diff --git "a/checkpoint-15743/trainer_state.json" "b/checkpoint-15743/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-15743/trainer_state.json"
@@ -0,0 +1,11052 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.021716375513544422,
+  "eval_steps": 500,
+  "global_step": 15743,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002922796166022129,
+      "grad_norm": 0.3604201674461365,
+      "learning_rate": 1.8e-06,
+      "loss": 0.6163836479187011,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005845592332044258,
+      "grad_norm": 0.307494193315506,
+      "learning_rate": 3.8e-06,
+      "loss": 0.6056640148162842,
+      "step": 20
+    },
+    {
+      "epoch": 0.00014613980830110646,
+      "grad_norm": 0.1861964464187622,
+      "learning_rate": 5.8e-06,
+      "loss": 0.5507771968841553,
+      "step": 30
+    },
+    {
+      "epoch": 0.00011691184664088517,
+      "grad_norm": 0.18063770234584808,
+      "learning_rate": 7.8e-06,
+      "loss": 0.5524736642837524,
+      "step": 40
+    },
+    {
+      "epoch": 0.0004091914632430981,
+      "grad_norm": 0.1831495314836502,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.5550893783569336,
+      "step": 50
+    },
+    {
+      "epoch": 0.000701471079845311,
+      "grad_norm": 0.2708336412906647,
+      "learning_rate": 1.18e-05,
+      "loss": 0.5027066707611084,
+      "step": 60
+    },
+    {
+      "epoch": 0.0009937506964475238,
+      "grad_norm": 0.16506414115428925,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.5025710105895996,
+      "step": 70
+    },
+    {
+      "epoch": 0.0012860303130497368,
+      "grad_norm": 0.7222521305084229,
+      "learning_rate": 1.58e-05,
+      "loss": 0.44574794769287107,
+      "step": 80
+    },
+    {
+      "epoch": 0.0015783099296519498,
+      "grad_norm": 0.3094443678855896,
+      "learning_rate": 1.78e-05,
+      "loss": 0.463105583190918,
+      "step": 90
+    },
+    {
+      "epoch": 0.0018705895462541628,
+      "grad_norm": 0.22881925106048584,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.47495498657226565,
+      "step": 100
+    },
+    {
+      "epoch": 0.0021628691628563755,
+      "grad_norm": 0.2994537055492401,
+      "learning_rate": 2.18e-05,
+      "loss": 0.44906439781188967,
+      "step": 110
+    },
+    {
+      "epoch": 0.0024551487794585887,
+      "grad_norm": 0.22821807861328125,
+      "learning_rate": 2.38e-05,
+      "loss": 0.4511695384979248,
+      "step": 120
+    },
+    {
+      "epoch": 0.0027474283960608015,
+      "grad_norm": 0.23172526061534882,
+      "learning_rate": 2.58e-05,
+      "loss": 0.45236430168151853,
+      "step": 130
+    },
+    {
+      "epoch": 0.003039708012663014,
+      "grad_norm": 0.25378280878067017,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.49311389923095705,
+      "step": 140
+    },
+    {
+      "epoch": 0.0033319876292652274,
+      "grad_norm": 0.30715328454971313,
+      "learning_rate": 2.98e-05,
+      "loss": 0.45659761428833007,
+      "step": 150
+    },
+    {
+      "epoch": 0.00362426724586744,
+      "grad_norm": 0.2318553328514099,
+      "learning_rate": 3.18e-05,
+      "loss": 0.48353352546691897,
+      "step": 160
+    },
+    {
+      "epoch": 0.003916546862469653,
+      "grad_norm": 0.3422481417655945,
+      "learning_rate": 3.38e-05,
+      "loss": 0.48243284225463867,
+      "step": 170
+    },
+    {
+      "epoch": 0.004208826479071866,
+      "grad_norm": 0.2627033591270447,
+      "learning_rate": 3.58e-05,
+      "loss": 0.44968676567077637,
+      "step": 180
+    },
+    {
+      "epoch": 0.004501106095674079,
+      "grad_norm": 0.270560622215271,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.43050732612609866,
+      "step": 190
+    },
+    {
+      "epoch": 0.004793385712276292,
+      "grad_norm": 0.29916077852249146,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.38156795501708984,
+      "step": 200
+    },
+    {
+      "epoch": 0.005085665328878505,
+      "grad_norm": 0.29618600010871887,
+      "learning_rate": 4.18e-05,
+      "loss": 0.41876654624938964,
+      "step": 210
+    },
+    {
+      "epoch": 0.005377944945480718,
+      "grad_norm": 0.24618731439113617,
+      "learning_rate": 4.38e-05,
+      "loss": 0.4200754642486572,
+      "step": 220
+    },
+    {
+      "epoch": 0.00567022456208293,
+      "grad_norm": 0.4028846323490143,
+      "learning_rate": 4.58e-05,
+      "loss": 0.4128837585449219,
+      "step": 230
+    },
+    {
+      "epoch": 0.0059625041786851435,
+      "grad_norm": 0.2899196445941925,
+      "learning_rate": 4.78e-05,
+      "loss": 0.4727485656738281,
+      "step": 240
+    },
+    {
+      "epoch": 0.006254783795287357,
+      "grad_norm": 0.372092068195343,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.3837599754333496,
+      "step": 250
+    },
+    {
+      "epoch": 0.00654706341188957,
+      "grad_norm": 0.27367591857910156,
+      "learning_rate": 5.1800000000000005e-05,
+      "loss": 0.4533690929412842,
+      "step": 260
+    },
+    {
+      "epoch": 0.006839343028491782,
+      "grad_norm": 0.2826191782951355,
+      "learning_rate": 5.380000000000001e-05,
+      "loss": 0.5056231021881104,
+      "step": 270
+    },
+    {
+      "epoch": 0.007131622645093995,
+      "grad_norm": 0.34035593271255493,
+      "learning_rate": 5.580000000000001e-05,
+      "loss": 0.4258269309997559,
+      "step": 280
+    },
+    {
+      "epoch": 0.0074239022616962086,
+      "grad_norm": 0.2767459750175476,
+      "learning_rate": 5.7799999999999995e-05,
+      "loss": 0.3629124641418457,
+      "step": 290
+    },
+    {
+      "epoch": 0.007716181878298421,
+      "grad_norm": 0.2805218994617462,
+      "learning_rate": 5.9800000000000003e-05,
+      "loss": 0.35598738193511964,
+      "step": 300
+    },
+    {
+      "epoch": 0.008008461494900634,
+      "grad_norm": 0.32115164399147034,
+      "learning_rate": 6.18e-05,
+      "loss": 0.45623183250427246,
+      "step": 310
+    },
+    {
+      "epoch": 0.008300741111502847,
+      "grad_norm": 0.25464335083961487,
+      "learning_rate": 6.38e-05,
+      "loss": 0.3766467809677124,
+      "step": 320
+    },
+    {
+      "epoch": 0.00859302072810506,
+      "grad_norm": 0.240229070186615,
+      "learning_rate": 6.58e-05,
+      "loss": 0.4206503391265869,
+      "step": 330
+    },
+    {
+      "epoch": 0.008885300344707274,
+      "grad_norm": 0.24640053510665894,
+      "learning_rate": 6.780000000000001e-05,
+      "loss": 0.42534866333007815,
+      "step": 340
+    },
+    {
+      "epoch": 0.009177579961309485,
+      "grad_norm": 0.22595764696598053,
+      "learning_rate": 6.98e-05,
+      "loss": 0.4169299602508545,
+      "step": 350
+    },
+    {
+      "epoch": 0.009469859577911698,
+      "grad_norm": 0.2035064995288849,
+      "learning_rate": 7.18e-05,
+      "loss": 0.36454784870147705,
+      "step": 360
+    },
+    {
+      "epoch": 0.009762139194513911,
+      "grad_norm": 0.21908830106258392,
+      "learning_rate": 7.38e-05,
+      "loss": 0.39443838596343994,
+      "step": 370
+    },
+    {
+      "epoch": 0.010054418811116125,
+      "grad_norm": 0.2039889395236969,
+      "learning_rate": 7.58e-05,
+      "loss": 0.4549224853515625,
+      "step": 380
+    },
+    {
+      "epoch": 0.010346698427718338,
+      "grad_norm": 0.24606066942214966,
+      "learning_rate": 7.780000000000001e-05,
+      "loss": 0.3763291835784912,
+      "step": 390
+    },
+    {
+      "epoch": 0.010638978044320551,
+      "grad_norm": 0.4349507689476013,
+      "learning_rate": 7.98e-05,
+      "loss": 0.42160859107971194,
+      "step": 400
+    },
+    {
+      "epoch": 0.010931257660922762,
+      "grad_norm": 0.2086913138628006,
+      "learning_rate": 8.18e-05,
+      "loss": 0.4335451126098633,
+      "step": 410
+    },
+    {
+      "epoch": 0.011223537277524976,
+      "grad_norm": 0.21672461926937103,
+      "learning_rate": 8.38e-05,
+      "loss": 0.39824934005737306,
+      "step": 420
+    },
+    {
+      "epoch": 0.011515816894127189,
+      "grad_norm": 0.34380897879600525,
+      "learning_rate": 8.58e-05,
+      "loss": 0.4544211387634277,
+      "step": 430
+    },
+    {
+      "epoch": 0.011808096510729402,
+      "grad_norm": 0.2024194449186325,
+      "learning_rate": 8.78e-05,
+      "loss": 0.4060193061828613,
+      "step": 440
+    },
+    {
+      "epoch": 0.012100376127331615,
+      "grad_norm": 0.22563715279102325,
+      "learning_rate": 8.98e-05,
+      "loss": 0.3707286834716797,
+      "step": 450
+    },
+    {
+      "epoch": 0.012392655743933828,
+      "grad_norm": 0.20325998961925507,
+      "learning_rate": 9.180000000000001e-05,
+      "loss": 0.37896702289581297,
+      "step": 460
+    },
+    {
+      "epoch": 0.012684935360536042,
+      "grad_norm": 0.20927369594573975,
+      "learning_rate": 9.38e-05,
+      "loss": 0.4257774829864502,
+      "step": 470
+    },
+    {
+      "epoch": 0.012977214977138253,
+      "grad_norm": 0.2156561017036438,
+      "learning_rate": 9.58e-05,
+      "loss": 0.33562607765197755,
+      "step": 480
+    },
+    {
+      "epoch": 0.013269494593740466,
+      "grad_norm": 0.17501753568649292,
+      "learning_rate": 9.78e-05,
+      "loss": 0.4147501945495605,
+      "step": 490
+    },
+    {
+      "epoch": 0.01356177421034268,
+      "grad_norm": 0.243160679936409,
+      "learning_rate": 9.98e-05,
+      "loss": 0.411363410949707,
+      "step": 500
+    },
+    {
+      "epoch": 0.013854053826944893,
+      "grad_norm": 0.2272544652223587,
+      "learning_rate": 0.00010180000000000001,
+      "loss": 0.36809728145599363,
+      "step": 510
+    },
+    {
+      "epoch": 0.014146333443547106,
+      "grad_norm": 0.2265583574771881,
+      "learning_rate": 0.0001038,
+      "loss": 0.352154016494751,
+      "step": 520
+    },
+    {
+      "epoch": 0.014438613060149319,
+      "grad_norm": 0.21968695521354675,
+      "learning_rate": 0.00010580000000000001,
+      "loss": 0.40874533653259276,
+      "step": 530
+    },
+    {
+      "epoch": 0.01473089267675153,
+      "grad_norm": 0.20146295428276062,
+      "learning_rate": 0.00010780000000000002,
+      "loss": 0.4024982452392578,
+      "step": 540
+    },
+    {
+      "epoch": 0.015023172293353744,
+      "grad_norm": 0.22711141407489777,
+      "learning_rate": 0.00010980000000000001,
+      "loss": 0.412246561050415,
+      "step": 550
+    },
+    {
+      "epoch": 0.015315451909955957,
+      "grad_norm": 0.2056141048669815,
+      "learning_rate": 0.00011180000000000002,
+      "loss": 0.4331079006195068,
+      "step": 560
+    },
+    {
+      "epoch": 0.01560773152655817,
+      "grad_norm": 0.18964128196239471,
+      "learning_rate": 0.0001138,
+      "loss": 0.4245738506317139,
+      "step": 570
+    },
+    {
+      "epoch": 0.01590001114316038,
+      "grad_norm": 0.18264859914779663,
+      "learning_rate": 0.0001158,
+      "loss": 0.3981050491333008,
+      "step": 580
+    },
+    {
+      "epoch": 0.016192290759762595,
+      "grad_norm": 0.198320671916008,
+      "learning_rate": 0.0001178,
+      "loss": 0.44843344688415526,
+      "step": 590
+    },
+    {
+      "epoch": 0.016484570376364808,
+      "grad_norm": 0.20583713054656982,
+      "learning_rate": 0.0001198,
+      "loss": 0.45852208137512207,
+      "step": 600
+    },
+    {
+      "epoch": 0.01677684999296702,
+      "grad_norm": 0.18764640390872955,
+      "learning_rate": 0.0001218,
+      "loss": 0.39210479259490966,
+      "step": 610
+    },
+    {
+      "epoch": 0.017069129609569234,
+      "grad_norm": 0.22744227945804596,
+      "learning_rate": 0.0001238,
+      "loss": 0.48337244987487793,
+      "step": 620
+    },
+    {
+      "epoch": 0.017361409226171447,
+      "grad_norm": 0.25332197546958923,
+      "learning_rate": 0.0001258,
+      "loss": 0.47956342697143556,
+      "step": 630
+    },
+    {
+      "epoch": 0.01765368884277366,
+      "grad_norm": 0.18094617128372192,
+      "learning_rate": 0.00012780000000000002,
+      "loss": 0.36708922386169435,
+      "step": 640
+    },
+    {
+      "epoch": 0.017945968459375874,
+      "grad_norm": 0.2056637406349182,
+      "learning_rate": 0.0001298,
+      "loss": 0.45494647026062013,
+      "step": 650
+    },
+    {
+      "epoch": 0.018238248075978087,
+      "grad_norm": 0.18681053817272186,
+      "learning_rate": 0.0001318,
+      "loss": 0.39863851070404055,
+      "step": 660
+    },
+    {
+      "epoch": 0.0185305276925803,
+      "grad_norm": 0.18398210406303406,
+      "learning_rate": 0.00013380000000000003,
+      "loss": 0.42660346031188967,
+      "step": 670
+    },
+    {
+      "epoch": 0.018822807309182513,
+      "grad_norm": 0.2018873244524002,
+      "learning_rate": 0.00013580000000000002,
+      "loss": 0.41592912673950194,
+      "step": 680
+    },
+    {
+      "epoch": 0.019115086925784727,
+      "grad_norm": 0.20749719440937042,
+      "learning_rate": 0.0001378,
+      "loss": 0.4529616355895996,
+      "step": 690
+    },
+    {
+      "epoch": 0.01940736654238694,
+      "grad_norm": 0.17765171825885773,
+      "learning_rate": 0.0001398,
+      "loss": 0.4367457389831543,
+      "step": 700
+    },
+    {
+      "epoch": 0.01969964615898915,
+      "grad_norm": 0.24419377744197845,
+      "learning_rate": 0.0001418,
+      "loss": 0.4324330806732178,
+      "step": 710
+    },
+    {
+      "epoch": 0.019991925775591363,
+      "grad_norm": 0.1319461613893509,
+      "learning_rate": 0.0001438,
+      "loss": 0.38980741500854493,
+      "step": 720
+    },
+    {
+      "epoch": 0.020284205392193576,
+      "grad_norm": 0.18414747714996338,
+      "learning_rate": 0.0001458,
+      "loss": 0.45140395164489744,
+      "step": 730
+    },
+    {
+      "epoch": 0.02057648500879579,
+      "grad_norm": 0.19065886735916138,
+      "learning_rate": 0.00014780000000000001,
+      "loss": 0.34149444103240967,
+      "step": 740
+    },
+    {
+      "epoch": 0.020868764625398002,
+      "grad_norm": 0.3047577738761902,
+      "learning_rate": 0.0001498,
+      "loss": 0.39748594760894773,
+      "step": 750
+    },
+    {
+      "epoch": 0.021161044242000215,
+      "grad_norm": 0.20547902584075928,
+      "learning_rate": 0.0001518,
+      "loss": 0.35707693099975585,
+      "step": 760
+    },
+    {
+      "epoch": 0.02145332385860243,
+      "grad_norm": 0.1835167109966278,
+      "learning_rate": 0.0001538,
+      "loss": 0.3866844177246094,
+      "step": 770
+    },
+    {
+      "epoch": 0.021745603475204642,
+      "grad_norm": 0.12297193706035614,
+      "learning_rate": 0.00015580000000000002,
+      "loss": 0.3809194803237915,
+      "step": 780
+    },
+    {
+      "epoch": 0.022037883091806855,
+      "grad_norm": 0.23474228382110596,
+      "learning_rate": 0.00015780000000000001,
+      "loss": 0.40447239875793456,
+      "step": 790
+    },
+    {
+      "epoch": 0.022330162708409068,
+      "grad_norm": 0.1950707882642746,
+      "learning_rate": 0.0001598,
+      "loss": 0.36057252883911134,
+      "step": 800
+    },
+    {
+      "epoch": 0.02262244232501128,
+      "grad_norm": 0.19598263502120972,
+      "learning_rate": 0.00016180000000000003,
+      "loss": 0.3532793760299683,
+      "step": 810
+    },
+    {
+      "epoch": 0.022914721941613495,
+      "grad_norm": 0.21619845926761627,
+      "learning_rate": 0.0001638,
+      "loss": 0.33423144817352296,
+      "step": 820
+    },
+    {
+      "epoch": 0.023207001558215708,
+      "grad_norm": 0.1934775710105896,
+      "learning_rate": 0.0001658,
+      "loss": 0.4006977558135986,
+      "step": 830
+    },
+    {
+      "epoch": 0.023499281174817917,
+      "grad_norm": 0.2345370650291443,
+      "learning_rate": 0.0001678,
+      "loss": 0.42209348678588865,
+      "step": 840
+    },
+    {
+      "epoch": 0.02379156079142013,
+      "grad_norm": 0.2699654698371887,
+      "learning_rate": 0.0001698,
+      "loss": 0.3334143400192261,
+      "step": 850
+    },
+    {
+      "epoch": 0.024083840408022344,
+      "grad_norm": 0.1839752197265625,
+      "learning_rate": 0.0001718,
+      "loss": 0.42084970474243166,
+      "step": 860
+    },
+    {
+      "epoch": 0.024376120024624557,
+      "grad_norm": 0.22491669654846191,
+      "learning_rate": 0.0001738,
+      "loss": 0.420766544342041,
+      "step": 870
+    },
+    {
+      "epoch": 0.02466839964122677,
+      "grad_norm": 0.18370424211025238,
+      "learning_rate": 0.00017580000000000002,
+      "loss": 0.3486522912979126,
+      "step": 880
+    },
+    {
+      "epoch": 0.024960679257828983,
+      "grad_norm": 0.23166073858737946,
+      "learning_rate": 0.0001778,
+      "loss": 0.42749767303466796,
+      "step": 890
+    },
+    {
+      "epoch": 0.025252958874431197,
+      "grad_norm": 0.22481808066368103,
+      "learning_rate": 0.0001798,
+      "loss": 0.380033016204834,
+      "step": 900
+    },
+    {
+      "epoch": 0.02554523849103341,
+      "grad_norm": 0.24065051972866058,
+      "learning_rate": 0.00018180000000000003,
+      "loss": 0.45965576171875,
+      "step": 910
+    },
+    {
+      "epoch": 0.025837518107635623,
+      "grad_norm": 0.190377876162529,
+      "learning_rate": 0.00018380000000000002,
+      "loss": 0.3290415287017822,
+      "step": 920
+    },
+    {
+      "epoch": 0.026129797724237836,
+      "grad_norm": 0.2699977159500122,
+      "learning_rate": 0.00018580000000000002,
+      "loss": 0.4352889537811279,
+      "step": 930
+    },
+    {
+      "epoch": 0.02642207734084005,
+      "grad_norm": 0.2756347060203552,
+      "learning_rate": 0.0001878,
+      "loss": 0.4288458347320557,
+      "step": 940
+    },
+    {
+      "epoch": 0.026714356957442262,
+      "grad_norm": 0.23565548658370972,
+      "learning_rate": 0.0001898,
+      "loss": 0.29646151065826415,
+      "step": 950
+    },
+    {
+      "epoch": 0.027006636574044476,
+      "grad_norm": 0.18665672838687897,
+      "learning_rate": 0.0001918,
+      "loss": 0.39093432426452634,
+      "step": 960
+    },
+    {
+      "epoch": 0.027298916190646685,
+      "grad_norm": 0.24153801798820496,
+      "learning_rate": 0.0001938,
+      "loss": 0.42369751930236815,
+      "step": 970
+    },
+    {
+      "epoch": 0.0275911958072489,
+      "grad_norm": 0.18671485781669617,
+      "learning_rate": 0.00019580000000000002,
+      "loss": 0.38338029384613037,
+      "step": 980
+    },
+    {
+      "epoch": 0.027883475423851112,
+      "grad_norm": 0.2615300118923187,
+      "learning_rate": 0.0001978,
+      "loss": 0.36945343017578125,
+      "step": 990
+    },
+    {
+      "epoch": 0.028175755040453325,
+      "grad_norm": 0.2992021143436432,
+      "learning_rate": 0.0001998,
+      "loss": 0.4174357891082764,
+      "step": 1000
+    },
+    {
+      "epoch": 0.028468034657055538,
+      "grad_norm": 0.27294614911079407,
+      "learning_rate": 0.00019999997371998945,
+      "loss": 0.38078644275665285,
+      "step": 1010
+    },
+    {
+      "epoch": 0.02876031427365775,
+      "grad_norm": 0.22782108187675476,
+      "learning_rate": 0.0001999998828755262,
+      "loss": 0.41697354316711427,
+      "step": 1020
+    },
+    {
+      "epoch": 0.029052593890259965,
+      "grad_norm": 0.20225366950035095,
+      "learning_rate": 0.00019999972714222465,
+      "loss": 0.38390820026397704,
+      "step": 1030
+    },
+    {
+      "epoch": 0.029344873506862178,
+      "grad_norm": 0.22823745012283325,
+      "learning_rate": 0.00019999950652018584,
+      "loss": 0.41745734214782715,
+      "step": 1040
+    },
+    {
+      "epoch": 0.02963715312346439,
+      "grad_norm": 0.2649468183517456,
+      "learning_rate": 0.0001999992210095529,
+      "loss": 0.4053181648254395,
+      "step": 1050
+    },
+    {
+      "epoch": 0.029929432740066604,
+      "grad_norm": 0.21082913875579834,
+      "learning_rate": 0.00019999887061051115,
+      "loss": 0.41599574089050295,
+      "step": 1060
+    },
+    {
+      "epoch": 0.030221712356668817,
+      "grad_norm": 0.2594168782234192,
+      "learning_rate": 0.00019999845532328792,
+      "loss": 0.4727328777313232,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03051399197327103,
+      "grad_norm": 0.19130517542362213,
+      "learning_rate": 0.0001999979751481527,
+      "loss": 0.42589612007141114,
+      "step": 1080
+    },
+    {
+      "epoch": 0.030806271589873244,
+      "grad_norm": 0.2338467836380005,
+      "learning_rate": 0.00019999743008541709,
+      "loss": 0.3404257774353027,
+      "step": 1090
+    },
+    {
+      "epoch": 0.031098551206475453,
+      "grad_norm": 0.2450529932975769,
+      "learning_rate": 0.00019999682013543474,
+      "loss": 0.35564963817596434,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03139083082307767,
+      "grad_norm": 0.24461933970451355,
+      "learning_rate": 0.00019999614529860144,
+      "loss": 0.39687209129333495,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03168311043967988,
+      "grad_norm": 0.23187988996505737,
+      "learning_rate": 0.0001999954055753551,
+      "loss": 0.41315088272094724,
+      "step": 1120
+    },
+    {
+      "epoch": 0.031975390056282096,
+      "grad_norm": 0.22851324081420898,
+      "learning_rate": 0.00019999460096617572,
+      "loss": 0.32306632995605467,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03226766967288431,
+      "grad_norm": 0.18886074423789978,
+      "learning_rate": 0.00019999373147158543,
+      "loss": 0.3556957244873047,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03255994928948652,
+      "grad_norm": 0.1941390335559845,
+      "learning_rate": 0.00019999279709214838,
+      "loss": 0.42602267265319826,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03285222890608873,
+      "grad_norm": 0.2191803753376007,
+      "learning_rate": 0.00019999179782847088,
+      "loss": 0.3481628656387329,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03314450852269094,
+      "grad_norm": 0.27078554034233093,
+      "learning_rate": 0.0001999907336812014,
+      "loss": 0.3957390785217285,
+      "step": 1170
+    },
+    {
+      "epoch": 0.033436788139293155,
+      "grad_norm": 0.31014931201934814,
+      "learning_rate": 0.00019998960465103042,
+      "loss": 0.3879474878311157,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03372906775589537,
+      "grad_norm": 0.1903751790523529,
+      "learning_rate": 0.00019998841073869055,
+      "loss": 0.42315473556518557,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03402134737249758,
+      "grad_norm": 0.2546682357788086,
+      "learning_rate": 0.0001999871519449565,
+      "loss": 0.36317920684814453,
+      "step": 1200
+    },
+    {
+      "epoch": 0.034313626989099795,
+      "grad_norm": 0.1552298665046692,
+      "learning_rate": 0.0001999858282706451,
+      "loss": 0.36159753799438477,
+      "step": 1210
+    },
+    {
+      "epoch": 0.03460590660570201,
+      "grad_norm": 0.19712704420089722,
+      "learning_rate": 0.00019998443971661525,
+      "loss": 0.3884085416793823,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03489818622230422,
+      "grad_norm": 0.2729942500591278,
+      "learning_rate": 0.000199982986283768,
+      "loss": 0.4681094169616699,
+      "step": 1230
+    },
+    {
+      "epoch": 0.035190465838906435,
+      "grad_norm": 0.1953604519367218,
+      "learning_rate": 0.00019998146797304642,
+      "loss": 0.34714839458465574,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03548274545550865,
+      "grad_norm": 0.2132028043270111,
+      "learning_rate": 0.0001999798847854358,
+      "loss": 0.4187171936035156,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03577502507211086,
+      "grad_norm": 0.21808749437332153,
+      "learning_rate": 0.0001999782367219634,
+      "loss": 0.4529675006866455,
+      "step": 1260
+    },
+    {
+      "epoch": 0.036067304688713074,
+      "grad_norm": 0.2665892541408539,
+      "learning_rate": 0.0001999765237836986,
+      "loss": 0.40354113578796386,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03635958430531529,
+      "grad_norm": 0.18756791949272156,
+      "learning_rate": 0.00019997474597175302,
+      "loss": 0.33424665927886965,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0366518639219175,
+      "grad_norm": 0.23514626920223236,
+      "learning_rate": 0.00019997290328728013,
+      "loss": 0.34480867385864256,
+      "step": 1290
+    },
+    {
+      "epoch": 0.036944143538519714,
+      "grad_norm": 0.24087919294834137,
+      "learning_rate": 0.0001999709957314757,
+      "loss": 0.40560035705566405,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03723642315512193,
+      "grad_norm": 0.1967712640762329,
+      "learning_rate": 0.0001999690233055775,
+      "loss": 0.4166306495666504,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03752870277172414,
+      "grad_norm": 0.22088399529457092,
+      "learning_rate": 0.00019996698601086545,
+      "loss": 0.4286780834197998,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03782098238832635,
+      "grad_norm": 0.21077096462249756,
+      "learning_rate": 0.0001999648838486615,
+      "loss": 0.3549200534820557,
+      "step": 1330
+    },
+    {
+      "epoch": 0.038113262004928566,
+      "grad_norm": 0.2842172682285309,
+      "learning_rate": 0.0001999627168203297,
+      "loss": 0.4626432418823242,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03840554162153078,
+      "grad_norm": 0.23517769575119019,
+      "learning_rate": 0.00019996048492727626,
+      "loss": 0.39797124862670896,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03869782123813299,
+      "grad_norm": 0.22531968355178833,
+      "learning_rate": 0.0001999581881709494,
+      "loss": 0.39006221294403076,
+      "step": 1360
+    },
+    {
+      "epoch": 0.038990100854735206,
+      "grad_norm": 0.1859465092420578,
+      "learning_rate": 0.00019995582655283944,
+      "loss": 0.3752875804901123,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03928238047133742,
+      "grad_norm": 0.1793965846300125,
+      "learning_rate": 0.00019995340007447884,
+      "loss": 0.3678503751754761,
+      "step": 1380
+    },
+    {
+      "epoch": 0.03957466008793963,
+      "grad_norm": 0.1874365508556366,
+      "learning_rate": 0.00019995090873744213,
+      "loss": 0.40687713623046873,
+      "step": 1390
+    },
+    {
+      "epoch": 0.039866939704541846,
+      "grad_norm": 0.22492875158786774,
+      "learning_rate": 0.00019994835254334592,
+      "loss": 0.3897545337677002,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04015921932114406,
+      "grad_norm": 0.29545730352401733,
+      "learning_rate": 0.00019994573149384882,
+      "loss": 0.4935605049133301,
+      "step": 1410
+    },
+    {
+      "epoch": 0.040451498937746265,
+      "grad_norm": 0.20854443311691284,
+      "learning_rate": 0.00019994304559065168,
+      "loss": 0.33220815658569336,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04074377855434848,
+      "grad_norm": 0.20468345284461975,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 0.3795356035232544,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04103605817095069,
+      "grad_norm": 0.29992666840553284,
+      "learning_rate": 0.00019993747923017065,
+      "loss": 0.43836030960083006,
+      "step": 1440
+    },
+    {
+      "epoch": 0.041328337787552905,
+      "grad_norm": 0.27834489941596985,
+      "learning_rate": 0.0001999345987764987,
+      "loss": 0.3854727029800415,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04162061740415512,
+      "grad_norm": 0.19028784334659576,
+      "learning_rate": 0.00019993165347635062,
+      "loss": 0.48855161666870117,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04191289702075733,
+      "grad_norm": 0.2719208002090454,
+      "learning_rate": 0.00019992864333163754,
+      "loss": 0.34480597972869875,
+      "step": 1470
+    },
+    {
+      "epoch": 0.042205176637359544,
+      "grad_norm": 0.2201714962720871,
+      "learning_rate": 0.0001999255683443127,
+      "loss": 0.45578484535217284,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04249745625396176,
+      "grad_norm": 0.2917408049106598,
+      "learning_rate": 0.00019992242851637145,
+      "loss": 0.45635452270507815,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04278973587056397,
+      "grad_norm": 0.20707593858242035,
+      "learning_rate": 0.00019991922384985115,
+      "loss": 0.4358846187591553,
+      "step": 1500
+    },
+    {
+      "epoch": 0.043082015487166184,
+      "grad_norm": 0.2693732976913452,
+      "learning_rate": 0.00019991595434683132,
+      "loss": 0.393514084815979,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0433742951037684,
+      "grad_norm": 0.192472904920578,
+      "learning_rate": 0.0001999126200094335,
+      "loss": 0.36677262783050535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04366657472037061,
+      "grad_norm": 0.1903146505355835,
+      "learning_rate": 0.00019990922083982127,
+      "loss": 0.3809571504592896,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04395885433697282,
+      "grad_norm": 0.18778850138187408,
+      "learning_rate": 0.00019990575684020035,
+      "loss": 0.3991628885269165,
+      "step": 1540
+    },
+    {
+      "epoch": 0.044251133953575036,
+      "grad_norm": 0.17295020818710327,
+      "learning_rate": 0.00019990222801281844,
+      "loss": 0.34786784648895264,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04454341357017725,
+      "grad_norm": 0.21790063381195068,
+      "learning_rate": 0.00019989863435996544,
+      "loss": 0.4168230533599854,
+      "step": 1560
+    },
+    {
+      "epoch": 0.04483569318677946,
+      "grad_norm": 0.1902916580438614,
+      "learning_rate": 0.00019989497588397318,
+      "loss": 0.3928706169128418,
+      "step": 1570
+    },
+    {
+      "epoch": 0.045127972803381676,
+      "grad_norm": 0.21699456870555878,
+      "learning_rate": 0.0001998912525872156,
+      "loss": 0.27401630878448485,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04542025241998389,
+      "grad_norm": 0.2811727821826935,
+      "learning_rate": 0.00019988746447210872,
+      "loss": 0.37927000522613524,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0457125320365861,
+      "grad_norm": 0.2877182364463806,
+      "learning_rate": 0.0001998836115411106,
+      "loss": 0.37084925174713135,
+      "step": 1600
+    },
+    {
+      "epoch": 0.046004811653188316,
+      "grad_norm": 0.30798202753067017,
+      "learning_rate": 0.00019987969379672138,
+      "loss": 0.3814183950424194,
+      "step": 1610
+    },
+    {
+      "epoch": 0.04629709126979053,
+      "grad_norm": 0.23142924904823303,
+      "learning_rate": 0.00019987571124148326,
+      "loss": 0.4828189849853516,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04658937088639274,
+      "grad_norm": 0.26372772455215454,
+      "learning_rate": 0.00019987166387798043,
+      "loss": 0.40070490837097167,
+      "step": 1630
+    },
+    {
+      "epoch": 0.046881650502994955,
+      "grad_norm": 0.20170119404792786,
+      "learning_rate": 0.00019986755170883918,
+      "loss": 0.4609227657318115,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04717393011959717,
+      "grad_norm": 0.22828887403011322,
+      "learning_rate": 0.0001998633747367279,
+      "loss": 0.3876612186431885,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04746620973619938,
+      "grad_norm": 0.28497815132141113,
+      "learning_rate": 0.00019985913296435693,
+      "loss": 0.4488811016082764,
+      "step": 1660
+    },
+    {
+      "epoch": 0.047758489352801595,
+      "grad_norm": 0.2250206023454666,
+      "learning_rate": 0.00019985482639447874,
+      "loss": 0.3711233615875244,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0480507689694038,
+      "grad_norm": 0.2564453184604645,
+      "learning_rate": 0.00019985045502988785,
+      "loss": 0.4409325122833252,
+      "step": 1680
+    },
+    {
+      "epoch": 0.048343048586006014,
+      "grad_norm": 0.22390565276145935,
+      "learning_rate": 0.00019984601887342073,
+      "loss": 0.3880199432373047,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04863532820260823,
+      "grad_norm": 0.18111835420131683,
+      "learning_rate": 0.00019984151792795595,
+      "loss": 0.3191307783126831,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04892760781921044,
+      "grad_norm": 0.27317941188812256,
+      "learning_rate": 0.00019983695219641417,
+      "loss": 0.36649317741394044,
+      "step": 1710
+    },
+    {
+      "epoch": 0.049219887435812654,
+      "grad_norm": 0.23256099224090576,
+      "learning_rate": 0.000199832321681758,
+      "loss": 0.4021451950073242,
+      "step": 1720
+    },
+    {
+      "epoch": 0.04951216705241487,
+      "grad_norm": 0.20527933537960052,
+      "learning_rate": 0.00019982762638699218,
+      "loss": 0.36731953620910646,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04980444666901708,
+      "grad_norm": 0.2149372100830078,
+      "learning_rate": 0.00019982286631516343,
+      "loss": 0.38255355358123777,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05009672628561929,
+      "grad_norm": 0.1913089156150818,
+      "learning_rate": 0.00019981804146936046,
+      "loss": 0.4117907524108887,
+      "step": 1750
+    },
+    {
+      "epoch": 0.050389005902221506,
+      "grad_norm": 0.2972666621208191,
+      "learning_rate": 0.0001998131518527141,
+      "loss": 0.39761054515838623,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05068128551882372,
+      "grad_norm": 0.21348360180854797,
+      "learning_rate": 0.00019980819746839715,
+      "loss": 0.4208479404449463,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05097356513542593,
+      "grad_norm": 0.21910355985164642,
+      "learning_rate": 0.00019980317831962447,
+      "loss": 0.45781354904174804,
+      "step": 1780
+    },
+    {
+      "epoch": 0.051265844752028146,
+      "grad_norm": 0.29217293858528137,
+      "learning_rate": 0.00019979809440965292,
+      "loss": 0.392667555809021,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05155812436863036,
+      "grad_norm": 0.2685600221157074,
+      "learning_rate": 0.0001997929457417814,
+      "loss": 0.3381932020187378,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05185040398523257,
+      "grad_norm": 0.208404079079628,
+      "learning_rate": 0.00019978773231935085,
+      "loss": 0.3530620574951172,
+      "step": 1810
+    },
+    {
+      "epoch": 0.052142683601834786,
+      "grad_norm": 0.23994538187980652,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 0.38266212940216066,
+      "step": 1820
+    },
+    {
+      "epoch": 0.052434963218437,
+      "grad_norm": 0.3314270079135895,
+      "learning_rate": 0.0001997771112243863,
+      "loss": 0.37249245643615725,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05272724283503921,
+      "grad_norm": 0.3160709738731384,
+      "learning_rate": 0.0001997717035587442,
+      "loss": 0.4115762233734131,
+      "step": 1840
+    },
+    {
+      "epoch": 0.053019522451641425,
+      "grad_norm": 0.20398829877376556,
+      "learning_rate": 0.00019976623115232692,
+      "loss": 0.4115797996520996,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05331180206824364,
+      "grad_norm": 0.24112626910209656,
+      "learning_rate": 0.00019976069400868534,
+      "loss": 0.3800527811050415,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05360408168484585,
+      "grad_norm": 0.3240782618522644,
+      "learning_rate": 0.00019975509213141255,
+      "loss": 0.4142290115356445,
+      "step": 1870
+    },
+    {
+      "epoch": 0.053896361301448065,
+      "grad_norm": 0.27430611848831177,
+      "learning_rate": 0.0001997494255241435,
+      "loss": 0.42593750953674314,
+      "step": 1880
+    },
+    {
+      "epoch": 0.05418864091805028,
+      "grad_norm": 0.2675989270210266,
+      "learning_rate": 0.00019974369419055514,
+      "loss": 0.35791664123535155,
+      "step": 1890
+    },
+    {
+      "epoch": 0.05448092053465249,
+      "grad_norm": 0.217862069606781,
+      "learning_rate": 0.00019973789813436656,
+      "loss": 0.42203755378723146,
+      "step": 1900
+    },
+    {
+      "epoch": 0.054773200151254704,
+      "grad_norm": 0.20300687849521637,
+      "learning_rate": 0.00019973203735933873,
+      "loss": 0.3735295534133911,
+      "step": 1910
+    },
+    {
+      "epoch": 0.05506547976785692,
+      "grad_norm": 0.24305196106433868,
+      "learning_rate": 0.0001997261118692746,
+      "loss": 0.3854319334030151,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05535775938445913,
+      "grad_norm": 0.17744389176368713,
+      "learning_rate": 0.0001997201216680192,
+      "loss": 0.3821631908416748,
+      "step": 1930
+    },
+    {
+      "epoch": 0.055650039001061344,
+      "grad_norm": 0.28994235396385193,
+      "learning_rate": 0.0001997140667594595,
+      "loss": 0.36383843421936035,
+      "step": 1940
+    },
+    {
+      "epoch": 0.05594231861766355,
+      "grad_norm": 0.21803227066993713,
+      "learning_rate": 0.00019970794714752445,
+      "loss": 0.4005745410919189,
+      "step": 1950
+    },
+    {
+      "epoch": 0.05623459823426576,
+      "grad_norm": 0.22061172127723694,
+      "learning_rate": 0.00019970176283618503,
+      "loss": 0.42282872200012206,
+      "step": 1960
+    },
+    {
+      "epoch": 0.056526877850867976,
+      "grad_norm": 0.20922382175922394,
+      "learning_rate": 0.00019969551382945414,
+      "loss": 0.34180996417999265,
+      "step": 1970
+    },
+    {
+      "epoch": 0.05681915746747019,
+      "grad_norm": 0.21188782155513763,
+      "learning_rate": 0.00019968920013138671,
+      "loss": 0.3581382751464844,
+      "step": 1980
+    },
+    {
+      "epoch": 0.0571114370840724,
+      "grad_norm": 0.19657376408576965,
+      "learning_rate": 0.00019968282174607962,
+      "loss": 0.35428762435913086,
+      "step": 1990
+    },
+    {
+      "epoch": 0.057403716700674616,
+      "grad_norm": 0.20557129383087158,
+      "learning_rate": 0.00019967637867767172,
+      "loss": 0.4382596969604492,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05769599631727683,
