diff --git "a/checkpoint-38001/trainer_state.json" "b/checkpoint-38001/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-38001/trainer_state.json"
@@ -0,0 +1,26634 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0000292279616603,
+  "eval_steps": 500,
+  "global_step": 38001,
+  "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
+    },
+    {
+      "epoch": 0.02192097124516597,
+      "grad_norm": 0.4186568856239319,
+      "learning_rate": 0.0001373436182617817,
+      "loss": 0.1427019238471985,
+      "step": 15750
+    },
+    {
+      "epoch": 0.02221325086176818,
+      "grad_norm": 0.3546358644962311,
+      "learning_rate": 0.00013726888008024575,
+      "loss": 0.1190258264541626,
+      "step": 15760
+    },
+    {
+      "epoch": 0.022505530478370395,
+      "grad_norm": 0.42977219820022583,
+      "learning_rate": 0.00013719411771533814,
+      "loss": 0.11482198238372802,
+      "step": 15770
+    },
+    {
+      "epoch": 0.022797810094972608,
+      "grad_norm": 0.4496024250984192,
+      "learning_rate": 0.00013711933121557135,
+      "loss": 0.09535525441169738,
+      "step": 15780
+    },
+    {
+      "epoch": 0.02309008971157482,
+      "grad_norm": 0.5403960943222046,
+      "learning_rate": 0.00013704452062947357,
+      "loss": 0.12732412815093994,
+      "step": 15790
+    },
+    {
+      "epoch": 0.023382369328177034,
+      "grad_norm": 0.37054574489593506,
+      "learning_rate": 0.0001369696860055885,
+      "loss": 0.14149919748306275,
+      "step": 15800
+    },
+    {
+      "epoch": 0.023674648944779247,
+      "grad_norm": 0.5391687750816345,
+      "learning_rate": 0.00013689482739247556,
+      "loss": 0.12390230894088745,
+      "step": 15810
+    },
+    {
+      "epoch": 0.02396692856138146,
+      "grad_norm": 0.5415944457054138,
+      "learning_rate": 0.0001368199448387097,
+      "loss": 0.13882920742034913,
+      "step": 15820
+    },
+    {
+      "epoch": 0.024259208177983674,
+      "grad_norm": 0.26245346665382385,
+      "learning_rate": 0.00013674503839288137,
+      "loss": 0.12357894182205201,
+      "step": 15830
+    },
+    {
+      "epoch": 0.024551487794585887,
+      "grad_norm": 0.46969351172447205,
+      "learning_rate": 0.00013667010810359658,
+      "loss": 0.12251794338226318,
+      "step": 15840
+    },
+    {
+      "epoch": 0.024843767411188097,
+      "grad_norm": 0.8196412920951843,
+      "learning_rate": 0.00013659515401947677,
+      "loss": 0.1509449601173401,
+      "step": 15850
+    },
+    {
+      "epoch": 0.02513604702779031,
+      "grad_norm": 0.37927013635635376,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.1330603241920471,
+      "step": 15860
+    },
+    {
+      "epoch": 0.025428326644392523,
+      "grad_norm": 0.5440946817398071,
+      "learning_rate": 0.000136445174661295,
+      "loss": 0.12676852941513062,
+      "step": 15870
+    },
+    {
+      "epoch": 0.025720606260994736,
+      "grad_norm": 0.5239357352256775,
+      "learning_rate": 0.00013637014948455302,
+      "loss": 0.12703244686126708,
+      "step": 15880
+    },
+    {
+      "epoch": 0.02601288587759695,
+      "grad_norm": 0.4245850741863251,
+      "learning_rate": 0.00013629510070761595,
+      "loss": 0.12266845703125,
+      "step": 15890
+    },
+    {
+      "epoch": 0.026305165494199163,
+      "grad_norm": 0.42964017391204834,
+      "learning_rate": 0.00013622002837918205,
+      "loss": 0.14235602617263793,
+      "step": 15900
+    },
+    {
+      "epoch": 0.026597445110801376,
+      "grad_norm": 0.42633870244026184,
+      "learning_rate": 0.00013614493254796497,
+      "loss": 0.10958071947097778,
+      "step": 15910
+    },
+    {
+      "epoch": 0.02688972472740359,
+      "grad_norm": 0.47604241967201233,
+      "learning_rate": 0.0001360698132626936,
+      "loss": 0.12668929100036622,
+      "step": 15920
+    },
+    {
+      "epoch": 0.027182004344005802,
+      "grad_norm": 0.5336122512817383,
+      "learning_rate": 0.00013599467057211198,
+      "loss": 0.13309048414230346,
+      "step": 15930
+    },
+    {
+      "epoch": 0.027474283960608015,
+      "grad_norm": 0.5357875823974609,
+      "learning_rate": 0.00013591950452497944,
+      "loss": 0.13534616231918334,
+      "step": 15940
+    },
+    {
+      "epoch": 0.02776656357721023,
+      "grad_norm": 0.4015578329563141,
+      "learning_rate": 0.0001358443151700704,
+      "loss": 0.13624606132507325,
+      "step": 15950
+    },
+    {
+      "epoch": 0.028058843193812442,
+      "grad_norm": 0.48634153604507446,
+      "learning_rate": 0.00013576910255617438,
+      "loss": 0.12898943424224854,
+      "step": 15960
+    },
+    {
+      "epoch": 0.028351122810414655,
+      "grad_norm": 0.3661693036556244,
+      "learning_rate": 0.00013569386673209602,
+      "loss": 0.12810779809951783,
+      "step": 15970
+    },
+    {
+      "epoch": 0.028643402427016865,
+      "grad_norm": 0.480487585067749,
+      "learning_rate": 0.0001356186077466551,
+      "loss": 0.1408245086669922,
+      "step": 15980
+    },
+    {
+      "epoch": 0.028935682043619078,
+      "grad_norm": 0.597980260848999,
+      "learning_rate": 0.0001355433256486863,
+      "loss": 0.11928675174713135,
+      "step": 15990
+    },
+    {
+      "epoch": 0.02922796166022129,
+      "grad_norm": 0.3814627230167389,
+      "learning_rate": 0.00013546802048703942,
+      "loss": 0.12503902912139891,
+      "step": 16000
+    },
+    {
+      "epoch": 0.029520241276823504,
+      "grad_norm": 0.49009934067726135,
+      "learning_rate": 0.00013539269231057908,
+      "loss": 0.12741589546203613,
+      "step": 16010
+    },
+    {
+      "epoch": 0.029812520893425717,
+      "grad_norm": 0.4525687098503113,
+      "learning_rate": 0.00013531734116818494,
+      "loss": 0.1689910411834717,
+      "step": 16020
+    },
+    {
+      "epoch": 0.03010480051002793,
+      "grad_norm": 0.36311495304107666,
+      "learning_rate": 0.00013524196710875156,
+      "loss": 0.14648044109344482,
+      "step": 16030
+    },
+    {
+      "epoch": 0.030397080126630144,
+      "grad_norm": 0.4981651306152344,
+      "learning_rate": 0.00013516657018118836,
+      "loss": 0.15659793615341186,
+      "step": 16040
+    },
+    {
+      "epoch": 0.030689359743232357,
+      "grad_norm": 0.3951387107372284,
+      "learning_rate": 0.00013509115043441953,
+      "loss": 0.11442358493804931,
+      "step": 16050
+    },
+    {
+      "epoch": 0.03098163935983457,
+      "grad_norm": 0.6192485094070435,
+      "learning_rate": 0.0001350157079173842,
+      "loss": 0.1294487237930298,
+      "step": 16060
+    },
+    {
+      "epoch": 0.03127391897643678,
+      "grad_norm": 0.6486902236938477,
+      "learning_rate": 0.00013494024267903613,
+      "loss": 0.13780689239501953,
+      "step": 16070
+    },
+    {
+      "epoch": 0.031566198593039,
+      "grad_norm": 0.4512653350830078,
+      "learning_rate": 0.00013486475476834394,
+      "loss": 0.1276625871658325,
+      "step": 16080
+    },
+    {
+      "epoch": 0.03185847820964121,
+      "grad_norm": 0.4696688950061798,
+      "learning_rate": 0.0001347892442342909,
+      "loss": 0.09121795296669007,
+      "step": 16090
+    },
+    {
+      "epoch": 0.03215075782624342,
+      "grad_norm": 0.49270835518836975,
+      "learning_rate": 0.00013471371112587502,
+      "loss": 0.11269618272781372,
+      "step": 16100
+    },
+    {
+      "epoch": 0.032443037442845636,
+      "grad_norm": 0.43599745631217957,
+      "learning_rate": 0.00013463815549210882,
+      "loss": 0.1315122961997986,
+      "step": 16110
+    },
+    {
+      "epoch": 0.03273531705944785,
+      "grad_norm": 0.38877591490745544,
+      "learning_rate": 0.00013456257738201957,
+      "loss": 0.12224313020706176,
+      "step": 16120
+    },
+    {
+      "epoch": 0.03302759667605006,
+      "grad_norm": 0.5652080178260803,
+      "learning_rate": 0.00013448697684464913,
+      "loss": 0.11750903129577636,
+      "step": 16130
+    },
+    {
+      "epoch": 0.033319876292652276,
+      "grad_norm": 0.492559015750885,
+      "learning_rate": 0.00013441135392905384,
+      "loss": 0.124018394947052,
+      "step": 16140
+    },
+    {
+      "epoch": 0.03361215590925449,
+      "grad_norm": 0.5575881600379944,
+      "learning_rate": 0.00013433570868430457,
+      "loss": 0.14741722345352173,
+      "step": 16150
+    },
+    {
+      "epoch": 0.0339044355258567,
+      "grad_norm": 0.4042343199253082,
+      "learning_rate": 0.0001342600411594867,
+      "loss": 0.1438590407371521,
+      "step": 16160
+    },
+    {
+      "epoch": 0.034196715142458915,
+      "grad_norm": 0.5482458472251892,
+      "learning_rate": 0.0001341843514037001,
+      "loss": 0.10708463191986084,
+      "step": 16170
+    },
+    {
+      "epoch": 0.03448899475906112,
+      "grad_norm": 0.4707915484905243,
+      "learning_rate": 0.00013410863946605901,
+      "loss": 0.12857834100723267,
+      "step": 16180
+    },
+    {
+      "epoch": 0.034781274375663335,
+      "grad_norm": 0.6200113296508789,
+      "learning_rate": 0.0001340329053956921,
+      "loss": 0.14773281812667846,
+      "step": 16190
+    },
+    {
+      "epoch": 0.03507355399226555,
+      "grad_norm": 0.4418794512748718,
+      "learning_rate": 0.00013395714924174231,
+      "loss": 0.11194784641265869,
+      "step": 16200
+    },
+    {
+      "epoch": 0.03536583360886776,
+      "grad_norm": 0.5243352055549622,
+      "learning_rate": 0.0001338813710533671,
+      "loss": 0.127723228931427,
+      "step": 16210
+    },
+    {
+      "epoch": 0.035658113225469974,
+      "grad_norm": 0.5773442387580872,
+      "learning_rate": 0.00013380557087973804,
+      "loss": 0.1729312300682068,
+      "step": 16220
+    },
+    {
+      "epoch": 0.03595039284207219,
+      "grad_norm": 0.37578070163726807,
+      "learning_rate": 0.00013372974877004103,
+      "loss": 0.12196640968322754,
+      "step": 16230
+    },
+    {
+      "epoch": 0.0362426724586744,
+      "grad_norm": 0.4779168367385864,
+      "learning_rate": 0.00013365390477347626,
+      "loss": 0.130403470993042,
+      "step": 16240
+    },
+    {
+      "epoch": 0.036534952075276614,
+      "grad_norm": 0.8030831813812256,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 0.093166184425354,
+      "step": 16250
+    },
+    {
+      "epoch": 0.03682723169187883,
+      "grad_norm": 0.3487117886543274,
+      "learning_rate": 0.00013350215131661494,
+      "loss": 0.1249579906463623,
+      "step": 16260
+    },
+    {
+      "epoch": 0.03711951130848104,
+      "grad_norm": 0.5379220247268677,
+      "learning_rate": 0.00013342624195478958,
+      "loss": 0.15377799272537232,
+      "step": 16270
+    },
+    {
+      "epoch": 0.03741179092508325,
+      "grad_norm": 0.5205707550048828,
+      "learning_rate": 0.00013335031090303874,
+      "loss": 0.13549846410751343,
+      "step": 16280
+    },
+    {
+      "epoch": 0.03770407054168547,
+      "grad_norm": 0.35249215364456177,
+      "learning_rate": 0.00013327435821063318,
+      "loss": 0.1513085126876831,
+      "step": 16290
+    },
+    {
+      "epoch": 0.03799635015828768,
+      "grad_norm": 0.4560491144657135,
+      "learning_rate": 0.00013319838392685784,
+      "loss": 0.13737787008285524,
+      "step": 16300
+    },
+    {
+      "epoch": 0.03828862977488989,
+      "grad_norm": 0.5328561067581177,
+      "learning_rate": 0.00013312238810101162,
+      "loss": 0.13795344829559325,
+      "step": 16310
+    },
+    {
+      "epoch": 0.038580909391492106,
+      "grad_norm": 0.6700594425201416,
+      "learning_rate": 0.00013304637078240735,
+      "loss": 0.1380225658416748,
+      "step": 16320
+    },
+    {
+      "epoch": 0.03887318900809432,
+      "grad_norm": 0.49719080328941345,
+      "learning_rate": 0.00013297033202037186,
+      "loss": 0.13025355339050293,
+      "step": 16330
+    },
+    {
+      "epoch": 0.03916546862469653,
+      "grad_norm": 0.4802429974079132,
+      "learning_rate": 0.00013289427186424587,
+      "loss": 0.10040791034698486,
+      "step": 16340
+    },
+    {
+      "epoch": 0.039457748241298746,
+      "grad_norm": 0.2794296145439148,
+      "learning_rate": 0.000132818190363384,
+      "loss": 0.151225483417511,
+      "step": 16350
+    },
+    {
+      "epoch": 0.03975002785790096,
+      "grad_norm": 0.5124934911727905,
+      "learning_rate": 0.00013274208756715474,
+      "loss": 0.13172839879989623,
+      "step": 16360
+    },
+    {
+      "epoch": 0.04004230747450317,
+      "grad_norm": 0.5249491930007935,
+      "learning_rate": 0.00013266596352494027,
+      "loss": 0.16215051412582399,
+      "step": 16370
+    },
+    {
+      "epoch": 0.040334587091105385,
+      "grad_norm": 0.4500700831413269,
+      "learning_rate": 0.00013258981828613678,
+      "loss": 0.12330397367477416,
+      "step": 16380
+    },
+    {
+      "epoch": 0.0406268667077076,
+      "grad_norm": 0.5716415047645569,
+      "learning_rate": 0.00013251365190015399,
+      "loss": 0.14464513063430787,
+      "step": 16390
+    },
+    {
+      "epoch": 0.04091914632430981,
+      "grad_norm": 0.5867295265197754,
+      "learning_rate": 0.00013243746441641546,
+      "loss": 0.13860440254211426,
+      "step": 16400
+    },
+    {
+      "epoch": 0.041211425940912025,
+      "grad_norm": 0.5596738457679749,
+      "learning_rate": 0.0001323612558843585,
+      "loss": 0.10580860376358033,
+      "step": 16410
+    },
+    {
+      "epoch": 0.04150370555751424,
+      "grad_norm": 0.4098498225212097,
+      "learning_rate": 0.00013228502635343386,
+      "loss": 0.17497682571411133,
+      "step": 16420
+    },
+    {
+      "epoch": 0.04179598517411645,
+      "grad_norm": 0.6621960401535034,
+      "learning_rate": 0.0001322087758731062,
+      "loss": 0.1301037549972534,
+      "step": 16430
+    },
+    {
+      "epoch": 0.04208826479071866,
+      "grad_norm": 0.4563891589641571,
+      "learning_rate": 0.00013213250449285355,
+      "loss": 0.13085787296295165,
+      "step": 16440
+    },
+    {
+      "epoch": 0.04238054440732087,
+      "grad_norm": 0.5588815212249756,
+      "learning_rate": 0.00013205621226216758,
+      "loss": 0.14839051961898803,
+      "step": 16450
+    },
+    {
+      "epoch": 0.042672824023923084,
+      "grad_norm": 0.3709474205970764,
+      "learning_rate": 0.00013197989923055347,
+      "loss": 0.1782838821411133,
+      "step": 16460
+    },
+    {
+      "epoch": 0.0429651036405253,
+      "grad_norm": 0.5441906452178955,
+      "learning_rate": 0.00013190356544753,
+      "loss": 0.15045090913772582,
+      "step": 16470
+    },
+    {
+      "epoch": 0.04325738325712751,
+      "grad_norm": 0.3722614347934723,
+      "learning_rate": 0.00013182721096262925,
+      "loss": 0.0921307623386383,
+      "step": 16480
+    },
+    {
+      "epoch": 0.04354966287372972,
+      "grad_norm": 0.5007218718528748,
+      "learning_rate": 0.00013175083582539688,
+      "loss": 0.13650132417678834,
+      "step": 16490
+    },
+    {
+      "epoch": 0.04384194249033194,
+      "grad_norm": 0.524430513381958,
+      "learning_rate": 0.00013167444008539185,
+      "loss": 0.15479122400283812,
+      "step": 16500
+    },
+    {
+      "epoch": 0.04413422210693415,
+      "grad_norm": 0.49985289573669434,
+      "learning_rate": 0.0001315980237921865,
+      "loss": 0.12488617897033691,
+      "step": 16510
+    },
+    {
+      "epoch": 0.04442650172353636,
+      "grad_norm": 0.5298058986663818,
+      "learning_rate": 0.0001315215869953666,
+      "loss": 0.13367410898208618,
+      "step": 16520
+    },
+    {
+      "epoch": 0.044718781340138576,
+      "grad_norm": 0.5753952264785767,
+      "learning_rate": 0.00013144512974453112,
+      "loss": 0.12412116527557374,
+      "step": 16530
+    },
+    {
+      "epoch": 0.04501106095674079,
+      "grad_norm": 0.4291190803050995,
+      "learning_rate": 0.00013136865208929235,
+      "loss": 0.09539954662322998,
+      "step": 16540
+    },
+    {
+      "epoch": 0.045303340573343,
+      "grad_norm": 0.7173647880554199,
+      "learning_rate": 0.00013129215407927576,
+      "loss": 0.1124645709991455,
+      "step": 16550
+    },
+    {
+      "epoch": 0.045595620189945216,
+      "grad_norm": 0.5498252511024475,
+      "learning_rate": 0.00013121563576412018,
+      "loss": 0.0975601851940155,
+      "step": 16560
+    },
+    {
+      "epoch": 0.04588789980654743,
+      "grad_norm": 0.7431814074516296,
+      "learning_rate": 0.00013113909719347742,
+      "loss": 0.12182284593582153,
+      "step": 16570
+    },
+    {
+      "epoch": 0.04618017942314964,
+      "grad_norm": 0.6450138092041016,
+      "learning_rate": 0.00013106253841701257,
+      "loss": 0.14334791898727417,
+      "step": 16580
+    },
+    {
+      "epoch": 0.046472459039751855,
+      "grad_norm": 0.30915600061416626,
+      "learning_rate": 0.0001309859594844038,
+      "loss": 0.16096971035003663,
+      "step": 16590
+    },
+    {
+      "epoch": 0.04676473865635407,
+      "grad_norm": 0.6087617874145508,
+      "learning_rate": 0.0001309093604453423,
+      "loss": 0.1394852042198181,
+      "step": 16600
+    },
+    {
+      "epoch": 0.04705701827295628,
+      "grad_norm": 0.4556034207344055,
+      "learning_rate": 0.0001308327413495324,
+      "loss": 0.13188859224319457,
+      "step": 16610
+    },
+    {
+      "epoch": 0.047349297889558495,
+      "grad_norm": 0.5532922744750977,
+      "learning_rate": 0.00013075610224669139,
+      "loss": 0.14009127616882325,
+      "step": 16620
+    },
+    {
+      "epoch": 0.04764157750616071,
+      "grad_norm": 0.44641000032424927,
+      "learning_rate": 0.00013067944318654953,
+      "loss": 0.11838617324829101,
+      "step": 16630
+    },
+    {
+      "epoch": 0.04793385712276292,
+      "grad_norm": 0.4414714276790619,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.1645753264427185,
+      "step": 16640
+    },
+    {
+      "epoch": 0.048226136739365134,
+      "grad_norm": 0.4849410951137543,
+      "learning_rate": 0.0001305260653933492,
+      "loss": 0.13441089391708375,
+      "step": 16650
+    },
+    {
+      "epoch": 0.04851841635596735,
+      "grad_norm": 0.45449644327163696,
+      "learning_rate": 0.00013044934675981586,
+      "loss": 0.11344125270843505,
+      "step": 16660
+    },
+    {
+      "epoch": 0.04881069597256956,
+      "grad_norm": 0.7107430696487427,
+      "learning_rate": 0.00013037260836803202,
+      "loss": 0.0925561785697937,
+      "step": 16670
+    },
+    {
+      "epoch": 0.049102975589171774,
+      "grad_norm": 0.4833727478981018,
+      "learning_rate": 0.00013029585026779234,
+      "loss": 0.11987150907516479,
+      "step": 16680
+    },
+    {
+      "epoch": 0.04939525520577399,
+      "grad_norm": 0.4320180416107178,
+      "learning_rate": 0.00013021907250890434,
+      "loss": 0.14564917087554932,
+      "step": 16690
+    },
+    {
+      "epoch": 0.04968753482237619,
+      "grad_norm": 0.3806993067264557,
+      "learning_rate": 0.00013014227514118832,
+      "loss": 0.09459810256958008,
+      "step": 16700
+    },
+    {
+      "epoch": 0.04997981443897841,
+      "grad_norm": 0.4729923903942108,
+      "learning_rate": 0.00013006545821447717,
+      "loss": 0.1615089774131775,
+      "step": 16710
+    },
+    {
+      "epoch": 0.05027209405558062,
+      "grad_norm": 0.46627670526504517,
+      "learning_rate": 0.0001299886217786166,
+      "loss": 0.13770408630371095,
+      "step": 16720
+    },
+    {
+      "epoch": 0.05056437367218283,
+      "grad_norm": 0.5141212940216064,
+      "learning_rate": 0.00012991176588346493,
+      "loss": 0.11769570112228393,
+      "step": 16730
+    },
+    {
+      "epoch": 0.050856653288785046,
+      "grad_norm": 0.42008060216903687,
+      "learning_rate": 0.00012983489057889313,
+      "loss": 0.1482585310935974,
+      "step": 16740
+    },
+    {
+      "epoch": 0.05114893290538726,
+      "grad_norm": 0.5725877285003662,
+      "learning_rate": 0.00012975799591478476,
+      "loss": 0.1301108717918396,
+      "step": 16750
+    },
+    {
+      "epoch": 0.05144121252198947,
+      "grad_norm": 0.5012943148612976,
+      "learning_rate": 0.00012968108194103594,
+      "loss": 0.10237042903900147,
+      "step": 16760
+    },
+    {
+      "epoch": 0.051733492138591686,
+      "grad_norm": 0.5335416793823242,
+      "learning_rate": 0.00012960414870755524,
+      "loss": 0.09824262261390686,
+      "step": 16770
+    },
+    {
+      "epoch": 0.0520257717551939,
+      "grad_norm": 0.7491273283958435,
+      "learning_rate": 0.0001295271962642639,
+      "loss": 0.1320204257965088,
+      "step": 16780
+    },
