 1.7299,
+      "step": 2450
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 6.789029669311551e-05,
+      "loss": 1.7505,
+      "step": 2455
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 6.74991968708906e-05,
+      "loss": 1.7365,
+      "step": 2460
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 6.710865210554518e-05,
+      "loss": 1.7553,
+      "step": 2465
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 6.671866906690085e-05,
+      "loss": 1.7333,
+      "step": 2470
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.632925441518593e-05,
+      "loss": 1.7383,
+      "step": 2475
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 6.59404148009216e-05,
+      "loss": 1.7512,
+      "step": 2480
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.555215686480847e-05,
+      "loss": 1.7351,
+      "step": 2485
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 6.516448723761315e-05,
+      "loss": 1.7418,
+      "step": 2490
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.477741254005484e-05,
+      "loss": 1.7204,
+      "step": 2495
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 6.43909393826925e-05,
+      "loss": 1.7271,
+      "step": 2500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.400507436581178e-05,
+      "loss": 1.7325,
+      "step": 2505
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 6.361982407931233e-05,
+      "loss": 1.7285,
+      "step": 2510
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 6.323519510259535e-05,
+      "loss": 1.7424,
+      "step": 2515
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 6.285119400445112e-05,
+      "loss": 1.7481,
+      "step": 2520
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 6.246782734294683e-05,
+      "loss": 1.7494,
+      "step": 2525
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 6.208510166531458e-05,
+      "loss": 1.7261,
+      "step": 2530
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 6.170302350783969e-05,
+      "loss": 1.7314,
+      "step": 2535
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 6.132159939574889e-05,
+      "loss": 1.7328,
+      "step": 2540
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 6.094083584309893e-05,
+      "loss": 1.747,
+      "step": 2545
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 6.056073935266547e-05,
+      "loss": 1.7404,
+      "step": 2550
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 6.01813164158318e-05,
+      "loss": 1.7293,
+      "step": 2555
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.980257351247818e-05,
+      "loss": 1.7255,
+      "step": 2560
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 5.942451711087102e-05,
+      "loss": 1.737,
+      "step": 2565
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.904715366755251e-05,
+      "loss": 1.7361,
+      "step": 2570
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 5.867048962723029e-05,
+      "loss": 1.7338,
+      "step": 2575
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.8294531422667544e-05,
+      "loss": 1.7351,
+      "step": 2580
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 5.791928547457279e-05,
+      "loss": 1.7371,
+      "step": 2585
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.754475819149072e-05,
+      "loss": 1.7261,
+      "step": 2590
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 5.7170955969692265e-05,
+      "loss": 1.7324,
+      "step": 2595
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.67978851930657e-05,
+      "loss": 1.7284,
+      "step": 2600
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 5.6425552233007454e-05,
+      "loss": 1.7296,
+      "step": 2605
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.605396344831331e-05,
+      "loss": 1.723,
+      "step": 2610
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 5.5683125185069906e-05,
+      "loss": 1.7224,
+      "step": 2615
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.531304377654623e-05,
+      "loss": 1.717,
+      "step": 2620
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 5.4943725543085544e-05,
+      "loss": 1.739,
+      "step": 2625
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.457517679199736e-05,
+      "loss": 1.7183,
+      "step": 2630
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 5.420740381744991e-05,
+      "loss": 1.7357,
+      "step": 2635
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.384041290036239e-05,
+      "loss": 1.7197,
+      "step": 2640
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.3474210308297846e-05,
+      "loss": 1.715,
+      "step": 2645
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.3108802295356155e-05,
+      "loss": 1.7214,
+      "step": 2650
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.2744195102067127e-05,
+      "loss": 1.7103,
+      "step": 2655
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.2380394955283996e-05,
+      "loss": 1.7168,
+      "step": 2660
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.2017408068077064e-05,
+      "loss": 1.7367,
+      "step": 2665
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.1655240639627554e-05,
+      "loss": 1.728,