+      "grad_norm": 0.2759098410606384,
+      "learning_rate": 0.0001996698709303439,
+      "loss": 0.38013901710510256,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05798827593387904,
+      "grad_norm": 0.2426660805940628,
+      "learning_rate": 0.00019966329850831894,
+      "loss": 0.4401369571685791,
+      "step": 2020
+    },
+    {
+      "epoch": 0.058280555550481256,
+      "grad_norm": 0.22679618000984192,
+      "learning_rate": 0.00019965666141586157,
+      "loss": 0.3150460243225098,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05857283516708347,
+      "grad_norm": 0.29090648889541626,
+      "learning_rate": 0.0001996499596572786,
+      "loss": 0.3733455896377563,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05886511478368568,
+      "grad_norm": 0.30679851770401,
+      "learning_rate": 0.00019964319323691868,
+      "loss": 0.443017578125,
+      "step": 2050
+    },
+    {
+      "epoch": 0.059157394400287895,
+      "grad_norm": 0.21110324561595917,
+      "learning_rate": 0.00019963636215917244,
+      "loss": 0.40988821983337403,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05944967401689011,
+      "grad_norm": 0.25294822454452515,
+      "learning_rate": 0.00019962946642847258,
+      "loss": 0.42485885620117186,
+      "step": 2070
+    },
+    {
+      "epoch": 0.05974195363349232,
+      "grad_norm": 0.24810326099395752,
+      "learning_rate": 0.0001996225060492936,
+      "loss": 0.3709405422210693,
+      "step": 2080
+    },
+    {
+      "epoch": 0.060034233250094535,
+      "grad_norm": 0.2301589697599411,
+      "learning_rate": 0.00019961548102615204,
+      "loss": 0.42913308143615725,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06032651286669675,
+      "grad_norm": 0.24390390515327454,
+      "learning_rate": 0.00019960839136360633,
+      "loss": 0.3363667011260986,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06061879248329896,
+      "grad_norm": 0.23716720938682556,
+      "learning_rate": 0.0001996012370662569,
+      "loss": 0.361738920211792,
+      "step": 2110
+    },
+    {
+      "epoch": 0.060911072099901174,
+      "grad_norm": 0.20727047324180603,
+      "learning_rate": 0.0001995940181387461,
+      "loss": 0.38850061893463134,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06120335171650339,
+      "grad_norm": 0.19902947545051575,
+      "learning_rate": 0.00019958673458575824,
+      "loss": 0.31249237060546875,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0614956313331056,
+      "grad_norm": 0.24424754083156586,
+      "learning_rate": 0.00019957938641201948,
+      "loss": 0.41524853706359866,
+      "step": 2140
+    },
+    {
+      "epoch": 0.061787910949707814,
+      "grad_norm": 0.24768763780593872,
+      "learning_rate": 0.00019957197362229797,
+      "loss": 0.39194750785827637,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06208019056631003,
+      "grad_norm": 0.3057568073272705,
+      "learning_rate": 0.00019956449622140384,
+      "loss": 0.31905083656311034,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06237247018291224,
+      "grad_norm": 0.3908090591430664,
+      "learning_rate": 0.00019955695421418907,
+      "loss": 0.4339137077331543,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06266474979951445,
+      "grad_norm": 0.21391141414642334,
+      "learning_rate": 0.00019954934760554757,
+      "loss": 0.3780949831008911,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06295702941611667,
+      "grad_norm": 0.16703805327415466,
+      "learning_rate": 0.00019954167640041522,
+      "loss": 0.40203108787536623,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06324930903271887,
+      "grad_norm": 0.2569977343082428,
+      "learning_rate": 0.00019953394060376973,
+      "loss": 0.42340788841247556,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06354158864932109,
+      "grad_norm": 0.2800414562225342,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 0.40265374183654784,
+      "step": 2210
+    },
+    {
+      "epoch": 0.0638338682659233,
+      "grad_norm": 0.21277093887329102,
+      "learning_rate": 0.0001995182752560601,
+      "loss": 0.3515303134918213,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06412614788252552,
+      "grad_norm": 0.3708363175392151,
+      "learning_rate": 0.00019951034571516097,
+      "loss": 0.3585836410522461,
+      "step": 2230
+    },
+    {
+      "epoch": 0.06441842749912773,
+      "grad_norm": 0.21470332145690918,
+      "learning_rate": 0.0001995023516030789,
+      "loss": 0.35557281970977783,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06471070711572995,
+      "grad_norm": 0.22539454698562622,
+      "learning_rate": 0.00019949429292500113,
+      "loss": 0.3998990058898926,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06500298673233215,
+      "grad_norm": 0.19151820242404938,
+      "learning_rate": 0.0001994861696861569,
+      "loss": 0.3600909948348999,
+      "step": 2260
+    },
+    {
+      "epoch": 0.06529526634893437,
+      "grad_norm": 0.21192649006843567,
+      "learning_rate": 0.00019947798189181727,
+      "loss": 0.34516177177429197,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06558754596553658,
+      "grad_norm": 0.2622006833553314,
+      "learning_rate": 0.00019946972954729519,
+      "loss": 0.3691245555877686,
+      "step": 2280
+    },
+    {
+      "epoch": 0.0658798255821388,
+      "grad_norm": 0.1908690184354782,
+      "learning_rate": 0.00019946141265794553,
+      "loss": 0.38103673458099363,
+      "step": 2290
+    },
+    {
+      "epoch": 0.066172105198741,
+      "grad_norm": 0.21785594522953033,
+      "learning_rate": 0.00019945303122916505,
+      "loss": 0.32585608959198,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06646438481534322,
+      "grad_norm": 0.31074655055999756,
+      "learning_rate": 0.00019944458526639235,
+      "loss": 0.4161215305328369,
+      "step": 2310
+    },
+    {
+      "epoch": 0.06675666443194543,
+      "grad_norm": 0.2080898880958557,
+      "learning_rate": 0.00019943607477510792,
+      "loss": 0.40408644676208494,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06704894404854765,
+      "grad_norm": 0.20214156806468964,
+      "learning_rate": 0.00019942749976083416,
+      "loss": 0.373272705078125,
+      "step": 2330
+    },
+    {
+      "epoch": 0.06734122366514986,
+      "grad_norm": 0.2070678323507309,
+      "learning_rate": 0.00019941886022913522,
+      "loss": 0.3865471363067627,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06763350328175206,
+      "grad_norm": 0.2521723210811615,
+      "learning_rate": 0.0001994101561856173,
+      "loss": 0.3574871063232422,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06792578289835428,
+      "grad_norm": 0.2097303718328476,
+      "learning_rate": 0.00019940138763592828,
+      "loss": 0.4145055770874023,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06821806251495649,
+      "grad_norm": 0.2956986725330353,
+      "learning_rate": 0.00019939255458575802,
+      "loss": 0.4034604072570801,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06851034213155871,
+      "grad_norm": 0.20965847373008728,
+      "learning_rate": 0.00019938365704083816,
+      "loss": 0.38535895347595217,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06880262174816092,
+      "grad_norm": 0.2410966157913208,
+      "learning_rate": 0.00019937469500694226,
+      "loss": 0.40244760513305666,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06909490136476314,
+      "grad_norm": 0.24631759524345398,
+      "learning_rate": 0.00019936566848988567,
+      "loss": 0.4188735008239746,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06938718098136534,
+      "grad_norm": 0.1852298080921173,
+      "learning_rate": 0.00019935657749552558,
+      "loss": 0.40854802131652834,
+      "step": 2410
+    },
+    {
+      "epoch": 0.06967946059796756,
+      "grad_norm": 0.24064451456069946,
+      "learning_rate": 0.00019934742202976107,
+      "loss": 0.35597121715545654,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06997174021456977,
+      "grad_norm": 0.2895231246948242,
+      "learning_rate": 0.00019933820209853297,
+      "loss": 0.39232187271118163,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07026401983117199,
+      "grad_norm": 0.19319918751716614,
+      "learning_rate": 0.00019932891770782404,
+      "loss": 0.3565209150314331,
+      "step": 2440
+    },
+    {
+      "epoch": 0.0705562994477742,
+      "grad_norm": 0.36313900351524353,
+      "learning_rate": 0.00019931956886365884,
+      "loss": 0.36571540832519533,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07084857906437642,
+      "grad_norm": 0.2750861346721649,
+      "learning_rate": 0.0001993101555721037,
+      "loss": 0.3466514587402344,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07114085868097862,
+      "grad_norm": 0.22644628584384918,
+      "learning_rate": 0.00019930067783926675,
+      "loss": 0.3429016828536987,
+      "step": 2470
+    },
+    {
+      "epoch": 0.07143313829758084,
+      "grad_norm": 0.25445985794067383,
+      "learning_rate": 0.0001992911356712981,
+      "loss": 0.4083054542541504,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07172541791418305,
+      "grad_norm": 0.2863784432411194,
+      "learning_rate": 0.00019928152907438946,
+      "loss": 0.38512208461761477,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07201769753078527,
+      "grad_norm": 0.3040670156478882,
+      "learning_rate": 0.0001992718580547745,
+      "loss": 0.3621720552444458,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07230997714738747,
+      "grad_norm": 0.21406209468841553,
+      "learning_rate": 0.00019926212261872866,
+      "loss": 0.39595961570739746,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0726022567639897,
+      "grad_norm": 0.185161292552948,
+      "learning_rate": 0.00019925232277256906,
+      "loss": 0.38945841789245605,
+      "step": 2520
+    },
+    {
+      "epoch": 0.0728945363805919,
+      "grad_norm": 0.35198086500167847,
+      "learning_rate": 0.00019924245852265485,
+      "loss": 0.38671183586120605,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07318681599719412,
+      "grad_norm": 0.2338062822818756,
+      "learning_rate": 0.0001992325298753867,
+      "loss": 0.40381150245666503,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07347909561379633,
+      "grad_norm": 0.22099274396896362,
+      "learning_rate": 0.00019922253683720732,
+      "loss": 0.39828627109527587,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07377137523039855,
+      "grad_norm": 0.2281026393175125,
+      "learning_rate": 0.000199212479414601,
+      "loss": 0.38828580379486083,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07406365484700075,
+      "grad_norm": 0.240388423204422,
+      "learning_rate": 0.00019920235761409393,
+      "loss": 0.3149177312850952,
+      "step": 2570
+    },
+    {
+      "epoch": 0.07435593446360297,
+      "grad_norm": 0.2369592785835266,
+      "learning_rate": 0.00019919217144225403,
+      "loss": 0.43828372955322265,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07464821408020518,
+      "grad_norm": 0.2708946168422699,
+      "learning_rate": 0.00019918192090569102,
+      "loss": 0.3750912666320801,
+      "step": 2590
+    },
+    {
+      "epoch": 0.07494049369680739,
+      "grad_norm": 0.25363683700561523,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 0.4024306297302246,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0752327733134096,
+      "grad_norm": 0.25413063168525696,
+      "learning_rate": 0.00019916122676504313,
+      "loss": 0.4374362945556641,
+      "step": 2610
+    },
+    {
+      "epoch": 0.07552505293001181,
+      "grad_norm": 0.2922676205635071,
+      "learning_rate": 0.00019915078317438652,
+      "loss": 0.4300537586212158,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07581733254661403,
+      "grad_norm": 0.23218266665935516,
+      "learning_rate": 0.00019914027524586315,
+      "loss": 0.30225658416748047,
+      "step": 2630
+    },
+    {
+      "epoch": 0.07610961216321624,
+      "grad_norm": 0.21953946352005005,
+      "learning_rate": 0.00019912970298629153,
+      "loss": 0.40011258125305177,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07640189177981846,
+      "grad_norm": 0.2304810881614685,
+      "learning_rate": 0.00019911906640253187,
+      "loss": 0.39437267780303953,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07669417139642067,
+      "grad_norm": 0.2002577781677246,
+      "learning_rate": 0.00019910836550148615,
+      "loss": 0.34653146266937257,
+      "step": 2660
+    },
+    {
+      "epoch": 0.07698645101302289,
+      "grad_norm": 0.26880741119384766,
+      "learning_rate": 0.00019909760029009806,
+      "loss": 0.38612444400787355,
+      "step": 2670
+    },
+    {
+      "epoch": 0.07727873062962509,
+      "grad_norm": 0.2160756140947342,
+      "learning_rate": 0.00019908677077535303,
+      "loss": 0.4106006145477295,
+      "step": 2680
+    },
+    {
+      "epoch": 0.07757101024622731,
+      "grad_norm": 0.2476145178079605,
+      "learning_rate": 0.00019907587696427821,
+      "loss": 0.4026120662689209,
+      "step": 2690
+    },
+    {
+      "epoch": 0.07786328986282952,
+      "grad_norm": 0.8202416300773621,
+      "learning_rate": 0.00019906491886394247,
+      "loss": 0.41068472862243655,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07815556947943174,
+      "grad_norm": 0.25270184874534607,
+      "learning_rate": 0.00019905389648145643,
+      "loss": 0.388895583152771,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07844784909603394,
+      "grad_norm": 0.2556864619255066,
+      "learning_rate": 0.00019904280982397236,
+      "loss": 0.46575565338134767,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07874012871263616,
+      "grad_norm": 0.22981390357017517,
+      "learning_rate": 0.0001990316588986843,
+      "loss": 0.3884773015975952,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07903240832923837,
+      "grad_norm": 0.18209731578826904,
+      "learning_rate": 0.00019902044371282793,
+      "loss": 0.3456049680709839,
+      "step": 2740
+    },
+    {
+      "epoch": 0.07932468794584059,
+      "grad_norm": 0.31520935893058777,
+      "learning_rate": 0.0001990091642736807,
+      "loss": 0.36254956722259524,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0796169675624428,
+      "grad_norm": 0.2921953797340393,
+      "learning_rate": 0.00019899782058856172,
+      "loss": 0.35902976989746094,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07990924717904502,
+      "grad_norm": 0.26231515407562256,
+      "learning_rate": 0.00019898641266483173,
+      "loss": 0.33397631645202636,
+      "step": 2770
+    },
+    {
+      "epoch": 0.08020152679564722,
+      "grad_norm": 0.2696617543697357,
+      "learning_rate": 0.00019897494050989326,
+      "loss": 0.3791987657546997,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08049380641224944,
+      "grad_norm": 0.24650204181671143,
+      "learning_rate": 0.00019896340413119043,
+      "loss": 0.4233126640319824,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08078608602885165,
+      "grad_norm": 0.2599400281906128,
+      "learning_rate": 0.00019895180353620911,
+      "loss": 0.4109656810760498,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08107836564545387,
+      "grad_norm": 0.24897293746471405,
+      "learning_rate": 0.00019894013873247678,
+      "loss": 0.38447999954223633,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08137064526205608,
+      "grad_norm": 0.20676682889461517,
+      "learning_rate": 0.0001989284097275626,
+      "loss": 0.42731261253356934,
+      "step": 2820
+    },
+    {
+      "epoch": 0.0816629248786583,
+      "grad_norm": 0.28797972202301025,
+      "learning_rate": 0.0001989166165290774,
+      "loss": 0.375295615196228,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0819552044952605,
+      "grad_norm": 0.24699875712394714,
+      "learning_rate": 0.00019890475914467366,
+      "loss": 0.37915284633636476,
+      "step": 2840
+    },
+    {
+      "epoch": 0.08224748411186272,
+      "grad_norm": 0.250649631023407,
+      "learning_rate": 0.0001988928375820455,
+      "loss": 0.3783683776855469,
+      "step": 2850
+    },
+    {
+      "epoch": 0.08253976372846493,
+      "grad_norm": 0.22644077241420746,
+      "learning_rate": 0.00019888085184892868,
+      "loss": 0.41778364181518557,
+      "step": 2860
+    },
+    {
+      "epoch": 0.08283204334506714,
+      "grad_norm": 0.2277672290802002,
+      "learning_rate": 0.00019886880195310065,
+      "loss": 0.36871073246002195,
+      "step": 2870
+    },
+    {
+      "epoch": 0.08312432296166936,
+      "grad_norm": 0.31684842705726624,
+      "learning_rate": 0.00019885668790238043,
+      "loss": 0.40143604278564454,
+      "step": 2880
+    },
+    {
+      "epoch": 0.08341660257827156,
+      "grad_norm": 0.33947882056236267,
+      "learning_rate": 0.0001988445097046287,
+      "loss": 0.41034741401672364,
+      "step": 2890
+    },
+    {
+      "epoch": 0.08370888219487378,
+      "grad_norm": 0.22324541211128235,
+      "learning_rate": 0.00019883226736774776,
+      "loss": 0.3964377403259277,
+      "step": 2900
+    },
+    {
+      "epoch": 0.08400116181147599,
+      "grad_norm": 0.23351380228996277,
+      "learning_rate": 0.00019881996089968156,
+      "loss": 0.41057047843933103,
+      "step": 2910
+    },
+    {
+      "epoch": 0.08429344142807821,
+      "grad_norm": 0.20214258134365082,
+      "learning_rate": 0.00019880759030841558,
+      "loss": 0.3924527406692505,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08458572104468041,
+      "grad_norm": 0.2522850036621094,
+      "learning_rate": 0.00019879515560197698,
+      "loss": 0.3682981252670288,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08487800066128263,
+      "grad_norm": 0.2607446014881134,
+      "learning_rate": 0.0001987826567884345,
+      "loss": 0.3693267345428467,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08517028027788484,
+      "grad_norm": 0.28362178802490234,
+      "learning_rate": 0.00019877009387589853,
+      "loss": 0.4493271350860596,
+      "step": 2950
+    },
+    {
+      "epoch": 0.08546255989448706,
+      "grad_norm": 0.20476898550987244,
+      "learning_rate": 0.00019875746687252096,
+      "loss": 0.4254180431365967,
+      "step": 2960
+    },
+    {
+      "epoch": 0.08575483951108927,
+      "grad_norm": 0.31387001276016235,
+      "learning_rate": 0.00019874477578649534,
+      "loss": 0.409135103225708,
+      "step": 2970
+    },
+    {
+      "epoch": 0.08604711912769149,
+      "grad_norm": 0.18842676281929016,
+      "learning_rate": 0.0001987320206260568,
+      "loss": 0.35662786960601806,
+      "step": 2980
+    },
+    {
+      "epoch": 0.0863393987442937,
+      "grad_norm": 0.28873375058174133,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 0.34955050945281985,
+      "step": 2990
+    },
+    {
+      "epoch": 0.08663167836089591,
+      "grad_norm": 0.2810521423816681,
+      "learning_rate": 0.00019870631811508909,
+      "loss": 0.4152508735656738,
+      "step": 3000
+    },
+    {
+      "epoch": 0.08692395797749812,
+      "grad_norm": 0.23523248732089996,
+      "learning_rate": 0.00019869337078123805,
+      "loss": 0.3674147367477417,
+      "step": 3010
+    },
+    {
+      "epoch": 0.08721623759410034,
+      "grad_norm": 0.22216211259365082,
+      "learning_rate": 0.0001986803594063302,
+      "loss": 0.3648097991943359,
+      "step": 3020
+    },
+    {
+      "epoch": 0.08750851721070255,
+      "grad_norm": 0.1880836933851242,
+      "learning_rate": 0.0001986672839988085,
+      "loss": 0.36497581005096436,
+      "step": 3030
+    },
+    {
+      "epoch": 0.08780079682730477,
+      "grad_norm": 0.270539790391922,
+      "learning_rate": 0.00019865414456715744,
+      "loss": 0.38210604190826414,
+      "step": 3040
+    },
+    {
+      "epoch": 0.08809307644390697,
+      "grad_norm": 0.30120202898979187,
+      "learning_rate": 0.00019864094111990302,
+      "loss": 0.45117645263671874,
+      "step": 3050
+    },
+    {
+      "epoch": 0.0883853560605092,
+      "grad_norm": 0.29563096165657043,
+      "learning_rate": 0.00019862767366561284,
+      "loss": 0.45659689903259276,
+      "step": 3060
+    },
+    {
+      "epoch": 0.0886776356771114,
+      "grad_norm": 0.3945213258266449,
+      "learning_rate": 0.00019861434221289602,
+      "loss": 0.3833597421646118,
+      "step": 3070
+    },
+    {
+      "epoch": 0.08896991529371362,
+      "grad_norm": 0.2667384743690491,
+      "learning_rate": 0.00019860094677040318,
+      "loss": 0.36721580028533934,
+      "step": 3080
+    },
+    {
+      "epoch": 0.08926219491031583,
+      "grad_norm": 0.23233367502689362,
+      "learning_rate": 0.00019858748734682646,
+      "loss": 0.4113668441772461,
+      "step": 3090
+    },
+    {
+      "epoch": 0.08955447452691805,
+      "grad_norm": 0.34524235129356384,
+      "learning_rate": 0.00019857396395089958,
+      "loss": 0.4031537055969238,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08984675414352025,
+      "grad_norm": 0.24106700718402863,
+      "learning_rate": 0.00019856037659139766,
+      "loss": 0.38901185989379883,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09013903376012246,
+      "grad_norm": 0.27765727043151855,
+      "learning_rate": 0.00019854672527713744,
+      "loss": 0.3432453155517578,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09043131337672468,
+      "grad_norm": 0.3085385859012604,
+      "learning_rate": 0.00019853301001697708,
+      "loss": 0.38917293548583987,
+      "step": 3130
+    },
+    {
+      "epoch": 0.09072359299332688,
+      "grad_norm": 0.23235079646110535,
+      "learning_rate": 0.0001985192308198163,
+      "loss": 0.3917238235473633,
+      "step": 3140
+    },
+    {
+      "epoch": 0.0910158726099291,
+      "grad_norm": 0.24645209312438965,
+      "learning_rate": 0.00019850538769459623,
+      "loss": 0.37504327297210693,
+      "step": 3150
+    },
+    {
+      "epoch": 0.09130815222653131,
+      "grad_norm": 0.2510511875152588,
+      "learning_rate": 0.00019849148065029953,
+      "loss": 0.33972527980804446,
+      "step": 3160
+    },
+    {
+      "epoch": 0.09160043184313353,
+      "grad_norm": 0.21637654304504395,
+      "learning_rate": 0.00019847750969595037,
+      "loss": 0.4131617069244385,
+      "step": 3170
+    },
+    {
+      "epoch": 0.09189271145973574,
+      "grad_norm": 0.2176114320755005,
+      "learning_rate": 0.00019846347484061433,
+      "loss": 0.3799577713012695,
+      "step": 3180
+    },
+    {
+      "epoch": 0.09218499107633796,
+      "grad_norm": 0.2536453902721405,
+      "learning_rate": 0.00019844937609339843,
+      "loss": 0.4001410484313965,
+      "step": 3190
+    },
+    {
+      "epoch": 0.09247727069294016,
+      "grad_norm": 0.23474979400634766,
+      "learning_rate": 0.00019843521346345127,
+      "loss": 0.3490931510925293,
+      "step": 3200
+    },
+    {
+      "epoch": 0.09276955030954238,
+      "grad_norm": 0.3026920258998871,
+      "learning_rate": 0.00019842098695996277,
+      "loss": 0.3579007387161255,
+      "step": 3210
+    },
+    {
+      "epoch": 0.09306182992614459,
+      "grad_norm": 0.20943433046340942,
+      "learning_rate": 0.0001984066965921644,
+      "loss": 0.3663020610809326,
+      "step": 3220
+    },
+    {
+      "epoch": 0.09335410954274681,
+      "grad_norm": 0.30826592445373535,
+      "learning_rate": 0.00019839234236932897,
+      "loss": 0.37556872367858884,
+      "step": 3230
+    },
+    {
+      "epoch": 0.09364638915934902,
+      "grad_norm": 0.22714120149612427,
+      "learning_rate": 0.0001983779243007708,
+      "loss": 0.3506728172302246,
+      "step": 3240
+    },
+    {
+      "epoch": 0.09393866877595124,
+      "grad_norm": 0.23695889115333557,
+      "learning_rate": 0.00019836344239584564,
+      "loss": 0.3726650238037109,
+      "step": 3250
+    },
+    {
+      "epoch": 0.09423094839255344,
+      "grad_norm": 0.2446599006652832,
+      "learning_rate": 0.00019834889666395063,
+      "loss": 0.3690079689025879,
+      "step": 3260
+    },
+    {
+      "epoch": 0.09452322800915566,
+      "grad_norm": 0.3368624746799469,
+      "learning_rate": 0.0001983342871145243,
+      "loss": 0.3897335767745972,
+      "step": 3270
+    },
+    {
+      "epoch": 0.09481550762575787,
+      "grad_norm": 0.1998128592967987,
+      "learning_rate": 0.0001983196137570467,
+      "loss": 0.4057304859161377,
+      "step": 3280
+    },
+    {
+      "epoch": 0.09510778724236009,
+      "grad_norm": 0.22602252662181854,
+      "learning_rate": 0.00019830487660103918,
+      "loss": 0.42441368103027344,
+      "step": 3290
+    },
+    {
+      "epoch": 0.0954000668589623,
+      "grad_norm": 0.17054006457328796,
+      "learning_rate": 0.00019829007565606448,
+      "loss": 0.30778021812438966,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09569234647556452,
+      "grad_norm": 0.22141747176647186,
+      "learning_rate": 0.00019827521093172682,
+      "loss": 0.35904369354248045,
+      "step": 3310
+    },
+    {
+      "epoch": 0.09598462609216672,
+      "grad_norm": 0.23470410704612732,
+      "learning_rate": 0.00019826028243767176,
+      "loss": 0.32441959381103513,
+      "step": 3320
+    },
+    {
+      "epoch": 0.09627690570876894,
+      "grad_norm": 0.29219353199005127,
+      "learning_rate": 0.0001982452901835862,
+      "loss": 0.3978474855422974,
+      "step": 3330
+    },
+    {
+      "epoch": 0.09656918532537115,
+      "grad_norm": 0.2738279402256012,
+      "learning_rate": 0.00019823023417919847,
+      "loss": 0.3899239063262939,
+      "step": 3340
+    },
+    {
+      "epoch": 0.09686146494197337,
+      "grad_norm": 0.3130064606666565,
+      "learning_rate": 0.00019821511443427828,
+      "loss": 0.43923206329345704,
+      "step": 3350
+    },
+    {
+      "epoch": 0.09715374455857557,
+      "grad_norm": 0.23388715088367462,
+      "learning_rate": 0.00019819993095863665,
+      "loss": 0.3272917985916138,
+      "step": 3360
+    },
+    {
+      "epoch": 0.0974460241751778,
+      "grad_norm": 0.18713881075382233,
+      "learning_rate": 0.00019818468376212592,
+      "loss": 0.3586859941482544,
+      "step": 3370
+    },
+    {
+      "epoch": 0.09773830379178,
+      "grad_norm": 0.26475366950035095,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 0.3713646173477173,
+      "step": 3380
+    },
+    {
+      "epoch": 0.09803058340838221,
+      "grad_norm": 0.27044302225112915,
+      "learning_rate": 0.00019815399824611363,
+      "loss": 0.3929482936859131,
+      "step": 3390
+    },
+    {
+      "epoch": 0.09832286302498443,
+      "grad_norm": 0.29278746247291565,
+      "learning_rate": 0.00019813855994652355,
+      "loss": 0.35873212814331057,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09861514264158663,
+      "grad_norm": 0.28539928793907166,
+      "learning_rate": 0.00019812305796588738,
+      "loss": 0.3577031135559082,
+      "step": 3410
+    },
+    {
+      "epoch": 0.09890742225818885,
+      "grad_norm": 0.24885481595993042,
+      "learning_rate": 0.00019810749231426422,
+      "loss": 0.34921112060546877,
+      "step": 3420
+    },
+    {
+      "epoch": 0.09919970187479106,
+      "grad_norm": 0.23236370086669922,
+      "learning_rate": 0.00019809186300175443,
+      "loss": 0.4110508918762207,
+      "step": 3430
+    },
+    {
+      "epoch": 0.09949198149139328,
+      "grad_norm": 0.3145415782928467,
+      "learning_rate": 0.00019807617003849968,
+      "loss": 0.4175889015197754,
+      "step": 3440
+    },
+    {
+      "epoch": 0.09978426110799549,
+      "grad_norm": 0.23489882051944733,
+      "learning_rate": 0.00019806041343468304,
+      "loss": 0.3190081834793091,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1000765407245977,
+      "grad_norm": 0.292187362909317,
+      "learning_rate": 0.00019804459320052874,
+      "loss": 0.4160562038421631,
+      "step": 3460
+    },
+    {
+      "epoch": 0.10036882034119991,
+      "grad_norm": 0.2864011526107788,
+      "learning_rate": 0.00019802870934630234,
+      "loss": 0.36125099658966064,
+      "step": 3470
+    },
+    {
+      "epoch": 0.10066109995780213,
+      "grad_norm": 0.23247593641281128,
+      "learning_rate": 0.00019801276188231075,
+      "loss": 0.34271228313446045,
+      "step": 3480
+    },
+    {
+      "epoch": 0.10095337957440434,
+      "grad_norm": 0.3038922846317291,
+      "learning_rate": 0.00019799675081890207,
+      "loss": 0.34409542083740235,
+      "step": 3490
+    },
+    {
+      "epoch": 0.10124565919100656,
+      "grad_norm": 0.306710422039032,
+      "learning_rate": 0.00019798067616646573,
+      "loss": 0.3873798131942749,
+      "step": 3500
+    },
+    {
+      "epoch": 0.10153793880760877,
+      "grad_norm": 0.206655815243721,
+      "learning_rate": 0.00019796453793543238,
+      "loss": 0.35344343185424804,
+      "step": 3510
+    },
+    {
+      "epoch": 0.10183021842421099,
+      "grad_norm": 0.2983470559120178,
+      "learning_rate": 0.00019794833613627394,
+      "loss": 0.3646138906478882,
+      "step": 3520
+    },
+    {
+      "epoch": 0.10212249804081319,
+      "grad_norm": 0.372056245803833,
+      "learning_rate": 0.0001979320707795036,
+      "loss": 0.3840024471282959,
+      "step": 3530
+    },
+    {
+      "epoch": 0.10241477765741541,
+      "grad_norm": 0.24439099431037903,
+      "learning_rate": 0.00019791574187567575,
+      "loss": 0.34989566802978517,
+      "step": 3540
+    },
+    {
+      "epoch": 0.10270705727401762,
+      "grad_norm": 0.26626816391944885,
+      "learning_rate": 0.00019789934943538606,
+      "loss": 0.39730632305145264,
+      "step": 3550
+    },
+    {
+      "epoch": 0.10299933689061984,
+      "grad_norm": 0.3126618564128876,
+      "learning_rate": 0.00019788289346927141,
+      "loss": 0.38080587387084963,
+      "step": 3560
+    },
+    {
+      "epoch": 0.10329161650722204,
+      "grad_norm": 0.2343582957983017,
+      "learning_rate": 0.00019786637398800988,
+      "loss": 0.4155257701873779,
+      "step": 3570
+    },
+    {
+      "epoch": 0.10358389612382427,
+      "grad_norm": 0.21754246950149536,
+      "learning_rate": 0.0001978497910023208,
+      "loss": 0.35129241943359374,
+      "step": 3580
+    },
+    {
+      "epoch": 0.10387617574042647,
+      "grad_norm": 0.3312534987926483,
+      "learning_rate": 0.00019783314452296466,
+      "loss": 0.39340853691101074,
+      "step": 3590
+    },
+    {
+      "epoch": 0.10416845535702869,
+      "grad_norm": 0.2296934276819229,
+      "learning_rate": 0.0001978164345607432,
+      "loss": 0.3211698293685913,
+      "step": 3600
+    },
+    {
+      "epoch": 0.1044607349736309,
+      "grad_norm": 0.2280420958995819,
+      "learning_rate": 0.00019779966112649935,
+      "loss": 0.31526205539703367,
+      "step": 3610
+    },
+    {
+      "epoch": 0.10475301459023312,
+      "grad_norm": 0.24602816998958588,
+      "learning_rate": 0.0001977828242311172,
+      "loss": 0.37086660861968995,
+      "step": 3620
+    },
+    {
+      "epoch": 0.10504529420683532,
+      "grad_norm": 0.27485036849975586,
+      "learning_rate": 0.000197765923885522,
+      "loss": 0.396453332901001,
+      "step": 3630
+    },
+    {
+      "epoch": 0.10533757382343753,
+      "grad_norm": 0.3155587911605835,
+      "learning_rate": 0.0001977489601006802,
+      "loss": 0.32536139488220217,
+      "step": 3640
+    },
+    {
+      "epoch": 0.10562985344003975,
+      "grad_norm": 0.30109748244285583,
+      "learning_rate": 0.00019773193288759945,
+      "loss": 0.4326496601104736,
+      "step": 3650
+    },
+    {
+      "epoch": 0.10592213305664196,
+      "grad_norm": 0.2719879746437073,
+      "learning_rate": 0.00019771484225732854,
+      "loss": 0.35510203838348386,
+      "step": 3660
+    },
+    {
+      "epoch": 0.10621441267324418,
+      "grad_norm": 0.22440284490585327,
+      "learning_rate": 0.00019769768822095732,
+      "loss": 0.32017083168029786,
+      "step": 3670
+    },
+    {
+      "epoch": 0.10650669228984638,
+      "grad_norm": 0.24922069907188416,
+      "learning_rate": 0.00019768047078961694,
+      "loss": 0.3830606460571289,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1067989719064486,
+      "grad_norm": 0.2627232074737549,
+      "learning_rate": 0.00019766318997447956,
+      "loss": 0.39596009254455566,
+      "step": 3690
+    },
+    {
+      "epoch": 0.10709125152305081,
+      "grad_norm": 0.2038666307926178,
+      "learning_rate": 0.00019764584578675848,
+      "loss": 0.3765840530395508,
+      "step": 3700
+    },
+    {
+      "epoch": 0.10738353113965303,
+      "grad_norm": 0.28827524185180664,
+      "learning_rate": 0.00019762843823770823,
+      "loss": 0.36150379180908204,
+      "step": 3710
+    },
+    {
+      "epoch": 0.10767581075625524,
+      "grad_norm": 0.2194090038537979,
+      "learning_rate": 0.00019761096733862435,
+      "loss": 0.3518270969390869,
+      "step": 3720
+    },
+    {
+      "epoch": 0.10796809037285746,
+      "grad_norm": 0.34698304533958435,
+      "learning_rate": 0.0001975934331008435,
+      "loss": 0.4450985431671143,
+      "step": 3730
+    },
+    {
+      "epoch": 0.10826036998945966,
+      "grad_norm": 0.2598106861114502,
+      "learning_rate": 0.00019757583553574345,
+      "loss": 0.4498758792877197,
+      "step": 3740
+    },
+    {
+      "epoch": 0.10855264960606188,
+      "grad_norm": 0.2896166145801544,
+      "learning_rate": 0.0001975581746547431,
+      "loss": 0.38241803646087646,
+      "step": 3750
+    },
+    {
+      "epoch": 0.10884492922266409,
+      "grad_norm": 0.2660869061946869,
+      "learning_rate": 0.00019754045046930239,
+      "loss": 0.3478813648223877,
+      "step": 3760
+    },
+    {
+      "epoch": 0.10913720883926631,
+      "grad_norm": 0.259555846452713,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 0.3579635858535767,
+      "step": 3770
+    },
+    {
+      "epoch": 0.10942948845586851,
+      "grad_norm": 0.3044942319393158,
+      "learning_rate": 0.0001975048122311451,
+      "loss": 0.4593231678009033,
+      "step": 3780
+    },
+    {
+      "epoch": 0.10972176807247074,
+      "grad_norm": 0.26130610704421997,
+      "learning_rate": 0.00019748689820155377,
+      "loss": 0.412431001663208,
+      "step": 3790
+    },
+    {
+      "epoch": 0.11001404768907294,
+      "grad_norm": 0.285075843334198,
+      "learning_rate": 0.0001974689209137726,
+      "loss": 0.36504647731781004,
+      "step": 3800
+    },
+    {
+      "epoch": 0.11030632730567516,
+      "grad_norm": 0.23923102021217346,
+      "learning_rate": 0.00019745088037946684,
+      "loss": 0.40529980659484866,
+      "step": 3810
+    },
+    {
+      "epoch": 0.11059860692227737,
+      "grad_norm": 0.28023964166641235,
+      "learning_rate": 0.00019743277661034286,
+      "loss": 0.4191715717315674,
+      "step": 3820
+    },
+    {
+      "epoch": 0.11089088653887959,
+      "grad_norm": 0.3470838665962219,
+      "learning_rate": 0.0001974146096181479,
+      "loss": 0.35360565185546877,
+      "step": 3830
+    },
+    {
+      "epoch": 0.1111831661554818,
+      "grad_norm": 0.2091302126646042,
+      "learning_rate": 0.0001973963794146704,
+      "loss": 0.37881152629852294,
+      "step": 3840
+    },
+    {
+      "epoch": 0.11147544577208401,
+      "grad_norm": 0.26603418588638306,
+      "learning_rate": 0.0001973780860117397,
+      "loss": 0.3370387315750122,
+      "step": 3850
+    },
+    {
+      "epoch": 0.11176772538868622,
+      "grad_norm": 0.23352454602718353,
+      "learning_rate": 0.0001973597294212262,
+      "loss": 0.40698604583740233,
+      "step": 3860
+    },
+    {
+      "epoch": 0.11206000500528844,
+      "grad_norm": 0.25433021783828735,
+      "learning_rate": 0.00019734130965504132,
+      "loss": 0.3438957929611206,
+      "step": 3870
+    },
+    {
+      "epoch": 0.00011691184664088517,
+      "grad_norm": 0.24003185331821442,
+      "learning_rate": 0.00019732282672513742,
+      "loss": 0.4005417823791504,
+      "step": 3880
+    },
+    {
+      "epoch": 0.0004091914632430981,
+      "grad_norm": 0.25248992443084717,
+      "learning_rate": 0.00019730428064350784,
+      "loss": 0.4049844741821289,
+      "step": 3890
+    },
+    {
+      "epoch": 0.000701471079845311,
+      "grad_norm": 0.3725515902042389,
+      "learning_rate": 0.00019728567142218703,
+      "loss": 0.36194710731506347,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0009937506964475238,
+      "grad_norm": 0.2143525928258896,
+      "learning_rate": 0.00019726699907325021,
+      "loss": 0.3787307500839233,
+      "step": 3910
+    },
+    {
+      "epoch": 0.0012860303130497368,
+      "grad_norm": 0.3514223098754883,
+      "learning_rate": 0.00019724826360881373,
+      "loss": 0.33108208179473875,
+      "step": 3920
+    },
+    {
+      "epoch": 0.0015783099296519498,
+      "grad_norm": 0.3794527053833008,
+      "learning_rate": 0.00019722946504103478,
+      "loss": 0.362584662437439,
+      "step": 3930
+    },
+    {
+      "epoch": 0.0018705895462541628,
+      "grad_norm": 0.2668137848377228,
+      "learning_rate": 0.00019721060338211155,
+      "loss": 0.37544841766357423,
+      "step": 3940
+    },
+    {
+      "epoch": 0.0021628691628563755,
+      "grad_norm": 0.286663681268692,
+      "learning_rate": 0.00019719167864428323,
+      "loss": 0.36108510494232177,
+      "step": 3950
+    },
+    {
+      "epoch": 0.0024551487794585887,
+      "grad_norm": 0.2159728705883026,
+      "learning_rate": 0.0001971726908398298,
+      "loss": 0.37459230422973633,
+      "step": 3960
+    },
+    {
+      "epoch": 0.0027474283960608015,
+      "grad_norm": 0.2895173132419586,
+      "learning_rate": 0.00019715363998107227,
+      "loss": 0.3811808109283447,
+      "step": 3970
+    },
+    {
+      "epoch": 0.003039708012663014,
+      "grad_norm": 0.23048807680606842,