+    {
+      "epoch": 0.05231805137179611,
+      "grad_norm": 0.4764960706233978,
+      "learning_rate": 0.00012945022466109545,
+      "loss": 0.10767765045166015,
+      "step": 16790
+    },
+    {
+      "epoch": 0.052610330988398325,
+      "grad_norm": 0.42410293221473694,
+      "learning_rate": 0.00012937323394799597,
+      "loss": 0.10336188077926636,
+      "step": 16800
+    },
+    {
+      "epoch": 0.05290261060500054,
+      "grad_norm": 0.6089711785316467,
+      "learning_rate": 0.00012929622417492387,
+      "loss": 0.17036283016204834,
+      "step": 16810
+    },
+    {
+      "epoch": 0.05319489022160275,
+      "grad_norm": 0.3166019022464752,
+      "learning_rate": 0.00012921919539184994,
+      "loss": 0.11705763339996338,
+      "step": 16820
+    },
+    {
+      "epoch": 0.053487169838204965,
+      "grad_norm": 0.5477255582809448,
+      "learning_rate": 0.00012914214764875737,
+      "loss": 0.10794316530227661,
+      "step": 16830
+    },
+    {
+      "epoch": 0.05377944945480718,
+      "grad_norm": 0.645714521408081,
+      "learning_rate": 0.0001290650809956416,
+      "loss": 0.14324039220809937,
+      "step": 16840
+    },
+    {
+      "epoch": 0.05407172907140939,
+      "grad_norm": 0.4442601799964905,
+      "learning_rate": 0.0001289879954825103,
+      "loss": 0.12037502527236939,
+      "step": 16850
+    },
+    {
+      "epoch": 0.054364008688011604,
+      "grad_norm": 0.9047551155090332,
+      "learning_rate": 0.0001289108911593835,
+      "loss": 0.1249876618385315,
+      "step": 16860
+    },
+    {
+      "epoch": 0.05465628830461382,
+      "grad_norm": 0.41341957449913025,
+      "learning_rate": 0.00012883376807629326,
+      "loss": 0.15676003694534302,
+      "step": 16870
+    },
+    {
+      "epoch": 0.05494856792121603,
+      "grad_norm": 0.3677040636539459,
+      "learning_rate": 0.00012875662628328397,
+      "loss": 0.1172824740409851,
+      "step": 16880
+    },
+    {
+      "epoch": 0.055240847537818244,
+      "grad_norm": 0.7432528138160706,
+      "learning_rate": 0.0001286794658304121,
+      "loss": 0.14492307901382445,
+      "step": 16890
+    },
+    {
+      "epoch": 0.05553312715442046,
+      "grad_norm": 0.44131892919540405,
+      "learning_rate": 0.0001286022867677462,
+      "loss": 0.10595977306365967,
+      "step": 16900
+    },
+    {
+      "epoch": 0.05582540677102267,
+      "grad_norm": 0.6255162358283997,
+      "learning_rate": 0.00012852508914536697,
+      "loss": 0.11324892044067383,
+      "step": 16910
+    },
+    {
+      "epoch": 0.056117686387624884,
+      "grad_norm": 0.692761242389679,
+      "learning_rate": 0.00012844787301336707,
+      "loss": 0.12741047143936157,
+      "step": 16920
+    },
+    {
+      "epoch": 0.0564099660042271,
+      "grad_norm": 0.47824689745903015,
+      "learning_rate": 0.0001283706384218512,
+      "loss": 0.11406828165054321,
+      "step": 16930
+    },
+    {
+      "epoch": 0.05670224562082931,
+      "grad_norm": 0.5563332438468933,
+      "learning_rate": 0.0001282933854209361,
+      "loss": 0.10179842710494995,
+      "step": 16940
+    },
+    {
+      "epoch": 0.05699452523743152,
+      "grad_norm": 0.48051705956459045,
+      "learning_rate": 0.00012821611406075044,
+      "loss": 0.13613823652267457,
+      "step": 16950
+    },
+    {
+      "epoch": 0.05728680485403373,
+      "grad_norm": 0.7206515073776245,
+      "learning_rate": 0.00012813882439143462,
+      "loss": 0.12540388107299805,
+      "step": 16960
+    },
+    {
+      "epoch": 0.05757908447063594,
+      "grad_norm": 0.3289095163345337,
+      "learning_rate": 0.00012806151646314117,
+      "loss": 0.11339770555496216,
+      "step": 16970
+    },
+    {
+      "epoch": 0.057871364087238156,
+      "grad_norm": 0.3510822653770447,
+      "learning_rate": 0.00012798419032603438,
+      "loss": 0.13783811330795287,
+      "step": 16980
+    },
+    {
+      "epoch": 0.05816364370384037,
+      "grad_norm": 0.45192140340805054,
+      "learning_rate": 0.00012790684603029026,
+      "loss": 0.12697784900665282,
+      "step": 16990
+    },
+    {
+      "epoch": 0.05845592332044258,
+      "grad_norm": 0.49669742584228516,
+      "learning_rate": 0.00012782948362609676,
+      "loss": 0.11173609495162964,
+      "step": 17000
+    },
+    {
+      "epoch": 0.058748202937044795,
+      "grad_norm": 0.585717499256134,
+      "learning_rate": 0.00012775210316365347,
+      "loss": 0.13749698400497437,
+      "step": 17010
+    },
+    {
+      "epoch": 0.05904048255364701,
+      "grad_norm": 0.5942949652671814,
+      "learning_rate": 0.00012767470469317172,
+      "loss": 0.17236855030059814,
+      "step": 17020
+    },
+    {
+      "epoch": 0.05933276217024922,
+      "grad_norm": 0.47945302724838257,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.10242586135864258,
+      "step": 17030
+    },
+    {
+      "epoch": 0.059625041786851435,
+      "grad_norm": 0.36522573232650757,
+      "learning_rate": 0.0001275198539289967,
+      "loss": 0.16563642024993896,
+      "step": 17040
+    },
+    {
+      "epoch": 0.05991732140345365,
+      "grad_norm": 0.4557415843009949,
+      "learning_rate": 0.00012744240173578436,
+      "loss": 0.09598564505577087,
+      "step": 17050
+    },
+    {
+      "epoch": 0.06020960102005586,
+      "grad_norm": 0.5454314947128296,
+      "learning_rate": 0.00012736493173549552,
+      "loss": 0.1102681279182434,
+      "step": 17060
+    },
+    {
+      "epoch": 0.060501880636658074,
+      "grad_norm": 0.38414788246154785,
+      "learning_rate": 0.00012728744397839953,
+      "loss": 0.11983218193054199,
+      "step": 17070
+    },
+    {
+      "epoch": 0.06079416025326029,
+      "grad_norm": 0.4613216817378998,
+      "learning_rate": 0.0001272099385147774,
+      "loss": 0.10324720144271851,
+      "step": 17080
+    },
+    {
+      "epoch": 0.0610864398698625,
+      "grad_norm": 0.3139341473579407,
+      "learning_rate": 0.00012713241539492163,
+      "loss": 0.1176465630531311,
+      "step": 17090
+    },
+    {
+      "epoch": 0.061378719486464714,
+      "grad_norm": 0.568040132522583,
+      "learning_rate": 0.000127054874669136,
+      "loss": 0.10955891609191895,
+      "step": 17100
+    },
+    {
+      "epoch": 0.06167099910306693,
+      "grad_norm": 0.4846459627151489,
+      "learning_rate": 0.000126977316387736,
+      "loss": 0.13272666931152344,
+      "step": 17110
+    },
+    {
+      "epoch": 0.06196327871966914,
+      "grad_norm": 0.5252483487129211,
+      "learning_rate": 0.00012689974060104827,
+      "loss": 0.11659430265426636,
+      "step": 17120
+    },
+    {
+      "epoch": 0.062255558336271354,
+      "grad_norm": 0.5926061868667603,
+      "learning_rate": 0.00012682214735941087,
+      "loss": 0.14632503986358641,
+      "step": 17130
+    },
+    {
+      "epoch": 0.06254783795287357,
+      "grad_norm": 0.6203616261482239,
+      "learning_rate": 0.0001267445367131733,
+      "loss": 0.18737653493881226,
+      "step": 17140
+    },
+    {
+      "epoch": 0.06284011756947577,
+      "grad_norm": 0.4781706929206848,
+      "learning_rate": 0.0001266669087126962,
+      "loss": 0.1547307252883911,
+      "step": 17150
+    },
+    {
+      "epoch": 0.063132397186078,
+      "grad_norm": 0.3953997492790222,
+      "learning_rate": 0.00012658926340835156,
+      "loss": 0.1776956796646118,
+      "step": 17160
+    },
+    {
+      "epoch": 0.0634246768026802,
+      "grad_norm": 0.44508683681488037,
+      "learning_rate": 0.00012651160085052257,
+      "loss": 0.16617000102996826,
+      "step": 17170
+    },
+    {
+      "epoch": 0.06371695641928242,
+      "grad_norm": 0.35256314277648926,
+      "learning_rate": 0.00012643392108960358,
+      "loss": 0.13866136074066163,
+      "step": 17180
+    },
+    {
+      "epoch": 0.06400923603588463,
+      "grad_norm": 0.429677814245224,
+      "learning_rate": 0.00012635622417600027,
+      "loss": 0.13087263107299804,
+      "step": 17190
+    },
+    {
+      "epoch": 0.06430151565248685,
+      "grad_norm": 0.4318985939025879,
+      "learning_rate": 0.00012627851016012916,
+      "loss": 0.16348389387130738,
+      "step": 17200
+    },
+    {
+      "epoch": 0.06459379526908905,
+      "grad_norm": 0.7080004811286926,
+      "learning_rate": 0.00012620077909241813,
+      "loss": 0.1620939254760742,
+      "step": 17210
+    },
+    {
+      "epoch": 0.06488607488569127,
+      "grad_norm": 0.4274366497993469,
+      "learning_rate": 0.00012612303102330603,
+      "loss": 0.12936292886734008,
+      "step": 17220
+    },
+    {
+      "epoch": 0.06517835450229348,
+      "grad_norm": 0.39892494678497314,
+      "learning_rate": 0.00012604526600324268,
+      "loss": 0.14804264307022094,
+      "step": 17230
+    },
+    {
+      "epoch": 0.0654706341188957,
+      "grad_norm": 0.3677639961242676,
+      "learning_rate": 0.000125967484082689,
+      "loss": 0.1572142004966736,
+      "step": 17240
+    },
+    {
+      "epoch": 0.0657629137354979,
+      "grad_norm": 0.45002129673957825,
+      "learning_rate": 0.00012588968531211686,
+      "loss": 0.12466174364089966,
+      "step": 17250
+    },
+    {
+      "epoch": 0.06605519335210012,
+      "grad_norm": 0.39502885937690735,
+      "learning_rate": 0.00012581186974200894,
+      "loss": 0.1583383560180664,
+      "step": 17260
+    },
+    {
+      "epoch": 0.06634747296870233,
+      "grad_norm": 0.46683114767074585,
+      "learning_rate": 0.00012573403742285901,
+      "loss": 0.14282358884811402,
+      "step": 17270
+    },
+    {
+      "epoch": 0.06663975258530455,
+      "grad_norm": 0.3787006139755249,
+      "learning_rate": 0.0001256561884051716,
+      "loss": 0.158794367313385,
+      "step": 17280
+    },
+    {
+      "epoch": 0.06693203220190676,
+      "grad_norm": 0.6142992377281189,
+      "learning_rate": 0.00012557832273946205,
+      "loss": 0.16118230819702148,
+      "step": 17290
+    },
+    {
+      "epoch": 0.06722431181850898,
+      "grad_norm": 0.614237368106842,
+      "learning_rate": 0.00012550044047625659,
+      "loss": 0.14131218194961548,
+      "step": 17300
+    },
+    {
+      "epoch": 0.06751659143511118,
+      "grad_norm": 0.6605455279350281,
+      "learning_rate": 0.00012542254166609214,
+      "loss": 0.17572733163833618,
+      "step": 17310
+    },
+    {
+      "epoch": 0.0678088710517134,
+      "grad_norm": 0.454250305891037,
+      "learning_rate": 0.00012534462635951642,
+      "loss": 0.3287865400314331,
+      "step": 17320
+    },
+    {
+      "epoch": 0.06810115066831561,
+      "grad_norm": 0.4073789715766907,
+      "learning_rate": 0.00012526669460708785,
+      "loss": 0.2643048048019409,
+      "step": 17330
+    },
+    {
+      "epoch": 0.06839343028491783,
+      "grad_norm": 0.46434909105300903,
+      "learning_rate": 0.0001251887464593754,
+      "loss": 0.26638553142547605,
+      "step": 17340
+    },
+    {
+      "epoch": 0.06868570990152004,
+      "grad_norm": 0.44712063670158386,
+      "learning_rate": 0.00012511078196695888,
+      "loss": 0.2876674890518188,
+      "step": 17350
+    },
+    {
+      "epoch": 0.06897798951812224,
+      "grad_norm": 0.2865538001060486,
+      "learning_rate": 0.0001250328011804286,
+      "loss": 0.2740046501159668,
+      "step": 17360
+    },
+    {
+      "epoch": 0.06927026913472446,
+      "grad_norm": 0.5578151345252991,
+      "learning_rate": 0.0001249548041503854,
+      "loss": 0.28751022815704347,
+      "step": 17370
+    },
+    {
+      "epoch": 0.06956254875132667,
+      "grad_norm": 0.6477270722389221,
+      "learning_rate": 0.00012487679092744068,
+      "loss": 0.24765028953552246,
+      "step": 17380
+    },
+    {
+      "epoch": 0.06985482836792889,
+      "grad_norm": 0.3667014539241791,
+      "learning_rate": 0.00012479876156221644,
+      "loss": 0.21680784225463867,
+      "step": 17390
+    },
+    {
+      "epoch": 0.0701471079845311,
+      "grad_norm": 0.5063636898994446,
+      "learning_rate": 0.00012472071610534504,
+      "loss": 0.2777646780014038,
+      "step": 17400
+    },
+    {
+      "epoch": 0.07043938760113332,
+      "grad_norm": 0.5982993245124817,
+      "learning_rate": 0.00012464265460746936,
+      "loss": 0.2317563533782959,
+      "step": 17410
+    },
+    {
+      "epoch": 0.07073166721773552,
+      "grad_norm": 0.49426621198654175,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 0.2235461950302124,
+      "step": 17420
+    },
+    {
+      "epoch": 0.07102394683433774,
+      "grad_norm": 0.3755197525024414,
+      "learning_rate": 0.00012448648369132854,
+      "loss": 0.21118478775024413,
+      "step": 17430
+    },
+    {
+      "epoch": 0.07131622645093995,
+      "grad_norm": 0.49265286326408386,
+      "learning_rate": 0.000124408374374401,
+      "loss": 0.21942646503448487,
+      "step": 17440
+    },
+    {
+      "epoch": 0.07160850606754217,
+      "grad_norm": 0.43882033228874207,
+      "learning_rate": 0.00012433024921914431,
+      "loss": 0.27982363700866697,
+      "step": 17450
+    },
+    {
+      "epoch": 0.07190078568414437,
+      "grad_norm": 0.5090575814247131,
+      "learning_rate": 0.00012425210827625308,
+      "loss": 0.26481919288635253,
+      "step": 17460
+    },
+    {
+      "epoch": 0.0721930653007466,
+      "grad_norm": 0.4228317439556122,
+      "learning_rate": 0.0001241739515964321,
+      "loss": 0.21948666572570802,
+      "step": 17470
+    },
+    {
+      "epoch": 0.0724853449173488,
+      "grad_norm": 0.37804725766181946,
+      "learning_rate": 0.00012409577923039634,
+      "loss": 0.2820150852203369,
+      "step": 17480
+    },
+    {
+      "epoch": 0.07277762453395102,
+      "grad_norm": 0.5117887854576111,
+      "learning_rate": 0.00012401759122887107,
+      "loss": 0.2367706775665283,
+      "step": 17490
+    },
+    {
+      "epoch": 0.07306990415055323,
+      "grad_norm": 0.4910002052783966,
+      "learning_rate": 0.0001239393876425916,
+      "loss": 0.2673805713653564,
+      "step": 17500
+    },
+    {
+      "epoch": 0.07336218376715545,
+      "grad_norm": 0.5312656164169312,
+      "learning_rate": 0.00012386116852230338,
+      "loss": 0.29644668102264404,
+      "step": 17510
+    },
+    {
+      "epoch": 0.07365446338375765,
+      "grad_norm": 0.4137590825557709,
+      "learning_rate": 0.000123782933918762,
+      "loss": 0.2519108057022095,
+      "step": 17520
+    },
+    {
+      "epoch": 0.07394674300035987,
+      "grad_norm": 0.5863052010536194,
+      "learning_rate": 0.000123704683882733,
+      "loss": 0.23267054557800293,
+      "step": 17530
+    },
+    {
+      "epoch": 0.07423902261696208,
+      "grad_norm": 0.4740700125694275,
+      "learning_rate": 0.00012362641846499196,
+      "loss": 0.27748005390167235,
+      "step": 17540
+    },
+    {
+      "epoch": 0.0745313022335643,
+      "grad_norm": 0.4319583773612976,
+      "learning_rate": 0.00012354813771632447,
+      "loss": 0.25856456756591795,
+      "step": 17550
+    },
+    {
+      "epoch": 0.0748235818501665,
+      "grad_norm": 0.5189881324768066,
+      "learning_rate": 0.00012346984168752612,
+      "loss": 0.2565593957901001,
+      "step": 17560
+    },
+    {
+      "epoch": 0.07511586146676873,
+      "grad_norm": 0.5176759362220764,
+      "learning_rate": 0.0001233915304294023,
+      "loss": 0.25342957973480223,
+      "step": 17570
+    },
+    {
+      "epoch": 0.07540814108337093,
+      "grad_norm": 0.526463508605957,
+      "learning_rate": 0.00012331320399276835,
+      "loss": 0.27576045989990233,
+      "step": 17580
+    },
+    {
+      "epoch": 0.07570042069997315,
+      "grad_norm": 0.3852822482585907,
+      "learning_rate": 0.0001232348624284494,
+      "loss": 0.24885883331298828,
+      "step": 17590
+    },
+    {
+      "epoch": 0.00017536776996132774,
+      "grad_norm": 0.32944732904434204,
+      "learning_rate": 0.00012315650578728052,
+      "loss": 0.15611624717712402,
+      "step": 17600
+    },
+    {
+      "epoch": 0.0004676473865635407,
+      "grad_norm": 0.40673837065696716,
+      "learning_rate": 0.00012307813412010642,
+      "loss": 0.13058433532714844,
+      "step": 17610
+    },
+    {
+      "epoch": 0.0007599270031657536,
+      "grad_norm": 0.4590607285499573,
+      "learning_rate": 0.00012299974747778162,
+      "loss": 0.10816874504089355,
+      "step": 17620
+    },
+    {
+      "epoch": 0.0010522066197679665,
+      "grad_norm": 0.6818011999130249,
+      "learning_rate": 0.00012292134591117043,
+      "loss": 0.0977628231048584,
+      "step": 17630
+    },
+    {
+      "epoch": 0.0013444862363701795,
+      "grad_norm": 0.3881150484085083,
+      "learning_rate": 0.00012284292947114674,
+      "loss": 0.08368622660636901,
+      "step": 17640
+    },
+    {
+      "epoch": 0.0016367658529723925,
+      "grad_norm": 0.6149243116378784,
+      "learning_rate": 0.00012276449820859407,
+      "loss": 0.09377326965332031,
+      "step": 17650
+    },
+    {
+      "epoch": 0.0019290454695746052,
+      "grad_norm": 0.6065635085105896,
+      "learning_rate": 0.00012268605217440566,
+      "loss": 0.10994265079498292,
+      "step": 17660
+    },
+    {
+      "epoch": 0.0022213250861768184,
+      "grad_norm": 0.5898813009262085,
+      "learning_rate": 0.00012260759141948432,
+      "loss": 0.08928636312484742,
+      "step": 17670
+    },
+    {
+      "epoch": 0.002513604702779031,
+      "grad_norm": 0.5225302577018738,
+      "learning_rate": 0.00012252911599474237,
+      "loss": 0.10890930891036987,
+      "step": 17680
+    },
+    {
+      "epoch": 0.002805884319381244,
+      "grad_norm": 0.5033912658691406,
+      "learning_rate": 0.00012245062595110163,
+      "loss": 0.08017398715019226,
+      "step": 17690
+    },
+    {
+      "epoch": 0.003098163935983457,
+      "grad_norm": 0.4297211468219757,
+      "learning_rate": 0.00012237212133949345,
+      "loss": 0.10672861337661743,
+      "step": 17700
+    },
+    {
+      "epoch": 0.00339044355258567,
+      "grad_norm": 0.798556923866272,
+      "learning_rate": 0.00012229360221085862,
+      "loss": 0.10870937108993531,
+      "step": 17710
+    },
+    {
+      "epoch": 0.0036827231691878826,
+      "grad_norm": 0.46279385685920715,
+      "learning_rate": 0.00012221506861614736,
+      "loss": 0.09393824934959412,
+      "step": 17720
+    },
+    {
+      "epoch": 0.003975002785790095,
+      "grad_norm": 0.5925435423851013,
+      "learning_rate": 0.00012213652060631927,
+      "loss": 0.11708329916000366,
+      "step": 17730
+    },
+    {
+      "epoch": 0.0042672824023923086,
+      "grad_norm": 0.3586753308773041,
+      "learning_rate": 0.00012205795823234327,
+      "loss": 0.10147005319595337,
+      "step": 17740
+    },
+    {
+      "epoch": 0.004559562018994522,
+      "grad_norm": 0.4583991467952728,
+      "learning_rate": 0.00012197938154519768,
+      "loss": 0.07401338815689087,
+      "step": 17750
+    },
+    {
+      "epoch": 0.004851841635596735,
+      "grad_norm": 0.5277538299560547,
+      "learning_rate": 0.00012190079059587,
+      "loss": 0.08578138947486877,
+      "step": 17760
+    },
+    {
+      "epoch": 0.005144121252198947,
+      "grad_norm": 0.4857586622238159,
+      "learning_rate": 0.00012182218543535708,
+      "loss": 0.07358272075653076,
+      "step": 17770
+    },
+    {
+      "epoch": 0.0054364008688011604,
+      "grad_norm": 0.6396073698997498,
+      "learning_rate": 0.00012174356611466497,
+      "loss": 0.09325509071350098,
+      "step": 17780
+    },
+    {
+      "epoch": 0.005728680485403374,
+      "grad_norm": 0.5275527238845825,
+      "learning_rate": 0.00012166493268480888,
+      "loss": 0.0789508044719696,
+      "step": 17790
+    },
+    {
+      "epoch": 0.006020960102005586,
+      "grad_norm": 0.3302583396434784,
+      "learning_rate": 0.00012158628519681317,
+      "loss": 0.10336633920669555,
+      "step": 17800
+    },
+    {
+      "epoch": 0.006313239718607799,
+      "grad_norm": 0.5686914324760437,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 0.08322003483772278,
+      "step": 17810
+    },
+    {
+      "epoch": 0.006605519335210012,
+      "grad_norm": 0.5167936682701111,
+      "learning_rate": 0.000121428948250546,
+      "loss": 0.10036357641220092,
+      "step": 17820
+    },
+    {
+      "epoch": 0.006897798951812225,
+      "grad_norm": 0.5632216930389404,
+      "learning_rate": 0.00012135025889436882,
+      "loss": 0.10954220294952392,
+      "step": 17830
+    },
+    {
+      "epoch": 0.007190078568414438,
+      "grad_norm": 0.7849739789962769,
+      "learning_rate": 0.00012127155568424043,