+      "step": 2670
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.129389885512175e-05,
+      "loss": 1.736,
+      "step": 2675
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.093338888564546e-05,
+      "loss": 1.7134,
+      "step": 2680
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.0573716888078494e-05,
+      "loss": 1.7038,
+      "step": 2685
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 5.021488900498954e-05,
+      "loss": 1.7214,
+      "step": 2690
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.9856911364531346e-05,
+      "loss": 1.7163,
+      "step": 2695
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.949979008033596e-05,
+      "loss": 1.7192,
+      "step": 2700
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.914353125141037e-05,
+      "loss": 1.7015,
+      "step": 2705
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.8788140962032345e-05,
+      "loss": 1.7358,
+      "step": 2710
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.8433625281646524e-05,
+      "loss": 1.7355,
+      "step": 2715
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.8079990264760686e-05,
+      "loss": 1.7125,
+      "step": 2720
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.7727241950842586e-05,
+      "loss": 1.7052,
+      "step": 2725
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.7375386364216465e-05,
+      "loss": 1.7299,
+      "step": 2730
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.7024429513960425e-05,
+      "loss": 1.7245,
+      "step": 2735
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.6674377393803715e-05,
+      "loss": 1.7031,
+      "step": 2740
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.632523598202435e-05,
+      "loss": 1.7232,
+      "step": 2745
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 4.597701124134708e-05,
+      "loss": 1.7184,
+      "step": 2750
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 4.562970911884148e-05,
+      "loss": 1.7215,
+      "step": 2755
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 4.528333554582044e-05,
+      "loss": 1.7049,
+      "step": 2760
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 4.493789643773881e-05,
+      "loss": 1.7042,
+      "step": 2765
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 4.459339769409252e-05,
+      "loss": 1.7046,
+      "step": 2770
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 4.4249845198317593e-05,
+      "loss": 1.717,
+      "step": 2775
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 4.3907244817689875e-05,
+      "loss": 1.7095,
+      "step": 2780
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 4.3565602403224695e-05,
+      "loss": 1.7151,
+      "step": 2785
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 4.3224923789577045e-05,
+      "loss": 1.6841,
+      "step": 2790
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 4.288521479494183e-05,
+      "loss": 1.7101,
+      "step": 2795
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 4.254648122095463e-05,
+      "loss": 1.7028,
+      "step": 2800
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 4.220872885259247e-05,
+      "loss": 1.7006,
+      "step": 2805
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 4.187196345807511e-05,
+      "loss": 1.717,
+      "step": 2810
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 4.153619078876664e-05,
+      "loss": 1.7117,
+      "step": 2815
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 4.120141657907698e-05,
+      "loss": 1.7073,
+      "step": 2820
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 4.086764654636419e-05,
+      "loss": 1.6958,
+      "step": 2825
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 4.0534886390836715e-05,
+      "loss": 1.7087,
+      "step": 2830
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 4.020314179545609e-05,
+      "loss": 1.7078,
+      "step": 2835
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.987241842583983e-05,
+      "loss": 1.69,
+      "step": 2840
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.9542721930164726e-05,
+      "loss": 1.7025,
+      "step": 2845
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.9214057939070324e-05,
+      "loss": 1.6981,
+      "step": 2850
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 3.888643206556285e-05,
+      "loss": 1.7104,
+      "step": 2855
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.8559849904919196e-05,
+      "loss": 1.7035,
+      "step": 2860
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 3.823431703459164e-05,
+      "loss": 1.6963,
+      "step": 2865
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.790983901411225e-05,
+      "loss": 1.728,
+      "step": 2870
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 3.758642138499819e-05,
+      "loss": 1.702,
+      "step": 2875
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.7264069670656886e-05,
+      "loss": 1.7027,
+      "step": 2880
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.694278937629197e-05,
+      "loss": 1.7041,
+      "step": 2885
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.6622585988809e-05,
+      "loss": 1.6936,