+      "learning_rate": 0.00019713452608037258,
+      "loss": 0.42634239196777346,
+      "step": 3980
+    },
+    {
+      "epoch": 0.0033319876292652274,
+      "grad_norm": 0.2901408076286316,
+      "learning_rate": 0.00019711534915013346,
+      "loss": 0.37782447338104247,
+      "step": 3990
+    },
+    {
+      "epoch": 0.00362426724586744,
+      "grad_norm": 0.19619092345237732,
+      "learning_rate": 0.00019709610920279867,
+      "loss": 0.3980048894882202,
+      "step": 4000
+    },
+    {
+      "epoch": 0.003916546862469653,
+      "grad_norm": 0.22392375767230988,
+      "learning_rate": 0.00019707680625085275,
+      "loss": 0.3944056987762451,
+      "step": 4010
+    },
+    {
+      "epoch": 0.004208826479071866,
+      "grad_norm": 0.2043333351612091,
+      "learning_rate": 0.0001970574403068212,
+      "loss": 0.3910648345947266,
+      "step": 4020
+    },
+    {
+      "epoch": 0.004501106095674079,
+      "grad_norm": 0.2545478641986847,
+      "learning_rate": 0.00019703801138327038,
+      "loss": 0.37434265613555906,
+      "step": 4030
+    },
+    {
+      "epoch": 0.004793385712276292,
+      "grad_norm": 0.21100255846977234,
+      "learning_rate": 0.0001970185194928075,
+      "loss": 0.3273258924484253,
+      "step": 4040
+    },
+    {
+      "epoch": 0.005085665328878505,
+      "grad_norm": 0.20953594148159027,
+      "learning_rate": 0.00019699896464808064,
+      "loss": 0.3656395673751831,
+      "step": 4050
+    },
+    {
+      "epoch": 0.005377944945480718,
+      "grad_norm": 0.19825880229473114,
+      "learning_rate": 0.0001969793468617787,
+      "loss": 0.3621396541595459,
+      "step": 4060
+    },
+    {
+      "epoch": 0.00567022456208293,
+      "grad_norm": 0.36366453766822815,
+      "learning_rate": 0.00019695966614663148,
+      "loss": 0.3576923131942749,
+      "step": 4070
+    },
+    {
+      "epoch": 0.0059625041786851435,
+      "grad_norm": 0.2716172933578491,
+      "learning_rate": 0.0001969399225154096,
+      "loss": 0.4180868625640869,
+      "step": 4080
+    },
+    {
+      "epoch": 0.006254783795287357,
+      "grad_norm": 0.3227168023586273,
+      "learning_rate": 0.00019692011598092444,
+      "loss": 0.335027289390564,
+      "step": 4090
+    },
+    {
+      "epoch": 0.00654706341188957,
+      "grad_norm": 0.2741522192955017,
+      "learning_rate": 0.00019690024655602827,
+      "loss": 0.39731588363647463,
+      "step": 4100
+    },
+    {
+      "epoch": 0.006839343028491782,
+      "grad_norm": 0.28186601400375366,
+      "learning_rate": 0.00019688031425361415,
+      "loss": 0.4415407657623291,
+      "step": 4110
+    },
+    {
+      "epoch": 0.007131622645093995,
+      "grad_norm": 0.30918681621551514,
+      "learning_rate": 0.0001968603190866159,
+      "loss": 0.381803035736084,
+      "step": 4120
+    },
+    {
+      "epoch": 0.0074239022616962086,
+      "grad_norm": 0.264800101518631,
+      "learning_rate": 0.0001968402610680082,
+      "loss": 0.3152090787887573,
+      "step": 4130
+    },
+    {
+      "epoch": 0.007716181878298421,
+      "grad_norm": 0.32021039724349976,
+      "learning_rate": 0.00019682014021080648,
+      "loss": 0.316313648223877,
+      "step": 4140
+    },
+    {
+      "epoch": 0.008008461494900634,
+      "grad_norm": 0.3371167778968811,
+      "learning_rate": 0.00019679995652806693,
+      "loss": 0.4139958381652832,
+      "step": 4150
+    },
+    {
+      "epoch": 0.008300741111502847,
+      "grad_norm": 0.2458915114402771,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 0.3321307897567749,
+      "step": 4160
+    },
+    {
+      "epoch": 0.00859302072810506,
+      "grad_norm": 0.2569628357887268,
+      "learning_rate": 0.00019675940073840304,
+      "loss": 0.3787196397781372,
+      "step": 4170
+    },
+    {
+      "epoch": 0.008885300344707274,
+      "grad_norm": 0.2238645851612091,
+      "learning_rate": 0.0001967390286577949,
+      "loss": 0.3773963212966919,
+      "step": 4180
+    },
+    {
+      "epoch": 0.009177579961309485,
+      "grad_norm": 0.24416768550872803,
+      "learning_rate": 0.00019671859380428135,
+      "loss": 0.3781037092208862,
+      "step": 4190
+    },
+    {
+      "epoch": 0.009469859577911698,
+      "grad_norm": 0.22842183709144592,
+      "learning_rate": 0.00019669809619112232,
+      "loss": 0.32853288650512696,
+      "step": 4200
+    },
+    {
+      "epoch": 0.009762139194513911,
+      "grad_norm": 0.2599496841430664,
+      "learning_rate": 0.0001966775358316185,
+      "loss": 0.3580941200256348,
+      "step": 4210
+    },
+    {
+      "epoch": 0.010054418811116125,
+      "grad_norm": 0.24886325001716614,
+      "learning_rate": 0.0001966569127391113,
+      "loss": 0.41251273155212403,
+      "step": 4220
+    },
+    {
+      "epoch": 0.010346698427718338,
+      "grad_norm": 0.2446001172065735,
+      "learning_rate": 0.00019663622692698285,
+      "loss": 0.34038605690002444,
+      "step": 4230
+    },
+    {
+      "epoch": 0.010638978044320551,
+      "grad_norm": 0.4652605950832367,
+      "learning_rate": 0.00019661547840865587,
+      "loss": 0.38447444438934325,
+      "step": 4240
+    },
+    {
+      "epoch": 0.010931257660922762,
+      "grad_norm": 0.2170983999967575,
+      "learning_rate": 0.0001965946671975939,
+      "loss": 0.3797696352005005,
+      "step": 4250
+    },
+    {
+      "epoch": 0.011223537277524976,
+      "grad_norm": 0.2702181041240692,
+      "learning_rate": 0.0001965737933073011,
+      "loss": 0.3525629758834839,
+      "step": 4260
+    },
+    {
+      "epoch": 0.011515816894127189,
+      "grad_norm": 0.3854679465293884,
+      "learning_rate": 0.00019655285675132232,
+      "loss": 0.4074872016906738,
+      "step": 4270
+    },
+    {
+      "epoch": 0.011808096510729402,
+      "grad_norm": 0.2575705349445343,
+      "learning_rate": 0.00019653185754324305,
+      "loss": 0.3649637222290039,
+      "step": 4280
+    },
+    {
+      "epoch": 0.012100376127331615,
+      "grad_norm": 0.2799028158187866,
+      "learning_rate": 0.00019651079569668945,
+      "loss": 0.32631697654724123,
+      "step": 4290
+    },
+    {
+      "epoch": 0.012392655743933828,
+      "grad_norm": 0.23203958570957184,
+      "learning_rate": 0.0001964896712253283,
+      "loss": 0.3405325174331665,
+      "step": 4300
+    },
+    {
+      "epoch": 0.012684935360536042,
+      "grad_norm": 0.2742440104484558,
+      "learning_rate": 0.0001964684841428671,
+      "loss": 0.380454421043396,
+      "step": 4310
+    },
+    {
+      "epoch": 0.012977214977138253,
+      "grad_norm": 0.24825844168663025,
+      "learning_rate": 0.00019644723446305386,
+      "loss": 0.3005128145217896,
+      "step": 4320
+    },
+    {
+      "epoch": 0.013269494593740466,
+      "grad_norm": 0.2241649329662323,
+      "learning_rate": 0.0001964259221996773,
+      "loss": 0.37446155548095705,
+      "step": 4330
+    },
+    {
+      "epoch": 0.01356177421034268,
+      "grad_norm": 0.23571254312992096,
+      "learning_rate": 0.0001964045473665667,
+      "loss": 0.36955699920654295,
+      "step": 4340
+    },
+    {
+      "epoch": 0.013854053826944893,
+      "grad_norm": 0.25073108077049255,
+      "learning_rate": 0.00019638310997759195,
+      "loss": 0.3190394639968872,
+      "step": 4350
+    },
+    {
+      "epoch": 0.014146333443547106,
+      "grad_norm": 0.3259909451007843,
+      "learning_rate": 0.00019636161004666356,
+      "loss": 0.3114713191986084,
+      "step": 4360
+    },
+    {
+      "epoch": 0.014438613060149319,
+      "grad_norm": 0.2959684729576111,
+      "learning_rate": 0.00019634004758773262,
+      "loss": 0.35817339420318606,
+      "step": 4370
+    },
+    {
+      "epoch": 0.01473089267675153,
+      "grad_norm": 0.24469250440597534,
+      "learning_rate": 0.00019631842261479072,
+      "loss": 0.35485110282897947,
+      "step": 4380
+    },
+    {
+      "epoch": 0.015023172293353744,
+      "grad_norm": 0.2899954617023468,
+      "learning_rate": 0.00019629673514187011,
+      "loss": 0.3684756517410278,
+      "step": 4390
+    },
+    {
+      "epoch": 0.015315451909955957,
+      "grad_norm": 0.2324911206960678,
+      "learning_rate": 0.00019627498518304354,
+      "loss": 0.3869553327560425,
+      "step": 4400
+    },
+    {
+      "epoch": 0.01560773152655817,
+      "grad_norm": 0.2527233064174652,
+      "learning_rate": 0.00019625317275242434,
+      "loss": 0.3787883520126343,
+      "step": 4410
+    },
+    {
+      "epoch": 0.01590001114316038,
+      "grad_norm": 0.2540733814239502,
+      "learning_rate": 0.00019623129786416635,
+      "loss": 0.3535881757736206,
+      "step": 4420
+    },
+    {
+      "epoch": 0.016192290759762595,
+      "grad_norm": 0.23976242542266846,
+      "learning_rate": 0.00019620936053246391,
+      "loss": 0.3950831890106201,
+      "step": 4430
+    },
+    {
+      "epoch": 0.016484570376364808,
+      "grad_norm": 0.2642195224761963,
+      "learning_rate": 0.000196187360771552,
+      "loss": 0.40725021362304686,
+      "step": 4440
+    },
+    {
+      "epoch": 0.01677684999296702,
+      "grad_norm": 0.23640264570713043,
+      "learning_rate": 0.00019616529859570593,
+      "loss": 0.3402501344680786,
+      "step": 4450
+    },
+    {
+      "epoch": 0.017069129609569234,
+      "grad_norm": 0.3188447058200836,
+      "learning_rate": 0.0001961431740192417,
+      "loss": 0.4290135383605957,
+      "step": 4460
+    },
+    {
+      "epoch": 0.017361409226171447,
+      "grad_norm": 0.37905675172805786,
+      "learning_rate": 0.00019612098705651564,
+      "loss": 0.42688684463500975,
+      "step": 4470
+    },
+    {
+      "epoch": 0.01765368884277366,
+      "grad_norm": 0.275573194026947,
+      "learning_rate": 0.0001960987377219246,
+      "loss": 0.31829161643981935,
+      "step": 4480
+    },
+    {
+      "epoch": 0.017945968459375874,
+      "grad_norm": 0.2890515923500061,
+      "learning_rate": 0.00019607642602990604,
+      "loss": 0.39717109203338624,
+      "step": 4490
+    },
+    {
+      "epoch": 0.018238248075978087,
+      "grad_norm": 0.27289193868637085,
+      "learning_rate": 0.00019605405199493768,
+      "loss": 0.3399673938751221,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0185305276925803,
+      "grad_norm": 0.2542383372783661,
+      "learning_rate": 0.00019603161563153788,
+      "loss": 0.3713631868362427,
+      "step": 4510
+    },
+    {
+      "epoch": 0.018822807309182513,
+      "grad_norm": 0.27196621894836426,
+      "learning_rate": 0.00019600911695426528,
+      "loss": 0.36151149272918703,
+      "step": 4520
+    },
+    {
+      "epoch": 0.019115086925784727,
+      "grad_norm": 0.32463765144348145,
+      "learning_rate": 0.000195986555977719,
+      "loss": 0.38517782688140867,
+      "step": 4530
+    },
+    {
+      "epoch": 0.01940736654238694,
+      "grad_norm": 0.22708411514759064,
+      "learning_rate": 0.00019596393271653867,
+      "loss": 0.3831400156021118,
+      "step": 4540
+    },
+    {
+      "epoch": 0.01969964615898915,
+      "grad_norm": 0.33081939816474915,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 0.3784955978393555,
+      "step": 4550
+    },
+    {
+      "epoch": 0.019991925775591363,
+      "grad_norm": 0.1940586268901825,
+      "learning_rate": 0.0001959184993990362,
+      "loss": 0.33077552318573,
+      "step": 4560
+    },
+    {
+      "epoch": 0.020284205392193576,
+      "grad_norm": 0.26001009345054626,
+      "learning_rate": 0.00019589568937219528,
+      "loss": 0.3919029474258423,
+      "step": 4570
+    },
+    {
+      "epoch": 0.02057648500879579,
+      "grad_norm": 0.23949424922466278,
+      "learning_rate": 0.00019587281711968258,
+      "loss": 0.2938846588134766,
+      "step": 4580
+    },
+    {
+      "epoch": 0.020868764625398002,
+      "grad_norm": 0.39603474736213684,
+      "learning_rate": 0.00019584988265633978,
+      "loss": 0.3431025505065918,
+      "step": 4590
+    },
+    {
+      "epoch": 0.021161044242000215,
+      "grad_norm": 0.29849958419799805,
+      "learning_rate": 0.00019582688599704872,
+      "loss": 0.3061420440673828,
+      "step": 4600
+    },
+    {
+      "epoch": 0.02145332385860243,
+      "grad_norm": 0.23199288547039032,
+      "learning_rate": 0.00019580382715673174,
+      "loss": 0.3330970287322998,
+      "step": 4610
+    },
+    {
+      "epoch": 0.021745603475204642,
+      "grad_norm": 0.21988391876220703,
+      "learning_rate": 0.00019578070615035144,
+      "loss": 0.32508776187896726,
+      "step": 4620
+    },
+    {
+      "epoch": 0.022037883091806855,
+      "grad_norm": 0.2894282639026642,
+      "learning_rate": 0.00019575752299291077,
+      "loss": 0.34811713695526125,
+      "step": 4630
+    },
+    {
+      "epoch": 0.022330162708409068,
+      "grad_norm": 0.2760803997516632,
+      "learning_rate": 0.00019573427769945307,
+      "loss": 0.29162118434906004,
+      "step": 4640
+    },
+    {
+      "epoch": 0.02262244232501128,
+      "grad_norm": 0.2863601744174957,
+      "learning_rate": 0.00019571097028506197,
+      "loss": 0.3005183458328247,
+      "step": 4650
+    },
+    {
+      "epoch": 0.022914721941613495,
+      "grad_norm": 0.32255980372428894,
+      "learning_rate": 0.00019568760076486133,
+      "loss": 0.27859556674957275,
+      "step": 4660
+    },
+    {
+      "epoch": 0.023207001558215708,
+      "grad_norm": 0.24672210216522217,
+      "learning_rate": 0.00019566416915401544,
+      "loss": 0.3426859378814697,
+      "step": 4670
+    },
+    {
+      "epoch": 0.023499281174817917,
+      "grad_norm": 0.34361395239830017,
+      "learning_rate": 0.00019564067546772878,
+      "loss": 0.3522571325302124,
+      "step": 4680
+    },
+    {
+      "epoch": 0.02379156079142013,
+      "grad_norm": 0.36237573623657227,
+      "learning_rate": 0.00019561711972124616,
+      "loss": 0.266711950302124,
+      "step": 4690
+    },
+    {
+      "epoch": 0.024083840408022344,
+      "grad_norm": 0.23873119056224823,
+      "learning_rate": 0.0001955935019298527,
+      "loss": 0.3517559289932251,
+      "step": 4700
+    },
+    {
+      "epoch": 0.024376120024624557,
+      "grad_norm": 0.33018001914024353,
+      "learning_rate": 0.00019556982210887364,
+      "loss": 0.3424465894699097,
+      "step": 4710
+    },
+    {
+      "epoch": 0.02466839964122677,
+      "grad_norm": 0.23929226398468018,
+      "learning_rate": 0.0001955460802736746,
+      "loss": 0.2846285581588745,
+      "step": 4720
+    },
+    {
+      "epoch": 0.024960679257828983,
+      "grad_norm": 0.32132086157798767,
+      "learning_rate": 0.00019552227643966142,
+      "loss": 0.35615015029907227,
+      "step": 4730
+    },
+    {
+      "epoch": 0.025252958874431197,
+      "grad_norm": 0.3130735754966736,
+      "learning_rate": 0.0001954984106222801,
+      "loss": 0.31338694095611574,
+      "step": 4740
+    },
+    {
+      "epoch": 0.02554523849103341,
+      "grad_norm": 0.39619776606559753,
+      "learning_rate": 0.00019547448283701698,
+      "loss": 0.37840611934661866,
+      "step": 4750
+    },
+    {
+      "epoch": 0.025837518107635623,
+      "grad_norm": 0.3082350492477417,
+      "learning_rate": 0.00019545049309939848,
+      "loss": 0.2591458082199097,
+      "step": 4760
+    },
+    {
+      "epoch": 0.026129797724237836,
+      "grad_norm": 0.35461875796318054,
+      "learning_rate": 0.0001954264414249913,
+      "loss": 0.35952343940734866,
+      "step": 4770
+    },
+    {
+      "epoch": 0.02642207734084005,
+      "grad_norm": 0.42175233364105225,
+      "learning_rate": 0.00019540232782940228,
+      "loss": 0.3499655246734619,
+      "step": 4780
+    },
+    {
+      "epoch": 0.026714356957442262,
+      "grad_norm": 0.3273662030696869,
+      "learning_rate": 0.00019537815232827853,
+      "loss": 0.23040478229522704,
+      "step": 4790
+    },
+    {
+      "epoch": 0.027006636574044476,
+      "grad_norm": 0.2945111095905304,
+      "learning_rate": 0.00019535391493730723,
+      "loss": 0.31700570583343507,
+      "step": 4800
+    },
+    {
+      "epoch": 0.027298916190646685,
+      "grad_norm": 0.30250272154808044,
+      "learning_rate": 0.00019532961567221576,
+      "loss": 0.3476430416107178,
+      "step": 4810
+    },
+    {
+      "epoch": 0.0275911958072489,
+      "grad_norm": 0.2619974911212921,
+      "learning_rate": 0.00019530525454877168,
+      "loss": 0.30992684364318845,
+      "step": 4820
+    },
+    {
+      "epoch": 0.027883475423851112,
+      "grad_norm": 0.3181731402873993,
+      "learning_rate": 0.00019528083158278263,
+      "loss": 0.3028948068618774,
+      "step": 4830
+    },
+    {
+      "epoch": 0.028175755040453325,
+      "grad_norm": 0.4383261203765869,
+      "learning_rate": 0.00019525634679009644,
+      "loss": 0.32675282955169677,
+      "step": 4840
+    },
+    {
+      "epoch": 0.028468034657055538,
+      "grad_norm": 0.328757643699646,
+      "learning_rate": 0.00019523180018660096,
+      "loss": 0.30401685237884524,
+      "step": 4850
+    },
+    {
+      "epoch": 0.02876031427365775,
+      "grad_norm": 0.3134559392929077,
+      "learning_rate": 0.00019520719178822428,
+      "loss": 0.3300363063812256,
+      "step": 4860
+    },
+    {
+      "epoch": 0.029052593890259965,
+      "grad_norm": 0.2833344042301178,
+      "learning_rate": 0.00019518252161093447,
+      "loss": 0.3005764245986938,
+      "step": 4870
+    },
+    {
+      "epoch": 0.029344873506862178,
+      "grad_norm": 0.31576523184776306,
+      "learning_rate": 0.0001951577896707398,
+      "loss": 0.33509833812713624,
+      "step": 4880
+    },
+    {
+      "epoch": 0.02963715312346439,
+      "grad_norm": 0.3154706656932831,
+      "learning_rate": 0.0001951329959836885,
+      "loss": 0.33233764171600344,
+      "step": 4890
+    },
+    {
+      "epoch": 0.029929432740066604,
+      "grad_norm": 0.2957933247089386,
+      "learning_rate": 0.00019510814056586897,
+      "loss": 0.33685870170593263,
+      "step": 4900
+    },
+    {
+      "epoch": 0.030221712356668817,
+      "grad_norm": 0.3386175036430359,
+      "learning_rate": 0.00019508322343340958,
+      "loss": 0.3845832824707031,
+      "step": 4910
+    },
+    {
+      "epoch": 0.03051399197327103,
+      "grad_norm": 0.2883608043193817,
+      "learning_rate": 0.00019505824460247886,
+      "loss": 0.34288792610168456,
+      "step": 4920
+    },
+    {
+      "epoch": 0.030806271589873244,
+      "grad_norm": 0.29853203892707825,
+      "learning_rate": 0.0001950332040892852,
+      "loss": 0.2630131483078003,
+      "step": 4930
+    },
+    {
+      "epoch": 0.031098551206475453,
+      "grad_norm": 0.3084704875946045,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 0.28366527557373045,
+      "step": 4940
+    },
+    {
+      "epoch": 0.03139083082307767,
+      "grad_norm": 0.310249388217926,
+      "learning_rate": 0.00019498293808114332,
+      "loss": 0.3226843595504761,
+      "step": 4950
+    },
+    {
+      "epoch": 0.03168311043967988,
+      "grad_norm": 0.32374560832977295,
+      "learning_rate": 0.00019495771261881216,
+      "loss": 0.32280118465423585,
+      "step": 4960
+    },
+    {
+      "epoch": 0.031975390056282096,
+      "grad_norm": 0.345403790473938,
+      "learning_rate": 0.0001949324255394522,
+      "loss": 0.23969836235046388,
+      "step": 4970
+    },
+    {
+      "epoch": 0.03226766967288431,
+      "grad_norm": 0.27165737748146057,
+      "learning_rate": 0.00019490707685947202,
+      "loss": 0.28694493770599366,
+      "step": 4980
+    },
+    {
+      "epoch": 0.03255994928948652,
+      "grad_norm": 0.24607214331626892,
+      "learning_rate": 0.00019488166659532,
+      "loss": 0.3478991985321045,
+      "step": 4990
+    },
+    {
+      "epoch": 0.03285222890608873,
+      "grad_norm": 0.3036491870880127,
+      "learning_rate": 0.0001948561947634847,
+      "loss": 0.2748469591140747,
+      "step": 5000
+    },
+    {
+      "epoch": 0.03314450852269094,
+      "grad_norm": 0.3728741407394409,
+      "learning_rate": 0.0001948306613804944,
+      "loss": 0.3124251842498779,
+      "step": 5010
+    },
+    {
+      "epoch": 0.033436788139293155,
+      "grad_norm": 0.41183504462242126,
+      "learning_rate": 0.00019480506646291752,
+      "loss": 0.3081713914871216,
+      "step": 5020
+    },
+    {
+      "epoch": 0.03372906775589537,
+      "grad_norm": 0.2485785037279129,
+      "learning_rate": 0.00019477941002736228,
+      "loss": 0.3440422773361206,
+      "step": 5030
+    },
+    {
+      "epoch": 0.03402134737249758,
+      "grad_norm": 0.3907031714916229,
+      "learning_rate": 0.00019475369209047693,
+      "loss": 0.29272844791412356,
+      "step": 5040
+    },
+    {
+      "epoch": 0.034313626989099795,
+      "grad_norm": 0.2438102811574936,
+      "learning_rate": 0.00019472791266894946,
+      "loss": 0.27734739780426027,
+      "step": 5050
+    },
+    {
+      "epoch": 0.03460590660570201,
+      "grad_norm": 0.24713954329490662,
+      "learning_rate": 0.00019470207177950792,
+      "loss": 0.3164196252822876,
+      "step": 5060
+    },
+    {
+      "epoch": 0.03489818622230422,
+      "grad_norm": 0.35341203212738037,
+      "learning_rate": 0.0001946761694389202,
+      "loss": 0.3532915830612183,
+      "step": 5070
+    },
+    {
+      "epoch": 0.035190465838906435,
+      "grad_norm": 0.25562161207199097,
+      "learning_rate": 0.00019465020566399393,
+      "loss": 0.26654715538024903,
+      "step": 5080
+    },
+    {
+      "epoch": 0.03548274545550865,
+      "grad_norm": 0.28326499462127686,
+      "learning_rate": 0.00019462418047157688,
+      "loss": 0.33709533214569093,
+      "step": 5090
+    },
+    {
+      "epoch": 0.03577502507211086,
+      "grad_norm": 0.2742396891117096,
+      "learning_rate": 0.00019459809387855641,
+      "loss": 0.3669175624847412,
+      "step": 5100
+    },
+    {
+      "epoch": 0.036067304688713074,
+      "grad_norm": 0.3157244026660919,
+      "learning_rate": 0.00019457194590185987,
+      "loss": 0.3240370273590088,
+      "step": 5110
+    },
+    {
+      "epoch": 0.03635958430531529,
+      "grad_norm": 0.28601714968681335,
+      "learning_rate": 0.00019454573655845438,
+      "loss": 0.24612798690795898,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0366518639219175,
+      "grad_norm": 0.28745201230049133,
+      "learning_rate": 0.0001945194658653469,
+      "loss": 0.2592262506484985,
+      "step": 5130
+    },
+    {
+      "epoch": 0.036944143538519714,
+      "grad_norm": 0.3364482820034027,
+      "learning_rate": 0.00019449313383958423,
+      "loss": 0.3229775667190552,
+      "step": 5140
+    },
+    {
+      "epoch": 0.03723642315512193,
+      "grad_norm": 0.26464542746543884,
+      "learning_rate": 0.00019446674049825288,
+      "loss": 0.3347605228424072,
+      "step": 5150
+    },
+    {
+      "epoch": 0.03752870277172414,
+      "grad_norm": 0.2748246490955353,
+      "learning_rate": 0.00019444028585847924,
+      "loss": 0.3474790096282959,
+      "step": 5160
+    },
+    {
+      "epoch": 0.03782098238832635,
+      "grad_norm": 0.29505252838134766,
+      "learning_rate": 0.00019441376993742946,
+      "loss": 0.2775235176086426,
+      "step": 5170
+    },
+    {
+      "epoch": 0.038113262004928566,
+      "grad_norm": 0.42069968581199646,
+      "learning_rate": 0.00019438719275230936,
+      "loss": 0.369965648651123,
+      "step": 5180
+    },
+    {
+      "epoch": 0.03840554162153078,
+      "grad_norm": 0.34723401069641113,
+      "learning_rate": 0.00019436055432036465,
+      "loss": 0.2996554136276245,
+      "step": 5190
+    },
+    {
+      "epoch": 0.03869782123813299,
+      "grad_norm": 0.3135107457637787,
+      "learning_rate": 0.0001943338546588807,
+      "loss": 0.30053389072418213,
+      "step": 5200
+    },
+    {
+      "epoch": 0.038990100854735206,
+      "grad_norm": 0.26011621952056885,
+      "learning_rate": 0.00019430709378518262,
+      "loss": 0.2881584405899048,
+      "step": 5210
+    },
+    {
+      "epoch": 0.03928238047133742,
+      "grad_norm": 0.23041443526744843,
+      "learning_rate": 0.00019428027171663527,
+      "loss": 0.2882877826690674,
+      "step": 5220
+    },
+    {
+      "epoch": 0.03957466008793963,
+      "grad_norm": 0.24758175015449524,
+      "learning_rate": 0.00019425338847064318,
+      "loss": 0.329343581199646,
+      "step": 5230
+    },
+    {
+      "epoch": 0.039866939704541846,
+      "grad_norm": 0.41276100277900696,
+      "learning_rate": 0.00019422644406465062,
+      "loss": 0.3012660026550293,
+      "step": 5240
+    },
+    {
+      "epoch": 0.04015921932114406,
+      "grad_norm": 0.34703317284584045,
+      "learning_rate": 0.00019419943851614152,
+      "loss": 0.4001108169555664,
+      "step": 5250
+    },
+    {
+      "epoch": 0.040451498937746265,
+      "grad_norm": 0.2764207124710083,
+      "learning_rate": 0.00019417237184263944,
+      "loss": 0.25676229000091555,
+      "step": 5260
+    },
+    {
+      "epoch": 0.04074377855434848,
+      "grad_norm": 0.3090943396091461,
+      "learning_rate": 0.0001941452440617077,
+      "loss": 0.3067681312561035,
+      "step": 5270
+    },
+    {
+      "epoch": 0.04103605817095069,
+      "grad_norm": 0.4089989960193634,
+      "learning_rate": 0.0001941180551909492,
+      "loss": 0.34026598930358887,
+      "step": 5280
+    },
+    {
+      "epoch": 0.041328337787552905,
+      "grad_norm": 0.3696897625923157,
+      "learning_rate": 0.00019409080524800655,
+      "loss": 0.294386625289917,
+      "step": 5290
+    },
+    {
+      "epoch": 0.04162061740415512,
+      "grad_norm": 0.26694053411483765,
+      "learning_rate": 0.00019406349425056188,
+      "loss": 0.391556978225708,
+      "step": 5300
+    },
+    {
+      "epoch": 0.04191289702075733,
+      "grad_norm": 0.3776414692401886,
+      "learning_rate": 0.000194036122216337,
+      "loss": 0.25552499294281006,
+      "step": 5310
+    },
+    {
+      "epoch": 0.042205176637359544,
+      "grad_norm": 0.2968269884586334,
+      "learning_rate": 0.00019400868916309336,
+      "loss": 0.36509559154510496,
+      "step": 5320
+    },
+    {
+      "epoch": 0.04249745625396176,
+      "grad_norm": 0.3928321599960327,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 0.35145699977874756,
+      "step": 5330
+    },
+    {
+      "epoch": 0.04278973587056397,
+      "grad_norm": 0.25441688299179077,
+      "learning_rate": 0.00019395364007079342,
+      "loss": 0.35477218627929685,
+      "step": 5340
+    },
+    {
+      "epoch": 0.043082015487166184,
+      "grad_norm": 0.35719209909439087,
+      "learning_rate": 0.00019392602406745785,
+      "loss": 0.3023816108703613,
+      "step": 5350
+    },
+    {
+      "epoch": 0.0433742951037684,
+      "grad_norm": 0.2842925488948822,
+      "learning_rate": 0.00019389834711654503,
+      "loss": 0.283978271484375,
+      "step": 5360
+    },
+    {
+      "epoch": 0.04366657472037061,
+      "grad_norm": 0.25824466347694397,
+      "learning_rate": 0.00019387060923601416,
+      "loss": 0.2891792058944702,
+      "step": 5370
+    },
+    {
+      "epoch": 0.04395885433697282,
+      "grad_norm": 0.24369360506534576,
+      "learning_rate": 0.00019384281044386415,
+      "loss": 0.3169957876205444,
+      "step": 5380
+    },
+    {
+      "epoch": 0.044251133953575036,
+      "grad_norm": 0.2803765833377838,
+      "learning_rate": 0.00019381495075813324,
+      "loss": 0.27735812664031984,
+      "step": 5390
+    },
+    {
+      "epoch": 0.04454341357017725,
+      "grad_norm": 0.2664027512073517,
+      "learning_rate": 0.00019378703019689932,
+      "loss": 0.32606966495513917,
+      "step": 5400
+    },
+    {
+      "epoch": 0.04483569318677946,
+      "grad_norm": 0.25099697709083557,
+      "learning_rate": 0.00019375904877827978,
+      "loss": 0.3093207597732544,
+      "step": 5410
+    },
+    {
+      "epoch": 0.045127972803381676,
+      "grad_norm": 0.29991742968559265,
+      "learning_rate": 0.00019373100652043138,
+      "loss": 0.20303540229797362,
+      "step": 5420
+    },
+    {
+      "epoch": 0.04542025241998389,
+      "grad_norm": 0.3771480917930603,
+      "learning_rate": 0.00019370290344155046,
+      "loss": 0.2821071147918701,
+      "step": 5430
+    },
+    {
+      "epoch": 0.0457125320365861,
+      "grad_norm": 0.3615019619464874,
+      "learning_rate": 0.00019367473955987283,
+      "loss": 0.2755707263946533,
+      "step": 5440
+    },
+    {
+      "epoch": 0.046004811653188316,
+      "grad_norm": 0.3952086269855499,
+      "learning_rate": 0.00019364651489367371,
+      "loss": 0.28334860801696776,
+      "step": 5450
+    },
+    {
+      "epoch": 0.04629709126979053,
+      "grad_norm": 0.3415733575820923,
+      "learning_rate": 0.0001936182294612678,
+      "loss": 0.38788304328918455,
+      "step": 5460
+    },
+    {
+      "epoch": 0.04658937088639274,
+      "grad_norm": 0.36460474133491516,
+      "learning_rate": 0.00019358988328100916,
+      "loss": 0.3156550645828247,
+      "step": 5470
+    },
+    {
+      "epoch": 0.046881650502994955,
+      "grad_norm": 0.24899375438690186,
+      "learning_rate": 0.00019356147637129137,
+      "loss": 0.35087759494781495,
+      "step": 5480
+    },
+    {
+      "epoch": 0.04717393011959717,
+      "grad_norm": 0.3454049229621887,
+      "learning_rate": 0.00019353300875054732,
+      "loss": 0.3055156946182251,
+      "step": 5490
+    },
+    {
+      "epoch": 0.04746620973619938,
+      "grad_norm": 0.36888831853866577,
+      "learning_rate": 0.00019350448043724941,
+      "loss": 0.35781381130218504,
+      "step": 5500
+    },
+    {
+      "epoch": 0.047758489352801595,
+      "grad_norm": 0.3067270517349243,
+      "learning_rate": 0.00019347589144990925,
+      "loss": 0.2821542263031006,
+      "step": 5510
+    },
+    {
+      "epoch": 0.0480507689694038,
+      "grad_norm": 0.3111253082752228,
+      "learning_rate": 0.00019344724180707802,
+      "loss": 0.35599780082702637,
+      "step": 5520
+    },
+    {
+      "epoch": 0.048343048586006014,
+      "grad_norm": 0.33388203382492065,
+      "learning_rate": 0.00019341853152734611,
+      "loss": 0.30465400218963623,
+      "step": 5530
+    },
+    {
+      "epoch": 0.04863532820260823,
+      "grad_norm": 0.2608368694782257,
+      "learning_rate": 0.00019338976062934334,
+      "loss": 0.24283418655395508,
+      "step": 5540
+    },
+    {
+      "epoch": 0.04892760781921044,
+      "grad_norm": 0.3442809283733368,
+      "learning_rate": 0.0001933609291317388,
+      "loss": 0.27430431842803954,
+      "step": 5550
+    },
+    {
+      "epoch": 0.049219887435812654,
+      "grad_norm": 0.3253636360168457,
+      "learning_rate": 0.00019333203705324095,
+      "loss": 0.31047348976135253,
+      "step": 5560
+    },
+    {
+      "epoch": 0.04951216705241487,
+      "grad_norm": 0.272540420293808,
+      "learning_rate": 0.00019330308441259755,
+      "loss": 0.2873514175415039,
+      "step": 5570
+    },
+    {
+      "epoch": 0.04980444666901708,
+      "grad_norm": 0.28369560837745667,
+      "learning_rate": 0.00019327407122859567,
+      "loss": 0.2900906324386597,
+      "step": 5580
+    },
+    {
+      "epoch": 0.05009672628561929,
+      "grad_norm": 0.2526591122150421,
+      "learning_rate": 0.0001932449975200616,
+      "loss": 0.32894172668457033,
+      "step": 5590
+    },
+    {
+      "epoch": 0.050389005902221506,
+      "grad_norm": 0.37079957127571106,
+      "learning_rate": 0.00019321586330586102,
+      "loss": 0.309123158454895,
+      "step": 5600
+    },
+    {
+      "epoch": 0.05068128551882372,
+      "grad_norm": 0.28275376558303833,
+      "learning_rate": 0.00019318666860489874,
+      "loss": 0.3356776714324951,
+      "step": 5610
+    },
+    {
+      "epoch": 0.05097356513542593,
+      "grad_norm": 0.3117635250091553,
+      "learning_rate": 0.00019315741343611893,
+      "loss": 0.34910783767700193,
+      "step": 5620
+    },
+    {
+      "epoch": 0.051265844752028146,
+      "grad_norm": 0.3331267833709717,
+      "learning_rate": 0.00019312809781850494,
+      "loss": 0.29843094348907473,
+      "step": 5630
+    },
+    {
+      "epoch": 0.05155812436863036,
+      "grad_norm": 0.3864763081073761,
+      "learning_rate": 0.00019309872177107935,
+      "loss": 0.236519455909729,
+      "step": 5640
+    },
+    {
+      "epoch": 0.05185040398523257,
+      "grad_norm": 0.27962878346443176,
+      "learning_rate": 0.00019306928531290395,
+      "loss": 0.255594801902771,
+      "step": 5650
+    },
+    {
+      "epoch": 0.052142683601834786,
+      "grad_norm": 0.2996375858783722,
+      "learning_rate": 0.00019303978846307979,
+      "loss": 0.2946054458618164,
+      "step": 5660
+    },
+    {
+      "epoch": 0.052434963218437,
+      "grad_norm": 0.4066357910633087,
+      "learning_rate": 0.00019301023124074695,
+      "loss": 0.28123533725738525,
+      "step": 5670
+    },
+    {
+      "epoch": 0.05272724283503921,
+      "grad_norm": 0.42517444491386414,
+      "learning_rate": 0.0001929806136650849,
+      "loss": 0.3094054937362671,
+      "step": 5680
+    },
+    {
+      "epoch": 0.053019522451641425,
+      "grad_norm": 0.2737913429737091,
+      "learning_rate": 0.00019295093575531208,
+      "loss": 0.3370650053024292,
+      "step": 5690
+    },
+    {
+      "epoch": 0.05331180206824364,
+      "grad_norm": 0.34213295578956604,
+      "learning_rate": 0.00019292119753068626,
+      "loss": 0.2837510108947754,
+      "step": 5700
+    },
+    {
+      "epoch": 0.05360408168484585,
+      "grad_norm": 0.41633471846580505,
+      "learning_rate": 0.00019289139901050416,
+      "loss": 0.3151435375213623,
+      "step": 5710
+    },
+    {
+      "epoch": 0.053896361301448065,
+      "grad_norm": 0.41348034143447876,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 0.3170133352279663,
+      "step": 5720
+    },
+    {
+      "epoch": 0.05418864091805028,
+      "grad_norm": 0.35909515619277954,
+      "learning_rate": 0.00019283162116085406,
+      "loss": 0.26323771476745605,
+      "step": 5730
+    },
+    {
+      "epoch": 0.05448092053465249,
+      "grad_norm": 0.32084155082702637,
+      "learning_rate": 0.0001928016418701753,
+      "loss": 0.3380338191986084,
+      "step": 5740
+    },
+    {
+      "epoch": 0.054773200151254704,
+      "grad_norm": 0.24091753363609314,
+      "learning_rate": 0.00019277160236151863,
+      "loss": 0.2938279628753662,
+      "step": 5750
+    },
+    {
+      "epoch": 0.05506547976785692,
+      "grad_norm": 0.33941394090652466,
+      "learning_rate": 0.00019274150265437642,
+      "loss": 0.2946443557739258,
+      "step": 5760
+    },
+    {
+      "epoch": 0.05535775938445913,
+      "grad_norm": 0.2492181956768036,
+      "learning_rate": 0.00019271134276828,
+      "loss": 0.3052149772644043,
+      "step": 5770
+    },
+    {
+      "epoch": 0.055650039001061344,
+      "grad_norm": 0.3869125247001648,
+      "learning_rate": 0.00019268112272279983,
+      "loss": 0.27257356643676756,
+      "step": 5780
+    },
+    {
+      "epoch": 0.05594231861766355,
+      "grad_norm": 0.31452256441116333,
+      "learning_rate": 0.0001926508425375453,
+      "loss": 0.30351474285125735,
+      "step": 5790
+    },
+    {
+      "epoch": 0.05623459823426576,
+      "grad_norm": 0.2884000241756439,
+      "learning_rate": 0.000192620502232165,
+      "loss": 0.3291373014450073,
+      "step": 5800
+    },
+    {
+      "epoch": 0.056526877850867976,
+      "grad_norm": 0.3090774118900299,
+      "learning_rate": 0.00019259010182634634,
+      "loss": 0.255063796043396,
+      "step": 5810
+    },
+    {
+      "epoch": 0.05681915746747019,
+      "grad_norm": 0.2950552999973297,
+      "learning_rate": 0.00019255964133981583,
+      "loss": 0.28303749561309816,
+      "step": 5820
+    },
+    {
+      "epoch": 0.0571114370840724,
+      "grad_norm": 0.2653549313545227,
+      "learning_rate": 0.00019252912079233898,
+      "loss": 0.2754636287689209,
+      "step": 5830
+    },
+    {
+      "epoch": 0.057403716700674616,
+      "grad_norm": 0.2716647982597351,
+      "learning_rate": 0.00019249854020372023,
+      "loss": 0.338680624961853,
+      "step": 5840
+    },
+    {
+      "epoch": 0.05769599631727683,
+      "grad_norm": 0.38978704810142517,
+      "learning_rate": 0.00019246789959380295,
+      "loss": 0.28801736831665037,
+      "step": 5850
+    },
+    {
+      "epoch": 0.05798827593387904,
+      "grad_norm": 0.3205948770046234,
+      "learning_rate": 0.00019243719898246955,
+      "loss": 0.33612334728240967,
+      "step": 5860
+    },
+    {
+      "epoch": 0.058280555550481256,
+      "grad_norm": 0.2750367522239685,
+      "learning_rate": 0.00019240643838964132,
+      "loss": 0.24042437076568604,
+      "step": 5870
+    },
+    {
+      "epoch": 0.05857283516708347,
+      "grad_norm": 0.39220261573791504,
+      "learning_rate": 0.00019237561783527846,
+      "loss": 0.2887096405029297,
+      "step": 5880
+    },
+    {
+      "epoch": 0.05886511478368568,
+      "grad_norm": 0.3734164237976074,
+      "learning_rate": 0.00019234473733938006,
+      "loss": 0.3519147396087646,
+      "step": 5890
+    },
+    {
+      "epoch": 0.059157394400287895,
+      "grad_norm": 0.259491503238678,
+      "learning_rate": 0.00019231379692198424,
+      "loss": 0.32158188819885253,
+      "step": 5900
+    },
+    {
+      "epoch": 0.05944967401689011,
+      "grad_norm": 0.34263959527015686,
+      "learning_rate": 0.00019228279660316776,
+      "loss": 0.3218578338623047,
+      "step": 5910
+    },
+    {
+      "epoch": 0.05974195363349232,
+      "grad_norm": 0.3145698010921478,
+      "learning_rate": 0.0001922517364030465,
+      "loss": 0.28389294147491456,
+      "step": 5920
+    },
+    {
+      "epoch": 0.060034233250094535,
+      "grad_norm": 0.3532608449459076,
+      "learning_rate": 0.00019222061634177506,
+      "loss": 0.31691362857818606,
+      "step": 5930
+    },
+    {
+      "epoch": 0.06032651286669675,
+      "grad_norm": 0.38111406564712524,
+      "learning_rate": 0.0001921894364395469,
+      "loss": 0.23623991012573242,
+      "step": 5940
+    },
+    {