+      "loss": 0.10286763906478882,
+      "step": 17840
+    },
+    {
+      "epoch": 0.007482358185016651,
+      "grad_norm": 0.32591718435287476,
+      "learning_rate": 0.00012119283867123047,
+      "loss": 0.06006983518600464,
+      "step": 17850
+    },
+    {
+      "epoch": 0.007774637801618863,
+      "grad_norm": 0.6355476379394531,
+      "learning_rate": 0.00012111410790641761,
+      "loss": 0.08210298418998718,
+      "step": 17860
+    },
+    {
+      "epoch": 0.008066917418221077,
+      "grad_norm": 0.39826828241348267,
+      "learning_rate": 0.00012103536344088937,
+      "loss": 0.08768460154533386,
+      "step": 17870
+    },
+    {
+      "epoch": 0.008359197034823289,
+      "grad_norm": 0.486331969499588,
+      "learning_rate": 0.00012095660532574214,
+      "loss": 0.0847106158733368,
+      "step": 17880
+    },
+    {
+      "epoch": 0.008651476651425502,
+      "grad_norm": 0.45923373103141785,
+      "learning_rate": 0.00012087783361208127,
+      "loss": 0.10063333511352539,
+      "step": 17890
+    },
+    {
+      "epoch": 0.008943756268027715,
+      "grad_norm": 0.634421169757843,
+      "learning_rate": 0.00012079904835102083,
+      "loss": 0.09054496884346008,
+      "step": 17900
+    },
+    {
+      "epoch": 0.009236035884629928,
+      "grad_norm": 0.3904540240764618,
+      "learning_rate": 0.00012072024959368369,
+      "loss": 0.08148691058158875,
+      "step": 17910
+    },
+    {
+      "epoch": 0.009528315501232142,
+      "grad_norm": 0.5390598773956299,
+      "learning_rate": 0.00012064143739120157,
+      "loss": 0.0779704511165619,
+      "step": 17920
+    },
+    {
+      "epoch": 0.009820595117834355,
+      "grad_norm": 0.5069109797477722,
+      "learning_rate": 0.00012056261179471484,
+      "loss": 0.08845973014831543,
+      "step": 17930
+    },
+    {
+      "epoch": 0.010112874734436566,
+      "grad_norm": 0.5046719908714294,
+      "learning_rate": 0.00012048377285537256,
+      "loss": 0.09100863933563233,
+      "step": 17940
+    },
+    {
+      "epoch": 0.01040515435103878,
+      "grad_norm": 0.5391289591789246,
+      "learning_rate": 0.00012040492062433245,
+      "loss": 0.06952707767486573,
+      "step": 17950
+    },
+    {
+      "epoch": 0.010697433967640993,
+      "grad_norm": 0.38306495547294617,
+      "learning_rate": 0.00012032605515276087,
+      "loss": 0.09656602144241333,
+      "step": 17960
+    },
+    {
+      "epoch": 0.010989713584243206,
+      "grad_norm": 0.3511393964290619,
+      "learning_rate": 0.00012024717649183282,
+      "loss": 0.08410727977752686,
+      "step": 17970
+    },
+    {
+      "epoch": 0.011281993200845419,
+      "grad_norm": 0.6148380637168884,
+      "learning_rate": 0.00012016828469273172,
+      "loss": 0.08414973020553589,
+      "step": 17980
+    },
+    {
+      "epoch": 0.011574272817447632,
+      "grad_norm": 0.5247045159339905,
+      "learning_rate": 0.00012008937980664965,
+      "loss": 0.09690540432929992,
+      "step": 17990
+    },
+    {
+      "epoch": 0.011866552434049845,
+      "grad_norm": 0.4626063406467438,
+      "learning_rate": 0.00012001046188478716,
+      "loss": 0.08502410650253296,
+      "step": 18000
+    },
+    {
+      "epoch": 0.012158832050652057,
+      "grad_norm": 0.3819771111011505,
+      "learning_rate": 0.00011993153097835319,
+      "loss": 0.07516132593154908,
+      "step": 18010
+    },
+    {
+      "epoch": 0.01245111166725427,
+      "grad_norm": 0.5130971074104309,
+      "learning_rate": 0.00011985258713856515,
+      "loss": 0.09254549145698547,
+      "step": 18020
+    },
+    {
+      "epoch": 0.012743391283856483,
+      "grad_norm": 0.47451257705688477,
+      "learning_rate": 0.00011977363041664889,
+      "loss": 0.08282625675201416,
+      "step": 18030
+    },
+    {
+      "epoch": 0.013035670900458696,
+      "grad_norm": 0.5941485166549683,
+      "learning_rate": 0.00011969466086383848,
+      "loss": 0.06811115145683289,
+      "step": 18040
+    },
+    {
+      "epoch": 0.01332795051706091,
+      "grad_norm": 0.6522603631019592,
+      "learning_rate": 0.0001196156785313765,
+      "loss": 0.09992296099662781,
+      "step": 18050
+    },
+    {
+      "epoch": 0.013620230133663123,
+      "grad_norm": 0.5348054766654968,
+      "learning_rate": 0.00011953668347051364,
+      "loss": 0.10149786472320557,
+      "step": 18060
+    },
+    {
+      "epoch": 0.013912509750265336,
+      "grad_norm": 0.40029653906822205,
+      "learning_rate": 0.00011945767573250903,
+      "loss": 0.06788808107376099,
+      "step": 18070
+    },
+    {
+      "epoch": 0.014204789366867547,
+      "grad_norm": 0.573715090751648,
+      "learning_rate": 0.00011937865536862987,
+      "loss": 0.07610483169555664,
+      "step": 18080
+    },
+    {
+      "epoch": 0.01449706898346976,
+      "grad_norm": 0.45399683713912964,
+      "learning_rate": 0.00011929962243015161,
+      "loss": 0.08041064143180847,
+      "step": 18090
+    },
+    {
+      "epoch": 0.014789348600071974,
+      "grad_norm": 0.5539518594741821,
+      "learning_rate": 0.00011922057696835788,
+      "loss": 0.08963637351989746,
+      "step": 18100
+    },
+    {
+      "epoch": 0.015081628216674187,
+      "grad_norm": 0.47155383229255676,
+      "learning_rate": 0.00011914151903454049,
+      "loss": 0.08405691385269165,
+      "step": 18110
+    },
+    {
+      "epoch": 0.0153739078332764,
+      "grad_norm": 0.7586756348609924,
+      "learning_rate": 0.00011906244867999918,
+      "loss": 0.1017306923866272,
+      "step": 18120
+    },
+    {
+      "epoch": 0.01566618744987861,
+      "grad_norm": 0.579651415348053,
+      "learning_rate": 0.00011898336595604187,
+      "loss": 0.08366804718971252,
+      "step": 18130
+    },
+    {
+      "epoch": 0.015958467066480825,
+      "grad_norm": 0.6099450588226318,
+      "learning_rate": 0.00011890427091398451,
+      "loss": 0.0918500542640686,
+      "step": 18140
+    },
+    {
+      "epoch": 0.016250746683083038,
+      "grad_norm": 0.5033301711082458,
+      "learning_rate": 0.00011882516360515097,
+      "loss": 0.09937824010848999,
+      "step": 18150
+    },
+    {
+      "epoch": 0.01654302629968525,
+      "grad_norm": 0.5654765367507935,
+      "learning_rate": 0.00011874604408087315,
+      "loss": 0.09792746305465698,
+      "step": 18160
+    },
+    {
+      "epoch": 0.016835305916287464,
+      "grad_norm": 0.7175434827804565,
+      "learning_rate": 0.0001186669123924909,
+      "loss": 0.08806752562522888,
+      "step": 18170
+    },
+    {
+      "epoch": 0.017127585532889678,
+      "grad_norm": 0.6218147873878479,
+      "learning_rate": 0.00011858776859135181,
+      "loss": 0.10286098718643188,
+      "step": 18180
+    },
+    {
+      "epoch": 0.01741986514949189,
+      "grad_norm": 0.34478169679641724,
+      "learning_rate": 0.00011850861272881148,
+      "loss": 0.1039006233215332,
+      "step": 18190
+    },
+    {
+      "epoch": 0.017712144766094104,
+      "grad_norm": 0.8133497834205627,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 0.07579670548439026,
+      "step": 18200
+    },
+    {
+      "epoch": 0.018004424382696317,
+      "grad_norm": 0.490291953086853,
+      "learning_rate": 0.00011835026502498849,
+      "loss": 0.08932452797889709,
+      "step": 18210
+    },
+    {
+      "epoch": 0.01829670399929853,
+      "grad_norm": 0.6248547434806824,
+      "learning_rate": 0.0001182710732864559,
+      "loss": 0.08163339495658875,
+      "step": 18220
+    },
+    {
+      "epoch": 0.018588983615900743,
+      "grad_norm": 0.5636594295501709,
+      "learning_rate": 0.00011819186969202222,
+      "loss": 0.09210932850837708,
+      "step": 18230
+    },
+    {
+      "epoch": 0.018881263232502953,
+      "grad_norm": 0.4976969361305237,
+      "learning_rate": 0.0001181126542930818,
+      "loss": 0.0911138415336609,
+      "step": 18240
+    },
+    {
+      "epoch": 0.019173542849105166,
+      "grad_norm": 0.5840553045272827,
+      "learning_rate": 0.00011803342714103667,
+      "loss": 0.09251430630683899,
+      "step": 18250
+    },
+    {
+      "epoch": 0.01946582246570738,
+      "grad_norm": 0.3408087193965912,
+      "learning_rate": 0.0001179541882872964,
+      "loss": 0.10071238279342651,
+      "step": 18260
+    },
+    {
+      "epoch": 0.019758102082309593,
+      "grad_norm": 0.6248094439506531,
+      "learning_rate": 0.00011787493778327832,
+      "loss": 0.0945007860660553,
+      "step": 18270
+    },
+    {
+      "epoch": 0.020050381698911806,
+      "grad_norm": 0.6405048370361328,
+      "learning_rate": 0.00011779567568040721,
+      "loss": 0.08274674415588379,
+      "step": 18280
+    },
+    {
+      "epoch": 0.02034266131551402,
+      "grad_norm": 0.3423234224319458,
+      "learning_rate": 0.00011771640203011532,
+      "loss": 0.08074001073837281,
+      "step": 18290
+    },
+    {
+      "epoch": 0.020634940932116232,
+      "grad_norm": 0.6958869099617004,
+      "learning_rate": 0.00011763711688384247,
+      "loss": 0.06366150379180908,
+      "step": 18300
+    },
+    {
+      "epoch": 0.020927220548718446,
+      "grad_norm": 0.451951801776886,
+      "learning_rate": 0.000117557820293036,
+      "loss": 0.08959733247756958,
+      "step": 18310
+    },
+    {
+      "epoch": 0.02121950016532066,
+      "grad_norm": 0.431623637676239,
+      "learning_rate": 0.00011747851230915052,
+      "loss": 0.06741684079170226,
+      "step": 18320
+    },
+    {
+      "epoch": 0.021511779781922872,
+      "grad_norm": 0.5109727382659912,
+      "learning_rate": 0.0001173991929836482,
+      "loss": 0.09104928374290466,
+      "step": 18330
+    },
+    {
+      "epoch": 0.021804059398525085,
+      "grad_norm": 0.6636129021644592,
+      "learning_rate": 0.00011731986236799848,
+      "loss": 0.09483992457389831,
+      "step": 18340
+    },
+    {
+      "epoch": 0.0220963390151273,
+      "grad_norm": 0.5643255710601807,
+      "learning_rate": 0.00011724052051367804,
+      "loss": 0.09266481399536133,
+      "step": 18350
+    },
+    {
+      "epoch": 0.02238861863172951,
+      "grad_norm": 0.4886293113231659,
+      "learning_rate": 0.00011716116747217107,
+      "loss": 0.06867415904998779,
+      "step": 18360
+    },
+    {
+      "epoch": 0.02268089824833172,
+      "grad_norm": 0.5121495127677917,
+      "learning_rate": 0.0001170818032949688,
+      "loss": 0.07524446249008179,
+      "step": 18370
+    },
+    {
+      "epoch": 0.022973177864933934,
+      "grad_norm": 0.3761802911758423,
+      "learning_rate": 0.00011700242803356984,
+      "loss": 0.08070670366287232,
+      "step": 18380
+    },
+    {
+      "epoch": 0.023265457481536148,
+      "grad_norm": 0.4995165765285492,
+      "learning_rate": 0.00011692304173947996,
+      "loss": 0.1004329800605774,
+      "step": 18390
+    },
+    {
+      "epoch": 0.02355773709813836,
+      "grad_norm": 0.5291754007339478,
+      "learning_rate": 0.00011684364446421201,
+      "loss": 0.09098554849624634,
+      "step": 18400
+    },
+    {
+      "epoch": 0.023850016714740574,
+      "grad_norm": 0.32648584246635437,
+      "learning_rate": 0.000116764236259286,
+      "loss": 0.08552274703979493,
+      "step": 18410
+    },
+    {
+      "epoch": 0.024142296331342787,
+      "grad_norm": 0.46827471256256104,
+      "learning_rate": 0.00011668481717622912,
+      "loss": 0.10650006532669068,
+      "step": 18420
+    },
+    {
+      "epoch": 0.024434575947945,
+      "grad_norm": 0.39092817902565,
+      "learning_rate": 0.00011660538726657551,
+      "loss": 0.0946243405342102,
+      "step": 18430
+    },
+    {
+      "epoch": 0.024726855564547214,
+      "grad_norm": 0.3989443778991699,
+      "learning_rate": 0.0001165259465818664,
+      "loss": 0.0793117880821228,
+      "step": 18440
+    },
+    {
+      "epoch": 0.025019135181149427,
+      "grad_norm": 0.5061081647872925,
+      "learning_rate": 0.00011644649517364996,
+      "loss": 0.11974735260009765,
+      "step": 18450
+    },
+    {
+      "epoch": 0.02531141479775164,
+      "grad_norm": 0.3200245201587677,
+      "learning_rate": 0.00011636703309348133,
+      "loss": 0.0845161497592926,
+      "step": 18460
+    },
+    {
+      "epoch": 0.025603694414353853,
+      "grad_norm": 0.4882155656814575,
+      "learning_rate": 0.00011628756039292268,
+      "loss": 0.10069631338119507,
+      "step": 18470
+    },
+    {
+      "epoch": 0.025895974030956066,
+      "grad_norm": 0.6742693781852722,
+      "learning_rate": 0.00011620807712354287,
+      "loss": 0.08431915640830993,
+      "step": 18480
+    },
+    {
+      "epoch": 0.02618825364755828,
+      "grad_norm": 0.6741700172424316,
+      "learning_rate": 0.00011612858333691779,
+      "loss": 0.10626136064529419,
+      "step": 18490
+    },
+    {
+      "epoch": 0.02648053326416049,
+      "grad_norm": 0.43489256501197815,
+      "learning_rate": 0.00011604907908463012,
+      "loss": 0.08305467367172241,
+      "step": 18500
+    },
+    {
+      "epoch": 0.026772812880762702,
+      "grad_norm": 0.7767978310585022,
+      "learning_rate": 0.00011596956441826924,
+      "loss": 0.07433313131332397,
+      "step": 18510
+    },
+    {
+      "epoch": 0.027065092497364916,
+      "grad_norm": 0.5814490914344788,
+      "learning_rate": 0.00011589003938943142,
+      "loss": 0.09056335091590881,
+      "step": 18520
+    },
+    {
+      "epoch": 0.02735737211396713,
+      "grad_norm": 0.5423917770385742,
+      "learning_rate": 0.00011581050404971954,
+      "loss": 0.09163503050804138,
+      "step": 18530
+    },
+    {
+      "epoch": 0.027649651730569342,
+      "grad_norm": 0.42647847533226013,
+      "learning_rate": 0.00011573095845074321,
+      "loss": 0.1142728328704834,
+      "step": 18540
+    },
+    {
+      "epoch": 0.027941931347171555,
+      "grad_norm": 0.3328469693660736,
+      "learning_rate": 0.00011565140264411878,
+      "loss": 0.08259868621826172,
+      "step": 18550
+    },
+    {
+      "epoch": 0.02823421096377377,
+      "grad_norm": 0.488360732793808,
+      "learning_rate": 0.00011557183668146907,
+      "loss": 0.10231245756149292,
+      "step": 18560
+    },
+    {
+      "epoch": 0.02852649058037598,
+      "grad_norm": 0.44326305389404297,
+      "learning_rate": 0.0001154922606144236,
+      "loss": 0.09847432374954224,
+      "step": 18570
+    },
+    {
+      "epoch": 0.028818770196978195,
+      "grad_norm": 0.48132699728012085,
+      "learning_rate": 0.00011541267449461841,
+      "loss": 0.09495120644569396,
+      "step": 18580
+    },
+    {
+      "epoch": 0.029111049813580408,
+      "grad_norm": 0.498384565114975,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.08849372267723084,
+      "step": 18590
+    },
+    {
+      "epoch": 0.00017536776996132774,
+      "grad_norm": 0.4812234938144684,
+      "learning_rate": 0.00011525347230330563,
+      "loss": 0.10695960124333699,
+      "step": 18600
+    },
+    {
+      "epoch": 0.0004676473865635407,
+      "grad_norm": 0.3547711968421936,
+      "learning_rate": 0.00011517385633510267,
+      "loss": 0.083427894115448,
+      "step": 18610
+    },
+    {
+      "epoch": 0.0007599270031657536,
+      "grad_norm": 0.410141259431839,
+      "learning_rate": 0.00011509423052074903,
+      "loss": 0.07211172580718994,
+      "step": 18620
+    },
+    {
+      "epoch": 0.0010522066197679665,
+      "grad_norm": 0.5318905711174011,
+      "learning_rate": 0.00011501459491191309,
+      "loss": 0.06363685131072998,
+      "step": 18630
+    },
+    {
+      "epoch": 0.0013444862363701795,
+      "grad_norm": 0.31552252173423767,
+      "learning_rate": 0.00011493494956026953,
+      "loss": 0.05316097140312195,
+      "step": 18640
+    },
+    {
+      "epoch": 0.0016367658529723925,
+      "grad_norm": 0.5953588485717773,
+      "learning_rate": 0.00011485529451749935,
+      "loss": 0.05686379075050354,
+      "step": 18650
+    },
+    {
+      "epoch": 0.0019290454695746052,
+      "grad_norm": 0.5396347641944885,
+      "learning_rate": 0.00011477562983528982,
+      "loss": 0.06358109712600708,
+      "step": 18660
+    },
+    {
+      "epoch": 0.0022213250861768184,
+      "grad_norm": 0.4352971911430359,
+      "learning_rate": 0.00011469595556533454,
+      "loss": 0.05197587013244629,
+      "step": 18670
+    },
+    {
+      "epoch": 0.002513604702779031,
+      "grad_norm": 0.46055442094802856,
+      "learning_rate": 0.0001146162717593332,
+      "loss": 0.06596101522445678,
+      "step": 18680
+    },
+    {
+      "epoch": 0.002805884319381244,
+      "grad_norm": 0.5026074647903442,
+      "learning_rate": 0.00011453657846899184,
+      "loss": 0.04540090560913086,
+      "step": 18690
+    },
+    {
+      "epoch": 0.003098163935983457,
+      "grad_norm": 0.396774023771286,
+      "learning_rate": 0.00011445687574602251,
+      "loss": 0.06005242466926575,
+      "step": 18700
+    },
+    {
+      "epoch": 0.00339044355258567,
+      "grad_norm": 0.6782927513122559,
+      "learning_rate": 0.00011437716364214346,
+      "loss": 0.0664262056350708,
+      "step": 18710
+    },
+    {
+      "epoch": 0.0036827231691878826,
+      "grad_norm": 0.3659330904483795,
+      "learning_rate": 0.00011429744220907903,
+      "loss": 0.0549392819404602,
+      "step": 18720
+    },
+    {
+      "epoch": 0.003975002785790095,
+      "grad_norm": 0.47903260588645935,
+      "learning_rate": 0.0001142177114985596,
+      "loss": 0.06735857129096985,
+      "step": 18730
+    },
+    {
+      "epoch": 0.0042672824023923086,
+      "grad_norm": 0.328568696975708,
+      "learning_rate": 0.00011413797156232149,
+      "loss": 0.061211001873016355,
+      "step": 18740
+    },
+    {
+      "epoch": 0.004559562018994522,
+      "grad_norm": 0.3654443919658661,
+      "learning_rate": 0.00011405822245210716,
+      "loss": 0.04326913356781006,
+      "step": 18750
+    },
+    {
+      "epoch": 0.004851841635596735,
+      "grad_norm": 0.44504234194755554,
+      "learning_rate": 0.00011397846421966488,
+      "loss": 0.049028819799423216,
+      "step": 18760
+    },
+    {
+      "epoch": 0.005144121252198947,
+      "grad_norm": 0.5135933756828308,
+      "learning_rate": 0.00011389869691674895,
+      "loss": 0.04076721668243408,
+      "step": 18770
+    },
+    {
+      "epoch": 0.0054364008688011604,
+      "grad_norm": 0.5516208410263062,
+      "learning_rate": 0.00011381892059511949,
+      "loss": 0.056097608804702756,
+      "step": 18780
+    },
+    {
+      "epoch": 0.005728680485403374,
+      "grad_norm": 0.4304979145526886,
+      "learning_rate": 0.00011373913530654248,
+      "loss": 0.04370307922363281,
+      "step": 18790
+    },
+    {
+      "epoch": 0.006020960102005586,
+      "grad_norm": 0.2555494010448456,
+      "learning_rate": 0.00011365934110278971,
+      "loss": 0.05759128332138062,
+      "step": 18800
+    },
+    {
+      "epoch": 0.006313239718607799,
+      "grad_norm": 0.5282100439071655,
+      "learning_rate": 0.00011357953803563883,
+      "loss": 0.04635744094848633,
+      "step": 18810
+    },
+    {
+      "epoch": 0.006605519335210012,
+      "grad_norm": 0.46182122826576233,
+      "learning_rate": 0.00011349972615687312,
+      "loss": 0.058734142780303956,
+      "step": 18820
+    },
+    {
+      "epoch": 0.006897798951812225,
+      "grad_norm": 0.47366786003112793,
+      "learning_rate": 0.0001134199055182817,
+      "loss": 0.0609737753868103,
+      "step": 18830
+    },
+    {
+      "epoch": 0.007190078568414438,
+      "grad_norm": 0.6910023093223572,
+      "learning_rate": 0.0001133400761716593,
+      "loss": 0.057688182592391966,
+      "step": 18840
+    },
+    {
+      "epoch": 0.007482358185016651,
+      "grad_norm": 0.22007109224796295,
+      "learning_rate": 0.00011326023816880625,
+      "loss": 0.035953018069267276,
+      "step": 18850
+    },
+    {
+      "epoch": 0.007774637801618863,
+      "grad_norm": 0.5203167796134949,
+      "learning_rate": 0.00011318039156152864,
+      "loss": 0.04725465178489685,
+      "step": 18860
+    },
+    {
+      "epoch": 0.008066917418221077,
+      "grad_norm": 0.3076709508895874,
+      "learning_rate": 0.00011310053640163804,
+      "loss": 0.049703779816627505,
+      "step": 18870
+    },
+    {
+      "epoch": 0.008359197034823289,
+      "grad_norm": 0.4022015333175659,
+      "learning_rate": 0.00011302067274095162,
+      "loss": 0.04838271141052246,
+      "step": 18880
+    },
+    {
+      "epoch": 0.008651476651425502,
+      "grad_norm": 0.36591270565986633,
+      "learning_rate": 0.00011294080063129202,
+      "loss": 0.058038800954818726,
+      "step": 18890
+    },