+      "step": 2890
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.6303464976721834e-05,
+      "loss": 1.711,
+      "step": 2895
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.5985431790059276e-05,
+      "loss": 1.6902,
+      "step": 2900
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.5668491860271945e-05,
+      "loss": 1.6984,
+      "step": 2905
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.535265060013965e-05,
+      "loss": 1.7125,
+      "step": 2910
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 3.50379134036787e-05,
+      "loss": 1.7151,
+      "step": 2915
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.472428564605006e-05,
+      "loss": 1.7007,
+      "step": 2920
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 3.441177268346724e-05,
+      "loss": 1.6791,
+      "step": 2925
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.41003798531052e-05,
+      "loss": 1.6978,
+      "step": 2930
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 3.379011247300889e-05,
+      "loss": 1.6956,
+      "step": 2935
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.348097584200258e-05,
+      "loss": 1.6897,
+      "step": 2940
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.317297523959927e-05,
+      "loss": 1.7143,
+      "step": 2945
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.2866115925910615e-05,
+      "loss": 1.6969,
+      "step": 2950
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.2560403141557084e-05,
+      "loss": 1.6883,
+      "step": 2955
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.225584210757838e-05,
+      "loss": 1.6873,
+      "step": 2960
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.195243802534438e-05,
+      "loss": 1.701,
+      "step": 2965
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 3.165019607646611e-05,
+      "loss": 1.6864,
+      "step": 2970
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.6935465335845947,
+      "eval_runtime": 113.7484,
+      "eval_samples_per_second": 117.092,
+      "eval_steps_per_second": 14.638,
+      "step": 2970
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.134912142270757e-05,
+      "loss": 1.9367,
+      "step": 2975
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 3.104921920589733e-05,
+      "loss": 1.6312,
+      "step": 2980
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.075049454784074e-05,
+      "loss": 1.6483,
+      "step": 2985
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 3.0452952550232562e-05,
+      "loss": 1.6396,
+      "step": 2990
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 3.0156598294569727e-05,
+      "loss": 1.6366,
+      "step": 2995
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.9861436842064727e-05,
+      "loss": 1.642,
+      "step": 3000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.956747323355895e-05,
+      "loss": 1.6344,
+      "step": 3005
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.9274712489436785e-05,
+      "loss": 1.6359,
+      "step": 3010
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.8983159609539635e-05,
+      "loss": 1.6309,
+      "step": 3015
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.8692819573080943e-05,
+      "loss": 1.6433,
+      "step": 3020
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.840369733856072e-05,
+      "loss": 1.6478,
+      "step": 3025
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.8115797843681124e-05,
+      "loss": 1.6372,
+      "step": 3030
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.7829126005262042e-05,
+      "loss": 1.638,
+      "step": 3035
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.7543686719157102e-05,
+      "loss": 1.6308,
+      "step": 3040
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.7259484860170203e-05,
+      "loss": 1.6396,
+      "step": 3045
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.6976525281972078e-05,
+      "loss": 1.6514,
+      "step": 3050
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.669481281701739e-05,
+      "loss": 1.6459,
+      "step": 3055
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.6414352276462417e-05,
+      "loss": 1.6357,
+      "step": 3060
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.613514845008275e-05,
+      "loss": 1.6362,
+      "step": 3065
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.5857206106191447e-05,
+      "loss": 1.6355,
+      "step": 3070
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.558052999155768e-05,
+      "loss": 1.6385,
+      "step": 3075
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.530512483132561e-05,
+      "loss": 1.6521,
+      "step": 3080
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.5030995328933726e-05,
+      "loss": 1.629,
+      "step": 3085
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.475814616603458e-05,
+      "loss": 1.6398,
+      "step": 3090
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.4486582002414716e-05,
+      "loss": 1.6354,
+      "step": 3095
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.4216307475915066e-05,
+      "loss": 1.6431,
+      "step": 3100