+      "epoch": 0.06061879248329896,
+      "grad_norm": 0.3058551549911499,
+      "learning_rate": 0.0001921581967165943,
+      "loss": 0.26392316818237305,
+      "step": 5950
+    },
+    {
+      "epoch": 0.060911072099901174,
+      "grad_norm": 0.2773318588733673,
+      "learning_rate": 0.0001921268971931884,
+      "loss": 0.3105238199234009,
+      "step": 5960
+    },
+    {
+      "epoch": 0.06120335171650339,
+      "grad_norm": 0.27138495445251465,
+      "learning_rate": 0.00019209553788963914,
+      "loss": 0.22701623439788818,
+      "step": 5970
+    },
+    {
+      "epoch": 0.0614956313331056,
+      "grad_norm": 0.32295671105384827,
+      "learning_rate": 0.00019206411882629517,
+      "loss": 0.3231555700302124,
+      "step": 5980
+    },
+    {
+      "epoch": 0.061787910949707814,
+      "grad_norm": 0.3322771489620209,
+      "learning_rate": 0.00019203264002354406,
+      "loss": 0.31249637603759767,
+      "step": 5990
+    },
+    {
+      "epoch": 0.06208019056631003,
+      "grad_norm": 0.34913370013237,
+      "learning_rate": 0.00019200110150181202,
+      "loss": 0.23186285495758058,
+      "step": 6000
+    },
+    {
+      "epoch": 0.0002922796166022129,
+      "grad_norm": 0.33879324793815613,
+      "learning_rate": 0.00019196950328156405,
+      "loss": 0.3232074022293091,
+      "step": 6010
+    },
+    {
+      "epoch": 0.0005845592332044258,
+      "grad_norm": 0.3840523660182953,
+      "learning_rate": 0.00019193784538330388,
+      "loss": 0.3079432725906372,
+      "step": 6020
+    },
+    {
+      "epoch": 0.0008768388498066388,
+      "grad_norm": 0.3380774259567261,
+      "learning_rate": 0.00019190612782757404,
+      "loss": 0.26655428409576415,
+      "step": 6030
+    },
+    {
+      "epoch": 0.0011691184664088517,
+      "grad_norm": 0.34157833456993103,
+      "learning_rate": 0.0001918743506349556,
+      "loss": 0.25333199501037595,
+      "step": 6040
+    },
+    {
+      "epoch": 0.0014613980830110646,
+      "grad_norm": 0.27082303166389465,
+      "learning_rate": 0.00019184251382606854,
+      "loss": 0.2442876100540161,
+      "step": 6050
+    },
+    {
+      "epoch": 0.0017536776996132776,
+      "grad_norm": 0.2830803394317627,
+      "learning_rate": 0.00019181061742157138,
+      "loss": 0.2704757213592529,
+      "step": 6060
+    },
+    {
+      "epoch": 0.0020459573162154906,
+      "grad_norm": 0.26303544640541077,
+      "learning_rate": 0.00019177866144216136,
+      "loss": 0.2877002000808716,
+      "step": 6070
+    },
+    {
+      "epoch": 0.0023382369328177033,
+      "grad_norm": 0.34139397740364075,
+      "learning_rate": 0.0001917466459085743,
+      "loss": 0.2769068956375122,
+      "step": 6080
+    },
+    {
+      "epoch": 0.002630516549419916,
+      "grad_norm": 0.32799649238586426,
+      "learning_rate": 0.00019171457084158483,
+      "loss": 0.279664421081543,
+      "step": 6090
+    },
+    {
+      "epoch": 0.0029227961660221293,
+      "grad_norm": 0.39274758100509644,
+      "learning_rate": 0.00019168243626200608,
+      "loss": 0.2821364402770996,
+      "step": 6100
+    },
+    {
+      "epoch": 0.003215075782624342,
+      "grad_norm": 0.34457409381866455,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 0.3095219612121582,
+      "step": 6110
+    },
+    {
+      "epoch": 0.0035073553992265552,
+      "grad_norm": 0.3495699167251587,
+      "learning_rate": 0.0001916179886485264,
+      "loss": 0.28960461616516114,
+      "step": 6120
+    },
+    {
+      "epoch": 0.003799635015828768,
+      "grad_norm": 0.28062158823013306,
+      "learning_rate": 0.00019158567565644489,
+      "loss": 0.26502583026885984,
+      "step": 6130
+    },
+    {
+      "epoch": 0.004091914632430981,
+      "grad_norm": 0.29114511609077454,
+      "learning_rate": 0.0001915533032354127,
+      "loss": 0.31997365951538087,
+      "step": 6140
+    },
+    {
+      "epoch": 0.0043841942490331935,
+      "grad_norm": 0.3683178126811981,
+      "learning_rate": 0.0001915208714064361,
+      "loss": 0.24852454662322998,
+      "step": 6150
+    },
+    {
+      "epoch": 0.004676473865635407,
+      "grad_norm": 0.3214621841907501,
+      "learning_rate": 0.00019148838019055958,
+      "loss": 0.23689863681793213,
+      "step": 6160
+    },
+    {
+      "epoch": 0.00496875348223762,
+      "grad_norm": 0.30240535736083984,
+      "learning_rate": 0.0001914558296088665,
+      "loss": 0.24038214683532716,
+      "step": 6170
+    },
+    {
+      "epoch": 0.005261033098839832,
+      "grad_norm": 0.3095683157444,
+      "learning_rate": 0.0001914232196824785,
+      "loss": 0.2682337284088135,
+      "step": 6180
+    },
+    {
+      "epoch": 0.005553312715442045,
+      "grad_norm": 0.3249179720878601,
+      "learning_rate": 0.0001913905504325558,
+      "loss": 0.25601763725280763,
+      "step": 6190
+    },
+    {
+      "epoch": 0.0058455923320442586,
+      "grad_norm": 0.39472144842147827,
+      "learning_rate": 0.00019135782188029713,
+      "loss": 0.28767030239105223,
+      "step": 6200
+    },
+    {
+      "epoch": 0.006137871948646472,
+      "grad_norm": 0.28574320673942566,
+      "learning_rate": 0.00019132503404693972,
+      "loss": 0.26905953884124756,
+      "step": 6210
+    },
+    {
+      "epoch": 0.006430151565248684,
+      "grad_norm": 0.28733396530151367,
+      "learning_rate": 0.00019129218695375922,
+      "loss": 0.26485185623168944,
+      "step": 6220
+    },
+    {
+      "epoch": 0.006722431181850897,
+      "grad_norm": 0.29587945342063904,
+      "learning_rate": 0.00019125928062206974,
+      "loss": 0.28349895477294923,
+      "step": 6230
+    },
+    {
+      "epoch": 0.0070147107984531104,
+      "grad_norm": 0.2649540603160858,
+      "learning_rate": 0.00019122631507322387,
+      "loss": 0.29269912242889407,
+      "step": 6240
+    },
+    {
+      "epoch": 0.007306990415055323,
+      "grad_norm": 0.25572097301483154,
+      "learning_rate": 0.00019119329032861256,
+      "loss": 0.2546536922454834,
+      "step": 6250
+    },
+    {
+      "epoch": 0.007599270031657536,
+      "grad_norm": 0.360807865858078,
+      "learning_rate": 0.00019116020640966524,
+      "loss": 0.19606984853744508,
+      "step": 6260
+    },
+    {
+      "epoch": 0.00789154964825975,
+      "grad_norm": 0.4035632014274597,
+      "learning_rate": 0.00019112706333784967,
+      "loss": 0.28473834991455077,
+      "step": 6270
+    },
+    {
+      "epoch": 0.008183829264861962,
+      "grad_norm": 0.306217223405838,
+      "learning_rate": 0.00019109386113467207,
+      "loss": 0.23012244701385498,
+      "step": 6280
+    },
+    {
+      "epoch": 0.008476108881464176,
+      "grad_norm": 0.6679860353469849,
+      "learning_rate": 0.00019106059982167698,
+      "loss": 0.27747671604156493,
+      "step": 6290
+    },
+    {
+      "epoch": 0.008768388498066387,
+      "grad_norm": 0.3663105070590973,
+      "learning_rate": 0.00019102727942044726,
+      "loss": 0.2566732406616211,
+      "step": 6300
+    },
+    {
+      "epoch": 0.0090606681146686,
+      "grad_norm": 0.29065728187561035,
+      "learning_rate": 0.00019099389995260423,
+      "loss": 0.2529840230941772,
+      "step": 6310
+    },
+    {
+      "epoch": 0.009352947731270813,
+      "grad_norm": 0.3425915539264679,
+      "learning_rate": 0.00019096046143980743,
+      "loss": 0.2695134401321411,
+      "step": 6320
+    },
+    {
+      "epoch": 0.009645227347873027,
+      "grad_norm": 0.49088072776794434,
+      "learning_rate": 0.0001909269639037547,
+      "loss": 0.2175128698348999,
+      "step": 6330
+    },
+    {
+      "epoch": 0.00993750696447524,
+      "grad_norm": 0.29592210054397583,
+      "learning_rate": 0.0001908934073661823,
+      "loss": 0.2875823974609375,
+      "step": 6340
+    },
+    {
+      "epoch": 0.010229786581077453,
+      "grad_norm": 0.3159569799900055,
+      "learning_rate": 0.0001908597918488647,
+      "loss": 0.24905283451080323,
+      "step": 6350
+    },
+    {
+      "epoch": 0.010522066197679664,
+      "grad_norm": 0.35035523772239685,
+      "learning_rate": 0.00019082611737361458,
+      "loss": 0.22769768238067628,
+      "step": 6360
+    },
+    {
+      "epoch": 0.010814345814281878,
+      "grad_norm": 0.28443869948387146,
+      "learning_rate": 0.000190792383962283,
+      "loss": 0.2708036661148071,
+      "step": 6370
+    },
+    {
+      "epoch": 0.01110662543088409,
+      "grad_norm": 0.4272829592227936,
+      "learning_rate": 0.0001907585916367592,
+      "loss": 0.27393512725830077,
+      "step": 6380
+    },
+    {
+      "epoch": 0.011398905047486304,
+      "grad_norm": 0.2440495491027832,
+      "learning_rate": 0.0001907247404189706,
+      "loss": 0.2520595073699951,
+      "step": 6390
+    },
+    {
+      "epoch": 0.011691184664088517,
+      "grad_norm": 0.3282374143600464,
+      "learning_rate": 0.00019069083033088294,
+      "loss": 0.29231948852539064,
+      "step": 6400
+    },
+    {
+      "epoch": 0.01198346428069073,
+      "grad_norm": 0.3208235204219818,
+      "learning_rate": 0.00019065686139450012,
+      "loss": 0.2337594985961914,
+      "step": 6410
+    },
+    {
+      "epoch": 0.012275743897292943,
+      "grad_norm": 0.2959216833114624,
+      "learning_rate": 0.00019062283363186416,
+      "loss": 0.2158582925796509,
+      "step": 6420
+    },
+    {
+      "epoch": 0.012568023513895155,
+      "grad_norm": 0.31310606002807617,
+      "learning_rate": 0.00019058874706505535,
+      "loss": 0.2612952947616577,
+      "step": 6430
+    },
+    {
+      "epoch": 0.012860303130497368,
+      "grad_norm": 0.37715741991996765,
+      "learning_rate": 0.00019055460171619207,
+      "loss": 0.21847269535064698,
+      "step": 6440
+    },
+    {
+      "epoch": 0.013152582747099581,
+      "grad_norm": 0.32075491547584534,
+      "learning_rate": 0.00019052039760743088,
+      "loss": 0.265203070640564,
+      "step": 6450
+    },
+    {
+      "epoch": 0.013444862363701795,
+      "grad_norm": 0.3275420069694519,
+      "learning_rate": 0.00019048613476096646,
+      "loss": 0.2622615098953247,
+      "step": 6460
+    },
+    {
+      "epoch": 0.013737141980304008,
+      "grad_norm": 0.29434168338775635,
+      "learning_rate": 0.00019045181319903156,
+      "loss": 0.24696013927459717,
+      "step": 6470
+    },
+    {
+      "epoch": 0.014029421596906221,
+      "grad_norm": 0.33702999353408813,
+      "learning_rate": 0.0001904174329438971,
+      "loss": 0.22126851081848145,
+      "step": 6480
+    },
+    {
+      "epoch": 0.014321701213508432,
+      "grad_norm": 0.34361857175827026,
+      "learning_rate": 0.00019038299401787206,
+      "loss": 0.20502333641052245,
+      "step": 6490
+    },
+    {
+      "epoch": 0.014613980830110646,
+      "grad_norm": 0.4263356924057007,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 0.2336045503616333,
+      "step": 6500
+    },
+    {
+      "epoch": 0.014906260446712859,
+      "grad_norm": 0.3184535503387451,
+      "learning_rate": 0.00019031394024257648,
+      "loss": 0.24427258968353271,
+      "step": 6510
+    },
+    {
+      "epoch": 0.015198540063315072,
+      "grad_norm": 0.3023635447025299,
+      "learning_rate": 0.00019027932543811416,
+      "loss": 0.28334190845489504,
+      "step": 6520
+    },
+    {
+      "epoch": 0.015490819679917285,
+      "grad_norm": 0.28802433609962463,
+      "learning_rate": 0.0001902446520523777,
+      "loss": 0.28285205364227295,
+      "step": 6530
+    },
+    {
+      "epoch": 0.0157830992965195,
+      "grad_norm": 0.345346063375473,
+      "learning_rate": 0.0001902099201078663,
+      "loss": 0.2579298496246338,
+      "step": 6540
+    },
+    {
+      "epoch": 0.01607537891312171,
+      "grad_norm": 0.3453812003135681,
+      "learning_rate": 0.00019017512962711712,
+      "loss": 0.2692086696624756,
+      "step": 6550
+    },
+    {
+      "epoch": 0.016367658529723925,
+      "grad_norm": 0.3395085632801056,
+      "learning_rate": 0.00019014028063270538,
+      "loss": 0.2863780498504639,
+      "step": 6560
+    },
+    {
+      "epoch": 0.016659938146326138,
+      "grad_norm": 0.2853766083717346,
+      "learning_rate": 0.00019010537314724415,
+      "loss": 0.2577486991882324,
+      "step": 6570
+    },
+    {
+      "epoch": 0.01695221776292835,
+      "grad_norm": 0.3124806880950928,
+      "learning_rate": 0.00019007040719338453,
+      "loss": 0.2790304899215698,
+      "step": 6580
+    },
+    {
+      "epoch": 0.01724449737953056,
+      "grad_norm": 0.34289953112602234,
+      "learning_rate": 0.0001900353827938156,
+      "loss": 0.2826568126678467,
+      "step": 6590
+    },
+    {
+      "epoch": 0.017536776996132774,
+      "grad_norm": 0.30955979228019714,
+      "learning_rate": 0.00019000029997126428,
+      "loss": 0.2861216068267822,
+      "step": 6600
+    },
+    {
+      "epoch": 0.017829056612734987,
+      "grad_norm": 0.33083009719848633,
+      "learning_rate": 0.00018996515874849542,
+      "loss": 0.23559873104095458,
+      "step": 6610
+    },
+    {
+      "epoch": 0.0181213362293372,
+      "grad_norm": 0.3657139241695404,
+      "learning_rate": 0.00018992995914831175,
+      "loss": 0.2663041830062866,
+      "step": 6620
+    },
+    {
+      "epoch": 0.018413615845939414,
+      "grad_norm": 0.40399572253227234,
+      "learning_rate": 0.00018989470119355398,
+      "loss": 0.22185773849487306,
+      "step": 6630
+    },
+    {
+      "epoch": 0.018705895462541627,
+      "grad_norm": 0.3088628351688385,
+      "learning_rate": 0.00018985938490710056,
+      "loss": 0.2567310094833374,
+      "step": 6640
+    },
+    {
+      "epoch": 0.01899817507914384,
+      "grad_norm": 0.41595155000686646,
+      "learning_rate": 0.00018982401031186787,
+      "loss": 0.2673985481262207,
+      "step": 6650
+    },
+    {
+      "epoch": 0.019290454695746053,
+      "grad_norm": 0.44323673844337463,
+      "learning_rate": 0.0001897885774308101,
+      "loss": 0.26906542778015136,
+      "step": 6660
+    },
+    {
+      "epoch": 0.019582734312348266,
+      "grad_norm": 0.3764394521713257,
+      "learning_rate": 0.00018975308628691923,
+      "loss": 0.2775614261627197,
+      "step": 6670
+    },
+    {
+      "epoch": 0.01987501392895048,
+      "grad_norm": 0.3118645250797272,
+      "learning_rate": 0.00018971753690322514,
+      "loss": 0.24204604625701903,
+      "step": 6680
+    },
+    {
+      "epoch": 0.020167293545552693,
+      "grad_norm": 0.37264496088027954,
+      "learning_rate": 0.0001896819293027954,
+      "loss": 0.2537872076034546,
+      "step": 6690
+    },
+    {
+      "epoch": 0.020459573162154906,
+      "grad_norm": 0.2669682204723358,
+      "learning_rate": 0.00018964626350873536,
+      "loss": 0.18956880569458007,
+      "step": 6700
+    },
+    {
+      "epoch": 0.02075185277875712,
+      "grad_norm": 0.3099878430366516,
+      "learning_rate": 0.00018961053954418824,
+      "loss": 0.2325613260269165,
+      "step": 6710
+    },
+    {
+      "epoch": 0.02104413239535933,
+      "grad_norm": 0.3033933639526367,
+      "learning_rate": 0.0001895747574323349,
+      "loss": 0.23195374011993408,
+      "step": 6720
+    },
+    {
+      "epoch": 0.021336412011961542,
+      "grad_norm": 0.3181861340999603,
+      "learning_rate": 0.00018953891719639395,
+      "loss": 0.1934121608734131,
+      "step": 6730
+    },
+    {
+      "epoch": 2.9227961660221293e-05,
+      "grad_norm": 0.3904183804988861,
+      "learning_rate": 0.00018950301885962173,
+      "loss": 0.14989452064037323,
+      "step": 6740
+    },
+    {
+      "epoch": 0.0003215075782624342,
+      "grad_norm": 0.3860507607460022,
+      "learning_rate": 0.0001894670624453123,
+      "loss": 0.2355290412902832,
+      "step": 6750
+    },
+    {
+      "epoch": 0.0006137871948646472,
+      "grad_norm": 0.22993847727775574,
+      "learning_rate": 0.0001894310479767974,
+      "loss": 0.17446482181549072,
+      "step": 6760
+    },
+    {
+      "epoch": 0.00090606681146686,
+      "grad_norm": 0.45022067427635193,
+      "learning_rate": 0.00018939497547744636,
+      "loss": 0.17517282962799072,
+      "step": 6770
+    },
+    {
+      "epoch": 0.001198346428069073,
+      "grad_norm": 0.35979047417640686,
+      "learning_rate": 0.00018935884497066628,
+      "loss": 0.1405559778213501,
+      "step": 6780
+    },
+    {
+      "epoch": 0.0014906260446712859,
+      "grad_norm": 0.3211457133293152,
+      "learning_rate": 0.0001893226564799019,
+      "loss": 0.1425009250640869,
+      "step": 6790
+    },
+    {
+      "epoch": 0.0017829056612734988,
+      "grad_norm": 0.4115179479122162,
+      "learning_rate": 0.00018928641002863546,
+      "loss": 0.16687273979187012,
+      "step": 6800
+    },
+    {
+      "epoch": 0.002075185277875712,
+      "grad_norm": 0.43852880597114563,
+      "learning_rate": 0.00018925010564038695,
+      "loss": 0.1651862621307373,
+      "step": 6810
+    },
+    {
+      "epoch": 0.0023674648944779246,
+      "grad_norm": 0.3372480869293213,
+      "learning_rate": 0.00018921374333871386,
+      "loss": 0.158694064617157,
+      "step": 6820
+    },
+    {
+      "epoch": 0.0026597445110801378,
+      "grad_norm": 0.3730330765247345,
+      "learning_rate": 0.00018917732314721128,
+      "loss": 0.1521228313446045,
+      "step": 6830
+    },
+    {
+      "epoch": 0.0029520241276823505,
+      "grad_norm": 0.43486854434013367,
+      "learning_rate": 0.00018914084508951187,
+      "loss": 0.16771166324615477,
+      "step": 6840
+    },
+    {
+      "epoch": 0.0032443037442845633,
+      "grad_norm": 0.40164175629615784,
+      "learning_rate": 0.00018910430918928593,
+      "loss": 0.16778719425201416,
+      "step": 6850
+    },
+    {
+      "epoch": 0.0035365833608867765,
+      "grad_norm": 0.4072563946247101,
+      "learning_rate": 0.00018906771547024112,
+      "loss": 0.17667644023895263,
+      "step": 6860
+    },
+    {
+      "epoch": 0.003828862977488989,
+      "grad_norm": 0.3342289626598358,
+      "learning_rate": 0.00018903106395612275,
+      "loss": 0.1560322165489197,
+      "step": 6870
+    },
+    {
+      "epoch": 0.004121142594091202,
+      "grad_norm": 0.39563751220703125,
+      "learning_rate": 0.00018899435467071353,
+      "loss": 0.1789429306983948,
+      "step": 6880
+    },
+    {
+      "epoch": 0.004413422210693415,
+      "grad_norm": 0.35209736227989197,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 0.14813270568847656,
+      "step": 6890
+    },
+    {
+      "epoch": 0.004705701827295628,
+      "grad_norm": 0.32932519912719727,
+      "learning_rate": 0.00018892076288134124,
+      "loss": 0.12475873231887817,
+      "step": 6900
+    },
+    {
+      "epoch": 0.004997981443897841,
+      "grad_norm": 0.48736169934272766,
+      "learning_rate": 0.00018888388042513106,
+      "loss": 0.1394341230392456,
+      "step": 6910
+    },
+    {
+      "epoch": 0.005290261060500054,
+      "grad_norm": 0.46268224716186523,
+      "learning_rate": 0.00018884694029313586,
+      "loss": 0.1412713885307312,
+      "step": 6920
+    },
+    {
+      "epoch": 0.005582540677102267,
+      "grad_norm": 0.3679901957511902,
+      "learning_rate": 0.0001888099425093257,
+      "loss": 0.1367684006690979,
+      "step": 6930
+    },
+    {
+      "epoch": 0.005874820293704479,
+      "grad_norm": 0.3641803562641144,
+      "learning_rate": 0.00018877288709770804,
+      "loss": 0.18430200815200806,
+      "step": 6940
+    },
+    {
+      "epoch": 0.0061670999103066925,
+      "grad_norm": 0.35064584016799927,
+      "learning_rate": 0.0001887357740823277,
+      "loss": 0.1364537477493286,
+      "step": 6950
+    },
+    {
+      "epoch": 0.006459379526908906,
+      "grad_norm": 0.270317941904068,
+      "learning_rate": 0.000188698603487267,
+      "loss": 0.1482532262802124,
+      "step": 6960
+    },
+    {
+      "epoch": 0.006751659143511119,
+      "grad_norm": 0.5113752484321594,
+      "learning_rate": 0.0001886613753366455,
+      "loss": 0.16875799894332885,
+      "step": 6970
+    },
+    {
+      "epoch": 0.007043938760113331,
+      "grad_norm": 0.5064671635627747,
+      "learning_rate": 0.00018862408965462007,
+      "loss": 0.1733761429786682,
+      "step": 6980
+    },
+    {
+      "epoch": 0.007336218376715544,
+      "grad_norm": 0.35405635833740234,
+      "learning_rate": 0.00018858674646538508,
+      "loss": 0.1359371781349182,
+      "step": 6990
+    },
+    {
+      "epoch": 0.007628497993317758,
+      "grad_norm": 0.3354845941066742,
+      "learning_rate": 0.00018854934579317205,
+      "loss": 0.11053211688995361,
+      "step": 7000
+    },
+    {
+      "epoch": 0.00792077760991997,
+      "grad_norm": 0.4442973732948303,
+      "learning_rate": 0.00018851188766224995,
+      "loss": 0.17160513401031494,
+      "step": 7010
+    },
+    {
+      "epoch": 0.008213057226522182,
+      "grad_norm": 0.29706066846847534,
+      "learning_rate": 0.00018847437209692486,
+      "loss": 0.1071900486946106,
+      "step": 7020
+    },
+    {
+      "epoch": 0.008505336843124395,
+      "grad_norm": 0.5107281804084778,
+      "learning_rate": 0.00018843679912154029,
+      "loss": 0.16834162473678588,
+      "step": 7030
+    },
+    {
+      "epoch": 0.008797616459726609,
+      "grad_norm": 0.4704040288925171,
+      "learning_rate": 0.0001883991687604769,
+      "loss": 0.16411263942718507,
+      "step": 7040
+    },
+    {
+      "epoch": 0.009089896076328822,
+      "grad_norm": 0.48398149013519287,
+      "learning_rate": 0.00018836148103815265,
+      "loss": 0.14024045467376708,
+      "step": 7050
+    },
+    {
+      "epoch": 0.009382175692931035,
+      "grad_norm": 0.3730473816394806,
+      "learning_rate": 0.00018832373597902266,
+      "loss": 0.14982374906539916,
+      "step": 7060
+    },
+    {
+      "epoch": 0.009674455309533248,
+      "grad_norm": 0.4151541590690613,
+      "learning_rate": 0.00018828593360757932,
+      "loss": 0.10983268022537232,
+      "step": 7070
+    },
+    {
+      "epoch": 0.009966734926135461,
+      "grad_norm": 0.5641967058181763,
+      "learning_rate": 0.00018824807394835218,
+      "loss": 0.1792724013328552,
+      "step": 7080
+    },
+    {
+      "epoch": 0.010259014542737673,
+      "grad_norm": 0.4540756642818451,
+      "learning_rate": 0.00018821015702590797,
+      "loss": 0.1320549726486206,
+      "step": 7090
+    },
+    {
+      "epoch": 0.010551294159339886,
+      "grad_norm": 0.3783016800880432,
+      "learning_rate": 0.00018817218286485053,
+      "loss": 0.12515761852264404,
+      "step": 7100
+    },
+    {
+      "epoch": 0.0108435737759421,
+      "grad_norm": 0.4021569490432739,
+      "learning_rate": 0.00018813415148982086,
+      "loss": 0.16211906671524048,
+      "step": 7110
+    },
+    {
+      "epoch": 0.011135853392544312,
+      "grad_norm": 0.6212558150291443,
+      "learning_rate": 0.0001880960629254972,
+      "loss": 0.1415063977241516,
+      "step": 7120
+    },
+    {
+      "epoch": 0.011428133009146526,
+      "grad_norm": 0.3575975298881531,
+      "learning_rate": 0.00018805791719659473,
+      "loss": 0.15111939907073973,
+      "step": 7130
+    },
+    {
+      "epoch": 0.011720412625748739,
+      "grad_norm": 0.39540404081344604,
+      "learning_rate": 0.00018801971432786584,
+      "loss": 0.19280412197113037,
+      "step": 7140
+    },
+    {
+      "epoch": 0.01201269224235095,
+      "grad_norm": 0.3093525171279907,
+      "learning_rate": 0.0001879814543440999,
+      "loss": 0.14252532720565797,
+      "step": 7150
+    },
+    {
+      "epoch": 0.012304971858953163,
+      "grad_norm": 0.4474407434463501,
+      "learning_rate": 0.0001879431372701235,
+      "loss": 0.12835644483566283,
+      "step": 7160
+    },
+    {
+      "epoch": 0.012597251475555377,
+      "grad_norm": 0.5637357831001282,
+      "learning_rate": 0.00018790476313080006,
+      "loss": 0.16645318269729614,
+      "step": 7170
+    },
+    {
+      "epoch": 0.01288953109215759,
+      "grad_norm": 0.5085288286209106,
+      "learning_rate": 0.00018786633195103021,
+      "loss": 0.10577408075332642,
+      "step": 7180
+    },
+    {
+      "epoch": 0.013181810708759803,
+      "grad_norm": 0.36965733766555786,
+      "learning_rate": 0.00018782784375575152,
+      "loss": 0.1587444543838501,
+      "step": 7190
+    },
+    {
+      "epoch": 0.013474090325362016,
+      "grad_norm": 0.288505494594574,
+      "learning_rate": 0.00018778929856993852,
+      "loss": 0.14752702713012694,
+      "step": 7200
+    },
+    {
+      "epoch": 0.01376636994196423,
+      "grad_norm": 0.31553298234939575,
+      "learning_rate": 0.0001877506964186028,
+      "loss": 0.14762223958969117,
+      "step": 7210
+    },
+    {
+      "epoch": 0.01405864955856644,
+      "grad_norm": 0.6161355376243591,
+      "learning_rate": 0.00018771203732679287,
+      "loss": 0.13177889585494995,
+      "step": 7220
+    },
+    {
+      "epoch": 0.014350929175168654,
+      "grad_norm": 0.47047898173332214,
+      "learning_rate": 0.0001876733213195942,
+      "loss": 0.12232577800750732,
+      "step": 7230
+    },
+    {
+      "epoch": 0.014643208791770867,
+      "grad_norm": 0.4503072500228882,
+      "learning_rate": 0.00018763454842212922,
+      "loss": 0.1304970860481262,
+      "step": 7240
+    },
+    {
+      "epoch": 0.01493548840837308,
+      "grad_norm": 0.36954185366630554,
+      "learning_rate": 0.00018759571865955717,
+      "loss": 0.14793392419815063,
+      "step": 7250
+    },
+    {
+      "epoch": 0.015227768024975294,
+      "grad_norm": 0.5747528672218323,
+      "learning_rate": 0.00018755683205707427,
+      "loss": 0.17184171676635743,
+      "step": 7260
+    },
+    {
+      "epoch": 0.015520047641577507,
+      "grad_norm": 0.46725842356681824,
+      "learning_rate": 0.0001875178886399137,
+      "loss": 0.14629058837890624,
+      "step": 7270
+    },
+    {
+      "epoch": 0.015812327258179718,
+      "grad_norm": 0.2947510778903961,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 0.15447490215301513,
+      "step": 7280
+    },
+    {
+      "epoch": 0.01610460687478193,
+      "grad_norm": 0.5247185826301575,
+      "learning_rate": 0.00018743983146267596,
+      "loss": 0.15706477165222169,
+      "step": 7290
+    },
+    {
+      "epoch": 0.016396886491384145,
+      "grad_norm": 0.42758095264434814,
+      "learning_rate": 0.0001874007177532493,
+      "loss": 0.18099625110626222,
+      "step": 7300
+    },
+    {
+      "epoch": 0.016689166107986358,
+      "grad_norm": 0.4924470782279968,
+      "learning_rate": 0.0001873615473304458,
+      "loss": 0.13735289573669435,
+      "step": 7310
+    },
+    {
+      "epoch": 0.01698144572458857,
+      "grad_norm": 0.4428713321685791,
+      "learning_rate": 0.00018732232021968269,
+      "loss": 0.1715406894683838,
+      "step": 7320
+    },
+    {
+      "epoch": 0.017273725341190784,
+      "grad_norm": 0.496646910905838,
+      "learning_rate": 0.00018728303644641403,
+      "loss": 0.17741830348968507,
+      "step": 7330
+    },
+    {
+      "epoch": 0.017566004957792997,
+      "grad_norm": 0.3481582999229431,
+      "learning_rate": 0.00018724369603613066,
+      "loss": 0.15302907228469848,
+      "step": 7340
+    },
+    {
+      "epoch": 0.01785828457439521,
+      "grad_norm": 0.3602871298789978,
+      "learning_rate": 0.00018720429901436013,
+      "loss": 0.12715169191360473,
+      "step": 7350
+    },
+    {
+      "epoch": 0.018150564190997424,
+      "grad_norm": 0.44130781292915344,
+      "learning_rate": 0.0001871648454066667,
+      "loss": 0.1618180513381958,
+      "step": 7360
+    },
+    {
+      "epoch": 0.018442843807599637,
+      "grad_norm": 0.5924626588821411,
+      "learning_rate": 0.00018712533523865146,
+      "loss": 0.12801215648651124,
+      "step": 7370
+    },
+    {
+      "epoch": 0.018735123424201847,
+      "grad_norm": 0.3802134096622467,
+      "learning_rate": 0.00018708576853595206,
+      "loss": 0.1363144874572754,
+      "step": 7380
+    },
+    {
+      "epoch": 0.01902740304080406,
+      "grad_norm": 0.4948204755783081,
+      "learning_rate": 0.00018704614532424294,
+      "loss": 0.17737267017364503,
+      "step": 7390
+    },
+    {
+      "epoch": 0.019319682657406273,
+      "grad_norm": 0.41467002034187317,
+      "learning_rate": 0.00018700646562923518,
+      "loss": 0.15785009860992433,
+      "step": 7400
+    },
+    {
+      "epoch": 0.019611962274008486,
+      "grad_norm": 0.535875141620636,
+      "learning_rate": 0.00018696672947667646,
+      "loss": 0.15521427392959594,
+      "step": 7410
+    },
+    {
+      "epoch": 0.0199042418906107,
+      "grad_norm": 0.487920343875885,
+      "learning_rate": 0.00018692693689235118,
+      "loss": 0.14909597635269164,
+      "step": 7420
+    },
+    {
+      "epoch": 0.020196521507212913,
+      "grad_norm": 0.649228036403656,
+      "learning_rate": 0.0001868870879020803,
+      "loss": 0.15439363718032836,
+      "step": 7430
+    },
+    {
+      "epoch": 0.020488801123815126,
+      "grad_norm": 0.40909361839294434,
+      "learning_rate": 0.00018684718253172142,
+      "loss": 0.08974369168281555,
+      "step": 7440
+    },
+    {
+      "epoch": 0.02078108074041734,
+      "grad_norm": 0.5107114315032959,
+      "learning_rate": 0.00018680722080716868,
+      "loss": 0.15838688611984253,
+      "step": 7450
+    },
+    {
+      "epoch": 0.021073360357019552,
+      "grad_norm": 0.36961808800697327,
+      "learning_rate": 0.00018676720275435279,
+      "loss": 0.1311396598815918,
+      "step": 7460
+    },
+    {
+      "epoch": 0.021365639973621765,
+      "grad_norm": 0.3621773421764374,
+      "learning_rate": 0.00018672712839924106,
+      "loss": 0.110987389087677,
+      "step": 7470
+    },
+    {
+      "epoch": 0.02165791959022398,
+      "grad_norm": 0.5529636144638062,
+      "learning_rate": 0.0001866869977678373,
+      "loss": 0.1936536192893982,
+      "step": 7480
+    },
+    {
+      "epoch": 0.02195019920682619,
+      "grad_norm": 0.3948989510536194,
+      "learning_rate": 0.0001866468108861818,
+      "loss": 0.268256402015686,
+      "step": 7490
+    },
+    {
+      "epoch": 0.022242478823428405,
+      "grad_norm": 0.3817199170589447,
+      "learning_rate": 0.00018660656778035146,
+      "loss": 0.24221539497375488,
+      "step": 7500
+    },
+    {
+      "epoch": 0.022534758440030615,
+      "grad_norm": 0.39620912075042725,
+      "learning_rate": 0.00018656626847645956,
+      "loss": 0.23234500885009765,
+      "step": 7510
+    },
+    {
+      "epoch": 0.022827038056632828,
+      "grad_norm": 0.28212642669677734,
+      "learning_rate": 0.0001865259130006559,
+      "loss": 0.19100323915481568,
+      "step": 7520
+    },
+    {
+      "epoch": 0.02311931767323504,
+      "grad_norm": 0.42614150047302246,
+      "learning_rate": 0.00018648550137912668,
+      "loss": 0.2577736616134644,
+      "step": 7530
+    },
+    {
+      "epoch": 0.023411597289837254,
+      "grad_norm": 0.3392845094203949,
+      "learning_rate": 0.00018644503363809457,
+      "loss": 0.252050518989563,
+      "step": 7540
+    },
+    {
+      "epoch": 0.023703876906439467,
+      "grad_norm": 0.4019984006881714,
+      "learning_rate": 0.00018640450980381864,
+      "loss": 0.22146039009094237,
+      "step": 7550
+    },
+    {
+      "epoch": 0.02399615652304168,
+      "grad_norm": 0.39968568086624146,
+      "learning_rate": 0.0001863639299025944,
+      "loss": 0.2448887825012207,
+      "step": 7560
+    },
+    {
+      "epoch": 0.024288436139643894,
+      "grad_norm": 0.4213276505470276,
+      "learning_rate": 0.00018632329396075367,
+      "loss": 0.2554018020629883,
+      "step": 7570
+    },
+    {
+      "epoch": 0.024580715756246107,
+      "grad_norm": 0.38602641224861145,
+      "learning_rate": 0.00018628260200466467,
+      "loss": 0.22209486961364747,
+      "step": 7580
+    },
+    {
+      "epoch": 0.02487299537284832,
+      "grad_norm": 0.39680173993110657,
+      "learning_rate": 0.00018624185406073202,
+      "loss": 0.2578996181488037,
+      "step": 7590
+    },
+    {
+      "epoch": 0.025165274989450533,
+      "grad_norm": 0.4088197946548462,
+      "learning_rate": 0.00018620105015539656,
+      "loss": 0.25044074058532717,
+      "step": 7600
+    },
+    {
+      "epoch": 0.025457554606052746,
+      "grad_norm": 0.5601091980934143,
+      "learning_rate": 0.00018616019031513555,
+      "loss": 0.24130692481994628,
+      "step": 7610
+    },
+    {
+      "epoch": 0.02574983422265496,
+      "grad_norm": 0.3851224184036255,
+      "learning_rate": 0.00018611927456646247,
+      "loss": 0.23012998104095458,
+      "step": 7620
+    },
+    {
+      "epoch": 0.026042113839257173,
+      "grad_norm": 0.3467167019844055,
+      "learning_rate": 0.00018607830293592708,
+      "loss": 0.24249653816223143,
+      "step": 7630
+    },
+    {
+      "epoch": 0.026334393455859383,
+      "grad_norm": 0.4822841286659241,
+      "learning_rate": 0.0001860372754501155,
+      "loss": 0.24729681015014648,
+      "step": 7640
+    },
+    {
+      "epoch": 0.026626673072461596,
+      "grad_norm": 0.3474031984806061,
+      "learning_rate": 0.00018599619213564993,
+      "loss": 0.18105405569076538,
+      "step": 7650
+    },
+    {
+      "epoch": 0.02691895268906381,
+      "grad_norm": 0.4408712685108185,
+      "learning_rate": 0.00018595505301918894,
+      "loss": 0.24149227142333984,
+      "step": 7660
+    },
+    {
+      "epoch": 0.027211232305666022,
+      "grad_norm": 0.41314172744750977,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 0.2502865791320801,
+      "step": 7670
+    },
+    {
+      "epoch": 0.027503511922268235,
+      "grad_norm": 0.4315735399723053,
+      "learning_rate": 0.0001858726074870958,
+      "loss": 0.2275188684463501,
+      "step": 7680
+    },
+    {
+      "epoch": 0.02779579153887045,
+      "grad_norm": 0.4159400463104248,
+      "learning_rate": 0.0001858313011249616,
+      "loss": 0.2392073392868042,
+      "step": 7690
+    },
+    {
+      "epoch": 0.02808807115547266,
+      "grad_norm": 0.5588905215263367,
+      "learning_rate": 0.000185789939067828,
+      "loss": 0.2303774356842041,
+      "step": 7700
+    },
+    {
+      "epoch": 0.028380350772074875,
+      "grad_norm": 0.45724934339523315,
+      "learning_rate": 0.00018574852134253434,
+      "loss": 0.25254509449005125,
+      "step": 7710
+    },
+    {
+      "epoch": 0.028672630388677088,
+      "grad_norm": 0.35972118377685547,
+      "learning_rate": 0.00018570704797595612,
+      "loss": 0.219134521484375,
+      "step": 7720
+    },
+    {
+      "epoch": 0.0289649100052793,
+      "grad_norm": 0.41628456115722656,
+      "learning_rate": 0.000185665518995005,
+      "loss": 0.2239521026611328,
+      "step": 7730
+    },
+    {
+      "epoch": 0.029257189621881514,
+      "grad_norm": 0.42209672927856445,
+      "learning_rate": 0.00018562393442662865,
+      "loss": 0.23624010086059571,
+      "step": 7740
+    },
+    {
+      "epoch": 0.029549469238483728,
+      "grad_norm": 0.45109790563583374,
+      "learning_rate": 0.00018558229429781087,
+      "loss": 0.23736159801483153,
+      "step": 7750
+    },
+    {
+      "epoch": 0.02984174885508594,
+      "grad_norm": 0.42826488614082336,
+      "learning_rate": 0.00018554059863557147,
+      "loss": 0.2701469898223877,
+      "step": 7760
+    },
+    {
+      "epoch": 0.03013402847168815,
+      "grad_norm": 0.37455710768699646,
+      "learning_rate": 0.00018549884746696632,
+      "loss": 0.259367561340332,
+      "step": 7770
+    },
+    {
+      "epoch": 0.030426308088290364,
+      "grad_norm": 0.5435960292816162,
+      "learning_rate": 0.00018545704081908727,
+      "loss": 0.28564670085906985,
+      "step": 7780
+    },
+    {
+      "epoch": 0.030718587704892577,
+      "grad_norm": 0.31798046827316284,
+      "learning_rate": 0.00018541517871906225,
+      "loss": 0.19342391490936278,
+      "step": 7790
+    },
+    {
+      "epoch": 0.03101086732149479,
+      "grad_norm": 0.339123398065567,
+      "learning_rate": 0.00018537326119405506,
+      "loss": 0.2209697484970093,
+      "step": 7800
+    },
+    {
+      "epoch": 0.03130314693809701,
+      "grad_norm": 0.3680548071861267,
+      "learning_rate": 0.00018533128827126561,
+      "loss": 0.25768260955810546,
+      "step": 7810
+    },
+    {
+      "epoch": 0.03159542655469922,
+      "grad_norm": 0.44600942730903625,
+      "learning_rate": 0.00018528925997792964,
+      "loss": 0.2178347587585449,
+      "step": 7820
+    },
+    {
+      "epoch": 0.03188770617130143,
+      "grad_norm": 0.33396708965301514,
+      "learning_rate": 0.0001852471763413188,
+      "loss": 0.1689058780670166,
+      "step": 7830
+    },
+    {
+      "epoch": 0.03217998578790364,
+      "grad_norm": 0.37988853454589844,
+      "learning_rate": 0.00018520503738874078,
+      "loss": 0.21625127792358398,
+      "step": 7840
+    },
+    {
+      "epoch": 0.03247226540450585,
+      "grad_norm": 0.43036121129989624,
+      "learning_rate": 0.00018516284314753904,
+      "loss": 0.25563740730285645,
+      "step": 7850
+    },
+    {
+      "epoch": 0.032764545021108066,
+      "grad_norm": 0.368350088596344,
+      "learning_rate": 0.00018512059364509298,
+      "loss": 0.19742240905761718,
+      "step": 7860
+    },
+    {
+      "epoch": 0.03305682463771028,
+      "grad_norm": 0.3769385516643524,
+      "learning_rate": 0.00018507828890881785,
+      "loss": 0.22123994827270507,
+      "step": 7870
+    },
+    {
+      "epoch": 0.03334910425431249,
+      "grad_norm": 0.4256688952445984,
+      "learning_rate": 0.00018503592896616474,
+      "loss": 0.22947709560394286,
+      "step": 7880
+    },
+    {
+      "epoch": 0.033641383870914705,
+      "grad_norm": 0.4441975951194763,
+      "learning_rate": 0.00018499351384462052,
+      "loss": 0.2533693313598633,
+      "step": 7890
+    },
+    {
+      "epoch": 0.03393366348751692,
+      "grad_norm": 0.2897224724292755,
+      "learning_rate": 0.00018495104357170794,
+      "loss": 0.2392364263534546,
+      "step": 7900
+    },
+    {
+      "epoch": 0.03422594310411913,