+    {
+      "epoch": 0.008943756268027715,
+      "grad_norm": 0.5593389272689819,
+      "learning_rate": 0.00011286092012448737,
+      "loss": 0.0486915796995163,
+      "step": 18900
+    },
+    {
+      "epoch": 0.009236035884629928,
+      "grad_norm": 0.3635571599006653,
+      "learning_rate": 0.00011278103127237128,
+      "loss": 0.044671615958213805,
+      "step": 18910
+    },
+    {
+      "epoch": 0.009528315501232142,
+      "grad_norm": 0.40763306617736816,
+      "learning_rate": 0.00011270113412678276,
+      "loss": 0.044573339819908145,
+      "step": 18920
+    },
+    {
+      "epoch": 0.009820595117834355,
+      "grad_norm": 0.4525219798088074,
+      "learning_rate": 0.00011262122873956619,
+      "loss": 0.050389724969863894,
+      "step": 18930
+    },
+    {
+      "epoch": 0.010112874734436566,
+      "grad_norm": 0.5440497994422913,
+      "learning_rate": 0.00011254131516257135,
+      "loss": 0.052079778909683225,
+      "step": 18940
+    },
+    {
+      "epoch": 0.01040515435103878,
+      "grad_norm": 0.2833980917930603,
+      "learning_rate": 0.00011246139344765325,
+      "loss": 0.038605299592018125,
+      "step": 18950
+    },
+    {
+      "epoch": 0.010697433967640993,
+      "grad_norm": 0.31991809606552124,
+      "learning_rate": 0.00011238146364667221,
+      "loss": 0.053519022464752194,
+      "step": 18960
+    },
+    {
+      "epoch": 0.010989713584243206,
+      "grad_norm": 0.3055783808231354,
+      "learning_rate": 0.00011230152581149387,
+      "loss": 0.05031843781471253,
+      "step": 18970
+    },
+    {
+      "epoch": 0.011281993200845419,
+      "grad_norm": 0.5740732550621033,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 0.04588623344898224,
+      "step": 18980
+    },
+    {
+      "epoch": 0.011574272817447632,
+      "grad_norm": 0.4770964980125427,
+      "learning_rate": 0.00011214162624603349,
+      "loss": 0.055134767293930055,
+      "step": 18990
+    },
+    {
+      "epoch": 0.011866552434049845,
+      "grad_norm": 0.3557377755641937,
+      "learning_rate": 0.00011206166461950855,
+      "loss": 0.05349679589271546,
+      "step": 19000
+    },
+    {
+      "epoch": 0.012158832050652057,
+      "grad_norm": 0.4615698456764221,
+      "learning_rate": 0.00011198169516630042,
+      "loss": 0.04118586480617523,
+      "step": 19010
+    },
+    {
+      "epoch": 0.01245111166725427,
+      "grad_norm": 0.38321709632873535,
+      "learning_rate": 0.00011190171793830033,
+      "loss": 0.05197300910949707,
+      "step": 19020
+    },
+    {
+      "epoch": 0.012743391283856483,
+      "grad_norm": 0.40673673152923584,
+      "learning_rate": 0.00011182173298740473,
+      "loss": 0.046470820903778076,
+      "step": 19030
+    },
+    {
+      "epoch": 0.013035670900458696,
+      "grad_norm": 0.5104904770851135,
+      "learning_rate": 0.00011174174036551492,
+      "loss": 0.038074332475662234,
+      "step": 19040
+    },
+    {
+      "epoch": 0.01332795051706091,
+      "grad_norm": 0.5641418695449829,
+      "learning_rate": 0.0001116617401245373,
+      "loss": 0.05830081701278687,
+      "step": 19050
+    },
+    {
+      "epoch": 0.013620230133663123,
+      "grad_norm": 0.4837978482246399,
+      "learning_rate": 0.00011158173231638309,
+      "loss": 0.059867626428604125,
+      "step": 19060
+    },
+    {
+      "epoch": 0.013912509750265336,
+      "grad_norm": 0.4095292091369629,
+      "learning_rate": 0.00011150171699296853,
+      "loss": 0.03946162462234497,
+      "step": 19070
+    },
+    {
+      "epoch": 0.014204789366867547,
+      "grad_norm": 0.6084176301956177,
+      "learning_rate": 0.00011142169420621469,
+      "loss": 0.04350184202194214,
+      "step": 19080
+    },
+    {
+      "epoch": 0.01449706898346976,
+      "grad_norm": 0.3639366924762726,
+      "learning_rate": 0.00011134166400804747,
+      "loss": 0.0470687061548233,
+      "step": 19090
+    },
+    {
+      "epoch": 0.014789348600071974,
+      "grad_norm": 0.5553967952728271,
+      "learning_rate": 0.00011126162645039761,
+      "loss": 0.0497103601694107,
+      "step": 19100
+    },
+    {
+      "epoch": 0.015081628216674187,
+      "grad_norm": 0.40798860788345337,
+      "learning_rate": 0.00011118158158520064,
+      "loss": 0.04873902499675751,
+      "step": 19110
+    },
+    {
+      "epoch": 0.0153739078332764,
+      "grad_norm": 0.5770225524902344,
+      "learning_rate": 0.00011110152946439674,
+      "loss": 0.05996111035346985,
+      "step": 19120
+    },
+    {
+      "epoch": 0.01566618744987861,
+      "grad_norm": 0.48800089955329895,
+      "learning_rate": 0.0001110214701399309,
+      "loss": 0.050528484582901004,
+      "step": 19130
+    },
+    {
+      "epoch": 0.015958467066480825,
+      "grad_norm": 0.6851457953453064,
+      "learning_rate": 0.00011094140366375277,
+      "loss": 0.05175333023071289,
+      "step": 19140
+    },
+    {
+      "epoch": 0.016250746683083038,
+      "grad_norm": 0.4330570101737976,
+      "learning_rate": 0.00011086133008781656,
+      "loss": 0.05753843784332276,
+      "step": 19150
+    },
+    {
+      "epoch": 0.01654302629968525,
+      "grad_norm": 0.4707883298397064,
+      "learning_rate": 0.00011078124946408121,
+      "loss": 0.055798256397247316,
+      "step": 19160
+    },
+    {
+      "epoch": 0.016835305916287464,
+      "grad_norm": 0.5940744876861572,
+      "learning_rate": 0.00011070116184451013,
+      "loss": 0.05048879384994507,
+      "step": 19170
+    },
+    {
+      "epoch": 0.017127585532889678,
+      "grad_norm": 0.5428313612937927,
+      "learning_rate": 0.0001106210672810713,
+      "loss": 0.057465237379074094,
+      "step": 19180
+    },
+    {
+      "epoch": 0.01741986514949189,
+      "grad_norm": 0.28007978200912476,
+      "learning_rate": 0.00011054096582573723,
+      "loss": 0.05806503891944885,
+      "step": 19190
+    },
+    {
+      "epoch": 0.017712144766094104,
+      "grad_norm": 0.6457118988037109,
+      "learning_rate": 0.00011046085753048485,
+      "loss": 0.041668251156806946,
+      "step": 19200
+    },
+    {
+      "epoch": 0.018004424382696317,
+      "grad_norm": 0.5596718192100525,
+      "learning_rate": 0.00011038074244729562,
+      "loss": 0.04973587095737457,
+      "step": 19210
+    },
+    {
+      "epoch": 0.01829670399929853,
+      "grad_norm": 0.5390315055847168,
+      "learning_rate": 0.00011030062062815531,
+      "loss": 0.049314957857131955,
+      "step": 19220
+    },
+    {
+      "epoch": 0.018588983615900743,
+      "grad_norm": 0.4929407835006714,
+      "learning_rate": 0.0001102204921250541,
+      "loss": 0.05299771428108215,
+      "step": 19230
+    },
+    {
+      "epoch": 0.018881263232502953,
+      "grad_norm": 0.4167720675468445,
+      "learning_rate": 0.00011014035698998651,
+      "loss": 0.0489980399608612,
+      "step": 19240
+    },
+    {
+      "epoch": 0.019173542849105166,
+      "grad_norm": 0.6488660573959351,
+      "learning_rate": 0.00011006021527495136,
+      "loss": 0.05308607816696167,
+      "step": 19250
+    },
+    {
+      "epoch": 0.01946582246570738,
+      "grad_norm": 0.33356109261512756,
+      "learning_rate": 0.00010998006703195177,
+      "loss": 0.05926038026809692,
+      "step": 19260
+    },
+    {
+      "epoch": 0.019758102082309593,
+      "grad_norm": 0.5040833353996277,
+      "learning_rate": 0.00010989991231299499,
+      "loss": 0.05293315649032593,
+      "step": 19270
+    },
+    {
+      "epoch": 0.020050381698911806,
+      "grad_norm": 0.5686571598052979,
+      "learning_rate": 0.00010981975117009263,
+      "loss": 0.047899335622787476,
+      "step": 19280
+    },
+    {
+      "epoch": 0.02034266131551402,
+      "grad_norm": 0.3700995147228241,
+      "learning_rate": 0.00010973958365526031,
+      "loss": 0.04503185451030731,
+      "step": 19290
+    },
+    {
+      "epoch": 0.020634940932116232,
+      "grad_norm": 0.5006086230278015,
+      "learning_rate": 0.00010965940982051795,
+      "loss": 0.03902796804904938,
+      "step": 19300
+    },
+    {
+      "epoch": 0.020927220548718446,
+      "grad_norm": 0.3904595971107483,
+      "learning_rate": 0.00010957922971788937,
+      "loss": 0.053290379047393796,
+      "step": 19310
+    },
+    {
+      "epoch": 0.02121950016532066,
+      "grad_norm": 0.37117278575897217,
+      "learning_rate": 0.00010949904339940267,
+      "loss": 0.04074969291687012,
+      "step": 19320
+    },
+    {
+      "epoch": 0.021511779781922872,
+      "grad_norm": 0.49424266815185547,
+      "learning_rate": 0.00010941885091708983,
+      "loss": 0.05467008948326111,
+      "step": 19330
+    },
+    {
+      "epoch": 0.021804059398525085,
+      "grad_norm": 0.5661629438400269,
+      "learning_rate": 0.00010933865232298686,
+      "loss": 0.05370916128158569,
+      "step": 19340
+    },
+    {
+      "epoch": 0.0220963390151273,
+      "grad_norm": 0.5256375074386597,
+      "learning_rate": 0.0001092584476691338,
+      "loss": 0.0527953028678894,
+      "step": 19350
+    },
+    {
+      "epoch": 0.02238861863172951,
+      "grad_norm": 0.44627049565315247,
+      "learning_rate": 0.0001091782370075746,
+      "loss": 0.03953578472137451,
+      "step": 19360
+    },
+    {
+      "epoch": 0.02268089824833172,
+      "grad_norm": 0.434017539024353,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.04138966798782349,
+      "step": 19370
+    },
+    {
+      "epoch": 0.022973177864933934,
+      "grad_norm": 0.34931954741477966,
+      "learning_rate": 0.00010901779786953276,
+      "loss": 0.04806713163852692,
+      "step": 19380
+    },
+    {
+      "epoch": 0.023265457481536148,
+      "grad_norm": 0.4455246329307556,
+      "learning_rate": 0.0001089375694971574,
+      "loss": 0.05707240104675293,
+      "step": 19390
+    },
+    {
+      "epoch": 0.02355773709813836,
+      "grad_norm": 0.4910562038421631,
+      "learning_rate": 0.00010885733532529023,
+      "loss": 0.05155366659164429,
+      "step": 19400
+    },
+    {
+      "epoch": 0.023850016714740574,
+      "grad_norm": 0.31981757283210754,
+      "learning_rate": 0.0001087770954059943,
+      "loss": 0.05193058848381042,
+      "step": 19410
+    },
+    {
+      "epoch": 0.024142296331342787,
+      "grad_norm": 0.30009976029396057,
+      "learning_rate": 0.00010869684979133649,
+      "loss": 0.05966759920120239,
+      "step": 19420
+    },
+    {
+      "epoch": 0.024434575947945,
+      "grad_norm": 0.3691764771938324,
+      "learning_rate": 0.00010861659853338726,
+      "loss": 0.053612631559371945,
+      "step": 19430
+    },
+    {
+      "epoch": 0.024726855564547214,
+      "grad_norm": 0.3629389703273773,
+      "learning_rate": 0.00010853634168422077,
+      "loss": 0.04587465822696686,
+      "step": 19440
+    },
+    {
+      "epoch": 0.025019135181149427,
+      "grad_norm": 0.49510499835014343,
+      "learning_rate": 0.00010845607929591488,
+      "loss": 0.0667344868183136,
+      "step": 19450
+    },
+    {
+      "epoch": 0.02531141479775164,
+      "grad_norm": 0.31854531168937683,
+      "learning_rate": 0.00010837581142055091,
+      "loss": 0.04679587483406067,
+      "step": 19460
+    },
+    {
+      "epoch": 0.025603694414353853,
+      "grad_norm": 0.4176080822944641,
+      "learning_rate": 0.00010829553811021385,
+      "loss": 0.05654631853103638,
+      "step": 19470
+    },
+    {
+      "epoch": 0.025895974030956066,
+      "grad_norm": 0.6177641153335571,
+      "learning_rate": 0.00010821525941699218,
+      "loss": 0.04781605005264282,
+      "step": 19480
+    },
+    {
+      "epoch": 0.02618825364755828,
+      "grad_norm": 0.6194220185279846,
+      "learning_rate": 0.00010813497539297786,
+      "loss": 0.060738807916641234,
+      "step": 19490
+    },
+    {
+      "epoch": 0.02648053326416049,
+      "grad_norm": 0.3684394955635071,
+      "learning_rate": 0.00010805468609026632,
+      "loss": 0.04900169372558594,
+      "step": 19500
+    },
+    {
+      "epoch": 0.026772812880762702,
+      "grad_norm": 0.7690474390983582,
+      "learning_rate": 0.00010797439156095644,
+      "loss": 0.04789839088916779,
+      "step": 19510
+    },
+    {
+      "epoch": 0.027065092497364916,
+      "grad_norm": 0.440841406583786,
+      "learning_rate": 0.00010789409185715042,
+      "loss": 0.05279890894889831,
+      "step": 19520
+    },
+    {
+      "epoch": 0.02735737211396713,
+      "grad_norm": 0.45056015253067017,
+      "learning_rate": 0.00010781378703095395,
+      "loss": 0.0551338791847229,
+      "step": 19530
+    },
+    {
+      "epoch": 0.027649651730569342,
+      "grad_norm": 0.3299720287322998,
+      "learning_rate": 0.00010773347713447588,
+      "loss": 0.0663510262966156,
+      "step": 19540
+    },
+    {
+      "epoch": 0.027941931347171555,
+      "grad_norm": 0.3492076098918915,
+      "learning_rate": 0.00010765316221982847,
+      "loss": 0.04370369017124176,
+      "step": 19550
+    },
+    {
+      "epoch": 0.02823421096377377,
+      "grad_norm": 0.3747090697288513,
+      "learning_rate": 0.00010757284233912717,
+      "loss": 0.059646451473236085,
+      "step": 19560
+    },
+    {
+      "epoch": 0.02852649058037598,
+      "grad_norm": 0.4360879957675934,
+      "learning_rate": 0.00010749251754449068,
+      "loss": 0.055461281538009645,
+      "step": 19570
+    },
+    {
+      "epoch": 0.028818770196978195,
+      "grad_norm": 0.5128139853477478,
+      "learning_rate": 0.00010741218788804092,
+      "loss": 0.0522183358669281,
+      "step": 19580
+    },
+    {
+      "epoch": 0.029111049813580408,
+      "grad_norm": 0.42321670055389404,
+      "learning_rate": 0.00010733185342190292,
+      "loss": 0.0505526602268219,
+      "step": 19590
+    },
+    {
+      "epoch": 0.02940332943018262,
+      "grad_norm": 0.6063235402107239,
+      "learning_rate": 0.00010725151419820483,
+      "loss": 0.09749404788017273,
+      "step": 19600
+    },
+    {
+      "epoch": 0.029695609046784834,
+      "grad_norm": 0.4130120277404785,
+      "learning_rate": 0.00010717117026907792,
+      "loss": 0.1309222936630249,
+      "step": 19610
+    },
+    {
+      "epoch": 0.029987888663387047,
+      "grad_norm": 0.44809237122535706,
+      "learning_rate": 0.00010709082168665645,
+      "loss": 0.1257900357246399,
+      "step": 19620
+    },
+    {
+      "epoch": 0.030280168279989257,
+      "grad_norm": 0.614315390586853,
+      "learning_rate": 0.00010701046850307777,
+      "loss": 0.1461053490638733,
+      "step": 19630
+    },
+    {
+      "epoch": 0.03057244789659147,
+      "grad_norm": 0.6700031161308289,
+      "learning_rate": 0.0001069301107704822,
+      "loss": 0.10423042774200439,
+      "step": 19640
+    },
+    {
+      "epoch": 0.030864727513193684,
+      "grad_norm": 0.4073422849178314,
+      "learning_rate": 0.00010684974854101302,
+      "loss": 0.09457704424858093,
+      "step": 19650
+    },
+    {
+      "epoch": 0.031157007129795897,
+      "grad_norm": 0.7099776864051819,
+      "learning_rate": 0.00010676938186681633,
+      "loss": 0.10997219085693359,
+      "step": 19660
+    },
+    {
+      "epoch": 0.03144928674639811,
+      "grad_norm": 0.533919632434845,
+      "learning_rate": 0.00010668901080004128,
+      "loss": 0.1042290210723877,
+      "step": 19670
+    },
+    {
+      "epoch": 0.03174156636300032,
+      "grad_norm": 0.5112128257751465,
+      "learning_rate": 0.00010660863539283972,
+      "loss": 0.10541698932647706,
+      "step": 19680
+    },
+    {
+      "epoch": 0.032033845979602536,
+      "grad_norm": 0.4949032962322235,
+      "learning_rate": 0.00010652825569736642,
+      "loss": 0.07829055190086365,
+      "step": 19690
+    },
+    {
+      "epoch": 0.03232612559620475,
+      "grad_norm": 0.7395544648170471,
+      "learning_rate": 0.00010644787176577886,
+      "loss": 0.09561094641685486,
+      "step": 19700
+    },
+    {
+      "epoch": 0.03261840521280696,
+      "grad_norm": 0.5629228949546814,
+      "learning_rate": 0.0001063674836502373,
+      "loss": 0.12351731061935425,
+      "step": 19710
+    },
+    {
+      "epoch": 0.032910684829409176,
+      "grad_norm": 0.5030995011329651,
+      "learning_rate": 0.00010628709140290476,
+      "loss": 0.09256659150123596,
+      "step": 19720
+    },
+    {
+      "epoch": 0.03320296444601139,
+      "grad_norm": 0.40227648615837097,
+      "learning_rate": 0.00010620669507594685,
+      "loss": 0.09367606639862061,
+      "step": 19730
+    },
+    {
+      "epoch": 0.0334952440626136,
+      "grad_norm": 0.6416395902633667,
+      "learning_rate": 0.00010612629472153188,
+      "loss": 0.10688693523406982,
+      "step": 19740
+    },
+    {
+      "epoch": 0.033787523679215815,
+      "grad_norm": 0.5081638693809509,
+      "learning_rate": 0.00010604589039183081,
+      "loss": 0.12269412279129029,
+      "step": 19750
+    },
+    {
+      "epoch": 0.03407980329581803,
+      "grad_norm": 0.4323157072067261,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 0.09958723187446594,
+      "step": 19760
+    },
+    {
+      "epoch": 0.03437208291242024,
+      "grad_norm": 0.5184201002120972,
+      "learning_rate": 0.00010588507001526676,
+      "loss": 0.08695968389511108,
+      "step": 19770
+    },
+    {
+      "epoch": 0.034664362529022455,
+      "grad_norm": 0.4592244327068329,
+      "learning_rate": 0.00010580465407275842,
+      "loss": 0.11281250715255738,
+      "step": 19780
+    },
+    {
+      "epoch": 0.03495664214562467,
+      "grad_norm": 0.5853420495986938,
+      "learning_rate": 0.00010572423436367302,
+      "loss": 0.1176479697227478,
+      "step": 19790
+    },
+    {
+      "epoch": 0.03524892176222688,
+      "grad_norm": 0.4518097937107086,
+      "learning_rate": 0.00010564381094019415,
+      "loss": 0.08045381903648377,
+      "step": 19800
+    },
+    {
+      "epoch": 0.035541201378829095,
+      "grad_norm": 0.48385971784591675,
+      "learning_rate": 0.00010556338385450762,
+      "loss": 0.12635680437088012,
+      "step": 19810
+    },
+    {
+      "epoch": 0.0358334809954313,
+      "grad_norm": 0.5627852082252502,
+      "learning_rate": 0.00010548295315880167,
+      "loss": 0.11994041204452514,
+      "step": 19820
+    },
+    {
+      "epoch": 0.036125760612033514,
+      "grad_norm": 0.43634769320487976,
+      "learning_rate": 0.00010540251890526697,
+      "loss": 0.11074424982070923,
+      "step": 19830
+    },
+    {
+      "epoch": 0.03641804022863573,
+      "grad_norm": 0.5890654921531677,
+      "learning_rate": 0.00010532208114609641,
+      "loss": 0.07576700448989868,
+      "step": 19840
+    },
+    {
+      "epoch": 0.03671031984523794,
+      "grad_norm": 0.4784446954727173,
+      "learning_rate": 0.00010524163993348517,
+      "loss": 0.08761775493621826,
+      "step": 19850
+    },
+    {
+      "epoch": 0.037002599461840154,
+      "grad_norm": 0.5902572870254517,
+      "learning_rate": 0.00010516119531963071,
+      "loss": 0.11185922622680664,
+      "step": 19860
+    },
+    {
+      "epoch": 0.03729487907844237,
+      "grad_norm": 0.374362051486969,
+      "learning_rate": 0.00010508074735673262,
+      "loss": 0.11105788946151733,
+      "step": 19870
+    },
+    {
+      "epoch": 0.03758715869504458,
+      "grad_norm": 0.4817831516265869,
+      "learning_rate": 0.00010500029609699269,
+      "loss": 0.1293437361717224,
+      "step": 19880
+    },
+    {
+      "epoch": 0.03787943831164679,
+      "grad_norm": 0.46657583117485046,
+      "learning_rate": 0.00010491984159261496,
+      "loss": 0.08614677786827088,
+      "step": 19890
+    },
+    {
+      "epoch": 0.038171717928249006,
+      "grad_norm": 0.3970106840133667,
+      "learning_rate": 0.00010483938389580541,
+      "loss": 0.13850605487823486,
+      "step": 19900
+    },
+    {
+      "epoch": 0.03846399754485122,
+      "grad_norm": 0.42310163378715515,
+      "learning_rate": 0.00010475892305877221,
+      "loss": 0.08650753498077393,
+      "step": 19910
+    },
+    {
+      "epoch": 0.03875627716145343,
+      "grad_norm": 0.7313628196716309,
+      "learning_rate": 0.00010467845913372552,
+      "loss": 0.12818144559860228,
+      "step": 19920
+    },
+    {
+      "epoch": 0.039048556778055646,
+      "grad_norm": 0.3901995122432709,
+      "learning_rate": 0.00010459799217287744,
+      "loss": 0.08122551441192627,
+      "step": 19930
+    },
+    {
+      "epoch": 0.03934083639465786,
+      "grad_norm": 0.7154889702796936,
+      "learning_rate": 0.00010451752222844223,
+      "loss": 0.1218881368637085,
+      "step": 19940
+    },
+    {
+      "epoch": 0.03963311601126007,