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.394732720235189e-05,
+      "loss": 1.6334,
+      "step": 3105
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.367964577543792e-05,
+      "loss": 1.6276,
+      "step": 3110
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.3413267766703773e-05,
+      "loss": 1.6438,
+      "step": 3115
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.3148197725419983e-05,
+      "loss": 1.6164,
+      "step": 3120
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 2.2884440178519305e-05,
+      "loss": 1.631,
+      "step": 3125
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 2.262199963051934e-05,
+      "loss": 1.6395,
+      "step": 3130
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 2.2360880563445764e-05,
+      "loss": 1.6389,
+      "step": 3135
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 2.2101087436755573e-05,
+      "loss": 1.63,
+      "step": 3140
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.184262468726098e-05,
+      "loss": 1.6276,
+      "step": 3145
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.1585496729053768e-05,
+      "loss": 1.6216,
+      "step": 3150
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 2.1329707953429822e-05,
+      "loss": 1.6282,
+      "step": 3155
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 2.1075262728814094e-05,
+      "loss": 1.6315,
+      "step": 3160
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 2.0822165400686035e-05,
+      "loss": 1.6443,
+      "step": 3165
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 2.0570420291505376e-05,
+      "loss": 1.6053,
+      "step": 3170
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.0320031700638352e-05,
+      "loss": 1.6289,
+      "step": 3175
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.0071003904284158e-05,
+      "loss": 1.6252,
+      "step": 3180
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.9823341155402153e-05,
+      "loss": 1.6193,
+      "step": 3185
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.9577047683638873e-05,
+      "loss": 1.6257,
+      "step": 3190
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.933212769525613e-05,
+      "loss": 1.6297,
+      "step": 3195
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.9088585373058976e-05,
+      "loss": 1.6248,
+      "step": 3200
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.884642487632442e-05,
+      "loss": 1.6381,
+      "step": 3205
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.8605650340730228e-05,
+      "loss": 1.6377,
+      "step": 3210
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.8366265878284382e-05,
+      "loss": 1.6152,
+      "step": 3215
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.8128275577254884e-05,
+      "loss": 1.6193,
+      "step": 3220
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.789168350209983e-05,
+      "loss": 1.6174,
+      "step": 3225
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.765649369339819e-05,
+      "loss": 1.6376,
+      "step": 3230
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.742271016778051e-05,
+      "loss": 1.622,
+      "step": 3235
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.7190336917860594e-05,
+      "loss": 1.63,
+      "step": 3240
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.695937791216716e-05,
+      "loss": 1.6347,
+      "step": 3245
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.672983709507615e-05,
+      "loss": 1.6179,
+      "step": 3250
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.650171838674328e-05,
+      "loss": 1.6227,
+      "step": 3255
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.6275025683037148e-05,
+      "loss": 1.6331,
+      "step": 3260
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.6049762855472685e-05,
+      "loss": 1.6197,
+      "step": 3265
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.5825933751145027e-05,
+      "loss": 1.6323,
+      "step": 3270
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.56035421926639e-05,
+      "loss": 1.6208,
+      "step": 3275
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.5382591978088144e-05,
+      "loss": 1.6216,
+      "step": 3280
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.5163086880861043e-05,
+      "loss": 1.6305,
+      "step": 3285
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.494503064974575e-05,
+      "loss": 1.6324,
+      "step": 3290
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.4728427008761402e-05,
+      "loss": 1.6317,
+      "step": 3295
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.4513279657119371e-05,
+      "loss": 1.6243,
+      "step": 3300
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.4299592269160166e-05,
+      "loss": 1.6287,
+      "step": 3305
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.4087368494290677e-05,
+      "loss": 1.6318,
+      "step": 3310
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.3876611956921793e-05,
+      "loss": 1.6273,
+      "step": 3315
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.3667326256406699e-05,
+      "loss": 1.6143,
+      "step": 3320
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.345951496697907e-05,