+      "grad_norm": 0.3767232298851013,
+      "learning_rate": 0.0001849085181749855,
+      "loss": 0.20210976600646974,
+      "step": 7910
+    },
+    {
+      "epoch": 0.034518222720721345,
+      "grad_norm": 0.4466973543167114,
+      "learning_rate": 0.00018486593768204742,
+      "loss": 0.23202342987060548,
+      "step": 7920
+    },
+    {
+      "epoch": 0.03481050233732356,
+      "grad_norm": 0.7618238925933838,
+      "learning_rate": 0.00018482330212052378,
+      "loss": 0.2601912498474121,
+      "step": 7930
+    },
+    {
+      "epoch": 0.03510278195392577,
+      "grad_norm": 0.5210285186767578,
+      "learning_rate": 0.00018478061151808032,
+      "loss": 0.1999301791191101,
+      "step": 7940
+    },
+    {
+      "epoch": 0.035395061570527984,
+      "grad_norm": 0.5094628930091858,
+      "learning_rate": 0.00018473786590241846,
+      "loss": 0.23355646133422853,
+      "step": 7950
+    },
+    {
+      "epoch": 0.0356873411871302,
+      "grad_norm": 0.378024160861969,
+      "learning_rate": 0.00018469506530127542,
+      "loss": 0.28238890171051023,
+      "step": 7960
+    },
+    {
+      "epoch": 0.03597962080373241,
+      "grad_norm": 0.392834335565567,
+      "learning_rate": 0.00018465220974242402,
+      "loss": 0.23338046073913574,
+      "step": 7970
+    },
+    {
+      "epoch": 0.036271900420334624,
+      "grad_norm": 0.38809776306152344,
+      "learning_rate": 0.0001846092992536728,
+      "loss": 0.20890545845031738,
+      "step": 7980
+    },
+    {
+      "epoch": 0.03656418003693684,
+      "grad_norm": 0.4836407005786896,
+      "learning_rate": 0.00018456633386286587,
+      "loss": 0.1863202929496765,
+      "step": 7990
+    },
+    {
+      "epoch": 0.03685645965353905,
+      "grad_norm": 0.5345965623855591,
+      "learning_rate": 0.000184523313597883,
+      "loss": 0.2104703426361084,
+      "step": 8000
+    },
+    {
+      "epoch": 0.037148739270141264,
+      "grad_norm": 0.5039252638816833,
+      "learning_rate": 0.0001844802384866396,
+      "loss": 0.28540492057800293,
+      "step": 8010
+    },
+    {
+      "epoch": 0.03744101888674348,
+      "grad_norm": 0.4195420444011688,
+      "learning_rate": 0.00018443710855708666,
+      "loss": 0.23810036182403566,
+      "step": 8020
+    },
+    {
+      "epoch": 0.03773329850334569,
+      "grad_norm": 0.42419466376304626,
+      "learning_rate": 0.00018439392383721068,
+      "loss": 0.22891428470611572,
+      "step": 8030
+    },
+    {
+      "epoch": 0.0380255781199479,
+      "grad_norm": 0.35419291257858276,
+      "learning_rate": 0.00018435068435503378,
+      "loss": 0.24456851482391356,
+      "step": 8040
+    },
+    {
+      "epoch": 0.038317857736550116,
+      "grad_norm": 0.41009142994880676,
+      "learning_rate": 0.0001843073901386136,
+      "loss": 0.24307987689971924,
+      "step": 8050
+    },
+    {
+      "epoch": 0.03861013735315233,
+      "grad_norm": 0.5681899785995483,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 0.24018590450286864,
+      "step": 8060
+    },
+    {
+      "epoch": 0.03890241696975454,
+      "grad_norm": 0.3411467671394348,
+      "learning_rate": 0.0001842206376154514,
+      "loss": 0.2108161211013794,
+      "step": 8070
+    },
+    {
+      "epoch": 0.039194696586356756,
+      "grad_norm": 0.4355998933315277,
+      "learning_rate": 0.00018417717936500215,
+      "loss": 0.1829616069793701,
+      "step": 8080
+    },
+    {
+      "epoch": 0.03948697620295897,
+      "grad_norm": 0.3300950527191162,
+      "learning_rate": 0.00018413366649289512,
+      "loss": 0.2595961332321167,
+      "step": 8090
+    },
+    {
+      "epoch": 0.039779255819561175,
+      "grad_norm": 0.5162358283996582,
+      "learning_rate": 0.00018409009902736534,
+      "loss": 0.23371126651763915,
+      "step": 8100
+    },
+    {
+      "epoch": 0.04007153543616339,
+      "grad_norm": 0.4499688148498535,
+      "learning_rate": 0.00018404647699668325,
+      "loss": 0.28339409828186035,
+      "step": 8110
+    },
+    {
+      "epoch": 0.0403638150527656,
+      "grad_norm": 0.4983923137187958,
+      "learning_rate": 0.0001840028004291547,
+      "loss": 0.20083823204040527,
+      "step": 8120
+    },
+    {
+      "epoch": 0.040656094669367815,
+      "grad_norm": 0.3119497001171112,
+      "learning_rate": 0.000183959069353121,
+      "loss": 0.24721684455871581,
+      "step": 8130
+    },
+    {
+      "epoch": 0.04094837428597003,
+      "grad_norm": 0.3148126006126404,
+      "learning_rate": 0.00018391528379695872,
+      "loss": 0.23982763290405273,
+      "step": 8140
+    },
+    {
+      "epoch": 0.04124065390257224,
+      "grad_norm": 0.3485938608646393,
+      "learning_rate": 0.0001838714437890798,
+      "loss": 0.1926335096359253,
+      "step": 8150
+    },
+    {
+      "epoch": 0.041532933519174454,
+      "grad_norm": 0.539906919002533,
+      "learning_rate": 0.0001838275493579316,
+      "loss": 0.3179804801940918,
+      "step": 8160
+    },
+    {
+      "epoch": 0.04182521313577667,
+      "grad_norm": 0.4284425377845764,
+      "learning_rate": 0.00018378360053199674,
+      "loss": 0.19903416633605958,
+      "step": 8170
+    },
+    {
+      "epoch": 0.04211749275237888,
+      "grad_norm": 0.34267932176589966,
+      "learning_rate": 0.00018373959733979312,
+      "loss": 0.22843327522277831,
+      "step": 8180
+    },
+    {
+      "epoch": 0.042409772368981094,
+      "grad_norm": 0.41015568375587463,
+      "learning_rate": 0.0001836955398098739,
+      "loss": 0.27080578804016114,
+      "step": 8190
+    },
+    {
+      "epoch": 0.04270205198558331,
+      "grad_norm": 0.363918662071228,
+      "learning_rate": 0.0001836514279708276,
+      "loss": 0.31085083484649656,
+      "step": 8200
+    },
+    {
+      "epoch": 0.04299433160218552,
+      "grad_norm": 0.4398233890533447,
+      "learning_rate": 0.00018360726185127788,
+      "loss": 0.2242945432662964,
+      "step": 8210
+    },
+    {
+      "epoch": 0.043286611218787734,
+      "grad_norm": 0.6542929410934448,
+      "learning_rate": 0.00018356304147988363,
+      "loss": 0.19122848510742188,
+      "step": 8220
+    },
+    {
+      "epoch": 0.04357889083538995,
+      "grad_norm": 0.42511820793151855,
+      "learning_rate": 0.00018351876688533903,
+      "loss": 0.22499797344207764,
+      "step": 8230
+    },
+    {
+      "epoch": 0.04387117045199216,
+      "grad_norm": 0.37980619072914124,
+      "learning_rate": 0.00018347443809637332,
+      "loss": 0.2509718418121338,
+      "step": 8240
+    },
+    {
+      "epoch": 0.04416345006859437,
+      "grad_norm": 0.3857591450214386,
+      "learning_rate": 0.000183430055141751,
+      "loss": 0.2113365888595581,
+      "step": 8250
+    },
+    {
+      "epoch": 0.044455729685196586,
+      "grad_norm": 0.4793608486652374,
+      "learning_rate": 0.0001833856180502717,
+      "loss": 0.23078553676605223,
+      "step": 8260
+    },
+    {
+      "epoch": 0.0447480093017988,
+      "grad_norm": 0.49650800228118896,
+      "learning_rate": 0.00018334112685077013,
+      "loss": 0.23284411430358887,
+      "step": 8270
+    },
+    {
+      "epoch": 0.04504028891840101,
+      "grad_norm": 0.42458006739616394,
+      "learning_rate": 0.0001832965815721162,
+      "loss": 0.16628366708755493,
+      "step": 8280
+    },
+    {
+      "epoch": 0.045332568535003226,
+      "grad_norm": 0.45042723417282104,
+      "learning_rate": 0.0001832519822432148,
+      "loss": 0.18936212062835694,
+      "step": 8290
+    },
+    {
+      "epoch": 0.04562484815160544,
+      "grad_norm": 0.46213051676750183,
+      "learning_rate": 0.000183207328893006,
+      "loss": 0.1903870224952698,
+      "step": 8300
+    },
+    {
+      "epoch": 0.04591712776820765,
+      "grad_norm": 0.4746432900428772,
+      "learning_rate": 0.00018316262155046483,
+      "loss": 0.2012183904647827,
+      "step": 8310
+    },
+    {
+      "epoch": 0.046209407384809865,
+      "grad_norm": 0.46848586201667786,
+      "learning_rate": 0.0001831178602446014,
+      "loss": 0.25582935810089114,
+      "step": 8320
+    },
+    {
+      "epoch": 0.04650168700141208,
+      "grad_norm": 0.46606945991516113,
+      "learning_rate": 0.00018307304500446092,
+      "loss": 0.2769018650054932,
+      "step": 8330
+    },
+    {
+      "epoch": 0.04679396661801429,
+      "grad_norm": 0.4793144166469574,
+      "learning_rate": 0.0001830281758591234,
+      "loss": 0.24349710941314698,
+      "step": 8340
+    },
+    {
+      "epoch": 0.047086246234616505,
+      "grad_norm": 0.4187564253807068,
+      "learning_rate": 0.00018298325283770399,
+      "loss": 0.2414229154586792,
+      "step": 8350
+    },
+    {
+      "epoch": 0.04737852585121871,
+      "grad_norm": 0.48000621795654297,
+      "learning_rate": 0.00018293827596935276,
+      "loss": 0.23797121047973632,
+      "step": 8360
+    },
+    {
+      "epoch": 0.047670805467820924,
+      "grad_norm": 0.46506500244140625,
+      "learning_rate": 0.0001828932452832547,
+      "loss": 0.22981607913970947,
+      "step": 8370
+    },
+    {
+      "epoch": 0.04796308508442314,
+      "grad_norm": 0.45346197485923767,
+      "learning_rate": 0.00018284816080862975,
+      "loss": 0.27556185722351073,
+      "step": 8380
+    },
+    {
+      "epoch": 0.04825536470102535,
+      "grad_norm": 0.3772803246974945,
+      "learning_rate": 0.00018280302257473268,
+      "loss": 0.2145310401916504,
+      "step": 8390
+    },
+    {
+      "epoch": 0.048547644317627564,
+      "grad_norm": 0.415454626083374,
+      "learning_rate": 0.00018275783061085326,
+      "loss": 0.20502667427062987,
+      "step": 8400
+    },
+    {
+      "epoch": 0.04883992393422978,
+      "grad_norm": 0.4649645984172821,
+      "learning_rate": 0.00018271258494631604,
+      "loss": 0.192656946182251,
+      "step": 8410
+    },
+    {
+      "epoch": 0.04913220355083199,
+      "grad_norm": 0.47047504782676697,
+      "learning_rate": 0.0001826672856104805,
+      "loss": 0.2049344778060913,
+      "step": 8420
+    },
+    {
+      "epoch": 0.049424483167434204,
+      "grad_norm": 0.455949604511261,
+      "learning_rate": 0.00018262193263274076,
+      "loss": 0.2255397319793701,
+      "step": 8430
+    },
+    {
+      "epoch": 0.04971676278403642,
+      "grad_norm": 0.5024589896202087,
+      "learning_rate": 0.00018257652604252597,
+      "loss": 0.20659456253051758,
+      "step": 8440
+    },
+    {
+      "epoch": 0.05000904240063863,
+      "grad_norm": 0.38048383593559265,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 0.2542705535888672,
+      "step": 8450
+    },
+    {
+      "epoch": 0.05030132201724084,
+      "grad_norm": 0.38060474395751953,
+      "learning_rate": 0.00018248555214256133,
+      "loss": 0.22730619907379152,
+      "step": 8460
+    },
+    {
+      "epoch": 0.050593601633843056,
+      "grad_norm": 0.4579905569553375,
+      "learning_rate": 0.00018243998489184346,
+      "loss": 0.24086904525756836,
+      "step": 8470
+    },
+    {
+      "epoch": 0.05088588125044527,
+      "grad_norm": 0.4802927076816559,
+      "learning_rate": 0.0001823943641467144,
+      "loss": 0.24286856651306152,
+      "step": 8480
+    },
+    {
+      "epoch": 0.05117816086704748,
+      "grad_norm": 0.5876162052154541,
+      "learning_rate": 0.00018234868993677704,
+      "loss": 0.2275693416595459,
+      "step": 8490
+    },
+    {
+      "epoch": 0.051470440483649696,
+      "grad_norm": 0.5871111750602722,
+      "learning_rate": 0.00018230296229166878,
+      "loss": 0.1766088128089905,
+      "step": 8500
+    },
+    {
+      "epoch": 0.05176272010025191,
+      "grad_norm": 0.5762088298797607,
+      "learning_rate": 0.00018225718124106189,
+      "loss": 0.1761952519416809,
+      "step": 8510
+    },
+    {
+      "epoch": 0.05205499971685412,
+      "grad_norm": 0.39905136823654175,
+      "learning_rate": 0.0001822113468146631,
+      "loss": 0.22955873012542724,
+      "step": 8520
+    },
+    {
+      "epoch": 0.052347279333456335,
+      "grad_norm": 0.3767608404159546,
+      "learning_rate": 0.0001821654590422139,
+      "loss": 0.2027675151824951,
+      "step": 8530
+    },
+    {
+      "epoch": 0.05263955895005855,
+      "grad_norm": 0.538753092288971,
+      "learning_rate": 0.00018211951795349044,
+      "loss": 0.20578184127807617,
+      "step": 8540
+    },
+    {
+      "epoch": 0.05293183856666076,
+      "grad_norm": 0.3167216181755066,
+      "learning_rate": 0.00018207352357830332,
+      "loss": 0.26151602268218993,
+      "step": 8550
+    },
+    {
+      "epoch": 0.053224118183262975,
+      "grad_norm": 0.41925323009490967,
+      "learning_rate": 0.00018202747594649781,
+      "loss": 0.21729581356048583,
+      "step": 8560
+    },
+    {
+      "epoch": 0.05351639779986519,
+      "grad_norm": 0.516607940196991,
+      "learning_rate": 0.0001819813750879537,
+      "loss": 0.19528690576553345,
+      "step": 8570
+    },
+    {
+      "epoch": 0.0538086774164674,
+      "grad_norm": 0.44117453694343567,
+      "learning_rate": 0.00018193522103258537,
+      "loss": 0.24370894432067872,
+      "step": 8580
+    },
+    {
+      "epoch": 0.054100957033069615,
+      "grad_norm": 0.4419977366924286,
+      "learning_rate": 0.00018188901381034164,
+      "loss": 0.20720908641815186,
+      "step": 8590
+    },
+    {
+      "epoch": 0.05439323664967183,
+      "grad_norm": 0.4751587510108948,
+      "learning_rate": 0.00018184275345120595,
+      "loss": 0.22340197563171388,
+      "step": 8600
+    },
+    {
+      "epoch": 0.05468551626627404,
+      "grad_norm": 0.37093105912208557,
+      "learning_rate": 0.00018179643998519608,
+      "loss": 0.24044995307922362,
+      "step": 8610
+    },
+    {
+      "epoch": 0.05497779588287625,
+      "grad_norm": 0.4716920852661133,
+      "learning_rate": 0.00018175007344236436,
+      "loss": 0.21340856552124024,
+      "step": 8620
+    },
+    {
+      "epoch": 0.05527007549947846,
+      "grad_norm": 0.49651649594306946,
+      "learning_rate": 0.00018170365385279753,
+      "loss": 0.2542828798294067,
+      "step": 8630
+    },
+    {
+      "epoch": 0.055562355116080674,
+      "grad_norm": 0.38708168268203735,
+      "learning_rate": 0.00018165718124661675,
+      "loss": 0.18677669763565063,
+      "step": 8640
+    },
+    {
+      "epoch": 0.05585463473268289,
+      "grad_norm": 0.353691041469574,
+      "learning_rate": 0.0001816106556539776,
+      "loss": 0.20028114318847656,
+      "step": 8650
+    },
+    {
+      "epoch": 0.0561469143492851,
+      "grad_norm": 0.4004218876361847,
+      "learning_rate": 0.00018156407710507005,
+      "loss": 0.24421725273132325,
+      "step": 8660
+    },
+    {
+      "epoch": 0.05643919396588731,
+      "grad_norm": 0.44456422328948975,
+      "learning_rate": 0.00018151744563011835,
+      "loss": 0.20305163860321046,
+      "step": 8670
+    },
+    {
+      "epoch": 0.056731473582489526,
+      "grad_norm": 0.3974579870700836,
+      "learning_rate": 0.00018147076125938125,
+      "loss": 0.20146026611328124,
+      "step": 8680
+    },
+    {
+      "epoch": 0.05702375319909174,
+      "grad_norm": 0.4056035280227661,
+      "learning_rate": 0.00018142402402315165,
+      "loss": 0.21869051456451416,
+      "step": 8690
+    },
+    {
+      "epoch": 0.05731603281569395,
+      "grad_norm": 0.603005051612854,
+      "learning_rate": 0.0001813772339517569,
+      "loss": 0.2276228427886963,
+      "step": 8700
+    },
+    {
+      "epoch": 0.057608312432296166,
+      "grad_norm": 0.4758279025554657,
+      "learning_rate": 0.00018133039107555852,
+      "loss": 0.20320589542388917,
+      "step": 8710
+    },
+    {
+      "epoch": 0.05790059204889838,
+      "grad_norm": 0.4756017327308655,
+      "learning_rate": 0.00018128349542495236,
+      "loss": 0.22813382148742675,
+      "step": 8720
+    },
+    {
+      "epoch": 0.05819287166550059,
+      "grad_norm": 0.4477347433567047,
+      "learning_rate": 0.00018123654703036852,
+      "loss": 0.21260826587677,
+      "step": 8730
+    },
+    {
+      "epoch": 0.058485151282102806,
+      "grad_norm": 0.4087035059928894,
+      "learning_rate": 0.00018118954592227126,
+      "loss": 0.2012578248977661,
+      "step": 8740
+    },
+    {
+      "epoch": 0.05877743089870502,
+      "grad_norm": 0.48341134190559387,
+      "learning_rate": 0.00018114249213115912,
+      "loss": 0.24345796108245848,
+      "step": 8750
+    },
+    {
+      "epoch": 0.05906971051530723,
+      "grad_norm": 0.4561239182949066,
+      "learning_rate": 0.0001810953856875648,
+      "loss": 0.2859776258468628,
+      "step": 8760
+    },
+    {
+      "epoch": 0.059361990131909445,
+      "grad_norm": 0.42124632000923157,
+      "learning_rate": 0.00018104822662205514,
+      "loss": 0.18385095596313478,
+      "step": 8770
+    },
+    {
+      "epoch": 0.05965426974851166,
+      "grad_norm": 0.402494877576828,
+      "learning_rate": 0.00018100101496523117,
+      "loss": 0.2651110887527466,
+      "step": 8780
+    },
+    {
+      "epoch": 0.05994654936511387,
+      "grad_norm": 0.5392200946807861,
+      "learning_rate": 0.00018095375074772798,
+      "loss": 0.19454458951950074,
+      "step": 8790
+    },
+    {
+      "epoch": 0.060238828981716085,
+      "grad_norm": 0.5405917763710022,
+      "learning_rate": 0.00018090643400021483,
+      "loss": 0.16877477169036864,
+      "step": 8800
+    },
+    {
+      "epoch": 0.0605311085983183,
+      "grad_norm": 0.4851214587688446,
+      "learning_rate": 0.00018085906475339507,
+      "loss": 0.21640102863311766,
+      "step": 8810
+    },
+    {
+      "epoch": 0.06082338821492051,
+      "grad_norm": 0.5040847659111023,
+      "learning_rate": 0.00018081164303800605,
+      "loss": 0.21950478553771974,
+      "step": 8820
+    },
+    {
+      "epoch": 0.061115667831522724,
+      "grad_norm": 0.3940047025680542,
+      "learning_rate": 0.00018076416888481922,
+      "loss": 0.17876697778701783,
+      "step": 8830
+    },
+    {
+      "epoch": 0.06140794744812494,
+      "grad_norm": 0.3477630317211151,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 0.18992202281951903,
+      "step": 8840
+    },
+    {
+      "epoch": 0.06170022706472715,
+      "grad_norm": 0.5437178015708923,
+      "learning_rate": 0.000180669063388308,
+      "loss": 0.24603896141052245,
+      "step": 8850
+    },
+    {
+      "epoch": 0.061992506681329364,
+      "grad_norm": 0.318469375371933,
+      "learning_rate": 0.00018062143210669654,
+      "loss": 0.2027036428451538,
+      "step": 8860
+    },
+    {
+      "epoch": 0.06228478629793158,
+      "grad_norm": 0.48581576347351074,
+      "learning_rate": 0.00018057374851071305,
+      "loss": 0.31111702919006345,
+      "step": 8870
+    },
+    {
+      "epoch": 0.06257706591453378,
+      "grad_norm": 0.4098704755306244,
+      "learning_rate": 0.00018052601263129895,
+      "loss": 0.36435375213623045,
+      "step": 8880
+    },
+    {
+      "epoch": 0.062869345531136,
+      "grad_norm": 0.41771751642227173,
+      "learning_rate": 0.0001804782244994295,
+      "loss": 0.34163413047790525,
+      "step": 8890
+    },
+    {
+      "epoch": 0.06316162514773821,
+      "grad_norm": 0.3779888451099396,
+      "learning_rate": 0.00018043038414611393,
+      "loss": 0.3448063135147095,
+      "step": 8900
+    },
+    {
+      "epoch": 0.06345390476434043,
+      "grad_norm": 0.35888513922691345,
+      "learning_rate": 0.00018038249160239527,
+      "loss": 0.36453568935394287,
+      "step": 8910
+    },
+    {
+      "epoch": 0.06374618438094264,
+      "grad_norm": 0.4041270315647125,
+      "learning_rate": 0.00018033454689935055,
+      "loss": 0.2908053398132324,
+      "step": 8920
+    },
+    {
+      "epoch": 0.06403846399754486,
+      "grad_norm": 0.3345756530761719,
+      "learning_rate": 0.00018028655006809054,
+      "loss": 0.2657099485397339,
+      "step": 8930
+    },
+    {
+      "epoch": 0.06433074361414706,
+      "grad_norm": 0.38582342863082886,
+      "learning_rate": 0.0001802385011397598,
+      "loss": 0.3580424070358276,
+      "step": 8940
+    },
+    {
+      "epoch": 0.06462302323074928,
+      "grad_norm": 0.3744290769100189,
+      "learning_rate": 0.00018019040014553684,
+      "loss": 0.31871423721313474,
+      "step": 8950
+    },
+    {
+      "epoch": 0.06491530284735149,
+      "grad_norm": 0.3543660044670105,
+      "learning_rate": 0.00018014224711663385,
+      "loss": 0.2808994770050049,
+      "step": 8960
+    },
+    {
+      "epoch": 0.06520758246395371,
+      "grad_norm": 0.39081060886383057,
+      "learning_rate": 0.00018009404208429677,
+      "loss": 0.31113102436065676,
+      "step": 8970
+    },
+    {
+      "epoch": 0.06549986208055592,
+      "grad_norm": 0.43403148651123047,
+      "learning_rate": 0.00018004578507980535,
+      "loss": 0.31454265117645264,
+      "step": 8980
+    },
+    {
+      "epoch": 0.06579214169715814,
+      "grad_norm": 0.40630385279655457,
+      "learning_rate": 0.00017999747613447307,
+      "loss": 0.2785582304000854,
+      "step": 8990
+    },
+    {
+      "epoch": 0.06608442131376034,
+      "grad_norm": 0.30130770802497864,
+      "learning_rate": 0.00017994911527964702,
+      "loss": 0.31431634426116944,
+      "step": 9000
+    },
+    {
+      "epoch": 0.06637670093036255,
+      "grad_norm": 0.4664454460144043,
+      "learning_rate": 0.00017990070254670805,
+      "loss": 0.303541374206543,
+      "step": 9010
+    },
+    {
+      "epoch": 0.06666898054696477,
+      "grad_norm": 0.5049722194671631,
+      "learning_rate": 0.0001798522379670707,
+      "loss": 0.35490193367004397,
+      "step": 9020
+    },
+    {
+      "epoch": 0.06696126016356697,
+      "grad_norm": 0.37954920530319214,
+      "learning_rate": 0.00017980372157218305,
+      "loss": 0.31733648777008056,
+      "step": 9030
+    },
+    {
+      "epoch": 0.0672535397801692,
+      "grad_norm": 0.48762884736061096,
+      "learning_rate": 0.0001797551533935269,
+      "loss": 0.32967276573181153,
+      "step": 9040
+    },
+    {
+      "epoch": 0.0675458193967714,
+      "grad_norm": 0.38147637248039246,
+      "learning_rate": 0.00017970653346261758,
+      "loss": 0.25834801197052004,
+      "step": 9050
+    },
+    {
+      "epoch": 0.06783809901337362,
+      "grad_norm": 0.43979838490486145,
+      "learning_rate": 0.00017965786181100408,
+      "loss": 0.390688419342041,
+      "step": 9060
+    },
+    {
+      "epoch": 0.06813037862997583,
+      "grad_norm": 0.4078044593334198,
+      "learning_rate": 0.00017960913847026887,
+      "loss": 0.3295466661453247,
+      "step": 9070
+    },
+    {
+      "epoch": 0.06842265824657805,
+      "grad_norm": 0.2929011285305023,
+      "learning_rate": 0.00017956036347202802,
+      "loss": 0.3083007335662842,
+      "step": 9080
+    },
+    {
+      "epoch": 0.06871493786318025,
+      "grad_norm": 0.3656540811061859,
+      "learning_rate": 0.0001795115368479311,
+      "loss": 0.3585763454437256,
+      "step": 9090
+    },
+    {
+      "epoch": 0.06900721747978247,
+      "grad_norm": 0.4311404824256897,
+      "learning_rate": 0.00017946265862966114,
+      "loss": 0.33038156032562255,
+      "step": 9100
+    },
+    {
+      "epoch": 0.06929949709638468,
+      "grad_norm": 0.38878104090690613,
+      "learning_rate": 0.00017941372884893472,
+      "loss": 0.3634531021118164,
+      "step": 9110
+    },
+    {
+      "epoch": 0.0695917767129869,
+      "grad_norm": 0.2877635657787323,
+      "learning_rate": 0.00017936474753750184,
+      "loss": 0.28738400936126707,
+      "step": 9120
+    },
+    {
+      "epoch": 0.0698840563295891,
+      "grad_norm": 0.3645118772983551,
+      "learning_rate": 0.00017931571472714594,
+      "loss": 0.3121068000793457,
+      "step": 9130
+    },
+    {
+      "epoch": 0.07017633594619133,
+      "grad_norm": 0.3508014678955078,
+      "learning_rate": 0.00017926663044968386,
+      "loss": 0.31461095809936523,
+      "step": 9140
+    },
+    {
+      "epoch": 0.07046861556279353,
+      "grad_norm": 0.39282628893852234,
+      "learning_rate": 0.00017921749473696585,
+      "loss": 0.2887994050979614,
+      "step": 9150
+    },
+    {
+      "epoch": 0.07076089517939575,
+      "grad_norm": 0.3809869587421417,
+      "learning_rate": 0.00017916830762087558,
+      "loss": 0.27705438137054444,
+      "step": 9160
+    },
+    {
+      "epoch": 0.07105317479599796,
+      "grad_norm": 0.4888908267021179,
+      "learning_rate": 0.00017911906913333,
+      "loss": 0.2871765375137329,
+      "step": 9170
+    },
+    {
+      "epoch": 0.07134545441260018,
+      "grad_norm": 0.4648347496986389,
+      "learning_rate": 0.00017906977930627944,
+      "loss": 0.28583788871765137,
+      "step": 9180
+    },
+    {
+      "epoch": 0.07163773402920239,
+      "grad_norm": 0.7212278246879578,
+      "learning_rate": 0.00017902043817170755,
+      "loss": 0.3232898235321045,
+      "step": 9190
+    },
+    {
+      "epoch": 0.0719300136458046,
+      "grad_norm": 0.437978595495224,
+      "learning_rate": 0.00017897104576163124,
+      "loss": 0.329192590713501,
+      "step": 9200
+    },
+    {
+      "epoch": 0.07222229326240681,
+      "grad_norm": 0.3960765302181244,
+      "learning_rate": 0.0001789216021081007,
+      "loss": 0.2738374948501587,
+      "step": 9210
+    },
+    {
+      "epoch": 0.07251457287900903,
+      "grad_norm": 0.34261566400527954,
+      "learning_rate": 0.0001788721072431994,
+      "loss": 0.3428427457809448,
+      "step": 9220
+    },
+    {
+      "epoch": 0.07280685249561124,
+      "grad_norm": 0.41202136874198914,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 0.32028656005859374,
+      "step": 9230
+    },
+    {
+      "epoch": 0.07309913211221346,
+      "grad_norm": 0.3411543667316437,
+      "learning_rate": 0.00017877296400778444,
+      "loss": 0.32127909660339354,
+      "step": 9240
+    },
+    {
+      "epoch": 0.07339141172881566,
+      "grad_norm": 0.2918577194213867,
+      "learning_rate": 0.0001787233157016037,
+      "loss": 0.3321556568145752,
+      "step": 9250
+    },
+    {
+      "epoch": 0.07368369134541788,
+      "grad_norm": 0.3458452820777893,
+      "learning_rate": 0.00017867361631271816,
+      "loss": 0.32092010974884033,
+      "step": 9260
+    },
+    {
+      "epoch": 0.07397597096202009,
+      "grad_norm": 0.3268740475177765,
+      "learning_rate": 0.00017862386587337708,
+      "loss": 0.2877511024475098,
+      "step": 9270
+    },
+    {
+      "epoch": 0.0742682505786223,
+      "grad_norm": 0.31670641899108887,
+      "learning_rate": 0.00017857406441586305,
+      "loss": 0.3533596992492676,
+      "step": 9280
+    },
+    {
+      "epoch": 0.07456053019522452,
+      "grad_norm": 0.4070427715778351,
+      "learning_rate": 0.00017852421197249175,
+      "loss": 0.3150961875915527,
+      "step": 9290
+    },
+    {
+      "epoch": 0.07485280981182672,
+      "grad_norm": 0.4110407531261444,
+      "learning_rate": 0.00017847430857561176,
+      "loss": 0.324555492401123,
+      "step": 9300
+    },
+    {
+      "epoch": 0.07514508942842894,
+      "grad_norm": 0.3359784483909607,
+      "learning_rate": 0.00017842435425760497,
+      "loss": 0.3238916635513306,
+      "step": 9310
+    },
+    {
+      "epoch": 0.07543736904503115,
+      "grad_norm": 0.439630925655365,
+      "learning_rate": 0.0001783743490508861,
+      "loss": 0.3546104907989502,
+      "step": 9320
+    },
+    {
+      "epoch": 0.07572964866163337,
+      "grad_norm": 0.3349355459213257,
+      "learning_rate": 0.00017832429298790303,
+      "loss": 0.28695437908172605,
+      "step": 9330
+    },
+    {
+      "epoch": 0.07602192827823558,
+      "grad_norm": 0.41388610005378723,
+      "learning_rate": 0.00017827418610113658,
+      "loss": 0.32479281425476075,
+      "step": 9340
+    },
+    {
+      "epoch": 0.0763142078948378,
+      "grad_norm": 0.434333860874176,
+      "learning_rate": 0.00017822402842310062,
+      "loss": 0.2958407402038574,
+      "step": 9350
+    },
+    {
+      "epoch": 0.07660648751144,
+      "grad_norm": 0.36904022097587585,
+      "learning_rate": 0.00017817381998634185,
+      "loss": 0.33018038272857664,
+      "step": 9360
+    },
+    {
+      "epoch": 0.07689876712804222,
+      "grad_norm": 0.300948828458786,
+      "learning_rate": 0.00017812356082344004,
+      "loss": 0.2697992563247681,
+      "step": 9370
+    },
+    {
+      "epoch": 0.07719104674464443,
+      "grad_norm": 0.3979049324989319,
+      "learning_rate": 0.00017807325096700778,
+      "loss": 0.3914919376373291,
+      "step": 9380
+    },
+    {
+      "epoch": 0.07748332636124665,
+      "grad_norm": 0.5494265556335449,
+      "learning_rate": 0.0001780228904496906,
+      "loss": 0.3006611347198486,
+      "step": 9390
+    },
+    {
+      "epoch": 0.07777560597784886,
+      "grad_norm": 0.36540231108665466,
+      "learning_rate": 0.00017797247930416687,
+      "loss": 0.3546081781387329,
+      "step": 9400
+    },
+    {
+      "epoch": 0.07806788559445108,
+      "grad_norm": 0.34306907653808594,
+      "learning_rate": 0.00017792201756314788,
+      "loss": 0.2969916820526123,
+      "step": 9410
+    },
+    {
+      "epoch": 0.07836016521105328,
+      "grad_norm": 0.4678438603878021,
+      "learning_rate": 0.00017787150525937765,
+      "loss": 0.37559340000152586,
+      "step": 9420
+    },
+    {
+      "epoch": 0.0786524448276555,
+      "grad_norm": 0.3107658624649048,
+      "learning_rate": 0.00017782094242563314,
+      "loss": 0.35972464084625244,
+      "step": 9430
+    },
+    {
+      "epoch": 0.07894472444425771,
+      "grad_norm": 0.2997909188270569,
+      "learning_rate": 0.00017777032909472397,
+      "loss": 0.27579784393310547,
+      "step": 9440
+    },
+    {
+      "epoch": 0.07923700406085993,
+      "grad_norm": 0.3486426770687103,
+      "learning_rate": 0.00017771966529949256,
+      "loss": 0.3102571487426758,
+      "step": 9450
+    },
+    {
+      "epoch": 0.07952928367746213,
+      "grad_norm": 0.36216917634010315,
+      "learning_rate": 0.00017766895107281414,
+      "loss": 0.29436097145080564,
+      "step": 9460
+    },
+    {
+      "epoch": 0.07982156329406435,
+      "grad_norm": 0.4105483591556549,
+      "learning_rate": 0.0001776181864475966,
+      "loss": 0.26256139278411866,
+      "step": 9470
+    },
+    {
+      "epoch": 0.08011384291066656,
+      "grad_norm": 0.34755370020866394,
+      "learning_rate": 0.00017756737145678057,
+      "loss": 0.30299837589263917,
+      "step": 9480
+    },
+    {
+      "epoch": 0.08040612252726878,
+      "grad_norm": 0.3453555405139923,
+      "learning_rate": 0.00017751650613333935,
+      "loss": 0.3280250310897827,
+      "step": 9490
+    },
+    {
+      "epoch": 0.08069840214387099,
+      "grad_norm": 0.2993573248386383,
+      "learning_rate": 0.00017746559051027885,
+      "loss": 0.3437278509140015,
+      "step": 9500
+    },
+    {
+      "epoch": 0.08099068176047321,
+      "grad_norm": 0.3788002133369446,
+      "learning_rate": 0.00017741462462063777,
+      "loss": 0.35432255268096924,
+      "step": 9510
+    },
+    {
+      "epoch": 0.08128296137707541,
+      "grad_norm": 0.43004506826400757,
+      "learning_rate": 0.0001773636084974872,
+      "loss": 0.38433198928833007,
+      "step": 9520
+    },
+    {
+      "epoch": 0.08157524099367762,
+      "grad_norm": 0.3504483997821808,
+      "learning_rate": 0.000177312542173931,
+      "loss": 0.2782111406326294,
+      "step": 9530
+    },
+    {
+      "epoch": 0.08186752061027984,
+      "grad_norm": 0.5386286377906799,
+      "learning_rate": 0.00017726142568310555,
+      "loss": 0.32076804637908934,
+      "step": 9540
+    },
+    {
+      "epoch": 0.08215980022688205,
+      "grad_norm": 0.4217930734157562,
+      "learning_rate": 0.0001772102590581798,
+      "loss": 0.28317766189575194,
+      "step": 9550
+    },
+    {
+      "epoch": 0.08245207984348427,
+      "grad_norm": 0.36821630597114563,
+      "learning_rate": 0.0001771590423323552,
+      "loss": 0.33262479305267334,
+      "step": 9560
+    },
+    {
+      "epoch": 0.08274435946008647,
+      "grad_norm": 0.393311470746994,
+      "learning_rate": 0.00017710777553886576,
+      "loss": 0.3143324851989746,
+      "step": 9570
+    },
+    {
+      "epoch": 0.08303663907668869,
+      "grad_norm": 0.4363000988960266,
+      "learning_rate": 0.0001770564587109779,
+      "loss": 0.3283381462097168,
+      "step": 9580
+    },
+    {
+      "epoch": 0.0833289186932909,
+      "grad_norm": 0.3231413662433624,
+      "learning_rate": 0.00017700509188199059,
+      "loss": 0.3241008520126343,
+      "step": 9590
+    },
+    {
+      "epoch": 0.08362119830989312,
+      "grad_norm": 0.40536484122276306,
+      "learning_rate": 0.00017695367508523515,
+      "loss": 0.33399035930633547,
+      "step": 9600
+    },
+    {
+      "epoch": 0.08391347792649533,
+      "grad_norm": 0.3365936875343323,
+      "learning_rate": 0.00017690220835407544,
+      "loss": 0.3233845949172974,
+      "step": 9610
+    },
+    {
+      "epoch": 0.08420575754309755,
+      "grad_norm": 0.36744701862335205,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 0.33194310665130616,
+      "step": 9620
+    },
+    {
+      "epoch": 0.08449803715969975,
+      "grad_norm": 0.3302032947540283,
+      "learning_rate": 0.00017679912522216039,
+      "loss": 0.28752167224884034,
+      "step": 9630
+    },
+    {
+      "epoch": 0.08479031677630197,
+      "grad_norm": 0.4165116250514984,
+      "learning_rate": 0.00017674750888829453,
+      "loss": 0.315144681930542,
+      "step": 9640
+    },
+    {
+      "epoch": 0.08508259639290418,
+      "grad_norm": 0.6105031967163086,
+      "learning_rate": 0.0001766958427538034,
+      "loss": 0.33243870735168457,
+      "step": 9650
+    },
+    {
+      "epoch": 0.0853748760095064,
+      "grad_norm": 0.4719133973121643,
+      "learning_rate": 0.0001766441268522126,
+      "loss": 0.3987681150436401,
+      "step": 9660
+    },
+    {
+      "epoch": 0.0856671556261086,
+      "grad_norm": 0.32340067625045776,
+      "learning_rate": 0.00017659236121707995,
+      "loss": 0.29813148975372317,
+      "step": 9670
+    },
+    {
+      "epoch": 0.08595943524271082,
+      "grad_norm": 0.38599932193756104,
+      "learning_rate": 0.00017654054588199568,
+      "loss": 0.3234374761581421,
+      "step": 9680
+    },
+    {
+      "epoch": 0.08625171485931303,
+      "grad_norm": 0.43444472551345825,
+      "learning_rate": 0.0001764886808805822,
+      "loss": 0.25468149185180666,
+      "step": 9690
+    },
+    {
+      "epoch": 0.08654399447591525,
+      "grad_norm": 0.3881465494632721,
+      "learning_rate": 0.00017643676624649407,
+      "loss": 0.324559760093689,
+      "step": 9700
+    },
+    {
+      "epoch": 0.08683627409251746,
+      "grad_norm": 0.4933682382106781,
+      "learning_rate": 0.00017638480201341824,
+      "loss": 0.29977123737335204,
+      "step": 9710
+    },
+    {
+      "epoch": 0.08712855370911968,
+      "grad_norm": 0.28426170349121094,
+      "learning_rate": 0.00017633278821507363,
+      "loss": 0.2909044027328491,
+      "step": 9720
+    },
+    {
+      "epoch": 0.08742083332572188,
+      "grad_norm": 0.4783095419406891,
+      "learning_rate": 0.00017628072488521155,
+      "loss": 0.3055413246154785,
+      "step": 9730
+    },
+    {
+      "epoch": 0.0877131129423241,
+      "grad_norm": 0.311156690120697,
+      "learning_rate": 0.0001762286120576152,
+      "loss": 0.2970812082290649,
+      "step": 9740
+    },
+    {
+      "epoch": 0.08800539255892631,
+      "grad_norm": 0.42607852816581726,
+      "learning_rate": 0.00017617644976610012,
+      "loss": 0.33449320793151854,
+      "step": 9750
+    },
+    {
+      "epoch": 0.08829767217552853,
+      "grad_norm": 0.40167832374572754,
+      "learning_rate": 0.00017612423804451378,
+      "loss": 0.39463398456573484,
+      "step": 9760
+    },
+    {
+      "epoch": 0.08858995179213074,
+      "grad_norm": 0.366714745759964,
+      "learning_rate": 0.00017607197692673588,
+      "loss": 0.309732985496521,
+      "step": 9770
+    },
+    {
+      "epoch": 0.08888223140873296,
+      "grad_norm": 0.4011765122413635,
+      "learning_rate": 0.00017601966644667803,
+      "loss": 0.3252063989639282,
+      "step": 9780
+    },
+    {
+      "epoch": 0.08917451102533516,
+      "grad_norm": 0.3573104739189148,
+      "learning_rate": 0.000175967306638284,
+      "loss": 0.3013187885284424,
+      "step": 9790
+    },
+    {
+      "epoch": 0.08946679064193737,
+      "grad_norm": 0.46793264150619507,
+      "learning_rate": 0.00017591489753552943,
+      "loss": 0.3776808500289917,
+      "step": 9800
+    },
+    {
+      "epoch": 0.08975907025853959,
+      "grad_norm": 0.3172089457511902,
+      "learning_rate": 0.00017586243917242205,
+      "loss": 0.30923995971679685,
+      "step": 9810
+    },
+    {
+      "epoch": 0.0900513498751418,
+      "grad_norm": 0.32853418588638306,
+      "learning_rate": 0.00017580993158300152,
+      "loss": 0.2775848865509033,
+      "step": 9820
+    },
+    {
+      "epoch": 0.09034362949174402,
+      "grad_norm": 0.4363964796066284,
+      "learning_rate": 0.00017575737480133942,
+      "loss": 0.28869173526763914,
+      "step": 9830
+    },
+    {
+      "epoch": 0.09063590910834622,
+      "grad_norm": 0.4401825964450836,
+      "learning_rate": 0.00017570476886153932,
+      "loss": 0.3337622880935669,
+      "step": 9840
+    },
+    {
+      "epoch": 0.09092818872494844,
+      "grad_norm": 0.2768244445323944,
+      "learning_rate": 0.00017565211379773663,
+      "loss": 0.271062970161438,
+      "step": 9850
+    },
+    {
+      "epoch": 0.09122046834155065,
+      "grad_norm": 0.2981739342212677,
+      "learning_rate": 0.0001755994096440986,
+      "loss": 0.28376638889312744,
+      "step": 9860
+    },
+    {
+      "epoch": 0.09151274795815287,
+      "grad_norm": 0.40843328833580017,
+      "learning_rate": 0.00017554665643482444,
+      "loss": 0.3219259262084961,
+      "step": 9870
+    },
+    {
+      "epoch": 0.09180502757475507,