+      "grad_norm": 0.48666444420814514,
+      "learning_rate": 0.00010443704935263583,
+      "loss": 0.10404480695724487,
+      "step": 19950
+    },
+    {
+      "epoch": 0.039925395627862285,
+      "grad_norm": 0.7362347841262817,
+      "learning_rate": 0.00010435657359767628,
+      "loss": 0.11382819414138794,
+      "step": 19960
+    },
+    {
+      "epoch": 0.0402176752444645,
+      "grad_norm": 0.47754883766174316,
+      "learning_rate": 0.00010427609501578344,
+      "loss": 0.12453900575637818,
+      "step": 19970
+    },
+    {
+      "epoch": 0.04050995486106671,
+      "grad_norm": 0.6992428302764893,
+      "learning_rate": 0.00010419561365917899,
+      "loss": 0.096640944480896,
+      "step": 19980
+    },
+    {
+      "epoch": 0.040802234477668925,
+      "grad_norm": 0.9672327637672424,
+      "learning_rate": 0.00010411512958008634,
+      "loss": 0.09723003506660462,
+      "step": 19990
+    },
+    {
+      "epoch": 0.04109451409427114,
+      "grad_norm": 0.44509005546569824,
+      "learning_rate": 0.00010403464283073081,
+      "loss": 0.1070358157157898,
+      "step": 20000
+    },
+    {
+      "epoch": 0.04138679371087335,
+      "grad_norm": 0.6379432678222656,
+      "learning_rate": 0.00010395415346333934,
+      "loss": 0.11554746627807617,
+      "step": 20010
+    },
+    {
+      "epoch": 0.041679073327475565,
+      "grad_norm": 0.3677731156349182,
+      "learning_rate": 0.00010387366153014062,
+      "loss": 0.12191895246505738,
+      "step": 20020
+    },
+    {
+      "epoch": 0.04197135294407778,
+      "grad_norm": 0.44732749462127686,
+      "learning_rate": 0.00010379316708336502,
+      "loss": 0.0947759985923767,
+      "step": 20030
+    },
+    {
+      "epoch": 0.04226363256067999,
+      "grad_norm": 0.5520200133323669,
+      "learning_rate": 0.00010371267017524444,
+      "loss": 0.13124908208847047,
+      "step": 20040
+    },
+    {
+      "epoch": 0.042555912177282204,
+      "grad_norm": 0.6720760464668274,
+      "learning_rate": 0.00010363217085801251,
+      "loss": 0.13020946979522705,
+      "step": 20050
+    },
+    {
+      "epoch": 0.04284819179388442,
+      "grad_norm": 0.5111295580863953,
+      "learning_rate": 0.00010355166918390436,
+      "loss": 0.1281100630760193,
+      "step": 20060
+    },
+    {
+      "epoch": 0.04314047141048663,
+      "grad_norm": 0.47320589423179626,
+      "learning_rate": 0.00010347116520515659,
+      "loss": 0.08569440841674805,
+      "step": 20070
+    },
+    {
+      "epoch": 0.043432751027088844,
+      "grad_norm": 0.5256133079528809,
+      "learning_rate": 0.00010339065897400744,
+      "loss": 0.08793003559112549,
+      "step": 20080
+    },
+    {
+      "epoch": 0.04372503064369105,
+      "grad_norm": 0.34225842356681824,
+      "learning_rate": 0.0001033101505426965,
+      "loss": 0.08962181210517883,
+      "step": 20090
+    },
+    {
+      "epoch": 0.04401731026029326,
+      "grad_norm": 0.31762269139289856,
+      "learning_rate": 0.00010322963996346477,
+      "loss": 0.1351110816001892,
+      "step": 20100
+    },
+    {
+      "epoch": 0.044309589876895476,
+      "grad_norm": 0.6208139657974243,
+      "learning_rate": 0.00010314912728855478,
+      "loss": 0.09876767992973327,
+      "step": 20110
+    },
+    {
+      "epoch": 0.04460186949349769,
+      "grad_norm": 0.4539724290370941,
+      "learning_rate": 0.00010306861257021025,
+      "loss": 0.09199790954589844,
+      "step": 20120
+    },
+    {
+      "epoch": 0.0448941491100999,
+      "grad_norm": 0.6438459753990173,
+      "learning_rate": 0.00010298809586067633,
+      "loss": 0.10594894886016845,
+      "step": 20130
+    },
+    {
+      "epoch": 0.045186428726702116,
+      "grad_norm": 0.7474523782730103,
+      "learning_rate": 0.0001029075772121995,
+      "loss": 0.06941388249397278,
+      "step": 20140
+    },
+    {
+      "epoch": 0.04547870834330433,
+      "grad_norm": 0.4501616358757019,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.06798421740531921,
+      "step": 20150
+    },
+    {
+      "epoch": 0.04577098795990654,
+      "grad_norm": 0.5744274258613586,
+      "learning_rate": 0.00010274653430740879,
+      "loss": 0.1025542140007019,
+      "step": 20160
+    },
+    {
+      "epoch": 0.046063267576508755,
+      "grad_norm": 0.5808742642402649,
+      "learning_rate": 0.00010266601015559396,
+      "loss": 0.08924067616462708,
+      "step": 20170
+    },
+    {
+      "epoch": 0.04635554719311097,
+      "grad_norm": 0.5310708284378052,
+      "learning_rate": 0.00010258548427383408,
+      "loss": 0.14584681987762452,
+      "step": 20180
+    },
+    {
+      "epoch": 0.04664782680971318,
+      "grad_norm": 0.6485773921012878,
+      "learning_rate": 0.00010250495671438154,
+      "loss": 0.10075128078460693,
+      "step": 20190
+    },
+    {
+      "epoch": 0.046940106426315395,
+      "grad_norm": 0.4725688397884369,
+      "learning_rate": 0.00010242442752948979,
+      "loss": 0.11341706514358521,
+      "step": 20200
+    },
+    {
+      "epoch": 0.04723238604291761,
+      "grad_norm": 0.39604341983795166,
+      "learning_rate": 0.00010234389677141332,
+      "loss": 0.09534388184547424,
+      "step": 20210
+    },
+    {
+      "epoch": 0.04752466565951982,
+      "grad_norm": 0.43284931778907776,
+      "learning_rate": 0.00010226336449240767,
+      "loss": 0.11433600187301636,
+      "step": 20220
+    },
+    {
+      "epoch": 0.047816945276122035,
+      "grad_norm": 0.6595618724822998,
+      "learning_rate": 0.00010218283074472939,
+      "loss": 0.1096876859664917,
+      "step": 20230
+    },
+    {
+      "epoch": 0.04810922489272425,
+      "grad_norm": 0.4113689661026001,
+      "learning_rate": 0.00010210229558063596,
+      "loss": 0.1113879919052124,
+      "step": 20240
+    },
+    {
+      "epoch": 0.04840150450932646,
+      "grad_norm": 0.45097818970680237,
+      "learning_rate": 0.00010202175905238573,
+      "loss": 0.09581774473190308,
+      "step": 20250
+    },
+    {
+      "epoch": 0.048693784125928674,
+      "grad_norm": 0.43855610489845276,
+      "learning_rate": 0.00010194122121223799,
+      "loss": 0.07291094064712525,
+      "step": 20260
+    },
+    {
+      "epoch": 0.04898606374253089,
+      "grad_norm": 0.6747995018959045,
+      "learning_rate": 0.00010186068211245288,
+      "loss": 0.09019469022750855,
+      "step": 20270
+    },
+    {
+      "epoch": 0.0492783433591331,
+      "grad_norm": 0.5385534763336182,
+      "learning_rate": 0.00010178014180529136,
+      "loss": 0.10373307466506958,
+      "step": 20280
+    },
+    {
+      "epoch": 0.049570622975735314,
+      "grad_norm": 0.4198622405529022,
+      "learning_rate": 0.00010169960034301513,
+      "loss": 0.08638992309570312,
+      "step": 20290
+    },
+    {
+      "epoch": 0.04986290259233753,
+      "grad_norm": 0.5938236713409424,
+      "learning_rate": 0.00010161905777788674,
+      "loss": 0.10053577423095703,
+      "step": 20300
+    },
+    {
+      "epoch": 0.05015518220893974,
+      "grad_norm": 0.5413650870323181,
+      "learning_rate": 0.0001015385141621693,
+      "loss": 0.11966345310211182,
+      "step": 20310
+    },
+    {
+      "epoch": 0.05044746182554195,
+      "grad_norm": 0.42238134145736694,
+      "learning_rate": 0.00010145796954812672,
+      "loss": 0.08828957080841064,
+      "step": 20320
+    },
+    {
+      "epoch": 0.050739741442144166,
+      "grad_norm": 0.6850259900093079,
+      "learning_rate": 0.00010137742398802355,
+      "loss": 0.12251119613647461,
+      "step": 20330
+    },
+    {
+      "epoch": 0.05103202105874638,
+      "grad_norm": 0.554640531539917,
+      "learning_rate": 0.00010129687753412489,
+      "loss": 0.10172643661499023,
+      "step": 20340
+    },
+    {
+      "epoch": 0.051324300675348586,
+      "grad_norm": 0.45570510625839233,
+      "learning_rate": 0.00010121633023869648,
+      "loss": 0.09071780443191528,
+      "step": 20350
+    },
+    {
+      "epoch": 0.0516165802919508,
+      "grad_norm": 0.4316807985305786,
+      "learning_rate": 0.00010113578215400461,
+      "loss": 0.07173900008201599,
+      "step": 20360
+    },
+    {
+      "epoch": 0.05190885990855301,
+      "grad_norm": 0.6063556671142578,
+      "learning_rate": 0.000101055233332316,
+      "loss": 0.07879880666732789,
+      "step": 20370
+    },
+    {
+      "epoch": 0.052201139525155225,
+      "grad_norm": 0.38370460271835327,
+      "learning_rate": 0.00010097468382589792,
+      "loss": 0.08907070755958557,
+      "step": 20380
+    },
+    {
+      "epoch": 0.05249341914175744,
+      "grad_norm": 0.6680926084518433,
+      "learning_rate": 0.00010089413368701812,
+      "loss": 0.09986951351165771,
+      "step": 20390
+    },
+    {
+      "epoch": 0.05278569875835965,
+      "grad_norm": 0.6887987852096558,
+      "learning_rate": 0.00010081358296794465,
+      "loss": 0.10064125061035156,
+      "step": 20400
+    },
+    {
+      "epoch": 0.053077978374961865,
+      "grad_norm": 0.453978955745697,
+      "learning_rate": 0.00010073303172094606,
+      "loss": 0.10356332063674926,
+      "step": 20410
+    },
+    {
+      "epoch": 0.05337025799156408,
+      "grad_norm": 0.4814583957195282,
+      "learning_rate": 0.00010065247999829115,
+      "loss": 0.09380494356155396,
+      "step": 20420
+    },
+    {
+      "epoch": 0.05366253760816629,
+      "grad_norm": 0.599068820476532,
+      "learning_rate": 0.00010057192785224902,
+      "loss": 0.09917624592781067,
+      "step": 20430
+    },
+    {
+      "epoch": 0.053954817224768505,
+      "grad_norm": 0.5479817986488342,
+      "learning_rate": 0.00010049137533508914,
+      "loss": 0.0935619056224823,
+      "step": 20440
+    },
+    {
+      "epoch": 0.05424709684137072,
+      "grad_norm": 0.2954467236995697,
+      "learning_rate": 0.00010041082249908111,
+      "loss": 0.06764938235282898,
+      "step": 20450
+    },
+    {
+      "epoch": 0.05453937645797293,
+      "grad_norm": 0.6653916835784912,
+      "learning_rate": 0.00010033026939649485,
+      "loss": 0.15330370664596557,
+      "step": 20460
+    },
+    {
+      "epoch": 0.054831656074575144,
+      "grad_norm": 0.3209666907787323,
+      "learning_rate": 0.00010024971607960037,
+      "loss": 0.09465944766998291,
+      "step": 20470
+    },
+    {
+      "epoch": 0.05512393569117736,
+      "grad_norm": 0.48317936062812805,
+      "learning_rate": 0.00010016916260066782,
+      "loss": 0.09779481291770935,
+      "step": 20480
+    },
+    {
+      "epoch": 0.05541621530777957,
+      "grad_norm": 0.6979444026947021,
+      "learning_rate": 0.00010008860901196748,
+      "loss": 0.09411098957061767,
+      "step": 20490
+    },
+    {
+      "epoch": 0.055708494924381784,
+      "grad_norm": 0.4678073823451996,
+      "learning_rate": 0.00010000805536576973,
+      "loss": 0.07406137585639953,
+      "step": 20500
+    },
+    {
+      "epoch": 0.056000774540984,
+      "grad_norm": 0.5124133825302124,
+      "learning_rate": 9.992750171434494e-05,
+      "loss": 0.1123774528503418,
+      "step": 20510
+    },
+    {
+      "epoch": 0.05629305415758621,
+      "grad_norm": 0.43516805768013,
+      "learning_rate": 9.984694810996351e-05,
+      "loss": 0.09069861769676209,
+      "step": 20520
+    },
+    {
+      "epoch": 0.05658533377418842,
+      "grad_norm": 0.3575292229652405,
+      "learning_rate": 9.976639460489576e-05,
+      "loss": 0.07993791103363038,
+      "step": 20530
+    },
+    {
+      "epoch": 0.056877613390790636,
+      "grad_norm": 0.6375089883804321,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 0.09009307026863098,
+      "step": 20540
+    },
+    {
+      "epoch": 0.05716989300739285,
+      "grad_norm": 0.5488107204437256,
+      "learning_rate": 9.960528810178254e-05,
+      "loss": 0.08873154520988465,
+      "step": 20550
+    },
+    {
+      "epoch": 0.05746217262399506,
+      "grad_norm": 0.41564804315567017,
+      "learning_rate": 9.952473520827727e-05,
+      "loss": 0.10190014839172364,
+      "step": 20560
+    },
+    {
+      "epoch": 0.057754452240597276,
+      "grad_norm": 0.5404197573661804,
+      "learning_rate": 9.944418262316619e-05,
+      "loss": 0.09907346963882446,
+      "step": 20570
+    },
+    {
+      "epoch": 0.05804673185719949,
+      "grad_norm": 0.49153250455856323,
+      "learning_rate": 9.936363039871895e-05,
+      "loss": 0.09962036609649658,
+      "step": 20580
+    },
+    {
+      "epoch": 0.0583390114738017,
+      "grad_norm": 0.4887259304523468,
+      "learning_rate": 9.928307858720505e-05,
+      "loss": 0.07296675443649292,
+      "step": 20590
+    },
+    {
+      "epoch": 0.058631291090403916,
+      "grad_norm": 0.5231586694717407,
+      "learning_rate": 9.920252724089367e-05,
+      "loss": 0.10240349769592286,
+      "step": 20600
+    },
+    {
+      "epoch": 0.05892357070700612,
+      "grad_norm": 0.6107679009437561,
+      "learning_rate": 9.912197641205368e-05,
+      "loss": 0.1289135217666626,
+      "step": 20610
+    },
+    {
+      "epoch": 0.059215850323608335,
+      "grad_norm": 0.4876420795917511,
+      "learning_rate": 9.90414261529537e-05,
+      "loss": 0.09926280379295349,
+      "step": 20620
+    },
+    {
+      "epoch": 0.05950812994021055,
+      "grad_norm": 0.5603168606758118,
+      "learning_rate": 9.896087651586183e-05,
+      "loss": 0.10614444017410278,
+      "step": 20630
+    },
+    {
+      "epoch": 0.05980040955681276,
+      "grad_norm": 0.41040411591529846,
+      "learning_rate": 9.888032755304597e-05,
+      "loss": 0.08651268482208252,
+      "step": 20640
+    },
+    {
+      "epoch": 0.060092689173414975,
+      "grad_norm": 0.5339871644973755,
+      "learning_rate": 9.879977931677339e-05,
+      "loss": 0.1002271056175232,
+      "step": 20650
+    },
+    {
+      "epoch": 0.06038496879001719,
+      "grad_norm": 0.7043265700340271,
+      "learning_rate": 9.871923185931094e-05,
+      "loss": 0.08061819672584533,
+      "step": 20660
+    },
+    {
+      "epoch": 0.0606772484066194,
+      "grad_norm": 0.5859558582305908,
+      "learning_rate": 9.863868523292509e-05,
+      "loss": 0.06576159596443176,
+      "step": 20670
+    },
+    {
+      "epoch": 0.060969528023221614,
+      "grad_norm": 0.4875462055206299,
+      "learning_rate": 9.855813948988155e-05,
+      "loss": 0.11136139631271362,
+      "step": 20680
+    },
+    {
+      "epoch": 0.06126180763982383,
+      "grad_norm": 0.40205755829811096,
+      "learning_rate": 9.847759468244569e-05,
+      "loss": 0.06584960222244263,
+      "step": 20690
+    },
+    {
+      "epoch": 0.06155408725642604,
+      "grad_norm": 0.44005072116851807,
+      "learning_rate": 9.839705086288206e-05,
+      "loss": 0.09514769911766052,
+      "step": 20700
+    },
+    {
+      "epoch": 0.061846366873028254,
+      "grad_norm": 0.4326503276824951,
+      "learning_rate": 9.831650808345478e-05,
+      "loss": 0.10229644775390626,
+      "step": 20710
+    },
+    {
+      "epoch": 0.06213864648963047,
+      "grad_norm": 0.5452603101730347,
+      "learning_rate": 9.823596639642711e-05,
+      "loss": 0.07525787949562072,
+      "step": 20720
+    },
+    {
+      "epoch": 0.06243092610623268,
+      "grad_norm": 0.5008875727653503,
+      "learning_rate": 9.815542585406167e-05,
+      "loss": 0.14297192096710204,
+      "step": 20730
+    },
+    {
+      "epoch": 0.0627232057228349,
+      "grad_norm": 0.4960061013698578,
+      "learning_rate": 9.80748865086204e-05,
+      "loss": 0.11484919786453247,
+      "step": 20740
+    },
+    {
+      "epoch": 0.0630154853394371,
+      "grad_norm": 0.5288963913917542,
+      "learning_rate": 9.799434841236436e-05,
+      "loss": 0.13866137266159057,
+      "step": 20750
+    },
+    {
+      "epoch": 0.06330776495603932,
+      "grad_norm": 0.559029221534729,
+      "learning_rate": 9.791381161755387e-05,
+      "loss": 0.11800849437713623,
+      "step": 20760
+    },
+    {
+      "epoch": 0.06360004457264153,
+      "grad_norm": 0.4200463593006134,
+      "learning_rate": 9.783327617644837e-05,
+      "loss": 0.10539906024932862,
+      "step": 20770
+    },
+    {
+      "epoch": 0.06389232418924375,
+      "grad_norm": 0.4462909400463104,
+      "learning_rate": 9.775274214130641e-05,
+      "loss": 0.10181390047073365,
+      "step": 20780
+    },
+    {
+      "epoch": 0.06418460380584595,
+      "grad_norm": 0.4182778298854828,
+      "learning_rate": 9.76722095643857e-05,
+      "loss": 0.10097401142120362,
+      "step": 20790
+    },
+    {
+      "epoch": 0.06447688342244817,
+      "grad_norm": 0.5911705493927002,
+      "learning_rate": 9.759167849794292e-05,
+      "loss": 0.11013593673706054,
+      "step": 20800
+    },
+    {
+      "epoch": 0.06476916303905038,
+      "grad_norm": 0.4312324523925781,
+      "learning_rate": 9.751114899423383e-05,
+      "loss": 0.11460916996002198,
+      "step": 20810
+    },
+    {
+      "epoch": 0.0650614426556526,
+      "grad_norm": 0.5479062795639038,
+      "learning_rate": 9.743062110551312e-05,
+      "loss": 0.10702826976776122,
+      "step": 20820
+    },
+    {
+      "epoch": 0.0653537222722548,
+      "grad_norm": 0.5323778986930847,
+      "learning_rate": 9.735009488403448e-05,
+      "loss": 0.09923620820045471,
+      "step": 20830
+    },
+    {
+      "epoch": 0.06564600188885703,
+      "grad_norm": 0.4252351224422455,
+      "learning_rate": 9.726957038205052e-05,
+      "loss": 0.11125737428665161,
+      "step": 20840
+    },
+    {
+      "epoch": 0.06593828150545923,
+      "grad_norm": 0.49175313115119934,
+      "learning_rate": 9.718904765181265e-05,
+      "loss": 0.11236046552658081,
+      "step": 20850
+    },
+    {
+      "epoch": 0.06623056112206145,
+      "grad_norm": 0.5369746685028076,
+      "learning_rate": 9.710852674557131e-05,
+      "loss": 0.08975967168807983,
+      "step": 20860
+    },
+    {
+      "epoch": 0.06652284073866366,
+      "grad_norm": 0.4693719148635864,
+      "learning_rate": 9.702800771557553e-05,
+      "loss": 0.13159210681915284,
+      "step": 20870
+    },
+    {
+      "epoch": 0.06681512035526588,
+      "grad_norm": 0.4856756627559662,
+      "learning_rate": 9.694749061407326e-05,
+      "loss": 0.09838325977325439,
+      "step": 20880
+    },
+    {
+      "epoch": 0.06710739997186808,
+      "grad_norm": 0.47578534483909607,
+      "learning_rate": 9.686697549331124e-05,
+      "loss": 0.0919321358203888,
+      "step": 20890
+    },
+    {
+      "epoch": 0.0673996795884703,
+      "grad_norm": 0.47636350989341736,
+      "learning_rate": 9.678646240553474e-05,
+      "loss": 0.11761914491653443,
+      "step": 20900
+    },
+    {
+      "epoch": 0.06769195920507251,
+      "grad_norm": 0.6965683698654175,
+      "learning_rate": 9.670595140298794e-05,
+      "loss": 0.1782463312149048,
+      "step": 20910
+    },
+    {
+      "epoch": 0.06798423882167473,
+      "grad_norm": 0.5787543058395386,
+      "learning_rate": 9.66254425379135e-05,
+      "loss": 0.19806710481643677,
+      "step": 20920
+    },
+    {
+      "epoch": 0.06827651843827694,
+      "grad_norm": 0.6710638403892517,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.15069743394851684,
+      "step": 20930
+    },
+    {
+      "epoch": 0.06856879805487916,
+      "grad_norm": 0.5006418228149414,
+      "learning_rate": 9.646443142914567e-05,
+      "loss": 0.1656505823135376,
+      "step": 20940
+    },
+    {
+      "epoch": 0.06886107767148136,
+      "grad_norm": 0.5066624879837036,
+      "learning_rate": 9.638392928993059e-05,
+      "loss": 0.20668559074401854,
+      "step": 20950
+    },
+    {
+      "epoch": 0.06915335728808358,
+      "grad_norm": 0.5155174732208252,
+      "learning_rate": 9.630342949714454e-05,
+      "loss": 0.19447414875030516,
+      "step": 20960
+    },
+    {
+      "epoch": 0.06944563690468579,
+      "grad_norm": 0.39817699790000916,
+      "learning_rate": 9.622293210302297e-05,
+      "loss": 0.15492986440658568,
+      "step": 20970
+    },
+    {
+      "epoch": 0.06973791652128801,
+      "grad_norm": 0.5270841717720032,
+      "learning_rate": 9.614243715979976e-05,
+      "loss": 0.1536700487136841,
+      "step": 20980
+    },
+    {
+      "epoch": 0.07003019613789022,
+      "grad_norm": 0.34332746267318726,
+      "learning_rate": 9.606194471970718e-05,
+      "loss": 0.1601084589958191,
+      "step": 20990
+    },
+    {
+      "epoch": 0.07032247575449242,
+      "grad_norm": 0.508064866065979,
+      "learning_rate": 9.598145483497592e-05,
+      "loss": 0.15211740732192994,
+      "step": 21000
+    },
+    {
+      "epoch": 0.07061475537109464,