+      "loss": 1.6209,
+      "step": 3325
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.3253181637692324e-05,
+      "loss": 1.6212,
+      "step": 3330
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.3048329792358883e-05,
+      "loss": 1.6277,
+      "step": 3335
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.2844962929490045e-05,
+      "loss": 1.6418,
+      "step": 3340
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.264308452223616e-05,
+      "loss": 1.6343,
+      "step": 3345
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.244269801832737e-05,
+      "loss": 1.6291,
+      "step": 3350
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.2243806840014726e-05,
+      "loss": 1.6088,
+      "step": 3355
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.2046414384011718e-05,
+      "loss": 1.6069,
+      "step": 3360
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.1850524021436337e-05,
+      "loss": 1.6121,
+      "step": 3365
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.1656139097753348e-05,
+      "loss": 1.6305,
+      "step": 3370
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.1463262932717323e-05,
+      "loss": 1.6394,
+      "step": 3375
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.127189882031584e-05,
+      "loss": 1.6238,
+      "step": 3380
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.1082050028713309e-05,
+      "loss": 1.6294,
+      "step": 3385
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0893719800195057e-05,
+      "loss": 1.622,
+      "step": 3390
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0706911351112037e-05,
+      "loss": 1.6231,
+      "step": 3395
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.052162787182588e-05,
+      "loss": 1.6262,
+      "step": 3400
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0337872526654358e-05,
+      "loss": 1.6081,
+      "step": 3405
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0155648453817435e-05,
+      "loss": 1.6253,
+      "step": 3410
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.974958765383591e-06,
+      "loss": 1.6118,
+      "step": 3415
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.795806547216701e-06,
+      "loss": 1.6339,
+      "step": 3420
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.618194858923346e-06,
+      "loss": 1.6173,
+      "step": 3425
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.442126733800572e-06,
+      "loss": 1.6203,
+      "step": 3430
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.267605178784033e-06,
+      "loss": 1.6263,
+      "step": 3435
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.094633174396694e-06,
+      "loss": 1.6188,
+      "step": 3440
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 8.923213674697884e-06,
+      "loss": 1.6063,
+      "step": 3445
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 8.753349607232864e-06,
+      "loss": 1.6181,
+      "step": 3450
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.585043872982868e-06,
+      "loss": 1.6265,
+      "step": 3455
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.418299346315483e-06,
+      "loss": 1.6186,
+      "step": 3460
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.253118874935629e-06,
+      "loss": 1.6172,
+      "step": 3465
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.089505279836873e-06,
+      "loss": 1.622,
+      "step": 3470
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 7.927461355253307e-06,
+      "loss": 1.5994,
+      "step": 3475
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 7.766989868611774e-06,
+      "loss": 1.6239,
+      "step": 3480
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 7.60809356048463e-06,
+      "loss": 1.6292,
+      "step": 3485
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 7.4507751445429315e-06,
+      "loss": 1.6292,
+      "step": 3490
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.2950373075101e-06,
+      "loss": 1.6154,
+      "step": 3495
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.140882709116048e-06,
+      "loss": 1.6178,
+      "step": 3500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 6.98831398205172e-06,
+      "loss": 1.6349,
+      "step": 3505
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 6.8373337319241555e-06,
+      "loss": 1.6319,
+      "step": 3510
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 6.687944537211988e-06,
+      "loss": 1.6126,
+      "step": 3515
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 6.540148949221403e-06,
+      "loss": 1.6207,
+      "step": 3520
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 6.393949492042606e-06,
+      "loss": 1.6072,
+      "step": 3525
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 6.249348662506627e-06,
+      "loss": 1.6298,
+      "step": 3530
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 6.106348930142758e-06,
+      "loss": 1.6197,
+      "step": 3535
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.964952737136353e-06,
+      "loss": 1.6168,
+      "step": 3540
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 5.825162498287096e-06,