+      "grad_norm": 0.3303707242012024,
+      "learning_rate": 0.00017549385420414514,
+      "loss": 0.3114008903503418,
+      "step": 9880
+    },
+    {
+      "epoch": 0.0920973071913573,
+      "grad_norm": 0.6027088165283203,
+      "learning_rate": 0.00017544100298632343,
+      "loss": 0.32715346813201907,
+      "step": 9890
+    },
+    {
+      "epoch": 0.0923895868079595,
+      "grad_norm": 0.405700147151947,
+      "learning_rate": 0.00017538810281565398,
+      "loss": 0.2934636831283569,
+      "step": 9900
+    },
+    {
+      "epoch": 0.09268186642456172,
+      "grad_norm": 0.34661969542503357,
+      "learning_rate": 0.00017533515372646308,
+      "loss": 0.27575581073760985,
+      "step": 9910
+    },
+    {
+      "epoch": 0.09297414604116393,
+      "grad_norm": 0.3284149169921875,
+      "learning_rate": 0.0001752821557531088,
+      "loss": 0.30129504203796387,
+      "step": 9920
+    },
+    {
+      "epoch": 0.09326642565776615,
+      "grad_norm": 0.4156268835067749,
+      "learning_rate": 0.000175229108929981,
+      "loss": 0.2920950651168823,
+      "step": 9930
+    },
+    {
+      "epoch": 0.09355870527436835,
+      "grad_norm": 0.3268771767616272,
+      "learning_rate": 0.00017517601329150115,
+      "loss": 0.2736488103866577,
+      "step": 9940
+    },
+    {
+      "epoch": 0.09385098489097057,
+      "grad_norm": 0.2907182276248932,
+      "learning_rate": 0.0001751228688721225,
+      "loss": 0.3027563333511353,
+      "step": 9950
+    },
+    {
+      "epoch": 0.09414326450757278,
+      "grad_norm": 0.3048081696033478,
+      "learning_rate": 0.00017506967570632978,
+      "loss": 0.30011472702026365,
+      "step": 9960
+    },
+    {
+      "epoch": 0.094435544124175,
+      "grad_norm": 0.5729956030845642,
+      "learning_rate": 0.00017501643382863952,
+      "loss": 0.31825685501098633,
+      "step": 9970
+    },
+    {
+      "epoch": 0.0947278237407772,
+      "grad_norm": 0.5286709070205688,
+      "learning_rate": 0.0001749631432735998,
+      "loss": 0.3022994756698608,
+      "step": 9980
+    },
+    {
+      "epoch": 0.09502010335737943,
+      "grad_norm": 0.3867145776748657,
+      "learning_rate": 0.0001749098040757903,
+      "loss": 0.36994185447692873,
+      "step": 9990
+    },
+    {
+      "epoch": 0.09531238297398163,
+      "grad_norm": 0.2982030212879181,
+      "learning_rate": 0.00017485641626982217,
+      "loss": 0.2684871435165405,
+      "step": 10000
+    },
+    {
+      "epoch": 0.09560466259058385,
+      "grad_norm": 0.4135000705718994,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 0.26283040046691897,
+      "step": 10010
+    },
+    {
+      "epoch": 0.09589694220718606,
+      "grad_norm": 0.38506409525871277,
+      "learning_rate": 0.00017474949497201282,
+      "loss": 0.272791314125061,
+      "step": 10020
+    },
+    {
+      "epoch": 0.09618922182378828,
+      "grad_norm": 0.3964892625808716,
+      "learning_rate": 0.00017469596154955164,
+      "loss": 0.3061663627624512,
+      "step": 10030
+    },
+    {
+      "epoch": 0.09648150144039049,
+      "grad_norm": 0.36178678274154663,
+      "learning_rate": 0.00017464237965769195,
+      "loss": 0.3407487154006958,
+      "step": 10040
+    },
+    {
+      "epoch": 0.0967737810569927,
+      "grad_norm": 0.2787010073661804,
+      "learning_rate": 0.0001745887493312025,
+      "loss": 0.29456703662872313,
+      "step": 10050
+    },
+    {
+      "epoch": 0.09706606067359491,
+      "grad_norm": 0.3525826930999756,
+      "learning_rate": 0.0001745350706048834,
+      "loss": 0.3412494659423828,
+      "step": 10060
+    },
+    {
+      "epoch": 0.09735834029019712,
+      "grad_norm": 0.35705459117889404,
+      "learning_rate": 0.0001744813435135662,
+      "loss": 0.2737375974655151,
+      "step": 10070
+    },
+    {
+      "epoch": 0.09765061990679934,
+      "grad_norm": 0.36130061745643616,
+      "learning_rate": 0.00017442756809211383,
+      "loss": 0.272174072265625,
+      "step": 10080
+    },
+    {
+      "epoch": 0.09794289952340154,
+      "grad_norm": 0.46362996101379395,
+      "learning_rate": 0.0001743737443754206,
+      "loss": 0.29140610694885255,
+      "step": 10090
+    },
+    {
+      "epoch": 0.09823517914000376,
+      "grad_norm": 0.37967824935913086,
+      "learning_rate": 0.00017431987239841206,
+      "loss": 0.3176786661148071,
+      "step": 10100
+    },
+    {
+      "epoch": 0.09852745875660597,
+      "grad_norm": 0.3551442325115204,
+      "learning_rate": 0.0001742659521960452,
+      "loss": 0.27897565364837645,
+      "step": 10110
+    },
+    {
+      "epoch": 0.09881973837320819,
+      "grad_norm": 0.4584147334098816,
+      "learning_rate": 0.00017421198380330829,
+      "loss": 0.2549800634384155,
+      "step": 10120
+    },
+    {
+      "epoch": 0.0991120179898104,
+      "grad_norm": 0.2789497375488281,
+      "learning_rate": 0.0001741579672552208,
+      "loss": 0.3275036334991455,
+      "step": 10130
+    },
+    {
+      "epoch": 0.09940429760641262,
+      "grad_norm": 0.3471240699291229,
+      "learning_rate": 0.00017410390258683345,
+      "loss": 0.3126558303833008,
+      "step": 10140
+    },
+    {
+      "epoch": 0.09969657722301482,
+      "grad_norm": 0.4327692687511444,
+      "learning_rate": 0.00017404978983322827,
+      "loss": 0.2968619346618652,
+      "step": 10150
+    },
+    {
+      "epoch": 0.09998885683961704,
+      "grad_norm": 0.44976386427879333,
+      "learning_rate": 0.0001739956290295184,
+      "loss": 0.2951622486114502,
+      "step": 10160
+    },
+    {
+      "epoch": 0.10028113645621925,
+      "grad_norm": 0.4886215925216675,
+      "learning_rate": 0.0001739414202108482,
+      "loss": 0.3082313537597656,
+      "step": 10170
+    },
+    {
+      "epoch": 0.10057341607282147,
+      "grad_norm": 0.39565685391426086,
+      "learning_rate": 0.00017388716341239322,
+      "loss": 0.27949323654174807,
+      "step": 10180
+    },
+    {
+      "epoch": 0.10086569568942368,
+      "grad_norm": 0.38107091188430786,
+      "learning_rate": 0.00017383285866936005,
+      "loss": 0.2645575046539307,
+      "step": 10190
+    },
+    {
+      "epoch": 0.1011579753060259,
+      "grad_norm": 0.35633140802383423,
+      "learning_rate": 0.00017377850601698653,
+      "loss": 0.3140716075897217,
+      "step": 10200
+    },
+    {
+      "epoch": 0.1014502549226281,
+      "grad_norm": 0.43105125427246094,
+      "learning_rate": 0.00017372410549054142,
+      "loss": 0.2868061065673828,
+      "step": 10210
+    },
+    {
+      "epoch": 0.10174253453923032,
+      "grad_norm": 0.402507483959198,
+      "learning_rate": 0.0001736696571253247,
+      "loss": 0.24577586650848388,
+      "step": 10220
+    },
+    {
+      "epoch": 0.10203481415583253,
+      "grad_norm": 0.3796424865722656,
+      "learning_rate": 0.00017361516095666725,
+      "loss": 0.34066386222839357,
+      "step": 10230
+    },
+    {
+      "epoch": 0.10232709377243475,
+      "grad_norm": 0.3590683937072754,
+      "learning_rate": 0.00017356061701993112,
+      "loss": 0.30033786296844484,
+      "step": 10240
+    },
+    {
+      "epoch": 0.10261937338903696,
+      "grad_norm": 0.36152827739715576,
+      "learning_rate": 0.0001735060253505092,
+      "loss": 0.31431779861450193,
+      "step": 10250
+    },
+    {
+      "epoch": 0.10291165300563918,
+      "grad_norm": 0.45164042711257935,
+      "learning_rate": 0.0001734513859838255,
+      "loss": 0.27492475509643555,
+      "step": 10260
+    },
+    {
+      "epoch": 0.10320393262224138,
+      "grad_norm": 0.47878938913345337,
+      "learning_rate": 0.0001733966989553349,
+      "loss": 0.3177743911743164,
+      "step": 10270
+    },
+    {
+      "epoch": 0.1034962122388436,
+      "grad_norm": 0.32319796085357666,
+      "learning_rate": 0.00017334196430052317,
+      "loss": 0.3067225694656372,
+      "step": 10280
+    },
+    {
+      "epoch": 0.10378849185544581,
+      "grad_norm": 0.35437121987342834,
+      "learning_rate": 0.00017328718205490712,
+      "loss": 0.29584062099456787,
+      "step": 10290
+    },
+    {
+      "epoch": 0.10408077147204803,
+      "grad_norm": 0.3970884680747986,
+      "learning_rate": 0.00017323235225403426,
+      "loss": 0.27340188026428225,
+      "step": 10300
+    },
+    {
+      "epoch": 0.10437305108865023,
+      "grad_norm": 0.3467753827571869,
+      "learning_rate": 0.0001731774749334831,
+      "loss": 0.2628843069076538,
+      "step": 10310
+    },
+    {
+      "epoch": 0.10466533070525244,
+      "grad_norm": 0.3711020052433014,
+      "learning_rate": 0.00017312255012886296,
+      "loss": 0.2869729518890381,
+      "step": 10320
+    },
+    {
+      "epoch": 0.10495761032185466,
+      "grad_norm": 0.37837013602256775,
+      "learning_rate": 0.00017306757787581388,
+      "loss": 0.29000446796417234,
+      "step": 10330
+    },
+    {
+      "epoch": 0.10524988993845687,
+      "grad_norm": 0.34540030360221863,
+      "learning_rate": 0.00017301255821000684,
+      "loss": 0.30840320587158204,
+      "step": 10340
+    },
+    {
+      "epoch": 0.10554216955505909,
+      "grad_norm": 0.4177977740764618,
+      "learning_rate": 0.00017295749116714345,
+      "loss": 0.3396437406539917,
+      "step": 10350
+    },
+    {
+      "epoch": 0.1058344491716613,
+      "grad_norm": 0.30129411816596985,
+      "learning_rate": 0.0001729023767829561,
+      "loss": 0.30897386074066163,
+      "step": 10360
+    },
+    {
+      "epoch": 0.10612672878826351,
+      "grad_norm": 0.42044156789779663,
+      "learning_rate": 0.00017284721509320795,
+      "loss": 0.23552024364471436,
+      "step": 10370
+    },
+    {
+      "epoch": 0.10641900840486572,
+      "grad_norm": 0.46460774540901184,
+      "learning_rate": 0.00017279200613369287,
+      "loss": 0.31631481647491455,
+      "step": 10380
+    },
+    {
+      "epoch": 0.10671128802146794,
+      "grad_norm": 0.3909938335418701,
+      "learning_rate": 0.00017273674994023523,
+      "loss": 0.29092400074005126,
+      "step": 10390
+    },
+    {
+      "epoch": 0.10700356763807015,
+      "grad_norm": 0.3790885806083679,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 0.30466766357421876,
+      "step": 10400
+    },
+    {
+      "epoch": 0.10729584725467237,
+      "grad_norm": 0.40464648604393005,
+      "learning_rate": 0.00017262609599494376,
+      "loss": 0.30627644062042236,
+      "step": 10410
+    },
+    {
+      "epoch": 0.10758812687127457,
+      "grad_norm": 0.6271727085113525,
+      "learning_rate": 0.00017257069831491204,
+      "loss": 0.26263678073883057,
+      "step": 10420
+    },
+    {
+      "epoch": 0.10788040648787679,
+      "grad_norm": 0.4266994595527649,
+      "learning_rate": 0.00017251525354454206,
+      "loss": 0.3327409505844116,
+      "step": 10430
+    },
+    {
+      "epoch": 0.108172686104479,
+      "grad_norm": 0.4452070891857147,
+      "learning_rate": 0.00017245976171981138,
+      "loss": 0.3669077634811401,
+      "step": 10440
+    },
+    {
+      "epoch": 0.10846496572108122,
+      "grad_norm": 0.3991874158382416,
+      "learning_rate": 0.00017240422287672798,
+      "loss": 0.3097517490386963,
+      "step": 10450
+    },
+    {
+      "epoch": 0.10875724533768343,
+      "grad_norm": 0.34101754426956177,
+      "learning_rate": 0.0001723486370513304,
+      "loss": 0.28102912902832033,
+      "step": 10460
+    },
+    {
+      "epoch": 0.10904952495428565,
+      "grad_norm": 0.39398688077926636,
+      "learning_rate": 0.00017229300427968775,
+      "loss": 0.2516593933105469,
+      "step": 10470
+    },
+    {
+      "epoch": 0.10934180457088785,
+      "grad_norm": 0.48973366618156433,
+      "learning_rate": 0.0001722373245978995,
+      "loss": 0.35596742630004885,
+      "step": 10480
+    },
+    {
+      "epoch": 0.10963408418749007,
+      "grad_norm": 0.4132535755634308,
+      "learning_rate": 0.00017218159804209557,
+      "loss": 0.33713736534118655,
+      "step": 10490
+    },
+    {
+      "epoch": 0.10992636380409228,
+      "grad_norm": 0.3700997829437256,
+      "learning_rate": 0.00017212582464843634,
+      "loss": 0.2989234447479248,
+      "step": 10500
+    },
+    {
+      "epoch": 0.1102186434206945,
+      "grad_norm": 0.36788684129714966,
+      "learning_rate": 0.00017207000445311257,
+      "loss": 0.31901462078094484,
+      "step": 10510
+    },
+    {
+      "epoch": 0.1105109230372967,
+      "grad_norm": 0.329478919506073,
+      "learning_rate": 0.00017201413749234537,
+      "loss": 0.3664107322692871,
+      "step": 10520
+    },
+    {
+      "epoch": 0.11080320265389892,
+      "grad_norm": 0.377465158700943,
+      "learning_rate": 0.00017195822380238615,
+      "loss": 0.28719477653503417,
+      "step": 10530
+    },
+    {
+      "epoch": 0.11109548227050113,
+      "grad_norm": 0.4122747480869293,
+      "learning_rate": 0.00017190226341951679,
+      "loss": 0.3151065826416016,
+      "step": 10540
+    },
+    {
+      "epoch": 0.11138776188710335,
+      "grad_norm": 0.33568012714385986,
+      "learning_rate": 0.00017184625638004932,
+      "loss": 0.291794490814209,
+      "step": 10550
+    },
+    {
+      "epoch": 0.11168004150370556,
+      "grad_norm": 0.3742920756340027,
+      "learning_rate": 0.0001717902027203261,
+      "loss": 0.2858127117156982,
+      "step": 10560
+    },
+    {
+      "epoch": 0.11197232112030778,
+      "grad_norm": 0.35819146037101746,
+      "learning_rate": 0.00017173410247671976,
+      "loss": 0.3074647903442383,
+      "step": 10570
+    },
+    {
+      "epoch": 0.11226460073690998,
+      "grad_norm": 0.38573023676872253,
+      "learning_rate": 0.0001716779556856331,
+      "loss": 0.2709053993225098,
+      "step": 10580
+    },
+    {
+      "epoch": 0.11255688035351219,
+      "grad_norm": 0.3329653739929199,
+      "learning_rate": 0.00017162176238349922,
+      "loss": 0.42522225379943845,
+      "step": 10590
+    },
+    {
+      "epoch": 0.11284915997011441,
+      "grad_norm": 0.3574890196323395,
+      "learning_rate": 0.0001715655226067813,
+      "loss": 0.36023573875427245,
+      "step": 10600
+    },
+    {
+      "epoch": 0.11314143958671662,
+      "grad_norm": 0.5043089389801025,
+      "learning_rate": 0.00017150923639197273,
+      "loss": 0.3962068557739258,
+      "step": 10610
+    },
+    {
+      "epoch": 0.11343371920331884,
+      "grad_norm": 0.3682282269001007,
+      "learning_rate": 0.00017145290377559706,
+      "loss": 0.34183344841003416,
+      "step": 10620
+    },
+    {
+      "epoch": 0.11372599881992104,
+      "grad_norm": 0.379218727350235,
+      "learning_rate": 0.00017139652479420786,
+      "loss": 0.3720597505569458,
+      "step": 10630
+    },
+    {
+      "epoch": 0.11401827843652326,
+      "grad_norm": 0.30579426884651184,
+      "learning_rate": 0.00017134009948438888,
+      "loss": 0.45345616340637207,
+      "step": 10640
+    },
+    {
+      "epoch": 0.11431055805312547,
+      "grad_norm": 0.30768728256225586,
+      "learning_rate": 0.00017128362788275384,
+      "loss": 0.3969339609146118,
+      "step": 10650
+    },
+    {
+      "epoch": 0.11460283766972769,
+      "grad_norm": 0.4029112458229065,
+      "learning_rate": 0.0001712271100259466,
+      "loss": 0.46826920509338377,
+      "step": 10660
+    },
+    {
+      "epoch": 0.1148951172863299,
+      "grad_norm": 0.4187719523906708,
+      "learning_rate": 0.00017117054595064098,
+      "loss": 0.3758350372314453,
+      "step": 10670
+    },
+    {
+      "epoch": 0.11518739690293212,
+      "grad_norm": 0.3466188311576843,
+      "learning_rate": 0.00017111393569354077,
+      "loss": 0.39464597702026366,
+      "step": 10680
+    },
+    {
+      "epoch": 0.11547967651953432,
+      "grad_norm": 0.3361622095108032,
+      "learning_rate": 0.00017105727929137975,
+      "loss": 0.38457086086273196,
+      "step": 10690
+    },
+    {
+      "epoch": 0.11577195613613654,
+      "grad_norm": 0.5668526887893677,
+      "learning_rate": 0.0001710005767809217,
+      "loss": 0.378650164604187,
+      "step": 10700
+    },
+    {
+      "epoch": 0.11606423575273875,
+      "grad_norm": 0.37848979234695435,
+      "learning_rate": 0.00017094382819896016,
+      "loss": 0.41998701095581054,
+      "step": 10710
+    },
+    {
+      "epoch": 0.11635651536934097,
+      "grad_norm": 0.4046863317489624,
+      "learning_rate": 0.00017088703358231878,
+      "loss": 0.40972108840942384,
+      "step": 10720
+    },
+    {
+      "epoch": 0.11664879498594317,
+      "grad_norm": 0.38377395272254944,
+      "learning_rate": 0.0001708301929678509,
+      "loss": 0.338122034072876,
+      "step": 10730
+    },
+    {
+      "epoch": 0.1169410746025454,
+      "grad_norm": 0.28298911452293396,
+      "learning_rate": 0.00017077330639243975,
+      "loss": 0.4060105323791504,
+      "step": 10740
+    },
+    {
+      "epoch": 0.1172333542191476,
+      "grad_norm": 0.31544896960258484,
+      "learning_rate": 0.0001707163738929985,
+      "loss": 0.3642102241516113,
+      "step": 10750
+    },
+    {
+      "epoch": 0.11752563383574982,
+      "grad_norm": 0.343009352684021,
+      "learning_rate": 0.00017065939550646998,
+      "loss": 0.38298349380493163,
+      "step": 10760
+    },
+    {
+      "epoch": 0.11781791345235203,
+      "grad_norm": 0.27471911907196045,
+      "learning_rate": 0.00017060237126982684,
+      "loss": 0.34847519397735593,
+      "step": 10770
+    },
+    {
+      "epoch": 0.11811019306895425,
+      "grad_norm": 0.4936792254447937,
+      "learning_rate": 0.00017054530122007154,
+      "loss": 0.4104027271270752,
+      "step": 10780
+    },
+    {
+      "epoch": 0.11840247268555645,
+      "grad_norm": 0.3444400727748871,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.3555137157440186,
+      "step": 10790
+    },
+    {
+      "epoch": 0.11869475230215867,
+      "grad_norm": 0.42564651370048523,
+      "learning_rate": 0.00017043102382938255,
+      "loss": 0.38498260974884035,
+      "step": 10800
+    },
+    {
+      "epoch": 0.11898703191876088,
+      "grad_norm": 0.32018008828163147,
+      "learning_rate": 0.00017037381656260223,
+      "loss": 0.4074843883514404,
+      "step": 10810
+    },
+    {
+      "epoch": 0.1192793115353631,
+      "grad_norm": 0.3715144395828247,
+      "learning_rate": 0.0001703165636310164,
+      "loss": 0.3738917112350464,
+      "step": 10820
+    },
+    {
+      "epoch": 0.1195715911519653,
+      "grad_norm": 0.43494096398353577,
+      "learning_rate": 0.00017025926507177583,
+      "loss": 0.42795720100402834,
+      "step": 10830
+    },
+    {
+      "epoch": 0.11986387076856751,
+      "grad_norm": 0.34510067105293274,
+      "learning_rate": 0.00017020192092206092,
+      "loss": 0.33625454902648927,
+      "step": 10840
+    },
+    {
+      "epoch": 0.12015615038516973,
+      "grad_norm": 0.3610187768936157,
+      "learning_rate": 0.00017014453121908175,
+      "loss": 0.3864196062088013,
+      "step": 10850
+    },
+    {
+      "epoch": 0.12044843000177194,
+      "grad_norm": 0.30097317695617676,
+      "learning_rate": 0.0001700870960000778,
+      "loss": 0.35089280605316164,
+      "step": 10860
+    },
+    {
+      "epoch": 0.12074070961837416,
+      "grad_norm": 0.34319519996643066,
+      "learning_rate": 0.00017002961530231816,
+      "loss": 0.3915343999862671,
+      "step": 10870
+    },
+    {
+      "epoch": 0.12103298923497637,
+      "grad_norm": 0.3472767174243927,
+      "learning_rate": 0.0001699720891631015,
+      "loss": 0.3753772497177124,
+      "step": 10880
+    },
+    {
+      "epoch": 0.12132526885157859,
+      "grad_norm": 0.431599497795105,
+      "learning_rate": 0.0001699145176197558,
+      "loss": 0.3457676410675049,
+      "step": 10890
+    },
+    {
+      "epoch": 0.12161754846818079,
+      "grad_norm": 0.2494288831949234,
+      "learning_rate": 0.0001698569007096387,
+      "loss": 0.32474460601806643,
+      "step": 10900
+    },
+    {
+      "epoch": 0.12190982808478301,
+      "grad_norm": 0.3083363175392151,
+      "learning_rate": 0.00016979923847013716,
+      "loss": 0.39173712730407717,
+      "step": 10910
+    },
+    {
+      "epoch": 0.12220210770138522,
+      "grad_norm": 0.35492876172065735,
+      "learning_rate": 0.00016974153093866757,
+      "loss": 0.4295250415802002,
+      "step": 10920
+    },
+    {
+      "epoch": 0.12249438731798744,
+      "grad_norm": 0.30018359422683716,
+      "learning_rate": 0.0001696837781526757,
+      "loss": 0.3731697082519531,
+      "step": 10930
+    },
+    {
+      "epoch": 0.12278666693458964,
+      "grad_norm": 0.3802156448364258,
+      "learning_rate": 0.00016962598014963676,
+      "loss": 0.3451104164123535,
+      "step": 10940
+    },
+    {
+      "epoch": 0.12307894655119186,
+      "grad_norm": 0.32948487997055054,
+      "learning_rate": 0.00016956813696705522,
+      "loss": 0.34248971939086914,
+      "step": 10950
+    },
+    {
+      "epoch": 0.12337122616779407,
+      "grad_norm": 0.39775627851486206,
+      "learning_rate": 0.0001695102486424649,
+      "loss": 0.4102600574493408,
+      "step": 10960
+    },
+    {
+      "epoch": 0.12366350578439629,
+      "grad_norm": 0.3884665071964264,
+      "learning_rate": 0.00016945231521342885,
+      "loss": 0.45301251411437987,
+      "step": 10970
+    },
+    {
+      "epoch": 0.1239557854009985,
+      "grad_norm": 0.3282186686992645,
+      "learning_rate": 0.00016939433671753955,
+      "loss": 0.42616887092590333,
+      "step": 10980
+    },
+    {
+      "epoch": 0.12424806501760072,
+      "grad_norm": 0.48224589228630066,
+      "learning_rate": 0.00016933631319241853,
+      "loss": 0.46968564987182615,
+      "step": 10990
+    },
+    {
+      "epoch": 0.12454034463420292,
+      "grad_norm": 0.34733015298843384,
+      "learning_rate": 0.00016927824467571667,
+      "loss": 0.4007678985595703,
+      "step": 11000
+    },
+    {
+      "epoch": 0.12483262425080514,
+      "grad_norm": 0.23755525052547455,
+      "learning_rate": 0.00016922013120511395,
+      "loss": 0.34498813152313235,
+      "step": 11010
+    },
+    {
+      "epoch": 0.12512490386740735,
+      "grad_norm": 0.41167688369750977,
+      "learning_rate": 0.00016916197281831964,
+      "loss": 0.40474462509155273,
+      "step": 11020
+    },
+    {
+      "epoch": 0.12541718348400957,
+      "grad_norm": 0.32988983392715454,
+      "learning_rate": 0.00016910376955307202,
+      "loss": 0.4097125053405762,
+      "step": 11030
+    },
+    {
+      "epoch": 0.1257094631006118,
+      "grad_norm": 0.3868061304092407,
+      "learning_rate": 0.0001690455214471386,
+      "loss": 0.45975518226623535,
+      "step": 11040
+    },
+    {
+      "epoch": 0.12600174271721398,
+      "grad_norm": 0.3485318422317505,
+      "learning_rate": 0.00016898722853831593,
+      "loss": 0.3942741632461548,
+      "step": 11050
+    },
+    {
+      "epoch": 0.1262940223338162,
+      "grad_norm": 0.3959900736808777,
+      "learning_rate": 0.00016892889086442964,
+      "loss": 0.37831666469573977,
+      "step": 11060
+    },
+    {
+      "epoch": 0.12658630195041842,
+      "grad_norm": 0.38046321272850037,
+      "learning_rate": 0.0001688705084633344,
+      "loss": 0.3793492317199707,
+      "step": 11070
+    },
+    {
+      "epoch": 0.12687858156702064,
+      "grad_norm": 0.30742865800857544,
+      "learning_rate": 0.00016881208137291395,
+      "loss": 0.44310817718505857,
+      "step": 11080
+    },
+    {
+      "epoch": 0.12717086118362284,
+      "grad_norm": 0.34402474761009216,
+      "learning_rate": 0.00016875360963108102,
+      "loss": 0.32990052700042727,
+      "step": 11090
+    },
+    {
+      "epoch": 0.12746314080022506,
+      "grad_norm": 0.285019189119339,
+      "learning_rate": 0.0001686950932757772,
+      "loss": 0.3506010055541992,
+      "step": 11100
+    },
+    {
+      "epoch": 0.12775542041682728,
+      "grad_norm": 0.4057694971561432,
+      "learning_rate": 0.00016863653234497317,
+      "loss": 0.4069084644317627,
+      "step": 11110
+    },
+    {
+      "epoch": 0.12804770003342947,
+      "grad_norm": 0.5622038841247559,
+      "learning_rate": 0.00016857792687666846,
+      "loss": 0.3701250314712524,
+      "step": 11120
+    },
+    {
+      "epoch": 0.1283399796500317,
+      "grad_norm": 0.3658289313316345,
+      "learning_rate": 0.00016851927690889157,
+      "loss": 0.3875246524810791,
+      "step": 11130
+    },
+    {
+      "epoch": 0.1286322592666339,
+      "grad_norm": 0.3033835291862488,
+      "learning_rate": 0.0001684605824796998,
+      "loss": 0.3205209732055664,
+      "step": 11140
+    },
+    {
+      "epoch": 0.12892453888323613,
+      "grad_norm": 0.4940462112426758,
+      "learning_rate": 0.00016840184362717933,
+      "loss": 0.4005399703979492,
+      "step": 11150
+    },
+    {
+      "epoch": 0.12921681849983832,
+      "grad_norm": 0.3256269097328186,
+      "learning_rate": 0.00016834306038944515,
+      "loss": 0.37451536655426027,
+      "step": 11160
+    },
+    {
+      "epoch": 0.12950909811644054,
+      "grad_norm": 0.3144507110118866,
+      "learning_rate": 0.00016828423280464106,
+      "loss": 0.4359837532043457,
+      "step": 11170
+    },
+    {
+      "epoch": 0.12980137773304276,
+      "grad_norm": 0.21178320050239563,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 0.4215986728668213,
+      "step": 11180
+    },
+    {
+      "epoch": 0.13009365734964498,
+      "grad_norm": 0.43970873951911926,
+      "learning_rate": 0.00016816644474654225,
+      "loss": 0.359506893157959,
+      "step": 11190
+    },
+    {
+      "epoch": 0.13038593696624717,
+      "grad_norm": 0.545147716999054,
+      "learning_rate": 0.00016810748434967898,
+      "loss": 0.3731210947036743,
+      "step": 11200
+    },
+    {
+      "epoch": 0.1306782165828494,
+      "grad_norm": 0.26596322655677795,
+      "learning_rate": 0.0001680484797586085,
+      "loss": 0.35032851696014405,
+      "step": 11210
+    },
+    {
+      "epoch": 0.1309704961994516,
+      "grad_norm": 0.28256499767303467,
+      "learning_rate": 0.0001679894310116183,
+      "loss": 0.40617709159851073,
+      "step": 11220
+    },
+    {
+      "epoch": 0.13126277581605383,
+      "grad_norm": 0.5122326016426086,
+      "learning_rate": 0.0001679303381470245,
+      "loss": 0.38198351860046387,
+      "step": 11230
+    },
+    {
+      "epoch": 0.13155505543265603,
+      "grad_norm": 0.5235885977745056,
+      "learning_rate": 0.00016787120120317174,
+      "loss": 0.3931105613708496,
+      "step": 11240
+    },
+    {
+      "epoch": 0.13184733504925825,
+      "grad_norm": 0.3486837148666382,
+      "learning_rate": 0.0001678120202184334,
+      "loss": 0.38572418689727783,
+      "step": 11250
+    },
+    {
+      "epoch": 0.13213961466586047,
+      "grad_norm": 0.2500704228878021,
+      "learning_rate": 0.00016775279523121138,
+      "loss": 0.3551050901412964,
+      "step": 11260
+    },
+    {
+      "epoch": 0.1324318942824627,
+      "grad_norm": 0.32847559452056885,
+      "learning_rate": 0.00016769352627993614,
+      "loss": 0.42284350395202636,
+      "step": 11270
+    },
+    {
+      "epoch": 0.13272417389906488,
+      "grad_norm": 0.3953547775745392,
+      "learning_rate": 0.00016763421340306657,
+      "loss": 0.44657316207885744,
+      "step": 11280
+    },
+    {
+      "epoch": 0.1330164535156671,
+      "grad_norm": 0.4074779748916626,
+      "learning_rate": 0.00016757485663909026,
+      "loss": 0.36473681926727297,
+      "step": 11290
+    },
+    {
+      "epoch": 0.13330873313226932,
+      "grad_norm": 0.3378523290157318,
+      "learning_rate": 0.00016751545602652313,
+      "loss": 0.3708677291870117,
+      "step": 11300
+    },
+    {
+      "epoch": 0.13360101274887154,
+      "grad_norm": 0.3703884184360504,
+      "learning_rate": 0.00016745601160390958,
+      "loss": 0.388303804397583,
+      "step": 11310
+    },
+    {
+      "epoch": 0.13389329236547373,
+      "grad_norm": 0.33747240900993347,
+      "learning_rate": 0.00016739652340982245,
+      "loss": 0.3949401617050171,
+      "step": 11320
+    },
+    {
+      "epoch": 0.13418557198207595,
+      "grad_norm": 0.30841803550720215,
+      "learning_rate": 0.00016733699148286299,
+      "loss": 0.38821983337402344,
+      "step": 11330
+    },
+    {
+      "epoch": 0.13447785159867817,
+      "grad_norm": 0.28988978266716003,
+      "learning_rate": 0.00016727741586166083,
+      "loss": 0.3583950996398926,
+      "step": 11340
+    },
+    {
+      "epoch": 0.1347701312152804,
+      "grad_norm": 0.2835023105144501,
+      "learning_rate": 0.00016721779658487393,
+      "loss": 0.3495969533920288,
+      "step": 11350
+    },
+    {
+      "epoch": 0.13506241083188258,
+      "grad_norm": 0.329822838306427,
+      "learning_rate": 0.00016715813369118861,
+      "loss": 0.37540647983551023,
+      "step": 11360
+    },
+    {
+      "epoch": 0.1353546904484848,
+      "grad_norm": 0.35542595386505127,
+      "learning_rate": 0.00016709842721931948,
+      "loss": 0.40444483757019045,
+      "step": 11370
+    },
+    {
+      "epoch": 0.13564697006508702,
+      "grad_norm": 0.3620280623435974,
+      "learning_rate": 0.0001670386772080094,
+      "loss": 0.43405852317810056,
+      "step": 11380
+    },
+    {
+      "epoch": 0.13593924968168922,
+      "grad_norm": 0.3063121736049652,
+      "learning_rate": 0.00016697888369602946,
+      "loss": 0.3646464586257935,
+      "step": 11390
+    },
+    {
+      "epoch": 0.13623152929829144,
+      "grad_norm": 0.3327750265598297,
+      "learning_rate": 0.0001669190467221791,
+      "loss": 0.37135269641876223,
+      "step": 11400
+    },
+    {
+      "epoch": 0.13652380891489366,
+      "grad_norm": 0.2765229642391205,
+      "learning_rate": 0.00016685916632528587,
+      "loss": 0.3316920042037964,
+      "step": 11410
+    },
+    {
+      "epoch": 0.13681608853149588,
+      "grad_norm": 0.3348167836666107,
+      "learning_rate": 0.0001667992425442055,
+      "loss": 0.3567049980163574,
+      "step": 11420
+    },
+    {
+      "epoch": 0.13710836814809807,
+      "grad_norm": 0.27585992217063904,
+      "learning_rate": 0.00016673927541782183,
+      "loss": 0.4191673755645752,
+      "step": 11430
+    },
+    {
+      "epoch": 0.1374006477647003,
+      "grad_norm": 0.2647240161895752,
+      "learning_rate": 0.00016667926498504696,
+      "loss": 0.35025711059570314,
+      "step": 11440
+    },
+    {
+      "epoch": 0.1376929273813025,
+      "grad_norm": 0.3229633867740631,
+      "learning_rate": 0.00016661921128482092,
+      "loss": 0.407482385635376,
+      "step": 11450
+    },
+    {
+      "epoch": 0.13798520699790473,
+      "grad_norm": 0.31561025977134705,
+      "learning_rate": 0.000166559114356112,
+      "loss": 0.37025392055511475,
+      "step": 11460
+    },
+    {
+      "epoch": 0.13827748661450692,
+      "grad_norm": 0.26758939027786255,
+      "learning_rate": 0.0001664989742379164,
+      "loss": 0.4734449863433838,
+      "step": 11470
+    },
+    {
+      "epoch": 0.13856976623110914,
+      "grad_norm": 0.32062336802482605,
+      "learning_rate": 0.0001664387909692583,
+      "loss": 0.42926297187805174,
+      "step": 11480
+    },
+    {
+      "epoch": 0.13886204584771136,
+      "grad_norm": 0.33874809741973877,
+      "learning_rate": 0.00016637856458919014,
+      "loss": 0.36198475360870364,
+      "step": 11490
+    },
+    {
+      "epoch": 0.13915432546431358,
+      "grad_norm": 0.4009056091308594,
+      "learning_rate": 0.00016631829513679206,
+      "loss": 0.3502147912979126,
+      "step": 11500
+    },
+    {
+      "epoch": 0.13944660508091578,
+      "grad_norm": 0.32171911001205444,
+      "learning_rate": 0.00016625798265117227,
+      "loss": 0.4310002326965332,
+      "step": 11510
+    },
+    {
+      "epoch": 0.139738884697518,
+      "grad_norm": 0.28301897644996643,
+      "learning_rate": 0.00016619762717146687,
+      "loss": 0.37462124824523924,
+      "step": 11520
+    },
+    {
+      "epoch": 0.14003116431412022,
+      "grad_norm": 0.3605278432369232,
+      "learning_rate": 0.00016613722873683995,
+      "loss": 0.48468642234802245,
+      "step": 11530
+    },
+    {
+      "epoch": 0.14032344393072244,
+      "grad_norm": 0.4420568645000458,
+      "learning_rate": 0.0001660767873864833,
+      "loss": 0.4313344478607178,
+      "step": 11540
+    },
+    {
+      "epoch": 0.14061572354732463,
+      "grad_norm": 0.30689021944999695,
+      "learning_rate": 0.0001660163031596167,
+      "loss": 0.3933013200759888,
+      "step": 11550
+    },
+    {
+      "epoch": 0.14090800316392685,
+      "grad_norm": 0.2765858471393585,
+      "learning_rate": 0.00016595577609548773,
+      "loss": 0.3849361419677734,
+      "step": 11560
+    },
+    {
+      "epoch": 0.14120028278052907,
+      "grad_norm": 0.3786977529525757,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 0.3608834981918335,
+      "step": 11570
+    },
+    {
+      "epoch": 0.1414925623971313,
+      "grad_norm": 0.31414663791656494,
+      "learning_rate": 0.00016583459361257176,
+      "loss": 0.399343204498291,
+      "step": 11580
+    },
+    {
+      "epoch": 0.14178484201373348,
+      "grad_norm": 0.28737974166870117,
+      "learning_rate": 0.0001657739382724188,
+      "loss": 0.37542450428009033,
+      "step": 11590
+    },
+    {
+      "epoch": 0.1420771216303357,
+      "grad_norm": 0.3305077850818634,
+      "learning_rate": 0.00016571324025227137,
+      "loss": 0.41343045234680176,
+      "step": 11600
+    },
+    {
+      "epoch": 0.14236940124693792,
+      "grad_norm": 0.319173663854599,
+      "learning_rate": 0.0001656524995915158,
+      "loss": 0.3659952163696289,
+      "step": 11610
+    },
+    {
+      "epoch": 0.1426616808635401,
+      "grad_norm": 0.39958882331848145,
+      "learning_rate": 0.00016559171632956604,
+      "loss": 0.4655965805053711,
+      "step": 11620
+    },
+    {
+      "epoch": 0.14295396048014233,
+      "grad_norm": 0.27597030997276306,
+      "learning_rate": 0.00016553089050586362,
+      "loss": 0.3778929471969604,
+      "step": 11630
+    },
+    {
+      "epoch": 0.14324624009674455,
+      "grad_norm": 0.3151189684867859,
+      "learning_rate": 0.00016547002215987785,
+      "loss": 0.3648439645767212,
+      "step": 11640
+    },
+    {
+      "epoch": 0.14353851971334677,
+      "grad_norm": 0.3472134470939636,
+      "learning_rate": 0.0001654091113311055,
+      "loss": 0.4437553405761719,
+      "step": 11650
+    },
+    {
+      "epoch": 0.14383079932994897,
+      "grad_norm": 0.3491552770137787,
+      "learning_rate": 0.0001653481580590709,
+      "loss": 0.4239807605743408,
+      "step": 11660
+    },
+    {
+      "epoch": 0.1441230789465512,
+      "grad_norm": 0.39337584376335144,
+      "learning_rate": 0.00016528716238332607,
+      "loss": 0.38459646701812744,
+      "step": 11670
+    },
+    {
+      "epoch": 0.1444153585631534,
+      "grad_norm": 0.27783700823783875,
+      "learning_rate": 0.00016522612434345034,
+      "loss": 0.40322322845458985,
+      "step": 11680
+    },
+    {
+      "epoch": 0.14470763817975563,
+      "grad_norm": 0.35323822498321533,
+      "learning_rate": 0.0001651650439790507,
+      "loss": 0.4304669857025146,
+      "step": 11690
+    },
+    {
+      "epoch": 0.14499991779635782,
+      "grad_norm": 0.3678315579891205,
+      "learning_rate": 0.0001651039213297615,
+      "loss": 0.3798572301864624,
+      "step": 11700
+    },
+    {
+      "epoch": 0.14529219741296004,
+      "grad_norm": 0.3504646122455597,
+      "learning_rate": 0.00016504275643524453,
+      "loss": 0.42955989837646485,
+      "step": 11710
+    },
+    {
+      "epoch": 0.14558447702956226,
+      "grad_norm": 0.3880676031112671,
+      "learning_rate": 0.00016498154933518912,
+      "loss": 0.3981906890869141,
+      "step": 11720
+    },
+    {
+      "epoch": 0.14587675664616448,
+      "grad_norm": 0.44784629344940186,
+      "learning_rate": 0.00016492030006931183,
+      "loss": 0.40975141525268555,
+      "step": 11730
+    },
+    {
+      "epoch": 0.14616903626276667,
+      "grad_norm": 0.3599526584148407,
+      "learning_rate": 0.00016485900867735666,
+      "loss": 0.37773587703704836,
+      "step": 11740
+    },
+    {
+      "epoch": 0.1464613158793689,
+      "grad_norm": 0.42367392778396606,
+      "learning_rate": 0.0001647976751990949,
+      "loss": 0.4041136264801025,
+      "step": 11750
+    },
+    {
+      "epoch": 0.1467535954959711,
+      "grad_norm": 0.3082481026649475,
+      "learning_rate": 0.00016473629967432522,
+      "loss": 0.39384486675262453,
+      "step": 11760
+    },
+    {
+      "epoch": 0.14704587511257333,
+      "grad_norm": 0.3220765292644501,
+      "learning_rate": 0.00016467488214287352,
+      "loss": 0.4166693687438965,
+      "step": 11770
+    },
+    {
+      "epoch": 0.14733815472917552,
+      "grad_norm": 0.25579124689102173,
+      "learning_rate": 0.00016461342264459297,
+      "loss": 0.4032891273498535,
+      "step": 11780
+    },
+    {
+      "epoch": 0.14763043434577774,
+      "grad_norm": 0.26783543825149536,
+      "learning_rate": 0.00016455192121936393,
+      "loss": 0.4460781097412109,
+      "step": 11790
+    },
+    {
+      "epoch": 0.14792271396237996,
+      "grad_norm": 0.37600019574165344,
+      "learning_rate": 0.00016449037790709404,
+      "loss": 0.39374067783355715,
+      "step": 11800
+    },
+    {
+      "epoch": 0.14821499357898218,
+      "grad_norm": 0.308377206325531,
+      "learning_rate": 0.00016442879274771813,
+      "loss": 0.3955789566040039,
+      "step": 11810
+    },
+    {
+      "epoch": 0.14850727319558438,
+      "grad_norm": 0.40771111845970154,
+      "learning_rate": 0.00016436716578119808,
+      "loss": 0.30364625453948973,
+      "step": 11820
+    },
+    {
+      "epoch": 0.1487995528121866,
+      "grad_norm": 0.3464655876159668,
+      "learning_rate": 0.00016430549704752294,
+      "loss": 0.3363095998764038,
+      "step": 11830
+    },
+    {
+      "epoch": 0.14909183242878882,
+      "grad_norm": 0.346879243850708,
+      "learning_rate": 0.00016424378658670897,