+      "grad_norm": 0.43824347853660583,
+      "learning_rate": 9.5900967557835e-05,
+      "loss": 0.14174535274505615,
+      "step": 21010
+    },
+    {
+      "epoch": 0.07090703498769685,
+      "grad_norm": 0.4774963855743408,
+      "learning_rate": 9.58204829405117e-05,
+      "loss": 0.14750251770019532,
+      "step": 21020
+    },
+    {
+      "epoch": 0.07119931460429907,
+      "grad_norm": 0.4684067368507385,
+      "learning_rate": 9.574000103523168e-05,
+      "loss": 0.11595194339752198,
+      "step": 21030
+    },
+    {
+      "epoch": 0.07149159422090128,
+      "grad_norm": 0.5429431796073914,
+      "learning_rate": 9.565952189421872e-05,
+      "loss": 0.17488476037979125,
+      "step": 21040
+    },
+    {
+      "epoch": 0.0717838738375035,
+      "grad_norm": 0.5405802130699158,
+      "learning_rate": 9.557904556969487e-05,
+      "loss": 0.1737047791481018,
+      "step": 21050
+    },
+    {
+      "epoch": 0.0720761534541057,
+      "grad_norm": 0.5695708990097046,
+      "learning_rate": 9.549857211388037e-05,
+      "loss": 0.1415093421936035,
+      "step": 21060
+    },
+    {
+      "epoch": 0.07236843307070792,
+      "grad_norm": 0.3601032793521881,
+      "learning_rate": 9.541810157899351e-05,
+      "loss": 0.1669877886772156,
+      "step": 21070
+    },
+    {
+      "epoch": 0.07266071268731013,
+      "grad_norm": 0.46882906556129456,
+      "learning_rate": 9.533763401725085e-05,
+      "loss": 0.18201847076416017,
+      "step": 21080
+    },
+    {
+      "epoch": 0.07295299230391235,
+      "grad_norm": 0.487192302942276,
+      "learning_rate": 9.525716948086685e-05,
+      "loss": 0.1555421233177185,
+      "step": 21090
+    },
+    {
+      "epoch": 0.07324527192051455,
+      "grad_norm": 0.49152183532714844,
+      "learning_rate": 9.517670802205403e-05,
+      "loss": 0.18150897026062013,
+      "step": 21100
+    },
+    {
+      "epoch": 0.07353755153711677,
+      "grad_norm": 0.5134358406066895,
+      "learning_rate": 9.509624969302308e-05,
+      "loss": 0.18555808067321777,
+      "step": 21110
+    },
+    {
+      "epoch": 0.07382983115371898,
+      "grad_norm": 0.4814067482948303,
+      "learning_rate": 9.501579454598239e-05,
+      "loss": 0.1541232943534851,
+      "step": 21120
+    },
+    {
+      "epoch": 0.0741221107703212,
+      "grad_norm": 0.5187480449676514,
+      "learning_rate": 9.493534263313855e-05,
+      "loss": 0.15459911823272704,
+      "step": 21130
+    },
+    {
+      "epoch": 0.0744143903869234,
+      "grad_norm": 0.4969259202480316,
+      "learning_rate": 9.48548940066959e-05,
+      "loss": 0.16813572645187377,
+      "step": 21140
+    },
+    {
+      "epoch": 0.07470667000352563,
+      "grad_norm": 0.5916034579277039,
+      "learning_rate": 9.477444871885664e-05,
+      "loss": 0.15696629285812377,
+      "step": 21150
+    },
+    {
+      "epoch": 0.07499894962012783,
+      "grad_norm": 0.4365858733654022,
+      "learning_rate": 9.469400682182088e-05,
+      "loss": 0.15727308988571168,
+      "step": 21160
+    },
+    {
+      "epoch": 0.07529122923673005,
+      "grad_norm": 0.4753931164741516,
+      "learning_rate": 9.46135683677865e-05,
+      "loss": 0.19089047908782958,
+      "step": 21170
+    },
+    {
+      "epoch": 0.07558350885333226,
+      "grad_norm": 0.3823680281639099,
+      "learning_rate": 9.453313340894913e-05,
+      "loss": 0.18855875730514526,
+      "step": 21180
+    },
+    {
+      "epoch": 0.07587578846993448,
+      "grad_norm": 0.6248448491096497,
+      "learning_rate": 9.445270199750212e-05,
+      "loss": 0.1365087866783142,
+      "step": 21190
+    },
+    {
+      "epoch": 0.07616806808653669,
+      "grad_norm": 0.6689746975898743,
+      "learning_rate": 9.437227418563657e-05,
+      "loss": 0.32422780990600586,
+      "step": 21200
+    },
+    {
+      "epoch": 0.0764603477031389,
+      "grad_norm": 0.33717310428619385,
+      "learning_rate": 9.42918500255412e-05,
+      "loss": 0.29150590896606443,
+      "step": 21210
+    },
+    {
+      "epoch": 0.07675262731974111,
+      "grad_norm": 0.38250431418418884,
+      "learning_rate": 9.421142956940237e-05,
+      "loss": 0.22583436965942383,
+      "step": 21220
+    },
+    {
+      "epoch": 0.07704490693634333,
+      "grad_norm": 0.606807291507721,
+      "learning_rate": 9.413101286940405e-05,
+      "loss": 0.2764282703399658,
+      "step": 21230
+    },
+    {
+      "epoch": 0.07733718655294554,
+      "grad_norm": 0.5241100192070007,
+      "learning_rate": 9.405059997772773e-05,
+      "loss": 0.29698355197906495,
+      "step": 21240
+    },
+    {
+      "epoch": 0.07762946616954776,
+      "grad_norm": 0.6197025179862976,
+      "learning_rate": 9.397019094655255e-05,
+      "loss": 0.31229093074798586,
+      "step": 21250
+    },
+    {
+      "epoch": 0.07792174578614997,
+      "grad_norm": 0.4762765169143677,
+      "learning_rate": 9.3889785828055e-05,
+      "loss": 0.26779415607452395,
+      "step": 21260
+    },
+    {
+      "epoch": 0.07821402540275217,
+      "grad_norm": 0.4839880168437958,
+      "learning_rate": 9.380938467440902e-05,
+      "loss": 0.282796573638916,
+      "step": 21270
+    },
+    {
+      "epoch": 0.07850630501935439,
+      "grad_norm": 0.6035153269767761,
+      "learning_rate": 9.372898753778619e-05,
+      "loss": 0.3550250768661499,
+      "step": 21280
+    },
+    {
+      "epoch": 0.0787985846359566,
+      "grad_norm": 0.4269202947616577,
+      "learning_rate": 9.36485944703552e-05,
+      "loss": 0.23325304985046386,
+      "step": 21290
+    },
+    {
+      "epoch": 0.07909086425255882,
+      "grad_norm": 0.4498310983181,
+      "learning_rate": 9.356820552428232e-05,
+      "loss": 0.24719438552856446,
+      "step": 21300
+    },
+    {
+      "epoch": 0.07938314386916102,
+      "grad_norm": 0.3350340723991394,
+      "learning_rate": 9.34878207517311e-05,
+      "loss": 0.2422471284866333,
+      "step": 21310
+    },
+    {
+      "epoch": 0.07967542348576324,
+      "grad_norm": 0.3384961187839508,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.2570115804672241,
+      "step": 21320
+    },
+    {
+      "epoch": 0.07996770310236545,
+      "grad_norm": 0.4409782290458679,
+      "learning_rate": 9.332706393583385e-05,
+      "loss": 0.24333391189575196,
+      "step": 21330
+    },
+    {
+      "epoch": 0.08025998271896767,
+      "grad_norm": 0.45633405447006226,
+      "learning_rate": 9.324669199680123e-05,
+      "loss": 0.25600931644439695,
+      "step": 21340
+    },
+    {
+      "epoch": 0.08055226233556988,
+      "grad_norm": 0.4886617958545685,
+      "learning_rate": 9.316632443991688e-05,
+      "loss": 0.2908273935317993,
+      "step": 21350
+    },
+    {
+      "epoch": 0.0808445419521721,
+      "grad_norm": 0.5876643657684326,
+      "learning_rate": 9.30859613173304e-05,
+      "loss": 0.31505000591278076,
+      "step": 21360
+    },
+    {
+      "epoch": 0.0811368215687743,
+      "grad_norm": 0.4913652837276459,
+      "learning_rate": 9.300560268118855e-05,
+      "loss": 0.2758547067642212,
+      "step": 21370
+    },
+    {
+      "epoch": 0.08142910118537652,
+      "grad_norm": 0.3628678023815155,
+      "learning_rate": 9.292524858363518e-05,
+      "loss": 0.29413058757781985,
+      "step": 21380
+    },
+    {
+      "epoch": 0.08172138080197873,
+      "grad_norm": 0.4525972902774811,
+      "learning_rate": 9.284489907681119e-05,
+      "loss": 0.2655428409576416,
+      "step": 21390
+    },
+    {
+      "epoch": 0.08201366041858095,
+      "grad_norm": 0.5577532052993774,
+      "learning_rate": 9.276455421285453e-05,
+      "loss": 0.2434457540512085,
+      "step": 21400
+    },
+    {
+      "epoch": 0.08230594003518316,
+      "grad_norm": 0.36672407388687134,
+      "learning_rate": 9.268421404390006e-05,
+      "loss": 0.2589518308639526,
+      "step": 21410
+    },
+    {
+      "epoch": 0.08259821965178538,
+      "grad_norm": 0.5585358738899231,
+      "learning_rate": 9.260387862207968e-05,
+      "loss": 0.28243749141693114,
+      "step": 21420
+    },
+    {
+      "epoch": 0.08289049926838758,
+      "grad_norm": 0.5466753244400024,
+      "learning_rate": 9.252354799952219e-05,
+      "loss": 0.25502901077270507,
+      "step": 21430
+    },
+    {
+      "epoch": 0.0831827788849898,
+      "grad_norm": 0.4764811396598816,
+      "learning_rate": 9.244322222835319e-05,
+      "loss": 0.24922637939453124,
+      "step": 21440
+    },
+    {
+      "epoch": 0.08347505850159201,
+      "grad_norm": 0.5183106660842896,
+      "learning_rate": 9.236290136069528e-05,
+      "loss": 0.288109564781189,
+      "step": 21450
+    },
+    {
+      "epoch": 0.08376733811819423,
+      "grad_norm": 0.5006160736083984,
+      "learning_rate": 9.22825854486677e-05,
+      "loss": 0.2728541851043701,
+      "step": 21460
+    },
+    {
+      "epoch": 0.08405961773479644,
+      "grad_norm": 0.5149085521697998,
+      "learning_rate": 9.220227454438672e-05,
+      "loss": 0.28053874969482423,
+      "step": 21470
+    },
+    {
+      "epoch": 0.08435189735139866,
+      "grad_norm": 0.3561913073062897,
+      "learning_rate": 9.21219686999651e-05,
+      "loss": 0.23067166805267333,
+      "step": 21480
+    },
+    {
+      "epoch": 0.08464417696800086,
+      "grad_norm": 0.416746586561203,
+      "learning_rate": 9.20416679675124e-05,
+      "loss": 0.2582935094833374,
+      "step": 21490
+    },
+    {
+      "epoch": 0.08493645658460308,
+      "grad_norm": 0.3843415677547455,
+      "learning_rate": 9.196137239913504e-05,
+      "loss": 0.26363263130187986,
+      "step": 21500
+    },
+    {
+      "epoch": 0.08522873620120529,
+      "grad_norm": 0.5268052816390991,
+      "learning_rate": 9.18810820469358e-05,
+      "loss": 0.3635519504547119,
+      "step": 21510
+    },
+    {
+      "epoch": 0.0855210158178075,
+      "grad_norm": 0.5326135754585266,
+      "learning_rate": 9.18007969630143e-05,
+      "loss": 0.275510573387146,
+      "step": 21520
+    },
+    {
+      "epoch": 0.08581329543440971,
+      "grad_norm": 0.4853200614452362,
+      "learning_rate": 9.172051719946665e-05,
+      "loss": 0.2593245983123779,
+      "step": 21530
+    },
+    {
+      "epoch": 0.08610557505101192,
+      "grad_norm": 0.41385698318481445,
+      "learning_rate": 9.164024280838544e-05,
+      "loss": 0.22775835990905763,
+      "step": 21540
+    },
+    {
+      "epoch": 0.08639785466761414,
+      "grad_norm": 0.4098763167858124,
+      "learning_rate": 9.155997384185997e-05,
+      "loss": 0.24704959392547607,
+      "step": 21550
+    },
+    {
+      "epoch": 0.08669013428421635,
+      "grad_norm": 0.3574011027812958,
+      "learning_rate": 9.14797103519758e-05,
+      "loss": 0.2609837532043457,
+      "step": 21560
+    },
+    {
+      "epoch": 0.08698241390081857,
+      "grad_norm": 0.48144325613975525,
+      "learning_rate": 9.139945239081511e-05,
+      "loss": 0.2421656608581543,
+      "step": 21570
+    },
+    {
+      "epoch": 0.08727469351742077,
+      "grad_norm": 0.4453323483467102,
+      "learning_rate": 9.131920001045638e-05,
+      "loss": 0.2405714988708496,
+      "step": 21580
+    },
+    {
+      "epoch": 0.087566973134023,
+      "grad_norm": 0.5210952758789062,
+      "learning_rate": 9.123895326297451e-05,
+      "loss": 0.21732499599456787,
+      "step": 21590
+    },
+    {
+      "epoch": 0.0878592527506252,
+      "grad_norm": 0.3638177812099457,
+      "learning_rate": 9.115871220044076e-05,
+      "loss": 0.27810077667236327,
+      "step": 21600
+    },
+    {
+      "epoch": 0.08815153236722742,
+      "grad_norm": 0.469762921333313,
+      "learning_rate": 9.107847687492267e-05,
+      "loss": 0.32399008274078367,
+      "step": 21610
+    },
+    {
+      "epoch": 0.08844381198382963,
+      "grad_norm": 0.46834689378738403,
+      "learning_rate": 9.099824733848408e-05,
+      "loss": 0.3033536672592163,
+      "step": 21620
+    },
+    {
+      "epoch": 0.08873609160043185,
+      "grad_norm": 0.44186925888061523,
+      "learning_rate": 9.091802364318507e-05,
+      "loss": 0.24690959453582764,
+      "step": 21630
+    },
+    {
+      "epoch": 0.08902837121703405,
+      "grad_norm": 0.4245443344116211,
+      "learning_rate": 9.083780584108189e-05,
+      "loss": 0.24903411865234376,
+      "step": 21640
+    },
+    {
+      "epoch": 0.08932065083363627,
+      "grad_norm": 0.5282859802246094,
+      "learning_rate": 9.075759398422705e-05,
+      "loss": 0.32451753616333007,
+      "step": 21650
+    },
+    {
+      "epoch": 0.08961293045023848,
+      "grad_norm": 0.4241032898426056,
+      "learning_rate": 9.067738812466912e-05,
+      "loss": 0.2569368600845337,
+      "step": 21660
+    },
+    {
+      "epoch": 0.0899052100668407,
+      "grad_norm": 0.5115688443183899,
+      "learning_rate": 9.059718831445283e-05,
+      "loss": 0.2512019872665405,
+      "step": 21670
+    },
+    {
+      "epoch": 0.0901974896834429,
+      "grad_norm": 0.3615689277648926,
+      "learning_rate": 9.05169946056189e-05,
+      "loss": 0.23099949359893798,
+      "step": 21680
+    },
+    {
+      "epoch": 0.09048976930004513,
+      "grad_norm": 0.5114195942878723,
+      "learning_rate": 9.043680705020431e-05,
+      "loss": 0.25421552658081054,
+      "step": 21690
+    },
+    {
+      "epoch": 0.09078204891664733,
+      "grad_norm": 0.5340844392776489,
+      "learning_rate": 9.035662570024177e-05,
+      "loss": 0.24773926734924318,
+      "step": 21700
+    },
+    {
+      "epoch": 0.09107432853324955,
+      "grad_norm": 0.45107701420783997,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.2503213882446289,
+      "step": 21710
+    },
+    {
+      "epoch": 0.09136660814985176,
+      "grad_norm": 0.4170847535133362,
+      "learning_rate": 9.019628182478405e-05,
+      "loss": 0.23387515544891357,
+      "step": 21720
+    },
+    {
+      "epoch": 0.09165888776645398,
+      "grad_norm": 0.4229448735713959,
+      "learning_rate": 9.011611940333428e-05,
+      "loss": 0.2676738739013672,
+      "step": 21730
+    },
+    {
+      "epoch": 0.09195116738305618,
+      "grad_norm": 0.6297507882118225,
+      "learning_rate": 9.003596339542735e-05,
+      "loss": 0.260935640335083,
+      "step": 21740
+    },
+    {
+      "epoch": 0.0922434469996584,
+      "grad_norm": 0.6706903576850891,
+      "learning_rate": 8.99558138530756e-05,
+      "loss": 0.2841729402542114,
+      "step": 21750
+    },
+    {
+      "epoch": 0.09253572661626061,
+      "grad_norm": 0.4275798797607422,
+      "learning_rate": 8.987567082828717e-05,
+      "loss": 0.21058928966522217,
+      "step": 21760
+    },
+    {
+      "epoch": 0.09282800623286283,
+      "grad_norm": 0.4024621248245239,
+      "learning_rate": 8.979553437306604e-05,
+      "loss": 0.22616417407989503,
+      "step": 21770
+    },
+    {
+      "epoch": 0.09312028584946504,
+      "grad_norm": 0.4971884489059448,
+      "learning_rate": 8.971540453941183e-05,
+      "loss": 0.26322891712188723,
+      "step": 21780
+    },
+    {
+      "epoch": 0.09341256546606724,
+      "grad_norm": 0.3559616804122925,
+      "learning_rate": 8.963528137932e-05,
+      "loss": 0.20896472930908203,
+      "step": 21790
+    },
+    {
+      "epoch": 0.09370484508266946,
+      "grad_norm": 0.4396965205669403,
+      "learning_rate": 8.955516494478155e-05,
+      "loss": 0.2522958993911743,
+      "step": 21800
+    },
+    {
+      "epoch": 0.09399712469927167,
+      "grad_norm": 0.4932815134525299,
+      "learning_rate": 8.947505528778315e-05,
+      "loss": 0.2599222183227539,
+      "step": 21810
+    },
+    {
+      "epoch": 0.09428940431587389,
+      "grad_norm": 0.3708024322986603,
+      "learning_rate": 8.939495246030712e-05,
+      "loss": 0.2251713514328003,
+      "step": 21820
+    },
+    {
+      "epoch": 0.0945816839324761,
+      "grad_norm": 0.5677950382232666,
+      "learning_rate": 8.93148565143313e-05,
+      "loss": 0.26872122287750244,
+      "step": 21830
+    },
+    {
+      "epoch": 0.09487396354907832,
+      "grad_norm": 0.5632646083831787,
+      "learning_rate": 8.923476750182908e-05,
+      "loss": 0.2765017509460449,
+      "step": 21840
+    },
+    {
+      "epoch": 0.09516624316568052,
+      "grad_norm": 0.3720809519290924,
+      "learning_rate": 8.915468547476936e-05,
+      "loss": 0.26583223342895507,
+      "step": 21850
+    },
+    {
+      "epoch": 0.09545852278228274,
+      "grad_norm": 0.4064728915691376,
+      "learning_rate": 8.907461048511646e-05,
+      "loss": 0.19693869352340698,
+      "step": 21860
+    },
+    {
+      "epoch": 0.09575080239888495,
+      "grad_norm": 0.5118590593338013,
+      "learning_rate": 8.899454258483024e-05,
+      "loss": 0.25725178718566893,
+      "step": 21870
+    },
+    {
+      "epoch": 0.09604308201548717,
+      "grad_norm": 0.39331305027008057,
+      "learning_rate": 8.89144818258658e-05,
+      "loss": 0.19522465467453004,
+      "step": 21880
+    },
+    {
+      "epoch": 0.09633536163208938,
+      "grad_norm": 0.37807610630989075,
+      "learning_rate": 8.883442826017381e-05,
+      "loss": 0.2611843585968018,
+      "step": 21890
+    },
+    {
+      "epoch": 0.0966276412486916,
+      "grad_norm": 0.4385603368282318,
+      "learning_rate": 8.875438193970003e-05,
+      "loss": 0.27183549404144286,
+      "step": 21900
+    },
+    {
+      "epoch": 0.0969199208652938,
+      "grad_norm": 0.34065866470336914,
+      "learning_rate": 8.867434291638577e-05,
+      "loss": 0.29387285709381106,
+      "step": 21910
+    },
+    {
+      "epoch": 0.09721220048189602,
+      "grad_norm": 0.5286417007446289,
+      "learning_rate": 8.859431124216745e-05,
+      "loss": 0.2083909273147583,
+      "step": 21920
+    },
+    {
+      "epoch": 0.09750448009849823,
+      "grad_norm": 0.5283611416816711,
+      "learning_rate": 8.851428696897667e-05,
+      "loss": 0.2495426893234253,
+      "step": 21930
+    },
+    {
+      "epoch": 0.09779675971510045,
+      "grad_norm": 0.490267813205719,
+      "learning_rate": 8.843427014874041e-05,
+      "loss": 0.22194223403930663,
+      "step": 21940
+    },
+    {
+      "epoch": 0.09808903933170265,
+      "grad_norm": 0.5040584802627563,
+      "learning_rate": 8.835426083338068e-05,
+      "loss": 0.2528690338134766,
+      "step": 21950
+    },
+    {
+      "epoch": 0.09838131894830487,
+      "grad_norm": 0.6567689776420593,
+      "learning_rate": 8.827425907481466e-05,
+      "loss": 0.2575475454330444,
+      "step": 21960
+    },
+    {
+      "epoch": 0.09867359856490708,
+      "grad_norm": 0.5918784141540527,
+      "learning_rate": 8.819426492495464e-05,
+      "loss": 0.1962011694908142,
+      "step": 21970
+    },
+    {
+      "epoch": 0.0989658781815093,
+      "grad_norm": 0.5246418714523315,
+      "learning_rate": 8.811427843570791e-05,
+      "loss": 0.248160457611084,
+      "step": 21980
+    },
+    {
+      "epoch": 0.09925815779811151,
+      "grad_norm": 0.7165157794952393,
+      "learning_rate": 8.803429965897688e-05,
+      "loss": 0.2671104192733765,
+      "step": 21990
+    },
+    {
+      "epoch": 0.09955043741471373,
+      "grad_norm": 0.471674382686615,
+      "learning_rate": 8.795432864665888e-05,
+      "loss": 0.24709248542785645,
+      "step": 22000
+    },
+    {
+      "epoch": 0.09984271703131593,
+      "grad_norm": 0.45702648162841797,
+      "learning_rate": 8.787436545064626e-05,
+      "loss": 0.19992809295654296,
+      "step": 22010
+    },
+    {
+      "epoch": 0.10013499664791815,
+      "grad_norm": 0.4930286407470703,
+      "learning_rate": 8.779441012282625e-05,
+      "loss": 0.29296419620513914,
+      "step": 22020
+    },
+    {
+      "epoch": 0.10042727626452036,
+      "grad_norm": 0.44137945771217346,
+      "learning_rate": 8.771446271508097e-05,
+      "loss": 0.21670725345611572,
+      "step": 22030
+    },
+    {
+      "epoch": 0.10071955588112257,
+      "grad_norm": 0.38750019669532776,
+      "learning_rate": 8.763452327928749e-05,
+      "loss": 0.20564742088317872,
+      "step": 22040
+    },
+    {
+      "epoch": 0.10101183549772479,
+      "grad_norm": 0.3885003626346588,
+      "learning_rate": 8.755459186731758e-05,
+      "loss": 0.2209414005279541,
+      "step": 22050
+    },
+    {
+      "epoch": 0.10130411511432699,
+      "grad_norm": 0.5545926690101624,
+      "learning_rate": 8.74746685310379e-05,
+      "loss": 0.2535659074783325,
+      "step": 22060
+    },
+    {
+      "epoch": 0.10159639473092921,
+      "grad_norm": 0.5099805593490601,
+      "learning_rate": 8.739475332230982e-05,
+      "loss": 0.2164079427719116,