+      "loss": 1.6102,
+      "step": 3545
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.686980600967817e-06,
+      "loss": 1.6184,
+      "step": 3550
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.550409405083657e-06,
+      "loss": 1.6228,
+      "step": 3555
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.415451243031811e-06,
+      "loss": 1.6221,
+      "step": 3560
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 5.282108419661646e-06,
+      "loss": 1.6303,
+      "step": 3565
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.150383212235421e-06,
+      "loss": 1.615,
+      "step": 3570
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 5.020277870389312e-06,
+      "loss": 1.6333,
+      "step": 3575
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.891794616095025e-06,
+      "loss": 1.6241,
+      "step": 3580
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.764935643621848e-06,
+      "loss": 1.6325,
+      "step": 3585
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.639703119499172e-06,
+      "loss": 1.6164,
+      "step": 3590
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.516099182479505e-06,
+      "loss": 1.6207,
+      "step": 3595
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.394125943501936e-06,
+      "loss": 1.6134,
+      "step": 3600
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.273785485656068e-06,
+      "loss": 1.6231,
+      "step": 3605
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.1550798641464605e-06,
+      "loss": 1.6069,
+      "step": 3610
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.038011106257555e-06,
+      "loss": 1.6091,
+      "step": 3615
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.922581211318999e-06,
+      "loss": 1.62,
+      "step": 3620
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.808792150671525e-06,
+      "loss": 1.6251,
+      "step": 3625
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.6966458676332973e-06,
+      "loss": 1.6327,
+      "step": 3630
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 3.5861442774667074e-06,
+      "loss": 1.6004,
+      "step": 3635
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.4772892673456626e-06,
+      "loss": 1.6086,
+      "step": 3640
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.3700826963233735e-06,
+      "loss": 1.625,
+      "step": 3645
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.2645263953005933e-06,
+      "loss": 1.6158,
+      "step": 3650
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 3.1606221669943206e-06,
+      "loss": 1.608,
+      "step": 3655
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 3.0583717859070883e-06,
+      "loss": 1.6005,
+      "step": 3660
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.9577769982965908e-06,
+      "loss": 1.6166,
+      "step": 3665
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 2.858839522145873e-06,
+      "loss": 1.6157,
+      "step": 3670
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 2.7615610471340093e-06,
+      "loss": 1.6033,
+      "step": 3675
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 2.6659432346072156e-06,
+      "loss": 1.6193,
+      "step": 3680
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 2.571987717550517e-06,
+      "loss": 1.62,
+      "step": 3685
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.4796961005598363e-06,
+      "loss": 1.619,
+      "step": 3690
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.389069959814594e-06,
+      "loss": 1.6346,
+      "step": 3695
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 2.3001108430507735e-06,
+      "loss": 1.6115,
+      "step": 3700
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 2.2128202695345324e-06,
+      "loss": 1.6129,
+      "step": 3705
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 2.1271997300362003e-06,
+      "loss": 1.5998,
+      "step": 3710
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 2.043250686804865e-06,
+      "loss": 1.6352,
+      "step": 3715
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.9609745735433614e-06,
+      "loss": 1.6213,
+      "step": 3720
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.8803727953838335e-06,
+      "loss": 1.6163,
+      "step": 3725
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.8014467288636872e-06,
+      "loss": 1.618,
+      "step": 3730
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.7241977219020988e-06,
+      "loss": 1.6189,
+      "step": 3735
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.6486270937770332e-06,
+      "loss": 1.6139,
+      "step": 3740
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.574736135102639e-06,
+      "loss": 1.6163,
+      "step": 3745
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.5025261078073005e-06,
+      "loss": 1.6152,
+      "step": 3750
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.4319982451119652e-06,
+      "loss": 1.6289,
+      "step": 3755
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.3631537515092163e-06,
+      "loss": 1.6263,
+      "step": 3760
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.29599380274259e-06,
+      "loss": 1.6108,