+      "loss": 0.3763661623001099,
+      "step": 11840
+    },
+    {
+      "epoch": 0.14938411204539104,
+      "grad_norm": 0.320578008890152,
+      "learning_rate": 0.00016418203443879933,
+      "loss": 0.36609616279602053,
+      "step": 11850
+    },
+    {
+      "epoch": 0.14967639166199323,
+      "grad_norm": 0.5568594932556152,
+      "learning_rate": 0.0001641202406438644,
+      "loss": 0.36088991165161133,
+      "step": 11860
+    },
+    {
+      "epoch": 0.14996867127859545,
+      "grad_norm": 0.34007784724235535,
+      "learning_rate": 0.00016405840524200143,
+      "loss": 0.42797341346740725,
+      "step": 11870
+    },
+    {
+      "epoch": 0.15026095089519767,
+      "grad_norm": 0.44615036249160767,
+      "learning_rate": 0.0001639965282733348,
+      "loss": 0.4170976638793945,
+      "step": 11880
+    },
+    {
+      "epoch": 0.15055323051179986,
+      "grad_norm": 0.286732941865921,
+      "learning_rate": 0.00016393460977801573,
+      "loss": 0.36138484477996824,
+      "step": 11890
+    },
+    {
+      "epoch": 0.15084551012840208,
+      "grad_norm": 0.3119763433933258,
+      "learning_rate": 0.00016387264979622255,
+      "loss": 0.36156315803527833,
+      "step": 11900
+    },
+    {
+      "epoch": 0.1511377897450043,
+      "grad_norm": 0.4090895354747772,
+      "learning_rate": 0.00016381064836816034,
+      "loss": 0.34096837043762207,
+      "step": 11910
+    },
+    {
+      "epoch": 0.15143006936160652,
+      "grad_norm": 0.34286850690841675,
+      "learning_rate": 0.0001637486055340612,
+      "loss": 0.345221209526062,
+      "step": 11920
+    },
+    {
+      "epoch": 0.15172234897820872,
+      "grad_norm": 0.28404855728149414,
+      "learning_rate": 0.00016368652133418404,
+      "loss": 0.3194119453430176,
+      "step": 11930
+    },
+    {
+      "epoch": 0.15201462859481094,
+      "grad_norm": 0.5382237434387207,
+      "learning_rate": 0.00016362439580881462,
+      "loss": 0.3892913579940796,
+      "step": 11940
+    },
+    {
+      "epoch": 0.15230690821141316,
+      "grad_norm": 0.5555295348167419,
+      "learning_rate": 0.0001635622289982655,
+      "loss": 0.36381146907806394,
+      "step": 11950
+    },
+    {
+      "epoch": 0.15259918782801538,
+      "grad_norm": 0.3906997740268707,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 0.3740372896194458,
+      "step": 11960
+    },
+    {
+      "epoch": 0.15289146744461757,
+      "grad_norm": 0.34010258316993713,
+      "learning_rate": 0.0001634377716830125,
+      "loss": 0.37568440437316897,
+      "step": 11970
+    },
+    {
+      "epoch": 0.1531837470612198,
+      "grad_norm": 0.3108118772506714,
+      "learning_rate": 0.0001633754812590676,
+      "loss": 0.33114032745361327,
+      "step": 11980
+    },
+    {
+      "epoch": 0.153476026677822,
+      "grad_norm": 0.33587223291397095,
+      "learning_rate": 0.00016331314971146092,
+      "loss": 0.419649600982666,
+      "step": 11990
+    },
+    {
+      "epoch": 0.15376830629442423,
+      "grad_norm": 0.3141031265258789,
+      "learning_rate": 0.0001632507770806388,
+      "loss": 0.3561969757080078,
+      "step": 12000
+    },
+    {
+      "epoch": 0.15406058591102642,
+      "grad_norm": 0.3614560663700104,
+      "learning_rate": 0.00016318836340707416,
+      "loss": 0.40635037422180176,
+      "step": 12010
+    },
+    {
+      "epoch": 0.15435286552762864,
+      "grad_norm": 0.3167742192745209,
+      "learning_rate": 0.00016312590873126655,
+      "loss": 0.41324801445007325,
+      "step": 12020
+    },
+    {
+      "epoch": 0.15464514514423086,
+      "grad_norm": 0.42213308811187744,
+      "learning_rate": 0.00016306341309374207,
+      "loss": 0.3753725528717041,
+      "step": 12030
+    },
+    {
+      "epoch": 0.15493742476083308,
+      "grad_norm": 0.2645309865474701,
+      "learning_rate": 0.0001630008765350535,
+      "loss": 0.38724603652954104,
+      "step": 12040
+    },
+    {
+      "epoch": 0.15522970437743527,
+      "grad_norm": 0.32701122760772705,
+      "learning_rate": 0.00016293829909578016,
+      "loss": 0.31719484329223635,
+      "step": 12050
+    },
+    {
+      "epoch": 0.1555219839940375,
+      "grad_norm": 0.31050342321395874,
+      "learning_rate": 0.00016287568081652784,
+      "loss": 0.39080009460449217,
+      "step": 12060
+    },
+    {
+      "epoch": 0.1558142636106397,
+      "grad_norm": 0.37763655185699463,
+      "learning_rate": 0.00016281302173792888,
+      "loss": 0.3775671720504761,
+      "step": 12070
+    },
+    {
+      "epoch": 0.15610654322724193,
+      "grad_norm": 0.3040887713432312,
+      "learning_rate": 0.00016275032190064205,
+      "loss": 0.37606053352355956,
+      "step": 12080
+    },
+    {
+      "epoch": 0.15639882284384413,
+      "grad_norm": 0.538139283657074,
+      "learning_rate": 0.0001626875813453526,
+      "loss": 0.37129499912261965,
+      "step": 12090
+    },
+    {
+      "epoch": 0.15669110246044635,
+      "grad_norm": 0.2923707962036133,
+      "learning_rate": 0.00016262480011277223,
+      "loss": 0.40114736557006836,
+      "step": 12100
+    },
+    {
+      "epoch": 0.15698338207704857,
+      "grad_norm": 0.33134588599205017,
+      "learning_rate": 0.00016256197824363896,
+      "loss": 0.3884502172470093,
+      "step": 12110
+    },
+    {
+      "epoch": 0.1572756616936508,
+      "grad_norm": 0.28084105253219604,
+      "learning_rate": 0.00016249911577871724,
+      "loss": 0.38241937160491946,
+      "step": 12120
+    },
+    {
+      "epoch": 0.15756794131025298,
+      "grad_norm": 0.6677065491676331,
+      "learning_rate": 0.00016243621275879784,
+      "loss": 0.4396663188934326,
+      "step": 12130
+    },
+    {
+      "epoch": 0.1578602209268552,
+      "grad_norm": 0.29335370659828186,
+      "learning_rate": 0.00016237326922469785,
+      "loss": 0.320920991897583,
+      "step": 12140
+    },
+    {
+      "epoch": 0.15815250054345742,
+      "grad_norm": 0.3213663101196289,
+      "learning_rate": 0.0001623102852172606,
+      "loss": 0.39825031757354734,
+      "step": 12150
+    },
+    {
+      "epoch": 0.1584447801600596,
+      "grad_norm": 0.2922569215297699,
+      "learning_rate": 0.00016224726077735578,
+      "loss": 0.3631072759628296,
+      "step": 12160
+    },
+    {
+      "epoch": 0.15873705977666183,
+      "grad_norm": 0.2652628421783447,
+      "learning_rate": 0.00016218419594587926,
+      "loss": 0.37177979946136475,
+      "step": 12170
+    },
+    {
+      "epoch": 0.15902933939326405,
+      "grad_norm": 0.5099869966506958,
+      "learning_rate": 0.0001621210907637531,
+      "loss": 0.3821613550186157,
+      "step": 12180
+    },
+    {
+      "epoch": 0.15932161900986627,
+      "grad_norm": 0.45877698063850403,
+      "learning_rate": 0.0001620579452719256,
+      "loss": 0.4426738262176514,
+      "step": 12190
+    },
+    {
+      "epoch": 0.15961389862646846,
+      "grad_norm": 0.35152602195739746,
+      "learning_rate": 0.00016199475951137116,
+      "loss": 0.4219931125640869,
+      "step": 12200
+    },
+    {
+      "epoch": 0.15990617824307068,
+      "grad_norm": 0.4421404004096985,
+      "learning_rate": 0.00016193153352309035,
+      "loss": 0.3285573959350586,
+      "step": 12210
+    },
+    {
+      "epoch": 0.1601984578596729,
+      "grad_norm": 0.3404903709888458,
+      "learning_rate": 0.00016186826734810979,
+      "loss": 0.3639610052108765,
+      "step": 12220
+    },
+    {
+      "epoch": 0.16049073747627512,
+      "grad_norm": 0.3349684774875641,
+      "learning_rate": 0.00016180496102748227,
+      "loss": 0.34041543006896974,
+      "step": 12230
+    },
+    {
+      "epoch": 0.16078301709287732,
+      "grad_norm": 0.2713918089866638,
+      "learning_rate": 0.00016174161460228654,
+      "loss": 0.42139177322387694,
+      "step": 12240
+    },
+    {
+      "epoch": 0.16107529670947954,
+      "grad_norm": 0.2780614495277405,
+      "learning_rate": 0.0001616782281136274,
+      "loss": 0.35974195003509524,
+      "step": 12250
+    },
+    {
+      "epoch": 0.16136757632608176,
+      "grad_norm": 0.3468208909034729,
+      "learning_rate": 0.00016161480160263567,
+      "loss": 0.4229461669921875,
+      "step": 12260
+    },
+    {
+      "epoch": 0.16165985594268398,
+      "grad_norm": 0.5275227427482605,
+      "learning_rate": 0.00016155133511046814,
+      "loss": 0.4648599624633789,
+      "step": 12270
+    },
+    {
+      "epoch": 0.16195213555928617,
+      "grad_norm": 0.24619515240192413,
+      "learning_rate": 0.0001614878286783075,
+      "loss": 0.38206264972686765,
+      "step": 12280
+    },
+    {
+      "epoch": 0.1622444151758884,
+      "grad_norm": 0.29844963550567627,
+      "learning_rate": 0.0001614242823473624,
+      "loss": 0.38174591064453123,
+      "step": 12290
+    },
+    {
+      "epoch": 0.1625366947924906,
+      "grad_norm": 0.3330749571323395,
+      "learning_rate": 0.00016136069615886736,
+      "loss": 0.3846591472625732,
+      "step": 12300
+    },
+    {
+      "epoch": 0.16282897440909283,
+      "grad_norm": 0.4110930860042572,
+      "learning_rate": 0.00016129707015408276,
+      "loss": 0.40579662322998045,
+      "step": 12310
+    },
+    {
+      "epoch": 0.16312125402569502,
+      "grad_norm": 0.3412657082080841,
+      "learning_rate": 0.0001612334043742948,
+      "loss": 0.3989887237548828,
+      "step": 12320
+    },
+    {
+      "epoch": 0.16341353364229724,
+      "grad_norm": 0.29700231552124023,
+      "learning_rate": 0.0001611696988608156,
+      "loss": 0.40211710929870603,
+      "step": 12330
+    },
+    {
+      "epoch": 0.16370581325889946,
+      "grad_norm": 0.2872560918331146,
+      "learning_rate": 0.00016110595365498287,
+      "loss": 0.37642407417297363,
+      "step": 12340
+    },
+    {
+      "epoch": 0.16399809287550168,
+      "grad_norm": 0.27716127038002014,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 0.36686077117919924,
+      "step": 12350
+    },
+    {
+      "epoch": 0.16429037249210388,
+      "grad_norm": 0.26968681812286377,
+      "learning_rate": 0.00016097834433173705,
+      "loss": 0.3480002164840698,
+      "step": 12360
+    },
+    {
+      "epoch": 0.1645826521087061,
+      "grad_norm": 0.4668227434158325,
+      "learning_rate": 0.0001609144802971282,
+      "loss": 0.39000298976898196,
+      "step": 12370
+    },
+    {
+      "epoch": 0.16487493172530832,
+      "grad_norm": 0.3059532046318054,
+      "learning_rate": 0.00016085057673577446,
+      "loss": 0.39782702922821045,
+      "step": 12380
+    },
+    {
+      "epoch": 0.16516721134191054,
+      "grad_norm": 0.26072707772254944,
+      "learning_rate": 0.0001607866336891421,
+      "loss": 0.35300729274749754,
+      "step": 12390
+    },
+    {
+      "epoch": 0.16545949095851273,
+      "grad_norm": 0.28154456615448,
+      "learning_rate": 0.0001607226511987231,
+      "loss": 0.34312262535095217,
+      "step": 12400
+    },
+    {
+      "epoch": 0.16575177057511495,
+      "grad_norm": 0.3773242235183716,
+      "learning_rate": 0.00016065862930603498,
+      "loss": 0.4181360721588135,
+      "step": 12410
+    },
+    {
+      "epoch": 0.16604405019171717,
+      "grad_norm": 0.3999667167663574,
+      "learning_rate": 0.00016059456805262088,
+      "loss": 0.4283918857574463,
+      "step": 12420
+    },
+    {
+      "epoch": 0.16633632980831936,
+      "grad_norm": 0.30317676067352295,
+      "learning_rate": 0.00016053046748004942,
+      "loss": 0.30436139106750487,
+      "step": 12430
+    },
+    {
+      "epoch": 0.16662860942492158,
+      "grad_norm": 0.30938252806663513,
+      "learning_rate": 0.0001604663276299148,
+      "loss": 0.35764341354370116,
+      "step": 12440
+    },
+    {
+      "epoch": 0.1669208890415238,
+      "grad_norm": 0.2666245698928833,
+      "learning_rate": 0.00016040214854383663,
+      "loss": 0.41030082702636717,
+      "step": 12450
+    },
+    {
+      "epoch": 0.16721316865812602,
+      "grad_norm": 0.4266664981842041,
+      "learning_rate": 0.0001603379302634601,
+      "loss": 0.3680065155029297,
+      "step": 12460
+    },
+    {
+      "epoch": 0.1675054482747282,
+      "grad_norm": 0.32077348232269287,
+      "learning_rate": 0.00016027367283045566,
+      "loss": 0.3224609851837158,
+      "step": 12470
+    },
+    {
+      "epoch": 0.16779772789133043,
+      "grad_norm": 0.37416574358940125,
+      "learning_rate": 0.00016020937628651927,
+      "loss": 0.3794640779495239,
+      "step": 12480
+    },
+    {
+      "epoch": 0.16809000750793265,
+      "grad_norm": 0.21655358374118805,
+      "learning_rate": 0.00016014504067337235,
+      "loss": 0.4269597053527832,
+      "step": 12490
+    },
+    {
+      "epoch": 0.16838228712453487,
+      "grad_norm": 0.3002743422985077,
+      "learning_rate": 0.0001600806660327615,
+      "loss": 0.3735721826553345,
+      "step": 12500
+    },
+    {
+      "epoch": 0.16867456674113707,
+      "grad_norm": 0.5429935455322266,
+      "learning_rate": 0.00016001625240645874,
+      "loss": 0.40839419364929197,
+      "step": 12510
+    },
+    {
+      "epoch": 0.1689668463577393,
+      "grad_norm": 0.3267802894115448,
+      "learning_rate": 0.0001599517998362614,
+      "loss": 0.40875792503356934,
+      "step": 12520
+    },
+    {
+      "epoch": 0.1692591259743415,
+      "grad_norm": 0.3141759932041168,
+      "learning_rate": 0.00015988730836399202,
+      "loss": 0.3485439300537109,
+      "step": 12530
+    },
+    {
+      "epoch": 0.16955140559094373,
+      "grad_norm": 0.240229532122612,
+      "learning_rate": 0.00015982277803149844,
+      "loss": 0.330309534072876,
+      "step": 12540
+    },
+    {
+      "epoch": 0.16984368520754592,
+      "grad_norm": 0.4100056290626526,
+      "learning_rate": 0.00015975820888065366,
+      "loss": 0.3462496280670166,
+      "step": 12550
+    },
+    {
+      "epoch": 0.17013596482414814,
+      "grad_norm": 0.3629114031791687,
+      "learning_rate": 0.00015969360095335596,
+      "loss": 0.32915968894958497,
+      "step": 12560
+    },
+    {
+      "epoch": 0.17042824444075036,
+      "grad_norm": 0.37758728861808777,
+      "learning_rate": 0.00015962895429152868,
+      "loss": 0.3499837636947632,
+      "step": 12570
+    },
+    {
+      "epoch": 0.17072052405735258,
+      "grad_norm": 0.3544098436832428,
+      "learning_rate": 0.0001595642689371203,
+      "loss": 0.3944863319396973,
+      "step": 12580
+    },
+    {
+      "epoch": 0.17101280367395477,
+      "grad_norm": 0.41380608081817627,
+      "learning_rate": 0.0001594995449321045,
+      "loss": 0.3458649396896362,
+      "step": 12590
+    },
+    {
+      "epoch": 0.171305083290557,
+      "grad_norm": 0.37134650349617004,
+      "learning_rate": 0.00015943478231847998,
+      "loss": 0.3552777528762817,
+      "step": 12600
+    },
+    {
+      "epoch": 0.1715973629071592,
+      "grad_norm": 0.4449710547924042,
+      "learning_rate": 0.00015936998113827048,
+      "loss": 0.3530218839645386,
+      "step": 12610
+    },
+    {
+      "epoch": 0.17188964252376143,
+      "grad_norm": 0.2860984802246094,
+      "learning_rate": 0.00015930514143352485,
+      "loss": 0.3369465351104736,
+      "step": 12620
+    },
+    {
+      "epoch": 0.17218192214036362,
+      "grad_norm": 0.3782135248184204,
+      "learning_rate": 0.00015924026324631673,
+      "loss": 0.38503026962280273,
+      "step": 12630
+    },
+    {
+      "epoch": 0.17247420175696584,
+      "grad_norm": 0.32457900047302246,
+      "learning_rate": 0.00015917534661874497,
+      "loss": 0.4050795078277588,
+      "step": 12640
+    },
+    {
+      "epoch": 0.17276648137356806,
+      "grad_norm": 0.33063048124313354,
+      "learning_rate": 0.00015911039159293328,
+      "loss": 0.357547926902771,
+      "step": 12650
+    },
+    {
+      "epoch": 0.17305876099017029,
+      "grad_norm": 0.4063471555709839,
+      "learning_rate": 0.00015904539821103022,
+      "loss": 0.39273886680603026,
+      "step": 12660
+    },
+    {
+      "epoch": 0.17335104060677248,
+      "grad_norm": 0.4146965444087982,
+      "learning_rate": 0.0001589803665152093,
+      "loss": 0.3372164011001587,
+      "step": 12670
+    },
+    {
+      "epoch": 0.1736433202233747,
+      "grad_norm": 0.3036330044269562,
+      "learning_rate": 0.0001589152965476689,
+      "loss": 0.4194971561431885,
+      "step": 12680
+    },
+    {
+      "epoch": 0.17393559983997692,
+      "grad_norm": 0.3907412588596344,
+      "learning_rate": 0.0001588501883506322,
+      "loss": 0.38817367553710935,
+      "step": 12690
+    },
+    {
+      "epoch": 0.00017536776996132774,
+      "grad_norm": 0.2965264320373535,
+      "learning_rate": 0.0001587850419663472,
+      "loss": 0.28757087389628094,
+      "step": 12700
+    },
+    {
+      "epoch": 0.0004676473865635407,
+      "grad_norm": 0.32146406173706055,
+      "learning_rate": 0.00015871985743708667,
+      "loss": 0.2646626472473145,
+      "step": 12710
+    },
+    {
+      "epoch": 0.0007599270031657536,
+      "grad_norm": 0.42033255100250244,
+      "learning_rate": 0.0001586546348051482,
+      "loss": 0.24724342823028564,
+      "step": 12720
+    },
+    {
+      "epoch": 0.0010522066197679665,
+      "grad_norm": 0.47541359066963196,
+      "learning_rate": 0.000158589374112854,
+      "loss": 0.21884760856628419,
+      "step": 12730
+    },
+    {
+      "epoch": 0.0013444862363701795,
+      "grad_norm": 0.40803465247154236,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.19054625034332276,
+      "step": 12740
+    },
+    {
+      "epoch": 0.0016367658529723925,
+      "grad_norm": 0.6147897243499756,
+      "learning_rate": 0.0001584587387166109,
+      "loss": 0.21417717933654784,
+      "step": 12750
+    },
+    {
+      "epoch": 0.0019290454695746052,
+      "grad_norm": 0.49164047837257385,
+      "learning_rate": 0.0001583933640974299,
+      "loss": 0.23636212348937988,
+      "step": 12760
+    },
+    {
+      "epoch": 0.0022213250861768184,
+      "grad_norm": 0.5567198991775513,
+      "learning_rate": 0.00015832795158742893,
+      "loss": 0.21207571029663086,
+      "step": 12770
+    },
+    {
+      "epoch": 0.002513604702779031,
+      "grad_norm": 0.43059054017066956,
+      "learning_rate": 0.0001582625012290534,
+      "loss": 0.2374553918838501,
+      "step": 12780
+    },
+    {
+      "epoch": 0.002805884319381244,
+      "grad_norm": 0.5089011788368225,
+      "learning_rate": 0.00015819701306477341,
+      "loss": 0.20542140007019044,
+      "step": 12790
+    },
+    {
+      "epoch": 0.003098163935983457,
+      "grad_norm": 0.4807959198951721,
+      "learning_rate": 0.00015813148713708346,
+      "loss": 0.24858999252319336,
+      "step": 12800
+    },
+    {
+      "epoch": 0.00339044355258567,
+      "grad_norm": 0.594322681427002,
+      "learning_rate": 0.00015806592348850259,
+      "loss": 0.246284556388855,
+      "step": 12810
+    },
+    {
+      "epoch": 0.0036827231691878826,
+      "grad_norm": 0.43168941140174866,
+      "learning_rate": 0.00015800032216157443,
+      "loss": 0.22252061367034912,
+      "step": 12820
+    },
+    {
+      "epoch": 0.003975002785790095,
+      "grad_norm": 0.491866797208786,
+      "learning_rate": 0.0001579346831988669,
+      "loss": 0.249434757232666,
+      "step": 12830
+    },
+    {
+      "epoch": 0.0042672824023923086,
+      "grad_norm": 0.4541182518005371,
+      "learning_rate": 0.00015786900664297246,
+      "loss": 0.22967674732208251,
+      "step": 12840
+    },
+    {
+      "epoch": 0.004559562018994522,
+      "grad_norm": 0.5304577946662903,
+      "learning_rate": 0.00015780329253650786,
+      "loss": 0.17900593280792237,
+      "step": 12850
+    },
+    {
+      "epoch": 0.004851841635596735,
+      "grad_norm": 0.4978583753108978,
+      "learning_rate": 0.00015773754092211428,
+      "loss": 0.20618371963500975,
+      "step": 12860
+    },
+    {
+      "epoch": 0.005144121252198947,
+      "grad_norm": 0.5236108303070068,
+      "learning_rate": 0.00015767175184245726,
+      "loss": 0.19417285919189453,
+      "step": 12870
+    },
+    {
+      "epoch": 0.0054364008688011604,
+      "grad_norm": 0.5180850625038147,
+      "learning_rate": 0.0001576059253402266,
+      "loss": 0.21911945343017578,
+      "step": 12880
+    },
+    {
+      "epoch": 0.005728680485403374,
+      "grad_norm": 0.4623672366142273,
+      "learning_rate": 0.00015754006145813642,
+      "loss": 0.1905503511428833,
+      "step": 12890
+    },
+    {
+      "epoch": 0.006020960102005586,
+      "grad_norm": 0.3613812327384949,
+      "learning_rate": 0.00015747416023892505,
+      "loss": 0.23439469337463378,
+      "step": 12900
+    },
+    {
+      "epoch": 0.006313239718607799,
+      "grad_norm": 0.4850548803806305,
+      "learning_rate": 0.00015740822172535508,
+      "loss": 0.20268757343292237,
+      "step": 12910
+    },
+    {
+      "epoch": 0.006605519335210012,
+      "grad_norm": 0.399964302778244,
+      "learning_rate": 0.0001573422459602133,
+      "loss": 0.22756083011627198,
+      "step": 12920
+    },
+    {
+      "epoch": 0.006897798951812225,
+      "grad_norm": 0.4820435643196106,
+      "learning_rate": 0.0001572762329863107,
+      "loss": 0.2592332363128662,
+      "step": 12930
+    },
+    {
+      "epoch": 0.007190078568414438,
+      "grad_norm": 0.7340999841690063,
+      "learning_rate": 0.00015721018284648233,
+      "loss": 0.2214745283126831,
+      "step": 12940
+    },
+    {
+      "epoch": 0.007482358185016651,
+      "grad_norm": 0.3809256851673126,
+      "learning_rate": 0.00015714409558358744,
+      "loss": 0.15033090114593506,
+      "step": 12950
+    },
+    {
+      "epoch": 0.007774637801618863,
+      "grad_norm": 0.5849661827087402,
+      "learning_rate": 0.00015707797124050934,
+      "loss": 0.197877836227417,
+      "step": 12960
+    },
+    {
+      "epoch": 0.008066917418221077,
+      "grad_norm": 0.43279632925987244,
+      "learning_rate": 0.00015701180986015537,
+      "loss": 0.21142268180847168,
+      "step": 12970
+    },
+    {
+      "epoch": 0.008359197034823289,
+      "grad_norm": 0.43526268005371094,
+      "learning_rate": 0.00015694561148545696,
+      "loss": 0.1974856972694397,
+      "step": 12980
+    },
+    {
+      "epoch": 0.008651476651425502,
+      "grad_norm": 0.417270690202713,
+      "learning_rate": 0.0001568793761593695,
+      "loss": 0.21758065223693848,
+      "step": 12990
+    },
+    {
+      "epoch": 0.008943756268027715,
+      "grad_norm": 0.5802550315856934,
+      "learning_rate": 0.0001568131039248724,
+      "loss": 0.21777856349945068,
+      "step": 13000
+    },
+    {
+      "epoch": 0.009236035884629928,
+      "grad_norm": 0.4395245909690857,
+      "learning_rate": 0.00015674679482496898,
+      "loss": 0.18997634649276735,
+      "step": 13010
+    },
+    {
+      "epoch": 0.009528315501232142,
+      "grad_norm": 0.47754234075546265,
+      "learning_rate": 0.00015668044890268648,
+      "loss": 0.18700885772705078,
+      "step": 13020
+    },
+    {
+      "epoch": 0.009820595117834355,
+      "grad_norm": 0.39656195044517517,
+      "learning_rate": 0.00015661406620107605,
+      "loss": 0.21086442470550537,
+      "step": 13030
+    },
+    {
+      "epoch": 0.010112874734436566,
+      "grad_norm": 0.4576987624168396,
+      "learning_rate": 0.0001565476467632127,
+      "loss": 0.2196120500564575,
+      "step": 13040
+    },
+    {
+      "epoch": 0.01040515435103878,
+      "grad_norm": 0.5417501330375671,
+      "learning_rate": 0.00015648119063219534,
+      "loss": 0.17074507474899292,
+      "step": 13050
+    },
+    {
+      "epoch": 0.010697433967640993,
+      "grad_norm": 0.3395984172821045,
+      "learning_rate": 0.00015641469785114654,
+      "loss": 0.23015992641448973,
+      "step": 13060
+    },
+    {
+      "epoch": 0.010989713584243206,
+      "grad_norm": 0.38607853651046753,
+      "learning_rate": 0.00015634816846321278,
+      "loss": 0.1890069842338562,
+      "step": 13070
+    },
+    {
+      "epoch": 0.011281993200845419,
+      "grad_norm": 0.8692626953125,
+      "learning_rate": 0.0001562816025115643,
+      "loss": 0.20842335224151612,
+      "step": 13080
+    },
+    {
+      "epoch": 0.011574272817447632,
+      "grad_norm": 0.4859260320663452,
+      "learning_rate": 0.00015621500003939497,
+      "loss": 0.22042946815490722,
+      "step": 13090
+    },
+    {
+      "epoch": 0.011866552434049845,
+      "grad_norm": 0.4462348222732544,
+      "learning_rate": 0.00015614836108992244,
+      "loss": 0.1920430541038513,
+      "step": 13100
+    },
+    {
+      "epoch": 0.012158832050652057,
+      "grad_norm": 0.40549933910369873,
+      "learning_rate": 0.000156081685706388,
+      "loss": 0.17254678010940552,
+      "step": 13110
+    },
+    {
+      "epoch": 0.01245111166725427,
+      "grad_norm": 0.45183098316192627,
+      "learning_rate": 0.00015601497393205659,
+      "loss": 0.20989205837249755,
+      "step": 13120
+    },
+    {
+      "epoch": 0.012743391283856483,
+      "grad_norm": 0.42247337102890015,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 0.19462287425994873,
+      "step": 13130
+    },
+    {
+      "epoch": 0.013035670900458696,
+      "grad_norm": 0.5868353843688965,
+      "learning_rate": 0.00015588144138418053,
+      "loss": 0.16435409784317018,
+      "step": 13140
+    },
+    {
+      "epoch": 0.01332795051706091,
+      "grad_norm": 0.5656533241271973,
+      "learning_rate": 0.00015581462069728374,
+      "loss": 0.2222140312194824,
+      "step": 13150
+    },
+    {
+      "epoch": 0.013620230133663123,
+      "grad_norm": 0.5332701206207275,
+      "learning_rate": 0.00015574776379288557,
+      "loss": 0.21317110061645508,
+      "step": 13160
+    },
+    {
+      "epoch": 0.013912509750265336,
+      "grad_norm": 0.3974187672138214,
+      "learning_rate": 0.0001556808707143687,
+      "loss": 0.1519344210624695,
+      "step": 13170
+    },
+    {
+      "epoch": 0.014204789366867547,
+      "grad_norm": 0.6250615119934082,
+      "learning_rate": 0.00015561394150513932,
+      "loss": 0.17461782693862915,
+      "step": 13180
+    },
+    {
+      "epoch": 0.01449706898346976,
+      "grad_norm": 0.44540587067604065,
+      "learning_rate": 0.0001555469762086271,
+      "loss": 0.19416953325271608,
+      "step": 13190
+    },
+    {
+      "epoch": 0.014789348600071974,
+      "grad_norm": 0.48378169536590576,
+      "learning_rate": 0.0001554799748682851,
+      "loss": 0.2017523765563965,
+      "step": 13200
+    },
+    {
+      "epoch": 0.015081628216674187,
+      "grad_norm": 0.4456275403499603,
+      "learning_rate": 0.00015541293752758972,
+      "loss": 0.1936066746711731,
+      "step": 13210
+    },
+    {
+      "epoch": 0.0153739078332764,
+      "grad_norm": 0.7115320563316345,
+      "learning_rate": 0.00015534586423004081,
+      "loss": 0.23387138843536376,
+      "step": 13220
+    },
+    {
+      "epoch": 0.01566618744987861,
+      "grad_norm": 0.45951545238494873,
+      "learning_rate": 0.0001552787550191615,
+      "loss": 0.19637324810028076,
+      "step": 13230
+    },
+    {
+      "epoch": 0.015958467066480825,
+      "grad_norm": 0.6550038456916809,
+      "learning_rate": 0.00015521160993849816,
+      "loss": 0.20966033935546874,
+      "step": 13240
+    },
+    {
+      "epoch": 0.016250746683083038,
+      "grad_norm": 0.5067435503005981,
+      "learning_rate": 0.00015514442903162058,
+      "loss": 0.21880669593811036,
+      "step": 13250
+    },
+    {
+      "epoch": 0.01654302629968525,
+      "grad_norm": 0.5030020475387573,
+      "learning_rate": 0.00015507721234212172,
+      "loss": 0.22664711475372315,
+      "step": 13260
+    },
+    {
+      "epoch": 0.016835305916287464,
+      "grad_norm": 0.6950615048408508,
+      "learning_rate": 0.0001550099599136177,
+      "loss": 0.2090200424194336,
+      "step": 13270
+    },
+    {
+      "epoch": 0.017127585532889678,
+      "grad_norm": 0.4828246533870697,
+      "learning_rate": 0.00015494267178974794,
+      "loss": 0.22880713939666747,
+      "step": 13280
+    },
+    {
+      "epoch": 0.01741986514949189,
+      "grad_norm": 0.3490094542503357,
+      "learning_rate": 0.00015487534801417497,
+      "loss": 0.23944876194000245,
+      "step": 13290
+    },
+    {
+      "epoch": 0.017712144766094104,
+      "grad_norm": 0.8854429721832275,
+      "learning_rate": 0.00015480798863058443,
+      "loss": 0.1837323546409607,
+      "step": 13300
+    },
+    {
+      "epoch": 0.018004424382696317,
+      "grad_norm": 0.4737114906311035,
+      "learning_rate": 0.00015474059368268512,
+      "loss": 0.20009281635284423,
+      "step": 13310
+    },
+    {
+      "epoch": 0.01829670399929853,
+      "grad_norm": 0.5213412046432495,
+      "learning_rate": 0.0001546731632142089,
+      "loss": 0.19314022064208985,
+      "step": 13320
+    },
+    {
+      "epoch": 0.018588983615900743,
+      "grad_norm": 0.5175877809524536,
+      "learning_rate": 0.00015460569726891062,
+      "loss": 0.20960526466369628,
+      "step": 13330
+    },
+    {
+      "epoch": 0.018881263232502953,
+      "grad_norm": 0.45757150650024414,
+      "learning_rate": 0.00015453819589056824,
+      "loss": 0.19740055799484252,
+      "step": 13340
+    },
+    {
+      "epoch": 0.019173542849105166,
+      "grad_norm": 0.6081610918045044,
+      "learning_rate": 0.00015447065912298265,
+      "loss": 0.20936164855957032,
+      "step": 13350
+    },
+    {
+      "epoch": 0.01946582246570738,
+      "grad_norm": 0.37390947341918945,
+      "learning_rate": 0.0001544030870099777,
+      "loss": 0.2155306816101074,
+      "step": 13360
+    },
+    {
+      "epoch": 0.019758102082309593,
+      "grad_norm": 0.6000289916992188,
+      "learning_rate": 0.00015433547959540028,
+      "loss": 0.22382841110229493,
+      "step": 13370
+    },
+    {
+      "epoch": 0.020050381698911806,
+      "grad_norm": 0.5844590067863464,
+      "learning_rate": 0.00015426783692312009,
+      "loss": 0.18855516910552977,
+      "step": 13380
+    },
+    {
+      "epoch": 0.02034266131551402,
+      "grad_norm": 0.38509050011634827,
+      "learning_rate": 0.00015420015903702962,
+      "loss": 0.1874496340751648,
+      "step": 13390
+    },
+    {
+      "epoch": 0.020634940932116232,
+      "grad_norm": 0.7230851650238037,
+      "learning_rate": 0.00015413244598104443,
+      "loss": 0.1505223274230957,
+      "step": 13400
+    },
+    {
+      "epoch": 0.020927220548718446,
+      "grad_norm": 0.4711814522743225,
+      "learning_rate": 0.00015406469779910277,
+      "loss": 0.20109641551971436,
+      "step": 13410
+    },
+    {
+      "epoch": 0.02121950016532066,
+      "grad_norm": 0.4168683886528015,
+      "learning_rate": 0.00015399691453516566,
+      "loss": 0.15834118127822877,
+      "step": 13420
+    },
+    {
+      "epoch": 0.021511779781922872,
+      "grad_norm": 0.4296996593475342,
+      "learning_rate": 0.00015392909623321695,
+      "loss": 0.19768688678741456,
+      "step": 13430
+    },
+    {
+      "epoch": 0.021804059398525085,
+      "grad_norm": 0.5149158239364624,
+      "learning_rate": 0.0001538612429372632,
+      "loss": 0.21142776012420655,
+      "step": 13440
+    },
+    {
+      "epoch": 0.0220963390151273,
+      "grad_norm": 0.580381453037262,
+      "learning_rate": 0.00015379335469133367,
+      "loss": 0.2325075387954712,
+      "step": 13450
+    },
+    {
+      "epoch": 0.02238861863172951,
+      "grad_norm": 0.4189280569553375,
+      "learning_rate": 0.00015372543153948032,
+      "loss": 0.17044657468795776,
+      "step": 13460
+    },
+    {
+      "epoch": 0.02268089824833172,
+      "grad_norm": 0.5113292336463928,
+      "learning_rate": 0.0001536574735257777,
+      "loss": 0.18655259609222413,
+      "step": 13470
+    },
+    {
+      "epoch": 0.022973177864933934,
+      "grad_norm": 0.3685324788093567,
+      "learning_rate": 0.0001535894806943231,
+      "loss": 0.18744727373123168,
+      "step": 13480
+    },
+    {
+      "epoch": 0.023265457481536148,
+      "grad_norm": 0.4151572287082672,
+      "learning_rate": 0.00015352145308923624,
+      "loss": 0.2236708164215088,
+      "step": 13490
+    },
+    {
+      "epoch": 0.02355773709813836,
+      "grad_norm": 0.5585652589797974,
+      "learning_rate": 0.0001534533907546596,
+      "loss": 0.21556272506713867,
+      "step": 13500
+    },
+    {
+      "epoch": 0.023850016714740574,
+      "grad_norm": 0.32925117015838623,
+      "learning_rate": 0.00015338529373475798,
+      "loss": 0.1861099123954773,
+      "step": 13510
+    },
+    {
+      "epoch": 0.024142296331342787,
+      "grad_norm": 0.44740787148475647,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.23113582134246827,
+      "step": 13520
+    },
+    {
+      "epoch": 0.024434575947945,
+      "grad_norm": 0.3342213034629822,
+      "learning_rate": 0.00015324899581575213,
+      "loss": 0.2114185571670532,
+      "step": 13530
+    },
+    {
+      "epoch": 0.024726855564547214,
+      "grad_norm": 0.4604058265686035,
+      "learning_rate": 0.00015318079500509018,
+      "loss": 0.18679240942001343,
+      "step": 13540
+    },
+    {
+      "epoch": 0.025019135181149427,
+      "grad_norm": 0.40558451414108276,
+      "learning_rate": 0.00015311255968598767,
+      "loss": 0.2514742374420166,
+      "step": 13550
+    },
+    {
+      "epoch": 0.02531141479775164,
+      "grad_norm": 0.48979073762893677,
+      "learning_rate": 0.00015304428990272183,
+      "loss": 0.20039637088775636,
+      "step": 13560
+    },
+    {
+      "epoch": 0.025603694414353853,
+      "grad_norm": 0.47555047273635864,
+      "learning_rate": 0.00015297598569959217,
+      "loss": 0.24015424251556397,
+      "step": 13570
+    },
+    {
+      "epoch": 0.025895974030956066,
+      "grad_norm": 0.5791113972663879,
+      "learning_rate": 0.00015290764712092055,
+      "loss": 0.19669321775436402,
+      "step": 13580
+    },
+    {
+      "epoch": 0.02618825364755828,
+      "grad_norm": 0.589297890663147,
+      "learning_rate": 0.0001528392742110511,
+      "loss": 0.2425452709197998,
+      "step": 13590
+    },
+    {
+      "epoch": 0.02648053326416049,
+      "grad_norm": 0.3458498418331146,
+      "learning_rate": 0.0001527708670143503,
+      "loss": 0.1887451171875,
+      "step": 13600
+    },
+    {
+      "epoch": 0.026772812880762702,
+      "grad_norm": 0.767162561416626,
+      "learning_rate": 0.00015270242557520683,
+      "loss": 0.1672801971435547,
+      "step": 13610
+    },
+    {
+      "epoch": 0.027065092497364916,
+      "grad_norm": 0.5664442777633667,
+      "learning_rate": 0.00015263394993803158,
+      "loss": 0.20956158638000488,
+      "step": 13620
+    },
+    {
+      "epoch": 0.02735737211396713,
+      "grad_norm": 0.41770660877227783,
+      "learning_rate": 0.00015256544014725763,
+      "loss": 0.20826971530914307,
+      "step": 13630
+    },
+    {
+      "epoch": 0.027649651730569342,
+      "grad_norm": 0.38560521602630615,
+      "learning_rate": 0.0001524968962473403,
+      "loss": 0.23094625473022462,
+      "step": 13640
+    },
+    {
+      "epoch": 0.027941931347171555,
+      "grad_norm": 0.41935762763023376,
+      "learning_rate": 0.00015242831828275692,
+      "loss": 0.19274245500564574,
+      "step": 13650
+    },
+    {
+      "epoch": 0.02823421096377377,
+      "grad_norm": 0.4347086250782013,
+      "learning_rate": 0.00015235970629800703,
+      "loss": 0.21260333061218262,
+      "step": 13660
+    },
+    {
+      "epoch": 0.02852649058037598,
+      "grad_norm": 0.39672279357910156,
+      "learning_rate": 0.00015229106033761217,
+      "loss": 0.21170544624328613,
+      "step": 13670
+    },
+    {
+      "epoch": 0.028818770196978195,
+      "grad_norm": 0.4290609359741211,
+      "learning_rate": 0.00015222238044611597,
+      "loss": 0.21978483200073243,
+      "step": 13680
+    },
+    {
+      "epoch": 0.029111049813580408,
+      "grad_norm": 0.46452850103378296,
+      "learning_rate": 0.00015215366666808405,
+      "loss": 0.19815123081207275,
+      "step": 13690
+    },
+    {
+      "epoch": 0.02940332943018262,
+      "grad_norm": 0.4070618152618408,
+      "learning_rate": 0.0001520849190481041,
+      "loss": 0.21544198989868163,
+      "step": 13700
+    },
+    {
+      "epoch": 0.029695609046784834,
+      "grad_norm": 0.3452356159687042,
+      "learning_rate": 0.00015201613763078564,
+      "loss": 0.22686522006988524,
+      "step": 13710
+    },
+    {
+      "epoch": 0.029987888663387047,
+      "grad_norm": 0.4308999478816986,
+      "learning_rate": 0.0001519473224607602,
+      "loss": 0.2282576560974121,
+      "step": 13720
+    },
+    {
+      "epoch": 0.030280168279989257,
+      "grad_norm": 0.5049606561660767,
+      "learning_rate": 0.0001518784735826812,
+      "loss": 0.27099735736846925,
+      "step": 13730
+    },
+    {
+      "epoch": 0.03057244789659147,
+      "grad_norm": 0.52198326587677,
+      "learning_rate": 0.00015180959104122392,
+      "loss": 0.20504209995269776,
+      "step": 13740
+    },
+    {
+      "epoch": 0.030864727513193684,
+      "grad_norm": 0.3714408576488495,
+      "learning_rate": 0.00015174067488108549,
+      "loss": 0.17479199171066284,
+      "step": 13750
+    },
+    {
+      "epoch": 0.031157007129795897,
+      "grad_norm": 0.8701677322387695,
+      "learning_rate": 0.00015167172514698492,
+      "loss": 0.20214717388153075,
+      "step": 13760
+    },
+    {
+      "epoch": 0.03144928674639811,
+      "grad_norm": 0.45611393451690674,
+      "learning_rate": 0.00015160274188366282,
+      "loss": 0.189096999168396,
+      "step": 13770
+    },
+    {
+      "epoch": 0.03174156636300032,
+      "grad_norm": 0.5112565755844116,
+      "learning_rate": 0.00015153372513588182,
+      "loss": 0.2118607997894287,
+      "step": 13780
+    },
+    {
+      "epoch": 0.032033845979602536,
+      "grad_norm": 0.40792563557624817,
+      "learning_rate": 0.00015146467494842603,
+      "loss": 0.1626830816268921,
+      "step": 13790
+    },
+    {
+      "epoch": 0.03232612559620475,
+      "grad_norm": 0.5236932635307312,