+      "step": 22070
+    },
+    {
+      "epoch": 0.10188867434753142,
+      "grad_norm": 0.4496231973171234,
+      "learning_rate": 8.731484629298943e-05,
+      "loss": 0.25582573413848875,
+      "step": 22080
+    },
+    {
+      "epoch": 0.10218095396413364,
+      "grad_norm": 0.3391959071159363,
+      "learning_rate": 8.72349474949276e-05,
+      "loss": 0.2223902940750122,
+      "step": 22090
+    },
+    {
+      "epoch": 0.10247323358073585,
+      "grad_norm": 0.4856221675872803,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.2591273069381714,
+      "step": 22100
+    },
+    {
+      "epoch": 0.10276551319733807,
+      "grad_norm": 0.604049026966095,
+      "learning_rate": 8.707517479995595e-05,
+      "loss": 0.2543874979019165,
+      "step": 22110
+    },
+    {
+      "epoch": 0.10305779281394027,
+      "grad_norm": 0.44738709926605225,
+      "learning_rate": 8.699530100672091e-05,
+      "loss": 0.22615392208099366,
+      "step": 22120
+    },
+    {
+      "epoch": 0.10335007243054249,
+      "grad_norm": 0.5753700733184814,
+      "learning_rate": 8.691543565209382e-05,
+      "loss": 0.26395750045776367,
+      "step": 22130
+    },
+    {
+      "epoch": 0.1036423520471447,
+      "grad_norm": 0.3914888799190521,
+      "learning_rate": 8.683557878789854e-05,
+      "loss": 0.2544043779373169,
+      "step": 22140
+    },
+    {
+      "epoch": 0.10393463166374692,
+      "grad_norm": 0.4133894741535187,
+      "learning_rate": 8.67557304659532e-05,
+      "loss": 0.25021748542785643,
+      "step": 22150
+    },
+    {
+      "epoch": 0.10422691128034912,
+      "grad_norm": 0.4265996813774109,
+      "learning_rate": 8.66758907380706e-05,
+      "loss": 0.19520456790924073,
+      "step": 22160
+    },
+    {
+      "epoch": 0.10451919089695134,
+      "grad_norm": 0.5155472159385681,
+      "learning_rate": 8.659605965605784e-05,
+      "loss": 0.22112524509429932,
+      "step": 22170
+    },
+    {
+      "epoch": 0.10481147051355355,
+      "grad_norm": 0.42998239398002625,
+      "learning_rate": 8.651623727171646e-05,
+      "loss": 0.22288107872009277,
+      "step": 22180
+    },
+    {
+      "epoch": 0.10510375013015577,
+      "grad_norm": 0.374395489692688,
+      "learning_rate": 8.643642363684232e-05,
+      "loss": 0.2607407808303833,
+      "step": 22190
+    },
+    {
+      "epoch": 0.10539602974675798,
+      "grad_norm": 0.43034663796424866,
+      "learning_rate": 8.63566188032256e-05,
+      "loss": 0.24510629177093507,
+      "step": 22200
+    },
+    {
+      "epoch": 0.1056883093633602,
+      "grad_norm": 0.378001868724823,
+      "learning_rate": 8.627682282265083e-05,
+      "loss": 0.2541597604751587,
+      "step": 22210
+    },
+    {
+      "epoch": 0.1059805889799624,
+      "grad_norm": 0.40215080976486206,
+      "learning_rate": 8.619703574689675e-05,
+      "loss": 0.23885085582733154,
+      "step": 22220
+    },
+    {
+      "epoch": 0.10627286859656462,
+      "grad_norm": 0.5046709775924683,
+      "learning_rate": 8.611725762773631e-05,
+      "loss": 0.1873058557510376,
+      "step": 22230
+    },
+    {
+      "epoch": 0.10656514821316683,
+      "grad_norm": 0.38730189204216003,
+      "learning_rate": 8.603748851693668e-05,
+      "loss": 0.24491121768951415,
+      "step": 22240
+    },
+    {
+      "epoch": 0.10685742782976905,
+      "grad_norm": 0.5529153347015381,
+      "learning_rate": 8.595772846625913e-05,
+      "loss": 0.26360013484954836,
+      "step": 22250
+    },
+    {
+      "epoch": 0.10714970744637126,
+      "grad_norm": 0.6121418476104736,
+      "learning_rate": 8.587797752745917e-05,
+      "loss": 0.2643434524536133,
+      "step": 22260
+    },
+    {
+      "epoch": 0.10744198706297348,
+      "grad_norm": 0.4332280457019806,
+      "learning_rate": 8.579823575228623e-05,
+      "loss": 0.19682024717330932,
+      "step": 22270
+    },
+    {
+      "epoch": 0.10773426667957568,
+      "grad_norm": 0.49118250608444214,
+      "learning_rate": 8.571850319248396e-05,
+      "loss": 0.24338388442993164,
+      "step": 22280
+    },
+    {
+      "epoch": 0.1080265462961779,
+      "grad_norm": 0.40187472105026245,
+      "learning_rate": 8.56387798997899e-05,
+      "loss": 0.30756092071533203,
+      "step": 22290
+    },
+    {
+      "epoch": 0.10831882591278011,
+      "grad_norm": 0.5171650648117065,
+      "learning_rate": 8.555906592593563e-05,
+      "loss": 0.290974497795105,
+      "step": 22300
+    },
+    {
+      "epoch": 0.10861110552938232,
+      "grad_norm": 0.5001437067985535,
+      "learning_rate": 8.547936132264673e-05,
+      "loss": 0.22287099361419677,
+      "step": 22310
+    },
+    {
+      "epoch": 0.10890338514598454,
+      "grad_norm": 0.4549509286880493,
+      "learning_rate": 8.539966614164257e-05,
+      "loss": 0.20466468334197999,
+      "step": 22320
+    },
+    {
+      "epoch": 0.10919566476258674,
+      "grad_norm": 0.5124779343605042,
+      "learning_rate": 8.53199804346366e-05,
+      "loss": 0.2701244592666626,
+      "step": 22330
+    },
+    {
+      "epoch": 0.10948794437918896,
+      "grad_norm": 0.5163180828094482,
+      "learning_rate": 8.524030425333591e-05,
+      "loss": 0.27376594543457033,
+      "step": 22340
+    },
+    {
+      "epoch": 0.10978022399579117,
+      "grad_norm": 0.41993916034698486,
+      "learning_rate": 8.516063764944153e-05,
+      "loss": 0.24432027339935303,
+      "step": 22350
+    },
+    {
+      "epoch": 0.11007250361239339,
+      "grad_norm": 0.4385879337787628,
+      "learning_rate": 8.508098067464832e-05,
+      "loss": 0.2540184259414673,
+      "step": 22360
+    },
+    {
+      "epoch": 0.1103647832289956,
+      "grad_norm": 0.5795029401779175,
+      "learning_rate": 8.500133338064471e-05,
+      "loss": 0.29016067981719973,
+      "step": 22370
+    },
+    {
+      "epoch": 0.11065706284559781,
+      "grad_norm": 0.40759894251823425,
+      "learning_rate": 8.492169581911308e-05,
+      "loss": 0.2696007966995239,
+      "step": 22380
+    },
+    {
+      "epoch": 0.11094934246220002,
+      "grad_norm": 0.4509391486644745,
+      "learning_rate": 8.484206804172931e-05,
+      "loss": 0.23639960289001466,
+      "step": 22390
+    },
+    {
+      "epoch": 0.11124162207880224,
+      "grad_norm": 0.4401426613330841,
+      "learning_rate": 8.476245010016303e-05,
+      "loss": 0.24333782196044923,
+      "step": 22400
+    },
+    {
+      "epoch": 0.11153390169540445,
+      "grad_norm": 0.425870418548584,
+      "learning_rate": 8.468284204607746e-05,
+      "loss": 0.20699169635772705,
+      "step": 22410
+    },
+    {
+      "epoch": 0.11182618131200667,
+      "grad_norm": 0.5457123517990112,
+      "learning_rate": 8.460324393112938e-05,
+      "loss": 0.2805339336395264,
+      "step": 22420
+    },
+    {
+      "epoch": 0.11211846092860887,
+      "grad_norm": 0.5705997943878174,
+      "learning_rate": 8.452365580696916e-05,
+      "loss": 0.18422719240188598,
+      "step": 22430
+    },
+    {
+      "epoch": 0.1124107405452111,
+      "grad_norm": 0.6635261178016663,
+      "learning_rate": 8.444407772524064e-05,
+      "loss": 0.28923327922821046,
+      "step": 22440
+    },
+    {
+      "epoch": 0.1127030201618133,
+      "grad_norm": 0.4814568758010864,
+      "learning_rate": 8.436450973758123e-05,
+      "loss": 0.2993224859237671,
+      "step": 22450
+    },
+    {
+      "epoch": 0.11299529977841552,
+      "grad_norm": 0.6957874298095703,
+      "learning_rate": 8.428495189562169e-05,
+      "loss": 0.28928070068359374,
+      "step": 22460
+    },
+    {
+      "epoch": 0.11328757939501773,
+      "grad_norm": 0.44914788007736206,
+      "learning_rate": 8.420540425098623e-05,
+      "loss": 0.23721640110015868,
+      "step": 22470
+    },
+    {
+      "epoch": 0.11357985901161995,
+      "grad_norm": 0.3936746120452881,
+      "learning_rate": 8.412586685529249e-05,
+      "loss": 0.2999537229537964,
+      "step": 22480
+    },
+    {
+      "epoch": 0.11387213862822215,
+      "grad_norm": 0.39298704266548157,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.321905255317688,
+      "step": 22490
+    },
+    {
+      "epoch": 0.11416441824482437,
+      "grad_norm": 0.49617722630500793,
+      "learning_rate": 8.396682301716722e-05,
+      "loss": 0.2993109703063965,
+      "step": 22500
+    },
+    {
+      "epoch": 0.11445669786142658,
+      "grad_norm": 0.4595164358615875,
+      "learning_rate": 8.388731667793753e-05,
+      "loss": 0.325481915473938,
+      "step": 22510
+    },
+    {
+      "epoch": 0.1147489774780288,
+      "grad_norm": 0.4199615716934204,
+      "learning_rate": 8.380782079405306e-05,
+      "loss": 0.32211430072784425,
+      "step": 22520
+    },
+    {
+      "epoch": 0.115041257094631,
+      "grad_norm": 0.5099579095840454,
+      "learning_rate": 8.372833541709797e-05,
+      "loss": 0.3178923845291138,
+      "step": 22530
+    },
+    {
+      "epoch": 0.11533353671123323,
+      "grad_norm": 0.4574131369590759,
+      "learning_rate": 8.364886059864929e-05,
+      "loss": 0.2803200721740723,
+      "step": 22540
+    },
+    {
+      "epoch": 0.11562581632783543,
+      "grad_norm": 0.491312712430954,
+      "learning_rate": 8.35693963902775e-05,
+      "loss": 0.2621448993682861,
+      "step": 22550
+    },
+    {
+      "epoch": 0.11591809594443764,
+      "grad_norm": 0.397571325302124,
+      "learning_rate": 8.348994284354603e-05,
+      "loss": 0.3059609651565552,
+      "step": 22560
+    },
+    {
+      "epoch": 0.11621037556103986,
+      "grad_norm": 0.42750972509384155,
+      "learning_rate": 8.341050001001133e-05,
+      "loss": 0.3178231239318848,
+      "step": 22570
+    },
+    {
+      "epoch": 0.11650265517764206,
+      "grad_norm": 0.42258918285369873,
+      "learning_rate": 8.33310679412231e-05,
+      "loss": 0.2862200975418091,
+      "step": 22580
+    },
+    {
+      "epoch": 0.11679493479424428,
+      "grad_norm": 0.4157114624977112,
+      "learning_rate": 8.32516466887239e-05,
+      "loss": 0.27230205535888674,
+      "step": 22590
+    },
+    {
+      "epoch": 0.11708721441084649,
+      "grad_norm": 0.4948621988296509,
+      "learning_rate": 8.317223630404935e-05,
+      "loss": 0.289802622795105,
+      "step": 22600
+    },
+    {
+      "epoch": 0.11737949402744871,
+      "grad_norm": 0.4887374937534332,
+      "learning_rate": 8.309283683872796e-05,
+      "loss": 0.2909388065338135,
+      "step": 22610
+    },
+    {
+      "epoch": 0.11767177364405092,
+      "grad_norm": 0.368778795003891,
+      "learning_rate": 8.301344834428116e-05,
+      "loss": 0.26004796028137206,
+      "step": 22620
+    },
+    {
+      "epoch": 0.11796405326065314,
+      "grad_norm": 0.41819992661476135,
+      "learning_rate": 8.293407087222335e-05,
+      "loss": 0.3301138639450073,
+      "step": 22630
+    },
+    {
+      "epoch": 0.11825633287725534,
+      "grad_norm": 0.34272080659866333,
+      "learning_rate": 8.285470447406162e-05,
+      "loss": 0.26231980323791504,
+      "step": 22640
+    },
+    {
+      "epoch": 0.11854861249385756,
+      "grad_norm": 0.3414085805416107,
+      "learning_rate": 8.277534920129606e-05,
+      "loss": 0.27947998046875,
+      "step": 22650
+    },
+    {
+      "epoch": 0.11884089211045977,
+      "grad_norm": 0.4434521198272705,
+      "learning_rate": 8.269600510541937e-05,
+      "loss": 0.3299944639205933,
+      "step": 22660
+    },
+    {
+      "epoch": 0.11913317172706199,
+      "grad_norm": 0.39193692803382874,
+      "learning_rate": 8.261667223791712e-05,
+      "loss": 0.2792419195175171,
+      "step": 22670
+    },
+    {
+      "epoch": 0.1194254513436642,
+      "grad_norm": 0.4870033264160156,
+      "learning_rate": 8.253735065026753e-05,
+      "loss": 0.32523226737976074,
+      "step": 22680
+    },
+    {
+      "epoch": 0.11971773096026642,
+      "grad_norm": 0.4635838568210602,
+      "learning_rate": 8.245804039394148e-05,
+      "loss": 0.2591820240020752,
+      "step": 22690
+    },
+    {
+      "epoch": 0.12001001057686862,
+      "grad_norm": 0.39743873476982117,
+      "learning_rate": 8.23787415204026e-05,
+      "loss": 0.2524199247360229,
+      "step": 22700
+    },
+    {
+      "epoch": 0.12030229019347084,
+      "grad_norm": 0.4468758702278137,
+      "learning_rate": 8.229945408110699e-05,
+      "loss": 0.2730138063430786,
+      "step": 22710
+    },
+    {
+      "epoch": 0.12059456981007305,
+      "grad_norm": 0.3939422369003296,
+      "learning_rate": 8.222017812750352e-05,
+      "loss": 0.2791752338409424,
+      "step": 22720
+    },
+    {
+      "epoch": 0.12088684942667527,
+      "grad_norm": 0.3863838016986847,
+      "learning_rate": 8.214091371103341e-05,
+      "loss": 0.31022751331329346,
+      "step": 22730
+    },
+    {
+      "epoch": 0.12117912904327748,
+      "grad_norm": 0.42934685945510864,
+      "learning_rate": 8.206166088313045e-05,
+      "loss": 0.27296009063720705,
+      "step": 22740
+    },
+    {
+      "epoch": 0.1214714086598797,
+      "grad_norm": 0.42042040824890137,
+      "learning_rate": 8.198241969522107e-05,
+      "loss": 0.2431617021560669,
+      "step": 22750
+    },
+    {
+      "epoch": 0.1217636882764819,
+      "grad_norm": 0.4736967384815216,
+      "learning_rate": 8.190319019872385e-05,
+      "loss": 0.267451548576355,
+      "step": 22760
+    },
+    {
+      "epoch": 0.12205596789308412,
+      "grad_norm": 0.396629273891449,
+      "learning_rate": 8.18239724450501e-05,
+      "loss": 0.3109576940536499,
+      "step": 22770
+    },
+    {
+      "epoch": 0.12234824750968633,
+      "grad_norm": 0.4852570593357086,
+      "learning_rate": 8.17447664856033e-05,
+      "loss": 0.3155143976211548,
+      "step": 22780
+    },
+    {
+      "epoch": 0.12264052712628855,
+      "grad_norm": 0.5345566272735596,
+      "learning_rate": 8.166557237177927e-05,
+      "loss": 0.23329670429229737,
+      "step": 22790
+    },
+    {
+      "epoch": 0.12293280674289075,
+      "grad_norm": 0.4613855481147766,
+      "learning_rate": 8.158639015496633e-05,
+      "loss": 0.28379886150360106,
+      "step": 22800
+    },
+    {
+      "epoch": 0.12322508635949297,
+      "grad_norm": 0.4406832754611969,
+      "learning_rate": 8.150721988654488e-05,
+      "loss": 0.28256797790527344,
+      "step": 22810
+    },
+    {
+      "epoch": 0.12351736597609518,
+      "grad_norm": 0.42703762650489807,
+      "learning_rate": 8.142806161788768e-05,
+      "loss": 0.3550808191299438,
+      "step": 22820
+    },
+    {
+      "epoch": 0.12380964559269739,
+      "grad_norm": 0.4377189576625824,
+      "learning_rate": 8.134891540035967e-05,
+      "loss": 0.3178999900817871,
+      "step": 22830
+    },
+    {
+      "epoch": 0.12410192520929961,
+      "grad_norm": 0.34141209721565247,
+      "learning_rate": 8.126978128531795e-05,
+      "loss": 0.2970437526702881,
+      "step": 22840
+    },
+    {
+      "epoch": 0.12439420482590181,
+      "grad_norm": 0.42445269227027893,
+      "learning_rate": 8.119065932411183e-05,
+      "loss": 0.37621397972106935,
+      "step": 22850
+    },
+    {
+      "epoch": 0.12468648444250403,
+      "grad_norm": 0.4631160795688629,
+      "learning_rate": 8.111154956808263e-05,
+      "loss": 0.294628643989563,
+      "step": 22860
+    },
+    {
+      "epoch": 0.12497876405910624,
+      "grad_norm": 0.4251672923564911,
+      "learning_rate": 8.103245206856386e-05,
+      "loss": 0.30714895725250246,
+      "step": 22870
+    },
+    {
+      "epoch": 0.12527104367570846,
+      "grad_norm": 0.3577144742012024,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 0.2903586864471436,
+      "step": 22880
+    },
+    {
+      "epoch": 0.12556332329231068,
+      "grad_norm": 0.4270741939544678,
+      "learning_rate": 8.087429404435162e-05,
+      "loss": 0.29819176197052,
+      "step": 22890
+    },
+    {
+      "epoch": 0.12585560290891287,
+      "grad_norm": 0.3734011650085449,
+      "learning_rate": 8.079523362228519e-05,
+      "loss": 0.3466984272003174,
+      "step": 22900
+    },
+    {
+      "epoch": 0.1261478825255151,
+      "grad_norm": 0.3908292353153229,
+      "learning_rate": 8.071618566198313e-05,
+      "loss": 0.2991729497909546,
+      "step": 22910
+    },
+    {
+      "epoch": 0.1264401621421173,
+      "grad_norm": 0.3335021734237671,
+      "learning_rate": 8.063715021473885e-05,
+      "loss": 0.2848759412765503,
+      "step": 22920
+    },
+    {
+      "epoch": 0.12673244175871953,
+      "grad_norm": 0.4307859539985657,
+      "learning_rate": 8.055812733183755e-05,
+      "loss": 0.37097415924072263,
+      "step": 22930
+    },
+    {
+      "epoch": 0.12702472137532173,
+      "grad_norm": 0.46101298928260803,
+      "learning_rate": 8.04791170645564e-05,
+      "loss": 0.2744488477706909,
+      "step": 22940
+    },
+    {
+      "epoch": 0.12731700099192395,
+      "grad_norm": 0.4156108796596527,
+      "learning_rate": 8.040011946416423e-05,
+      "loss": 0.2462397813796997,
+      "step": 22950
+    },
+    {
+      "epoch": 0.12760928060852617,
+      "grad_norm": 0.4069465100765228,
+      "learning_rate": 8.03211345819217e-05,
+      "loss": 0.2732763528823853,
+      "step": 22960
+    },
+    {
+      "epoch": 0.12790156022512839,
+      "grad_norm": 0.5664887428283691,
+      "learning_rate": 8.024216246908136e-05,
+      "loss": 0.31092813014984133,
+      "step": 22970
+    },
+    {
+      "epoch": 0.12819383984173058,
+      "grad_norm": 0.7392646670341492,
+      "learning_rate": 8.016320317688722e-05,
+      "loss": 0.27896509170532224,
+      "step": 22980
+    },
+    {
+      "epoch": 0.1284861194583328,
+      "grad_norm": 0.4116365611553192,
+      "learning_rate": 8.008425675657522e-05,
+      "loss": 0.24654555320739746,
+      "step": 22990
+    },
+    {
+      "epoch": 0.12877839907493502,
+      "grad_norm": 0.524607241153717,
+      "learning_rate": 8.000532325937278e-05,
+      "loss": 0.3084801197052002,
+      "step": 23000
+    },
+    {
+      "epoch": 0.12907067869153724,
+      "grad_norm": 0.45251375436782837,
+      "learning_rate": 7.992640273649898e-05,
+      "loss": 0.23576037883758544,
+      "step": 23010
+    },
+    {
+      "epoch": 0.12936295830813943,
+      "grad_norm": 0.3694298267364502,
+      "learning_rate": 7.984749523916453e-05,
+      "loss": 0.3187490701675415,
+      "step": 23020
+    },
+    {
+      "epoch": 0.12965523792474165,
+      "grad_norm": 0.5533820986747742,
+      "learning_rate": 7.97686008185716e-05,
+      "loss": 0.3504955291748047,
+      "step": 23030
+    },
+    {
+      "epoch": 0.12994751754134387,
+      "grad_norm": 0.4420567750930786,
+      "learning_rate": 7.968971952591399e-05,
+      "loss": 0.26173017024993894,
+      "step": 23040
+    },
+    {
+      "epoch": 0.1302397971579461,
+      "grad_norm": 0.42212194204330444,
+      "learning_rate": 7.961085141237686e-05,
+      "loss": 0.2616743087768555,
+      "step": 23050
+    },
+    {
+      "epoch": 0.13053207677454828,
+      "grad_norm": 0.410424143075943,
+      "learning_rate": 7.953199652913686e-05,
+      "loss": 0.30044851303100584,
+      "step": 23060
+    },
+    {
+      "epoch": 0.1308243563911505,
+      "grad_norm": 0.5561143755912781,
+      "learning_rate": 7.945315492736213e-05,
+      "loss": 0.24086339473724366,
+      "step": 23070
+    },
+    {
+      "epoch": 0.13111663600775272,
+      "grad_norm": 0.3605208098888397,
+      "learning_rate": 7.937432665821207e-05,
+      "loss": 0.31395468711853025,
+      "step": 23080
+    },
+    {
+      "epoch": 0.13140891562435492,
+      "grad_norm": 0.3601417541503906,
+      "learning_rate": 7.929551177283752e-05,
+      "loss": 0.28766019344329835,
+      "step": 23090
+    },
+    {
+      "epoch": 0.13170119524095714,
+      "grad_norm": 0.37903833389282227,
+      "learning_rate": 7.921671032238061e-05,
+      "loss": 0.3364447593688965,
+      "step": 23100
+    },
+    {
+      "epoch": 0.13199347485755936,
+      "grad_norm": 0.3952993154525757,
+      "learning_rate": 7.91379223579747e-05,
+      "loss": 0.2406454563140869,
+      "step": 23110
+    },
+    {
+      "epoch": 0.13228575447416158,
+      "grad_norm": 0.35304901003837585,
+      "learning_rate": 7.90591479307445e-05,
+      "loss": 0.32697529792785646,
+      "step": 23120
+    },
+    {
+      "epoch": 0.13257803409076377,
+      "grad_norm": 0.4308299720287323,
+      "learning_rate": 7.898038709180581e-05,
+      "loss": 0.30465385913848875,
+      "step": 23130
+    },
+    {
+      "epoch": 0.132870313707366,
+      "grad_norm": 0.35119277238845825,