+      "step": 3765
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.2305195457865681e-06,
+      "loss": 1.6016,
+      "step": 3770
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.1667320988269614e-06,
+      "loss": 1.6129,
+      "step": 3775
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.104632551241802e-06,
+      "loss": 1.6075,
+      "step": 3780
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.0442219635827587e-06,
+      "loss": 1.6188,
+      "step": 3785
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.85501367557029e-07,
+      "loss": 1.6018,
+      "step": 3790
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.284717660097086e-07,
+      "loss": 1.6072,
+      "step": 3795
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.7313413290665e-07,
+      "loss": 1.6053,
+      "step": 3800
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 8.194894133178532e-07,
+      "loss": 1.6125,
+      "step": 3805
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.675385234013011e-07,
+      "loss": 1.616,
+      "step": 3810
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 7.172823503873605e-07,
+      "loss": 1.6151,
+      "step": 3815
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.687217525635614e-07,
+      "loss": 1.637,
+      "step": 3820
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 6.218575592600084e-07,
+      "loss": 1.6135,
+      "step": 3825
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.766905708351478e-07,
+      "loss": 1.6159,
+      "step": 3830
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 5.332215586621669e-07,
+      "loss": 1.6136,
+      "step": 3835
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.91451265115761e-07,
+      "loss": 1.6354,
+      "step": 3840
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.513804035595093e-07,
+      "loss": 1.6142,
+      "step": 3845
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.130096583336407e-07,
+      "loss": 1.6209,
+      "step": 3850
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 3.763396847433875e-07,
+      "loss": 1.6133,
+      "step": 3855
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.413711090477722e-07,
+      "loss": 1.6025,
+      "step": 3860
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.081045284488937e-07,
+      "loss": 1.6299,
+      "step": 3865
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.76540511081802e-07,
+      "loss": 1.624,
+      "step": 3870
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.4667959600468414e-07,
+      "loss": 1.6148,
+      "step": 3875
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.1852229318977125e-07,
+      "loss": 1.6089,
+      "step": 3880
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.9206908351453446e-07,
+      "loss": 1.6341,
+      "step": 3885
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.6732041875354709e-07,
+      "loss": 1.6184,
+      "step": 3890
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.4427672157072414e-07,
+      "loss": 1.5931,
+      "step": 3895
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.229383855121058e-07,
+      "loss": 1.605,
+      "step": 3900
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.0330577499915173e-07,
+      "loss": 1.6359,
+      "step": 3905
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 8.537922532252384e-08,
+      "loss": 1.5941,
+      "step": 3910
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.915904263634643e-08,
+      "loss": 1.6207,
+      "step": 3915
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 5.464550395295476e-08,
+      "loss": 1.6085,
+      "step": 3920
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.183885713822111e-08,
+      "loss": 1.6145,
+      "step": 3925
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 3.0739320907269276e-08,
+      "loss": 1.6102,
+      "step": 3930
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 2.134708482077752e-08,
+      "loss": 1.6245,
+      "step": 3935
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.3662309281714525e-08,
+      "loss": 1.6035,
+      "step": 3940
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 7.685125532586047e-09,
+      "loss": 1.6038,
+      "step": 3945
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.415635653258864e-09,
+      "loss": 1.6323,
+      "step": 3950
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.539125591511266e-10,
+      "loss": 1.6162,
+      "step": 3955
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.0,
+      "loss": 1.6228,
+      "step": 3960
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.66698157787323,
+      "eval_runtime": 111.3625,
+      "eval_samples_per_second": 119.6,
+      "eval_steps_per_second": 14.951,
+      "step": 3960
+    },
+    {
+      "epoch": 4.0,
+      "step": 3960,
+      "total_flos": 9.454097668453171e+17,
+      "train_loss": 1.9931427570304485,
+      "train_runtime": 30194.781,
+      "train_samples_per_second": 33.608,
+      "train_steps_per_second": 0.131
+    }
+  ],
+  "max_steps": 3960,
+  "num_train_epochs": 4,
+  "total_flos": 9.454097668453171e+17,
+  "trial_name": null,
+  "trial_params": null
+}