+      "learning_rate": 0.00015139559136610146,
+      "loss": 0.19841558933258058,
+      "step": 13800
+    },
+    {
+      "epoch": 0.03261840521280696,
+      "grad_norm": 0.45573049783706665,
+      "learning_rate": 0.0001513264744337356,
+      "loss": 0.23235554695129396,
+      "step": 13810
+    },
+    {
+      "epoch": 0.032910684829409176,
+      "grad_norm": 0.4380190968513489,
+      "learning_rate": 0.00015125732419617776,
+      "loss": 0.181347119808197,
+      "step": 13820
+    },
+    {
+      "epoch": 0.03320296444601139,
+      "grad_norm": 0.3415079712867737,
+      "learning_rate": 0.00015118814069829873,
+      "loss": 0.19460681676864625,
+      "step": 13830
+    },
+    {
+      "epoch": 0.0334952440626136,
+      "grad_norm": 0.46894368529319763,
+      "learning_rate": 0.0001511189239849909,
+      "loss": 0.22168433666229248,
+      "step": 13840
+    },
+    {
+      "epoch": 0.033787523679215815,
+      "grad_norm": 0.39394792914390564,
+      "learning_rate": 0.0001510496741011683,
+      "loss": 0.2261059284210205,
+      "step": 13850
+    },
+    {
+      "epoch": 0.03407980329581803,
+      "grad_norm": 0.48252561688423157,
+      "learning_rate": 0.00015098039109176647,
+      "loss": 0.19224013090133668,
+      "step": 13860
+    },
+    {
+      "epoch": 0.03437208291242024,
+      "grad_norm": 0.47888413071632385,
+      "learning_rate": 0.0001509110750017423,
+      "loss": 0.1769060492515564,
+      "step": 13870
+    },
+    {
+      "epoch": 0.034664362529022455,
+      "grad_norm": 0.38338038325309753,
+      "learning_rate": 0.0001508417258760743,
+      "loss": 0.21865322589874267,
+      "step": 13880
+    },
+    {
+      "epoch": 0.03495664214562467,
+      "grad_norm": 0.5309614539146423,
+      "learning_rate": 0.00015077234375976238,
+      "loss": 0.2368218183517456,
+      "step": 13890
+    },
+    {
+      "epoch": 0.03524892176222688,
+      "grad_norm": 0.3875935673713684,
+      "learning_rate": 0.0001507029286978278,
+      "loss": 0.17107208967208862,
+      "step": 13900
+    },
+    {
+      "epoch": 0.035541201378829095,
+      "grad_norm": 0.39589938521385193,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.2395782470703125,
+      "step": 13910
+    },
+    {
+      "epoch": 0.0358334809954313,
+      "grad_norm": 0.4878383278846741,
+      "learning_rate": 0.0001505639999172828,
+      "loss": 0.2368070363998413,
+      "step": 13920
+    },
+    {
+      "epoch": 0.036125760612033514,
+      "grad_norm": 0.5454540848731995,
+      "learning_rate": 0.00015049448628882178,
+      "loss": 0.21010217666625977,
+      "step": 13930
+    },
+    {
+      "epoch": 0.03641804022863573,
+      "grad_norm": 0.5538231730461121,
+      "learning_rate": 0.00015042493989503677,
+      "loss": 0.15809463262557982,
+      "step": 13940
+    },
+    {
+      "epoch": 0.03671031984523794,
+      "grad_norm": 0.40927162766456604,
+      "learning_rate": 0.00015035536078105577,
+      "loss": 0.1732885718345642,
+      "step": 13950
+    },
+    {
+      "epoch": 0.037002599461840154,
+      "grad_norm": 0.4968807101249695,
+      "learning_rate": 0.0001502857489920278,
+      "loss": 0.2206113815307617,
+      "step": 13960
+    },
+    {
+      "epoch": 0.03729487907844237,
+      "grad_norm": 0.36112022399902344,
+      "learning_rate": 0.00015021610457312333,
+      "loss": 0.21986114978790283,
+      "step": 13970
+    },
+    {
+      "epoch": 0.03758715869504458,
+      "grad_norm": 0.46643736958503723,
+      "learning_rate": 0.00015014642756953372,
+      "loss": 0.2392876148223877,
+      "step": 13980
+    },
+    {
+      "epoch": 0.03787943831164679,
+      "grad_norm": 0.4169055223464966,
+      "learning_rate": 0.0001500767180264717,
+      "loss": 0.179047155380249,
+      "step": 13990
+    },
+    {
+      "epoch": 0.038171717928249006,
+      "grad_norm": 0.48698920011520386,
+      "learning_rate": 0.00015000697598917107,
+      "loss": 0.2633962631225586,
+      "step": 14000
+    },
+    {
+      "epoch": 0.03846399754485122,
+      "grad_norm": 0.3935231566429138,
+      "learning_rate": 0.0001499372015028866,
+      "loss": 0.18749451637268066,
+      "step": 14010
+    },
+    {
+      "epoch": 0.03875627716145343,
+      "grad_norm": 0.5970826148986816,
+      "learning_rate": 0.00014986739461289426,
+      "loss": 0.23365018367767335,
+      "step": 14020
+    },
+    {
+      "epoch": 0.039048556778055646,
+      "grad_norm": 0.4059792757034302,
+      "learning_rate": 0.00014979755536449096,
+      "loss": 0.1624300956726074,
+      "step": 14030
+    },
+    {
+      "epoch": 0.03934083639465786,
+      "grad_norm": 0.5249799489974976,
+      "learning_rate": 0.0001497276838029946,
+      "loss": 0.22115600109100342,
+      "step": 14040
+    },
+    {
+      "epoch": 0.03963311601126007,
+      "grad_norm": 0.48356330394744873,
+      "learning_rate": 0.00014965777997374415,
+      "loss": 0.20501933097839356,
+      "step": 14050
+    },
+    {
+      "epoch": 0.039925395627862285,
+      "grad_norm": 0.5692484974861145,
+      "learning_rate": 0.00014958784392209937,
+      "loss": 0.2200528860092163,
+      "step": 14060
+    },
+    {
+      "epoch": 0.0402176752444645,
+      "grad_norm": 0.46216651797294617,
+      "learning_rate": 0.00014951787569344104,
+      "loss": 0.23633096218109131,
+      "step": 14070
+    },
+    {
+      "epoch": 0.04050995486106671,
+      "grad_norm": 0.5820059776306152,
+      "learning_rate": 0.00014944787533317078,
+      "loss": 0.19459173679351807,
+      "step": 14080
+    },
+    {
+      "epoch": 0.040802234477668925,
+      "grad_norm": 0.8841289281845093,
+      "learning_rate": 0.00014937784288671107,
+      "loss": 0.1946939468383789,
+      "step": 14090
+    },
+    {
+      "epoch": 0.04109451409427114,
+      "grad_norm": 0.3671911954879761,
+      "learning_rate": 0.0001493077783995052,
+      "loss": 0.21439237594604493,
+      "step": 14100
+    },
+    {
+      "epoch": 0.04138679371087335,
+      "grad_norm": 0.48408985137939453,
+      "learning_rate": 0.00014923768191701725,
+      "loss": 0.2266697645187378,
+      "step": 14110
+    },
+    {
+      "epoch": 0.041679073327475565,
+      "grad_norm": 0.3516557812690735,
+      "learning_rate": 0.00014916755348473205,
+      "loss": 0.24147336483001708,
+      "step": 14120
+    },
+    {
+      "epoch": 0.04197135294407778,
+      "grad_norm": 0.40993836522102356,
+      "learning_rate": 0.00014909739314815522,
+      "loss": 0.18642300367355347,
+      "step": 14130
+    },
+    {
+      "epoch": 0.04226363256067999,
+      "grad_norm": 0.48947736620903015,
+      "learning_rate": 0.00014902720095281305,
+      "loss": 0.23755452632904053,
+      "step": 14140
+    },
+    {
+      "epoch": 0.042555912177282204,
+      "grad_norm": 0.5798397660255432,
+      "learning_rate": 0.0001489569769442524,
+      "loss": 0.24685332775115967,
+      "step": 14150
+    },
+    {
+      "epoch": 0.04284819179388442,
+      "grad_norm": 0.39642032980918884,
+      "learning_rate": 0.00014888672116804096,
+      "loss": 0.24382400512695312,
+      "step": 14160
+    },
+    {
+      "epoch": 0.04314047141048663,
+      "grad_norm": 0.4151468873023987,
+      "learning_rate": 0.00014881643366976692,
+      "loss": 0.18123635053634643,
+      "step": 14170
+    },
+    {
+      "epoch": 0.043432751027088844,
+      "grad_norm": 0.5330682396888733,
+      "learning_rate": 0.00014874611449503905,
+      "loss": 0.18556181192398072,
+      "step": 14180
+    },
+    {
+      "epoch": 0.04372503064369105,
+      "grad_norm": 0.3223675787448883,
+      "learning_rate": 0.0001486757636894867,
+      "loss": 0.177858567237854,
+      "step": 14190
+    },
+    {
+      "epoch": 0.04401731026029326,
+      "grad_norm": 0.30147096514701843,
+      "learning_rate": 0.0001486053812987598,
+      "loss": 0.24367544651031495,
+      "step": 14200
+    },
+    {
+      "epoch": 0.044309589876895476,
+      "grad_norm": 0.49423977732658386,
+      "learning_rate": 0.00014853496736852862,
+      "loss": 0.19779525995254515,
+      "step": 14210
+    },
+    {
+      "epoch": 0.04460186949349769,
+      "grad_norm": 0.35623109340667725,
+      "learning_rate": 0.00014846452194448411,
+      "loss": 0.18698278665542603,
+      "step": 14220
+    },
+    {
+      "epoch": 0.0448941491100999,
+      "grad_norm": 0.49302974343299866,
+      "learning_rate": 0.00014839404507233743,
+      "loss": 0.21252613067626952,
+      "step": 14230
+    },
+    {
+      "epoch": 0.045186428726702116,
+      "grad_norm": 0.5952360033988953,
+      "learning_rate": 0.00014832353679782031,
+      "loss": 0.1472646951675415,
+      "step": 14240
+    },
+    {
+      "epoch": 0.04547870834330433,
+      "grad_norm": 0.42707258462905884,
+      "learning_rate": 0.00014825299716668487,
+      "loss": 0.15044156312942505,
+      "step": 14250
+    },
+    {
+      "epoch": 0.04577098795990654,
+      "grad_norm": 0.45512667298316956,
+      "learning_rate": 0.00014818242622470338,
+      "loss": 0.20248939990997314,
+      "step": 14260
+    },
+    {
+      "epoch": 0.046063267576508755,
+      "grad_norm": 0.5003895163536072,
+      "learning_rate": 0.0001481118240176686,
+      "loss": 0.187678062915802,
+      "step": 14270
+    },
+    {
+      "epoch": 0.04635554719311097,
+      "grad_norm": 0.4616633653640747,
+      "learning_rate": 0.00014804119059139362,
+      "loss": 0.2720855712890625,
+      "step": 14280
+    },
+    {
+      "epoch": 0.04664782680971318,
+      "grad_norm": 0.5666295289993286,
+      "learning_rate": 0.00014797052599171162,
+      "loss": 0.2122922420501709,
+      "step": 14290
+    },
+    {
+      "epoch": 0.046940106426315395,
+      "grad_norm": 0.4432585537433624,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 0.22646279335021974,
+      "step": 14300
+    },
+    {
+      "epoch": 0.04723238604291761,
+      "grad_norm": 0.37214380502700806,
+      "learning_rate": 0.0001478291034555608,
+      "loss": 0.1915860056877136,
+      "step": 14310
+    },
+    {
+      "epoch": 0.04752466565951982,
+      "grad_norm": 0.3837926685810089,
+      "learning_rate": 0.00014775834561085952,
+      "loss": 0.23169612884521484,
+      "step": 14320
+    },
+    {
+      "epoch": 0.047816945276122035,
+      "grad_norm": 0.5426556468009949,
+      "learning_rate": 0.00014768755677628626,
+      "loss": 0.21559467315673828,
+      "step": 14330
+    },
+    {
+      "epoch": 0.04810922489272425,
+      "grad_norm": 0.5345485806465149,
+      "learning_rate": 0.00014761673699777514,
+      "loss": 0.21412506103515624,
+      "step": 14340
+    },
+    {
+      "epoch": 0.04840150450932646,
+      "grad_norm": 0.41953203082084656,
+      "learning_rate": 0.00014754588632128038,
+      "loss": 0.20062601566314697,
+      "step": 14350
+    },
+    {
+      "epoch": 0.048693784125928674,
+      "grad_norm": 0.41677215695381165,
+      "learning_rate": 0.00014747500479277617,
+      "loss": 0.1592464804649353,
+      "step": 14360
+    },
+    {
+      "epoch": 0.04898606374253089,
+      "grad_norm": 0.5690479874610901,
+      "learning_rate": 0.00014740409245825678,
+      "loss": 0.18383355140686036,
+      "step": 14370
+    },
+    {
+      "epoch": 0.0492783433591331,
+      "grad_norm": 0.5026178956031799,
+      "learning_rate": 0.00014733314936373645,
+      "loss": 0.2009049654006958,
+      "step": 14380
+    },
+    {
+      "epoch": 0.049570622975735314,
+      "grad_norm": 0.40158507227897644,
+      "learning_rate": 0.00014726217555524937,
+      "loss": 0.17864898443222046,
+      "step": 14390
+    },
+    {
+      "epoch": 0.04986290259233753,
+      "grad_norm": 0.5003887414932251,
+      "learning_rate": 0.0001471911710788497,
+      "loss": 0.2057699203491211,
+      "step": 14400
+    },
+    {
+      "epoch": 0.05015518220893974,
+      "grad_norm": 0.45541295409202576,
+      "learning_rate": 0.0001471201359806115,
+      "loss": 0.2345907211303711,
+      "step": 14410
+    },
+    {
+      "epoch": 0.05044746182554195,
+      "grad_norm": 0.405276358127594,
+      "learning_rate": 0.0001470490703066286,
+      "loss": 0.18805484771728515,
+      "step": 14420
+    },
+    {
+      "epoch": 0.050739741442144166,
+      "grad_norm": 0.5377190709114075,
+      "learning_rate": 0.0001469779741030148,
+      "loss": 0.24354803562164307,
+      "step": 14430
+    },
+    {
+      "epoch": 0.05103202105874638,
+      "grad_norm": 0.4767274856567383,
+      "learning_rate": 0.00014690684741590362,
+      "loss": 0.2150782585144043,
+      "step": 14440
+    },
+    {
+      "epoch": 0.051324300675348586,
+      "grad_norm": 0.41643649339675903,
+      "learning_rate": 0.0001468356902914484,
+      "loss": 0.18937363624572753,
+      "step": 14450
+    },
+    {
+      "epoch": 0.0516165802919508,
+      "grad_norm": 0.3812100887298584,
+      "learning_rate": 0.00014676450277582224,
+      "loss": 0.1526624321937561,
+      "step": 14460
+    },
+    {
+      "epoch": 0.05190885990855301,
+      "grad_norm": 0.5975919961929321,
+      "learning_rate": 0.000146693284915218,
+      "loss": 0.1669335961341858,
+      "step": 14470
+    },
+    {
+      "epoch": 0.052201139525155225,
+      "grad_norm": 0.43654942512512207,
+      "learning_rate": 0.0001466220367558481,
+      "loss": 0.18291643857955933,
+      "step": 14480
+    },
+    {
+      "epoch": 0.05249341914175744,
+      "grad_norm": 0.5156868100166321,
+      "learning_rate": 0.00014655075834394467,
+      "loss": 0.19642175436019899,
+      "step": 14490
+    },
+    {
+      "epoch": 0.05278569875835965,
+      "grad_norm": 0.5537710189819336,
+      "learning_rate": 0.0001464794497257596,
+      "loss": 0.20847535133361816,
+      "step": 14500
+    },
+    {
+      "epoch": 0.053077978374961865,
+      "grad_norm": 0.4609392583370209,
+      "learning_rate": 0.00014640811094756418,
+      "loss": 0.2045342206954956,
+      "step": 14510
+    },
+    {
+      "epoch": 0.05337025799156408,
+      "grad_norm": 0.4139176905155182,
+      "learning_rate": 0.00014633674205564943,
+      "loss": 0.1959432601928711,
+      "step": 14520
+    },
+    {
+      "epoch": 0.05366253760816629,
+      "grad_norm": 0.5219272971153259,
+      "learning_rate": 0.00014626534309632584,
+      "loss": 0.20467445850372315,
+      "step": 14530
+    },
+    {
+      "epoch": 0.053954817224768505,
+      "grad_norm": 0.5077035427093506,
+      "learning_rate": 0.0001461939141159234,
+      "loss": 0.20045676231384277,
+      "step": 14540
+    },
+    {
+      "epoch": 0.05424709684137072,
+      "grad_norm": 0.31908538937568665,
+      "learning_rate": 0.0001461224551607916,
+      "loss": 0.14442721605300904,
+      "step": 14550
+    },
+    {
+      "epoch": 0.05453937645797293,
+      "grad_norm": 0.5112200379371643,
+      "learning_rate": 0.0001460509662772994,
+      "loss": 0.2712923288345337,
+      "step": 14560
+    },
+    {
+      "epoch": 0.054831656074575144,
+      "grad_norm": 0.30819952487945557,
+      "learning_rate": 0.00014597944751183514,
+      "loss": 0.18166859149932862,
+      "step": 14570
+    },
+    {
+      "epoch": 0.05512393569117736,
+      "grad_norm": 0.42283931374549866,
+      "learning_rate": 0.0001459078989108066,
+      "loss": 0.2021012544631958,
+      "step": 14580
+    },
+    {
+      "epoch": 0.05541621530777957,
+      "grad_norm": 0.5170047879219055,
+      "learning_rate": 0.0001458363205206409,
+      "loss": 0.1975739121437073,
+      "step": 14590
+    },
+    {
+      "epoch": 0.055708494924381784,
+      "grad_norm": 0.6347094774246216,
+      "learning_rate": 0.0001457647123877844,
+      "loss": 0.15992212295532227,
+      "step": 14600
+    },
+    {
+      "epoch": 0.056000774540984,
+      "grad_norm": 0.4478125274181366,
+      "learning_rate": 0.00014569307455870292,
+      "loss": 0.2281398057937622,
+      "step": 14610
+    },
+    {
+      "epoch": 0.05629305415758621,
+      "grad_norm": 0.4383460283279419,
+      "learning_rate": 0.0001456214070798814,
+      "loss": 0.1910434365272522,
+      "step": 14620
+    },
+    {
+      "epoch": 0.05658533377418842,
+      "grad_norm": 0.3631839454174042,
+      "learning_rate": 0.0001455497099978242,
+      "loss": 0.16486721038818358,
+      "step": 14630
+    },
+    {
+      "epoch": 0.056877613390790636,
+      "grad_norm": 0.492399662733078,
+      "learning_rate": 0.00014547798335905467,
+      "loss": 0.18056585788726806,
+      "step": 14640
+    },
+    {
+      "epoch": 0.05716989300739285,
+      "grad_norm": 0.5111512541770935,
+      "learning_rate": 0.00014540622721011546,
+      "loss": 0.1827410101890564,
+      "step": 14650
+    },
+    {
+      "epoch": 0.05746217262399506,
+      "grad_norm": 0.4539899528026581,
+      "learning_rate": 0.0001453344415975684,
+      "loss": 0.21658852100372314,
+      "step": 14660
+    },
+    {
+      "epoch": 0.057754452240597276,
+      "grad_norm": 0.5448858141899109,
+      "learning_rate": 0.0001452626265679944,
+      "loss": 0.2032935619354248,
+      "step": 14670
+    },
+    {
+      "epoch": 0.05804673185719949,
+      "grad_norm": 0.4533732235431671,
+      "learning_rate": 0.00014519078216799342,
+      "loss": 0.2064185619354248,
+      "step": 14680
+    },
+    {
+      "epoch": 0.0583390114738017,
+      "grad_norm": 0.4278698265552521,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 0.1551392674446106,
+      "step": 14690
+    },
+    {
+      "epoch": 0.058631291090403916,
+      "grad_norm": 0.4973037838935852,
+      "learning_rate": 0.00014504700544320576,
+      "loss": 0.19894620180130004,
+      "step": 14700
+    },
+    {
+      "epoch": 0.05892357070700612,
+      "grad_norm": 0.4972725510597229,
+      "learning_rate": 0.00014497507321171424,
+      "loss": 0.2594135046005249,
+      "step": 14710
+    },
+    {
+      "epoch": 0.059215850323608335,
+      "grad_norm": 0.39482206106185913,
+      "learning_rate": 0.00014490311179638602,
+      "loss": 0.19961639642715454,
+      "step": 14720
+    },
+    {
+      "epoch": 0.05950812994021055,
+      "grad_norm": 0.46872907876968384,
+      "learning_rate": 0.00014483112124391605,
+      "loss": 0.2166891574859619,
+      "step": 14730
+    },
+    {
+      "epoch": 0.05980040955681276,
+      "grad_norm": 0.35188496112823486,
+      "learning_rate": 0.00014475910160101824,
+      "loss": 0.17436405420303344,
+      "step": 14740
+    },
+    {
+      "epoch": 0.060092689173414975,
+      "grad_norm": 0.5011220574378967,
+      "learning_rate": 0.00014468705291442534,
+      "loss": 0.1993683934211731,
+      "step": 14750
+    },
+    {
+      "epoch": 0.06038496879001719,
+      "grad_norm": 0.655901312828064,
+      "learning_rate": 0.00014461497523088896,
+      "loss": 0.17413907051086425,
+      "step": 14760
+    },
+    {
+      "epoch": 0.0606772484066194,
+      "grad_norm": 0.5140685439109802,
+      "learning_rate": 0.0001445428685971795,
+      "loss": 0.14213814735412597,
+      "step": 14770
+    },
+    {
+      "epoch": 0.060969528023221614,
+      "grad_norm": 0.43015050888061523,
+      "learning_rate": 0.00014447073306008624,
+      "loss": 0.21556131839752196,
+      "step": 14780
+    },
+    {
+      "epoch": 0.06126180763982383,
+      "grad_norm": 0.4331928789615631,
+      "learning_rate": 0.0001443985686664171,
+      "loss": 0.14444485902786255,
+      "step": 14790
+    },
+    {
+      "epoch": 0.06155408725642604,
+      "grad_norm": 0.51627117395401,
+      "learning_rate": 0.00014432637546299879,
+      "loss": 0.19857748746871948,
+      "step": 14800
+    },
+    {
+      "epoch": 0.061846366873028254,
+      "grad_norm": 0.39337605237960815,
+      "learning_rate": 0.00014425415349667668,
+      "loss": 0.2117015838623047,
+      "step": 14810
+    },
+    {
+      "epoch": 0.06213864648963047,
+      "grad_norm": 0.5243445634841919,
+      "learning_rate": 0.00014418190281431482,
+      "loss": 0.1673484206199646,
+      "step": 14820
+    },
+    {
+      "epoch": 0.06243092610623268,
+      "grad_norm": 0.41369903087615967,
+      "learning_rate": 0.00014410962346279592,
+      "loss": 0.28499248027801516,
+      "step": 14830
+    },
+    {
+      "epoch": 0.0627232057228349,
+      "grad_norm": 0.4046480357646942,
+      "learning_rate": 0.0001440373154890212,
+      "loss": 0.24999215602874755,
+      "step": 14840
+    },
+    {
+      "epoch": 0.0630154853394371,
+      "grad_norm": 0.4554937183856964,
+      "learning_rate": 0.00014396497893991064,
+      "loss": 0.2884443998336792,
+      "step": 14850
+    },
+    {
+      "epoch": 0.06330776495603932,
+      "grad_norm": 0.3988826870918274,
+      "learning_rate": 0.00014389261386240252,
+      "loss": 0.25861554145812987,
+      "step": 14860
+    },
+    {
+      "epoch": 0.06360004457264153,
+      "grad_norm": 0.3818569779396057,
+      "learning_rate": 0.0001438202203034538,
+      "loss": 0.24147512912750244,
+      "step": 14870
+    },
+    {
+      "epoch": 0.06389232418924375,
+      "grad_norm": 0.39010724425315857,
+      "learning_rate": 0.00014374779831003985,
+      "loss": 0.22250096797943114,
+      "step": 14880
+    },
+    {
+      "epoch": 0.06418460380584595,
+      "grad_norm": 0.3751218616962433,
+      "learning_rate": 0.00014367534792915457,
+      "loss": 0.228631854057312,
+      "step": 14890
+    },
+    {
+      "epoch": 0.06447688342244817,
+      "grad_norm": 0.46549203991889954,
+      "learning_rate": 0.0001436028692078102,
+      "loss": 0.24026453495025635,
+      "step": 14900
+    },
+    {
+      "epoch": 0.06476916303905038,
+      "grad_norm": 0.3721192181110382,
+      "learning_rate": 0.00014353036219303737,
+      "loss": 0.22989704608917236,
+      "step": 14910
+    },
+    {
+      "epoch": 0.0650614426556526,
+      "grad_norm": 0.41028088331222534,
+      "learning_rate": 0.00014345782693188516,
+      "loss": 0.23178772926330565,
+      "step": 14920
+    },
+    {
+      "epoch": 0.0653537222722548,
+      "grad_norm": 0.40676048398017883,
+      "learning_rate": 0.0001433852634714208,
+      "loss": 0.22049374580383302,
+      "step": 14930
+    },
+    {
+      "epoch": 0.06564600188885703,
+      "grad_norm": 0.3672485053539276,
+      "learning_rate": 0.00014331267185873004,
+      "loss": 0.22965245246887206,
+      "step": 14940
+    },
+    {
+      "epoch": 0.06593828150545923,
+      "grad_norm": 0.45485424995422363,
+      "learning_rate": 0.00014324005214091676,
+      "loss": 0.23864634037017823,
+      "step": 14950
+    },
+    {
+      "epoch": 0.06623056112206145,
+      "grad_norm": 0.5851138234138489,
+      "learning_rate": 0.0001431674043651031,
+      "loss": 0.20639584064483643,
+      "step": 14960
+    },
+    {
+      "epoch": 0.06652284073866366,
+      "grad_norm": 0.4605776369571686,
+      "learning_rate": 0.00014309472857842938,
+      "loss": 0.2729495048522949,
+      "step": 14970
+    },
+    {
+      "epoch": 0.06681512035526588,
+      "grad_norm": 0.38282501697540283,
+      "learning_rate": 0.00014302202482805416,
+      "loss": 0.22896134853363037,
+      "step": 14980
+    },
+    {
+      "epoch": 0.06710739997186808,
+      "grad_norm": 0.4912174642086029,
+      "learning_rate": 0.00014294929316115411,
+      "loss": 0.2096320390701294,
+      "step": 14990
+    },
+    {
+      "epoch": 0.0673996795884703,
+      "grad_norm": 0.4156826436519623,
+      "learning_rate": 0.00014287653362492407,
+      "loss": 0.2500204563140869,
+      "step": 15000
+    },
+    {
+      "epoch": 0.0002922796166022129,
+      "grad_norm": 0.42408451437950134,
+      "learning_rate": 0.0001428037462665768,
+      "loss": 0.19309383630752563,
+      "step": 15010
+    },
+    {
+      "epoch": 0.0005845592332044258,
+      "grad_norm": 0.5625495910644531,
+      "learning_rate": 0.00014273093113334332,
+      "loss": 0.16390117406845092,
+      "step": 15020
+    },
+    {
+      "epoch": 0.0008768388498066388,
+      "grad_norm": 0.4544651210308075,
+      "learning_rate": 0.00014265808827247257,
+      "loss": 0.13237448930740356,
+      "step": 15030
+    },
+    {
+      "epoch": 0.0011691184664088517,
+      "grad_norm": 0.501217246055603,
+      "learning_rate": 0.00014258521773123143,
+      "loss": 0.1310074210166931,
+      "step": 15040
+    },
+    {
+      "epoch": 0.0014613980830110646,
+      "grad_norm": 0.40772339701652527,
+      "learning_rate": 0.00014251231955690488,
+      "loss": 0.11214158535003663,
+      "step": 15050
+    },
+    {
+      "epoch": 0.0017536776996132776,
+      "grad_norm": 0.329706609249115,
+      "learning_rate": 0.0001424393937967957,
+      "loss": 0.13458265066146852,
+      "step": 15060
+    },
+    {
+      "epoch": 0.0020459573162154906,
+      "grad_norm": 0.3190118074417114,
+      "learning_rate": 0.00014236644049822465,
+      "loss": 0.14507204294204712,
+      "step": 15070
+    },
+    {
+      "epoch": 0.0023382369328177033,
+      "grad_norm": 0.5345238447189331,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.1323810338973999,
+      "step": 15080
+    },
+    {
+      "epoch": 0.002630516549419916,
+      "grad_norm": 0.5074023604393005,
+      "learning_rate": 0.0001422204514750692,
+      "loss": 0.13561161756515502,
+      "step": 15090
+    },
+    {
+      "epoch": 0.0029227961660221293,
+      "grad_norm": 0.5453847646713257,
+      "learning_rate": 0.00014214741584521546,
+      "loss": 0.13266334533691407,
+      "step": 15100
+    },
+    {
+      "epoch": 0.003215075782624342,
+      "grad_norm": 0.495571494102478,
+      "learning_rate": 0.00014207435286636114,
+      "loss": 0.14747509956359864,
+      "step": 15110
+    },
+    {
+      "epoch": 0.0035073553992265552,
+      "grad_norm": 0.5273986458778381,
+      "learning_rate": 0.00014200126258591612,
+      "loss": 0.1343027114868164,
+      "step": 15120
+    },
+    {
+      "epoch": 0.003799635015828768,
+      "grad_norm": 0.45895904302597046,
+      "learning_rate": 0.00014192814505130773,
+      "loss": 0.12265520095825196,
+      "step": 15130
+    },
+    {
+      "epoch": 0.004091914632430981,
+      "grad_norm": 0.36669641733169556,
+      "learning_rate": 0.00014185500030998126,
+      "loss": 0.1610312581062317,
+      "step": 15140
+    },
+    {
+      "epoch": 0.0043841942490331935,
+      "grad_norm": 0.464180588722229,
+      "learning_rate": 0.00014178182840939949,
+      "loss": 0.11420994997024536,
+      "step": 15150
+    },
+    {
+      "epoch": 0.004676473865635407,
+      "grad_norm": 0.4995633363723755,
+      "learning_rate": 0.00014170862939704289,
+      "loss": 0.10678664445877076,
+      "step": 15160
+    },
+    {
+      "epoch": 0.00496875348223762,
+      "grad_norm": 0.33280885219573975,
+      "learning_rate": 0.00014163540332040947,
+      "loss": 0.1105881929397583,
+      "step": 15170
+    },
+    {
+      "epoch": 0.005261033098839832,
+      "grad_norm": 0.44710686802864075,
+      "learning_rate": 0.0001415621502270149,
+      "loss": 0.11454238891601562,
+      "step": 15180
+    },
+    {
+      "epoch": 0.005553312715442045,
+      "grad_norm": 0.47306498885154724,
+      "learning_rate": 0.0001414888701643923,
+      "loss": 0.11980700492858887,
+      "step": 15190
+    },
+    {
+      "epoch": 0.0058455923320442586,
+      "grad_norm": 0.6128643155097961,
+      "learning_rate": 0.00014141556318009226,
+      "loss": 0.13866671323776245,
+      "step": 15200
+    },
+    {
+      "epoch": 0.006137871948646472,
+      "grad_norm": 0.38859984278678894,
+      "learning_rate": 0.0001413422293216829,
+      "loss": 0.12619128227233886,
+      "step": 15210
+    },
+    {
+      "epoch": 0.006430151565248684,
+      "grad_norm": 0.3652571141719818,
+      "learning_rate": 0.0001412688686367498,
+      "loss": 0.12176882028579712,
+      "step": 15220
+    },
+    {
+      "epoch": 0.006722431181850897,
+      "grad_norm": 0.4183637201786041,
+      "learning_rate": 0.0001411954811728959,
+      "loss": 0.12742173671722412,
+      "step": 15230
+    },
+    {
+      "epoch": 0.0070147107984531104,
+      "grad_norm": 0.3554188907146454,
+      "learning_rate": 0.0001411220669777415,
+      "loss": 0.1413712978363037,
+      "step": 15240
+    },
+    {
+      "epoch": 0.007306990415055323,
+      "grad_norm": 0.36271801590919495,
+      "learning_rate": 0.00014104862609892432,
+      "loss": 0.12233136892318726,
+      "step": 15250
+    },
+    {
+      "epoch": 0.007599270031657536,
+      "grad_norm": 0.3980635404586792,
+      "learning_rate": 0.0001409751585840993,
+      "loss": 0.08514137864112854,
+      "step": 15260
+    },
+    {
+      "epoch": 0.00789154964825975,
+      "grad_norm": 0.6249545812606812,
+      "learning_rate": 0.00014090166448093872,
+      "loss": 0.133427894115448,
+      "step": 15270
+    },
+    {
+      "epoch": 0.008183829264861962,
+      "grad_norm": 0.40774139761924744,
+      "learning_rate": 0.00014082814383713213,
+      "loss": 0.10439600944519042,
+      "step": 15280
+    },
+    {
+      "epoch": 0.008476108881464176,
+      "grad_norm": 0.6646661162376404,
+      "learning_rate": 0.0001407545967003863,
+      "loss": 0.1308104872703552,
+      "step": 15290
+    },
+    {
+      "epoch": 0.008768388498066387,
+      "grad_norm": 0.6027724742889404,
+      "learning_rate": 0.0001406810231184251,
+      "loss": 0.11796362400054931,
+      "step": 15300
+    },
+    {
+      "epoch": 0.0090606681146686,
+      "grad_norm": 0.32626479864120483,
+      "learning_rate": 0.0001406074231389897,
+      "loss": 0.11564600467681885,
+      "step": 15310
+    },
+    {
+      "epoch": 0.009352947731270813,
+      "grad_norm": 0.4703027904033661,
+      "learning_rate": 0.00014053379680983828,
+      "loss": 0.12225807905197143,
+      "step": 15320
+    },
+    {
+      "epoch": 0.009645227347873027,
+      "grad_norm": 0.48698291182518005,
+      "learning_rate": 0.00014046014417874617,
+      "loss": 0.08796515464782714,
+      "step": 15330
+    },
+    {
+      "epoch": 0.00993750696447524,
+      "grad_norm": 0.45816943049430847,
+      "learning_rate": 0.00014038646529350579,
+      "loss": 0.13419857025146484,
+      "step": 15340
+    },
+    {
+      "epoch": 0.010229786581077453,
+      "grad_norm": 0.47064435482025146,
+      "learning_rate": 0.00014031276020192652,
+      "loss": 0.11092145442962646,
+      "step": 15350
+    },
+    {
+      "epoch": 0.010522066197679664,
+      "grad_norm": 0.5047792196273804,
+      "learning_rate": 0.00014023902895183484,
+      "loss": 0.09699101448059082,
+      "step": 15360
+    },
+    {
+      "epoch": 0.010814345814281878,
+      "grad_norm": 0.38016995787620544,
+      "learning_rate": 0.00014016527159107416,
+      "loss": 0.11821193695068359,
+      "step": 15370
+    },
+    {
+      "epoch": 0.01110662543088409,
+      "grad_norm": 0.5109589695930481,
+      "learning_rate": 0.00014009148816750478,
+      "loss": 0.12638478279113768,
+      "step": 15380
+    },
+    {
+      "epoch": 0.011398905047486304,
+      "grad_norm": 0.3145149350166321,
+      "learning_rate": 0.00014001767872900403,
+      "loss": 0.11602962017059326,
+      "step": 15390
+    },
+    {
+      "epoch": 0.011691184664088517,
+      "grad_norm": 0.3991243243217468,
+      "learning_rate": 0.000139943843323466,
+      "loss": 0.1362460732460022,
+      "step": 15400
+    },
+    {
+      "epoch": 0.01198346428069073,
+      "grad_norm": 0.42630577087402344,
+      "learning_rate": 0.00013986998199880168,
+      "loss": 0.10832134485244752,
+      "step": 15410
+    },
+    {
+      "epoch": 0.012275743897292943,
+      "grad_norm": 0.4114335775375366,
+      "learning_rate": 0.00013979609480293894,
+      "loss": 0.0968300998210907,
+      "step": 15420
+    },
+    {
+      "epoch": 0.012568023513895155,
+      "grad_norm": 0.439420610666275,
+      "learning_rate": 0.00013972218178382232,
+      "loss": 0.12040265798568725,
+      "step": 15430
+    },
+    {
+      "epoch": 0.012860303130497368,
+      "grad_norm": 0.37576794624328613,
+      "learning_rate": 0.0001396482429894132,
+      "loss": 0.0988838791847229,
+      "step": 15440
+    },
+    {
+      "epoch": 0.013152582747099581,
+      "grad_norm": 0.47005695104599,
+      "learning_rate": 0.0001395742784676897,
+      "loss": 0.1276581883430481,
+      "step": 15450
+    },
+    {
+      "epoch": 0.013444862363701795,
+      "grad_norm": 0.5016588568687439,
+      "learning_rate": 0.0001395002882666465,
+      "loss": 0.12103404998779296,
+      "step": 15460
+    },
+    {
+      "epoch": 0.013737141980304008,
+      "grad_norm": 0.36553123593330383,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 0.12047480344772339,
+      "step": 15470
+    },
+    {
+      "epoch": 0.014029421596906221,
+      "grad_norm": 0.4887537360191345,
+      "learning_rate": 0.0001393522310186636,
+      "loss": 0.10463247299194336,
+      "step": 15480
+    },
+    {
+      "epoch": 0.014321701213508432,
+      "grad_norm": 0.5245959162712097,
+      "learning_rate": 0.00013927816406779657,
+      "loss": 0.08904374241828919,
+      "step": 15490
+    },
+    {
+      "epoch": 0.014613980830110646,
+      "grad_norm": 0.6415546536445618,
+      "learning_rate": 0.00013920407162975538,
+      "loss": 0.10240782499313354,
+      "step": 15500
+    },
+    {
+      "epoch": 0.014906260446712859,
+      "grad_norm": 0.3800129294395447,
+      "learning_rate": 0.0001391299537526177,
+      "loss": 0.11430275440216064,
+      "step": 15510
+    },
+    {
+      "epoch": 0.015198540063315072,
+      "grad_norm": 0.4069846570491791,
+      "learning_rate": 0.00013905581048447786,
+      "loss": 0.13011096715927123,
+      "step": 15520
+    },
+    {
+      "epoch": 0.015490819679917285,
+      "grad_norm": 0.3222843408584595,
+      "learning_rate": 0.00013898164187344664,
+      "loss": 0.1252989649772644,
+      "step": 15530
+    },
+    {
+      "epoch": 0.0157830992965195,
+      "grad_norm": 0.5734412670135498,
+      "learning_rate": 0.00013890744796765122,
+      "loss": 0.12262144088745117,
+      "step": 15540
+    },
+    {
+      "epoch": 0.01607537891312171,
+      "grad_norm": 0.408058762550354,
+      "learning_rate": 0.00013883322881523522,
+      "loss": 0.1195600152015686,
+      "step": 15550
+    },
+    {
+      "epoch": 0.016367658529723925,
+      "grad_norm": 0.40590909123420715,
+      "learning_rate": 0.00013875898446435867,
+      "loss": 0.1388644576072693,
+      "step": 15560
+    },
+    {
+      "epoch": 0.016659938146326138,
+      "grad_norm": 0.319157212972641,
+      "learning_rate": 0.00013868471496319786,
+      "loss": 0.11914747953414917,
+      "step": 15570
+    },
+    {
+      "epoch": 0.01695221776292835,
+      "grad_norm": 0.444359689950943,
+      "learning_rate": 0.00013861042035994553,
+      "loss": 0.13284153938293458,
+      "step": 15580
+    },
+    {
+      "epoch": 0.01724449737953056,
+      "grad_norm": 0.4568331837654114,
+      "learning_rate": 0.0001385361007028106,
+      "loss": 0.13507847785949706,
+      "step": 15590
+    },
+    {
+      "epoch": 0.017536776996132774,
+      "grad_norm": 0.4915296137332916,
+      "learning_rate": 0.0001384617560400183,
+      "loss": 0.12999931573867798,
+      "step": 15600
+    },
+    {
+      "epoch": 0.017829056612734987,
+      "grad_norm": 0.47444307804107666,
+      "learning_rate": 0.00013838738641981012,
+      "loss": 0.1029810905456543,
+      "step": 15610
+    },
+    {
+      "epoch": 0.0181213362293372,
+      "grad_norm": 0.49894392490386963,
+      "learning_rate": 0.00013831299189044364,
+      "loss": 0.12314598560333252,
+      "step": 15620
+    },
+    {
+      "epoch": 0.018413615845939414,
+      "grad_norm": 0.5075197815895081,
+      "learning_rate": 0.00013823857250019265,
+      "loss": 0.09839518666267395,
+      "step": 15630
+    },
+    {
+      "epoch": 0.018705895462541627,
+      "grad_norm": 0.3502114713191986,
+      "learning_rate": 0.00013816412829734714,
+      "loss": 0.12280040979385376,
+      "step": 15640
+    },
+    {
+      "epoch": 0.01899817507914384,
+      "grad_norm": 0.5903626084327698,
+      "learning_rate": 0.0001380896593302131,
+      "loss": 0.12850145101547242,
+      "step": 15650
+    },
+    {
+      "epoch": 0.019290454695746053,
+      "grad_norm": 0.6615191102027893,
+      "learning_rate": 0.0001380151656471127,
+      "loss": 0.13114123344421386,
+      "step": 15660
+    },
+    {
+      "epoch": 0.019582734312348266,
+      "grad_norm": 0.5432319641113281,
+      "learning_rate": 0.000137940647296384,
+      "loss": 0.13208657503128052,
+      "step": 15670
+    },
+    {
+      "epoch": 0.01987501392895048,
+      "grad_norm": 0.35975706577301025,
+      "learning_rate": 0.0001378661043263812,
+      "loss": 0.10644789934158325,
+      "step": 15680
+    },
+    {
+      "epoch": 0.020167293545552693,
+      "grad_norm": 0.4836406111717224,
+      "learning_rate": 0.0001377915367854744,
+      "loss": 0.12172921895980834,
+      "step": 15690
+    },
+    {
+      "epoch": 0.020459573162154906,
+      "grad_norm": 0.27832433581352234,
+      "learning_rate": 0.0001377169447220497,
+      "loss": 0.0808457612991333,
+      "step": 15700
+    },
+    {
+      "epoch": 0.02075185277875712,
+      "grad_norm": 0.4474148750305176,
+      "learning_rate": 0.000137642328184509,
+      "loss": 0.11133733987808228,
+      "step": 15710
+    },
+    {
+      "epoch": 0.02104413239535933,
+      "grad_norm": 0.3808692395687103,
+      "learning_rate": 0.00013756768722127024,
+      "loss": 0.10502228736877442,
+      "step": 15720
+    },
+    {
+      "epoch": 0.021336412011961542,
+      "grad_norm": 0.35518625378608704,
+      "learning_rate": 0.00013749302188076717,
+      "loss": 0.08781594634056092,
+      "step": 15730
+    },
+    {
+      "epoch": 0.021628691628563755,
+      "grad_norm": 0.5510978102684021,
+      "learning_rate": 0.00013741833221144918,
+      "loss": 0.13099899291992187,
+      "step": 15740
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 40000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.678859518001283e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}