+      "learning_rate": 7.89016398922658e-05,
+      "loss": 0.32311019897460935,
+      "step": 23140
+    },
+    {
+      "epoch": 0.1331625933239682,
+      "grad_norm": 0.5565539002418518,
+      "learning_rate": 7.882290638322253e-05,
+      "loss": 0.2558394432067871,
+      "step": 23150
+    },
+    {
+      "epoch": 0.13345487294057043,
+      "grad_norm": 0.5316254496574402,
+      "learning_rate": 7.874418661576543e-05,
+      "loss": 0.30089993476867677,
+      "step": 23160
+    },
+    {
+      "epoch": 0.13374715255717262,
+      "grad_norm": 0.821410596370697,
+      "learning_rate": 7.86654806409749e-05,
+      "loss": 0.31371493339538575,
+      "step": 23170
+    },
+    {
+      "epoch": 0.13403943217377484,
+      "grad_norm": 0.3491901457309723,
+      "learning_rate": 7.85867885099223e-05,
+      "loss": 0.2663713932037354,
+      "step": 23180
+    },
+    {
+      "epoch": 0.13433171179037706,
+      "grad_norm": 0.34341737627983093,
+      "learning_rate": 7.850811027367017e-05,
+      "loss": 0.2644458532333374,
+      "step": 23190
+    },
+    {
+      "epoch": 0.13462399140697928,
+      "grad_norm": 0.39471107721328735,
+      "learning_rate": 7.842944598327195e-05,
+      "loss": 0.28630890846252444,
+      "step": 23200
+    },
+    {
+      "epoch": 0.13491627102358147,
+      "grad_norm": 0.4434603154659271,
+      "learning_rate": 7.835079568977207e-05,
+      "loss": 0.2714740514755249,
+      "step": 23210
+    },
+    {
+      "epoch": 0.1352085506401837,
+      "grad_norm": 0.4811277687549591,
+      "learning_rate": 7.82721594442058e-05,
+      "loss": 0.2684128522872925,
+      "step": 23220
+    },
+    {
+      "epoch": 0.13550083025678591,
+      "grad_norm": 0.48829028010368347,
+      "learning_rate": 7.819353729759936e-05,
+      "loss": 0.3625423192977905,
+      "step": 23230
+    },
+    {
+      "epoch": 0.13579310987338813,
+      "grad_norm": 0.3815400302410126,
+      "learning_rate": 7.811492930096984e-05,
+      "loss": 0.2830656051635742,
+      "step": 23240
+    },
+    {
+      "epoch": 0.13608538948999033,
+      "grad_norm": 0.4248940646648407,
+      "learning_rate": 7.803633550532507e-05,
+      "loss": 0.28103759288787844,
+      "step": 23250
+    },
+    {
+      "epoch": 0.13637766910659255,
+      "grad_norm": 0.4027959406375885,
+      "learning_rate": 7.795775596166376e-05,
+      "loss": 0.24977524280548097,
+      "step": 23260
+    },
+    {
+      "epoch": 0.13666994872319477,
+      "grad_norm": 0.3476473093032837,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.24661388397216796,
+      "step": 23270
+    },
+    {
+      "epoch": 0.136962228339797,
+      "grad_norm": 0.5637992024421692,
+      "learning_rate": 7.78006398342398e-05,
+      "loss": 0.3302830934524536,
+      "step": 23280
+    },
+    {
+      "epoch": 0.13725450795639918,
+      "grad_norm": 0.3477632999420166,
+      "learning_rate": 7.772210335242811e-05,
+      "loss": 0.2610548257827759,
+      "step": 23290
+    },
+    {
+      "epoch": 0.1375467875730014,
+      "grad_norm": 0.36640655994415283,
+      "learning_rate": 7.764358132650167e-05,
+      "loss": 0.30639872550964353,
+      "step": 23300
+    },
+    {
+      "epoch": 0.13783906718960362,
+      "grad_norm": 0.46990543603897095,
+      "learning_rate": 7.756507380741261e-05,
+      "loss": 0.27899532318115233,
+      "step": 23310
+    },
+    {
+      "epoch": 0.13813134680620584,
+      "grad_norm": 0.5981518626213074,
+      "learning_rate": 7.748658084610358e-05,
+      "loss": 0.34725563526153563,
+      "step": 23320
+    },
+    {
+      "epoch": 0.13842362642280803,
+      "grad_norm": 0.6311739087104797,
+      "learning_rate": 7.74081024935078e-05,
+      "loss": 0.38969509601593016,
+      "step": 23330
+    },
+    {
+      "epoch": 0.13871590603941025,
+      "grad_norm": 0.39023062586784363,
+      "learning_rate": 7.732963880054904e-05,
+      "loss": 0.23270516395568847,
+      "step": 23340
+    },
+    {
+      "epoch": 0.13900818565601247,
+      "grad_norm": 0.35662397742271423,
+      "learning_rate": 7.725118981814151e-05,
+      "loss": 0.2492440700531006,
+      "step": 23350
+    },
+    {
+      "epoch": 0.13930046527261467,
+      "grad_norm": 0.40297189354896545,
+      "learning_rate": 7.717275559718999e-05,
+      "loss": 0.33704442977905275,
+      "step": 23360
+    },
+    {
+      "epoch": 0.13959274488921689,
+      "grad_norm": 0.48780739307403564,
+      "learning_rate": 7.709433618858945e-05,
+      "loss": 0.25130453109741213,
+      "step": 23370
+    },
+    {
+      "epoch": 0.1398850245058191,
+      "grad_norm": 0.45817744731903076,
+      "learning_rate": 7.701593164322553e-05,
+      "loss": 0.37227559089660645,
+      "step": 23380
+    },
+    {
+      "epoch": 0.14017730412242133,
+      "grad_norm": 0.45841383934020996,
+      "learning_rate": 7.693754201197405e-05,
+      "loss": 0.33836328983306885,
+      "step": 23390
+    },
+    {
+      "epoch": 0.14046958373902352,
+      "grad_norm": 0.42280739545822144,
+      "learning_rate": 7.685916734570112e-05,
+      "loss": 0.31687211990356445,
+      "step": 23400
+    },
+    {
+      "epoch": 0.14076186335562574,
+      "grad_norm": 0.4889216423034668,
+      "learning_rate": 7.678080769526329e-05,
+      "loss": 0.28561785221099856,
+      "step": 23410
+    },
+    {
+      "epoch": 0.14105414297222796,
+      "grad_norm": 0.4177890717983246,
+      "learning_rate": 7.670246311150724e-05,
+      "loss": 0.2947185516357422,
+      "step": 23420
+    },
+    {
+      "epoch": 0.14134642258883018,
+      "grad_norm": 0.35976889729499817,
+      "learning_rate": 7.662413364526996e-05,
+      "loss": 0.2869388580322266,
+      "step": 23430
+    },
+    {
+      "epoch": 0.14163870220543237,
+      "grad_norm": 0.4119426906108856,
+      "learning_rate": 7.654581934737857e-05,
+      "loss": 0.29015545845031737,
+      "step": 23440
+    },
+    {
+      "epoch": 0.1419309818220346,
+      "grad_norm": 0.3500671088695526,
+      "learning_rate": 7.646752026865034e-05,
+      "loss": 0.3236443519592285,
+      "step": 23450
+    },
+    {
+      "epoch": 0.1422232614386368,
+      "grad_norm": 0.36175280809402466,
+      "learning_rate": 7.638923645989275e-05,
+      "loss": 0.2728204965591431,
+      "step": 23460
+    },
+    {
+      "epoch": 0.14251554105523903,
+      "grad_norm": 0.40692171454429626,
+      "learning_rate": 7.631096797190327e-05,
+      "loss": 0.32873854637145994,
+      "step": 23470
+    },
+    {
+      "epoch": 0.14280782067184122,
+      "grad_norm": 0.35549959540367126,
+      "learning_rate": 7.62327148554695e-05,
+      "loss": 0.31428894996643064,
+      "step": 23480
+    },
+    {
+      "epoch": 0.14310010028844344,
+      "grad_norm": 0.424144446849823,
+      "learning_rate": 7.615447716136901e-05,
+      "loss": 0.29877779483795164,
+      "step": 23490
+    },
+    {
+      "epoch": 0.14339237990504566,
+      "grad_norm": 0.33184027671813965,
+      "learning_rate": 7.60762549403694e-05,
+      "loss": 0.2769747495651245,
+      "step": 23500
+    },
+    {
+      "epoch": 0.14368465952164788,
+      "grad_norm": 0.4286607503890991,
+      "learning_rate": 7.599804824322823e-05,
+      "loss": 0.41048274040222166,
+      "step": 23510
+    },
+    {
+      "epoch": 0.14397693913825008,
+      "grad_norm": 0.37445783615112305,
+      "learning_rate": 7.591985712069295e-05,
+      "loss": 0.29165737628936766,
+      "step": 23520
+    },
+    {
+      "epoch": 0.1442692187548523,
+      "grad_norm": 0.38956761360168457,
+      "learning_rate": 7.584168162350098e-05,
+      "loss": 0.2995201349258423,
+      "step": 23530
+    },
+    {
+      "epoch": 0.14456149837145452,
+      "grad_norm": 0.317703515291214,
+      "learning_rate": 7.576352180237952e-05,
+      "loss": 0.29964461326599123,
+      "step": 23540
+    },
+    {
+      "epoch": 0.14485377798805674,
+      "grad_norm": 0.34540048241615295,
+      "learning_rate": 7.56853777080456e-05,
+      "loss": 0.34931671619415283,
+      "step": 23550
+    },
+    {
+      "epoch": 0.14514605760465893,
+      "grad_norm": 0.45366546511650085,
+      "learning_rate": 7.560724939120613e-05,
+      "loss": 0.32204089164733884,
+      "step": 23560
+    },
+    {
+      "epoch": 0.14543833722126115,
+      "grad_norm": 0.4753202795982361,
+      "learning_rate": 7.552913690255765e-05,
+      "loss": 0.29108362197875975,
+      "step": 23570
+    },
+    {
+      "epoch": 0.14573061683786337,
+      "grad_norm": 0.33317831158638,
+      "learning_rate": 7.545104029278665e-05,
+      "loss": 0.3161603927612305,
+      "step": 23580
+    },
+    {
+      "epoch": 0.1460228964544656,
+      "grad_norm": 0.539392352104187,
+      "learning_rate": 7.537295961256902e-05,
+      "loss": 0.31974081993103026,
+      "step": 23590
+    },
+    {
+      "epoch": 0.14631517607106778,
+      "grad_norm": 0.43225207924842834,
+      "learning_rate": 7.529489491257051e-05,
+      "loss": 0.2690849542617798,
+      "step": 23600
+    },
+    {
+      "epoch": 0.14660745568767,
+      "grad_norm": 0.7462082505226135,
+      "learning_rate": 7.521684624344651e-05,
+      "loss": 0.33043417930603025,
+      "step": 23610
+    },
+    {
+      "epoch": 0.14689973530427222,
+      "grad_norm": 0.31991690397262573,
+      "learning_rate": 7.51388136558419e-05,
+      "loss": 0.3021215915679932,
+      "step": 23620
+    },
+    {
+      "epoch": 0.14719201492087441,
+      "grad_norm": 0.33238911628723145,
+      "learning_rate": 7.506079720039121e-05,
+      "loss": 0.30872747898101804,
+      "step": 23630
+    },
+    {
+      "epoch": 0.14748429453747663,
+      "grad_norm": 0.3633115887641907,
+      "learning_rate": 7.498279692771842e-05,
+      "loss": 0.3323874235153198,
+      "step": 23640
+    },
+    {
+      "epoch": 0.14777657415407885,
+      "grad_norm": 0.3906925618648529,
+      "learning_rate": 7.490481288843713e-05,
+      "loss": 0.33025100231170657,
+      "step": 23650
+    },
+    {
+      "epoch": 0.14806885377068107,
+      "grad_norm": 0.43061619997024536,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.3064288854598999,
+      "step": 23660
+    },
+    {
+      "epoch": 0.14836113338728327,
+      "grad_norm": 0.2776353061199188,
+      "learning_rate": 7.474889371245035e-05,
+      "loss": 0.23351268768310546,
+      "step": 23670
+    },
+    {
+      "epoch": 0.1486534130038855,
+      "grad_norm": 0.36395132541656494,
+      "learning_rate": 7.467095867691913e-05,
+      "loss": 0.2552823305130005,
+      "step": 23680
+    },
+    {
+      "epoch": 0.1489456926204877,
+      "grad_norm": 0.38996076583862305,
+      "learning_rate": 7.459304007712782e-05,
+      "loss": 0.2691317558288574,
+      "step": 23690
+    },
+    {
+      "epoch": 0.14923797223708993,
+      "grad_norm": 0.36750954389572144,
+      "learning_rate": 7.451513796363699e-05,
+      "loss": 0.27944235801696776,
+      "step": 23700
+    },
+    {
+      "epoch": 0.14953025185369212,
+      "grad_norm": 0.305603563785553,
+      "learning_rate": 7.443725238699644e-05,
+      "loss": 0.2889078617095947,
+      "step": 23710
+    },
+    {
+      "epoch": 0.14982253147029434,
+      "grad_norm": 0.41687682271003723,
+      "learning_rate": 7.435938339774526e-05,
+      "loss": 0.29069340229034424,
+      "step": 23720
+    },
+    {
+      "epoch": 0.15011481108689656,
+      "grad_norm": 0.32410717010498047,
+      "learning_rate": 7.428153104641184e-05,
+      "loss": 0.2986827611923218,
+      "step": 23730
+    },
+    {
+      "epoch": 0.15040709070349878,
+      "grad_norm": 0.46407151222229004,
+      "learning_rate": 7.420369538351368e-05,
+      "loss": 0.35066375732421873,
+      "step": 23740
+    },
+    {
+      "epoch": 0.15069937032010097,
+      "grad_norm": 0.49871572852134705,
+      "learning_rate": 7.412587645955756e-05,
+      "loss": 0.26365346908569337,
+      "step": 23750
+    },
+    {
+      "epoch": 0.1509916499367032,
+      "grad_norm": 0.3395722508430481,
+      "learning_rate": 7.404807432503928e-05,
+      "loss": 0.2627224922180176,
+      "step": 23760
+    },
+    {
+      "epoch": 0.1512839295533054,
+      "grad_norm": 0.31821733713150024,
+      "learning_rate": 7.397028903044377e-05,
+      "loss": 0.22571127414703368,
+      "step": 23770
+    },
+    {
+      "epoch": 0.15157620916990763,
+      "grad_norm": 0.41738808155059814,
+      "learning_rate": 7.389252062624519e-05,
+      "loss": 0.27345297336578367,
+      "step": 23780
+    },
+    {
+      "epoch": 0.15186848878650983,
+      "grad_norm": 0.41709476709365845,
+      "learning_rate": 7.381476916290644e-05,
+      "loss": 0.24168214797973633,
+      "step": 23790
+    },
+    {
+      "epoch": 0.15216076840311205,
+      "grad_norm": 0.3823196291923523,
+      "learning_rate": 7.373703469087974e-05,
+      "loss": 0.2707072257995605,
+      "step": 23800
+    },
+    {
+      "epoch": 0.15245304801971427,
+      "grad_norm": 0.42271918058395386,
+      "learning_rate": 7.365931726060613e-05,
+      "loss": 0.2766434669494629,
+      "step": 23810
+    },
+    {
+      "epoch": 0.15274532763631649,
+      "grad_norm": 0.37058496475219727,
+      "learning_rate": 7.35816169225155e-05,
+      "loss": 0.26467232704162597,
+      "step": 23820
+    },
+    {
+      "epoch": 5.8455923320442586e-05,
+      "grad_norm": 0.3033806085586548,
+      "learning_rate": 7.350393372702687e-05,
+      "loss": 0.17150813341140747,
+      "step": 23830
+    },
+    {
+      "epoch": 0.0003507355399226555,
+      "grad_norm": 0.3767378628253937,
+      "learning_rate": 7.342626772454797e-05,
+      "loss": 0.1403013586997986,
+      "step": 23840
+    },
+    {
+      "epoch": 0.0006430151565248684,
+      "grad_norm": 0.449060320854187,
+      "learning_rate": 7.334861896547545e-05,
+      "loss": 0.08162813782691955,
+      "step": 23850
+    },
+    {
+      "epoch": 0.0009352947731270814,
+      "grad_norm": 0.3904483914375305,
+      "learning_rate": 7.32709875001947e-05,
+      "loss": 0.0868235170841217,
+      "step": 23860
+    },
+    {
+      "epoch": 0.0012275743897292943,
+      "grad_norm": 0.314506471157074,
+      "learning_rate": 7.319337337908002e-05,
+      "loss": 0.06524655222892761,
+      "step": 23870
+    },
+    {
+      "epoch": 0.001519854006331507,
+      "grad_norm": 0.367190420627594,
+      "learning_rate": 7.311577665249429e-05,
+      "loss": 0.07309747338294983,
+      "step": 23880
+    },
+    {
+      "epoch": 0.00181213362293372,
+      "grad_norm": 0.6414294242858887,
+      "learning_rate": 7.30381973707892e-05,
+      "loss": 0.0894463300704956,
+      "step": 23890
+    },
+    {
+      "epoch": 0.002104413239535933,
+      "grad_norm": 0.26352372765541077,
+      "learning_rate": 7.296063558430514e-05,
+      "loss": 0.07071267366409302,
+      "step": 23900
+    },
+    {
+      "epoch": 0.002396692856138146,
+      "grad_norm": 0.34883931279182434,
+      "learning_rate": 7.288309134337106e-05,
+      "loss": 0.07626253366470337,
+      "step": 23910
+    },
+    {
+      "epoch": 0.002688972472740359,
+      "grad_norm": 0.5369346737861633,
+      "learning_rate": 7.280556469830464e-05,
+      "loss": 0.07025394439697266,
+      "step": 23920
+    },
+    {
+      "epoch": 0.0029812520893425717,
+      "grad_norm": 0.5184519290924072,
+      "learning_rate": 7.272805569941203e-05,
+      "loss": 0.076833575963974,
+      "step": 23930
+    },
+    {
+      "epoch": 0.003273531705944785,
+      "grad_norm": 0.3384343981742859,
+      "learning_rate": 7.265056439698796e-05,
+      "loss": 0.06525980830192565,
+      "step": 23940
+    },
+    {
+      "epoch": 0.0035658113225469977,
+      "grad_norm": 0.49539029598236084,
+      "learning_rate": 7.257309084131576e-05,
+      "loss": 0.0821267008781433,
+      "step": 23950
+    },
+    {
+      "epoch": 0.0038580909391492104,
+      "grad_norm": 0.5111546516418457,
+      "learning_rate": 7.24956350826671e-05,
+      "loss": 0.07291510701179504,
+      "step": 23960
+    },
+    {
+      "epoch": 0.004150370555751424,
+      "grad_norm": 0.5053063631057739,
+      "learning_rate": 7.241819717130227e-05,
+      "loss": 0.07973734736442566,
+      "step": 23970
+    },
+    {
+      "epoch": 0.004442650172353637,
+      "grad_norm": 0.592573881149292,
+      "learning_rate": 7.234077715746985e-05,
+      "loss": 0.06683561801910401,
+      "step": 23980
+    },
+    {
+      "epoch": 0.004734929788955849,
+      "grad_norm": 0.5798428654670715,
+      "learning_rate": 7.226337509140678e-05,
+      "loss": 0.050779479742050174,
+      "step": 23990
+    },
+    {
+      "epoch": 0.005027209405558062,
+      "grad_norm": 0.6188417673110962,
+      "learning_rate": 7.218599102333855e-05,
+      "loss": 0.05950241088867188,
+      "step": 24000
+    },
+    {
+      "epoch": 0.0053194890221602755,
+      "grad_norm": 0.3991755545139313,
+      "learning_rate": 7.210862500347871e-05,
+      "loss": 0.06148055791854858,
+      "step": 24010
+    },
+    {
+      "epoch": 0.005611768638762488,
+      "grad_norm": 0.5449476838111877,
+      "learning_rate": 7.203127708202932e-05,
+      "loss": 0.058894121646881105,
+      "step": 24020
+    },
+    {
+      "epoch": 0.005904048255364701,
+      "grad_norm": 0.45845234394073486,
+      "learning_rate": 7.195394730918062e-05,
+      "loss": 0.08477458953857422,
+      "step": 24030
+    },
+    {
+      "epoch": 0.006196327871966914,
+      "grad_norm": 0.5977475047111511,
+      "learning_rate": 7.187663573511096e-05,
+      "loss": 0.052658849954605104,
+      "step": 24040
+    },
+    {
+      "epoch": 0.0064886074885691265,
+      "grad_norm": 0.7027382850646973,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.06697238683700561,
+      "step": 24050
+    },
+    {
+      "epoch": 0.00678088710517134,
+      "grad_norm": 0.48088330030441284,
+      "learning_rate": 7.17220673839637e-05,
+      "loss": 0.0706555724143982,
+      "step": 24060
+    },
+    {
+      "epoch": 0.007073166721773553,
+      "grad_norm": 0.39975523948669434,
+      "learning_rate": 7.164481070718382e-05,
+      "loss": 0.07680798768997192,
+      "step": 24070
+    },
+    {
+      "epoch": 0.007365446338375765,
+      "grad_norm": 0.43049493432044983,
+      "learning_rate": 7.156757242977842e-05,
+      "loss": 0.06296321153640747,
+      "step": 24080
+    },
+    {
+      "epoch": 0.007657725954977978,
+      "grad_norm": 0.6044527292251587,
+      "learning_rate": 7.149035260186656e-05,
+      "loss": 0.045496830344200136,
+      "step": 24090
+    },
+    {
+      "epoch": 0.00795000557158019,
+      "grad_norm": 0.5679788589477539,
+      "learning_rate": 7.141315127355541e-05,
+      "loss": 0.07253391742706299,
+      "step": 24100
+    },
+    {
+      "epoch": 0.008242285188182404,
+      "grad_norm": 0.5709264874458313,
+      "learning_rate": 7.133596849494e-05,
+      "loss": 0.047528055310249326,
+      "step": 24110
+    },
+    {
+      "epoch": 0.008534564804784617,
+      "grad_norm": 0.4692271649837494,
+      "learning_rate": 7.125880431610345e-05,
+      "loss": 0.0647103250026703,
+      "step": 24120
+    },
+    {
+      "epoch": 0.00882684442138683,
+      "grad_norm": 0.561491847038269,
+      "learning_rate": 7.118165878711673e-05,
+      "loss": 0.06678470373153686,
+      "step": 24130
+    },
+    {
+      "epoch": 0.009119124037989043,
+      "grad_norm": 0.35136979818344116,
+      "learning_rate": 7.110453195803876e-05,
+      "loss": 0.05238202214241028,
+      "step": 24140
+    },
+    {
+      "epoch": 0.009411403654591257,
+      "grad_norm": 0.4195147156715393,
+      "learning_rate": 7.10274238789163e-05,
+      "loss": 0.06432364583015442,
+      "step": 24150
+    },
+    {
+      "epoch": 0.00970368327119347,
+      "grad_norm": 0.5664438605308533,
+      "learning_rate": 7.09503345997839e-05,
+      "loss": 0.048707437515258786,
+      "step": 24160
+    },
+    {
+      "epoch": 0.009995962887795681,
+      "grad_norm": 0.6620915532112122,
+      "learning_rate": 7.087326417066402e-05,
+      "loss": 0.07986681461334229,
+      "step": 24170
+    },
+    {
+      "epoch": 0.010288242504397895,
+      "grad_norm": 0.425568163394928,
+      "learning_rate": 7.079621264156675e-05,
+      "loss": 0.04848785400390625,
+      "step": 24180
+    },
+    {
+      "epoch": 0.010580522121000108,
+      "grad_norm": 0.5818431973457336,
+      "learning_rate": 7.07191800624901e-05,
+      "loss": 0.04921112358570099,
+      "step": 24190
+    },
+    {
+      "epoch": 0.010872801737602321,
+      "grad_norm": 0.743874192237854,
+      "learning_rate": 7.064216648341959e-05,
+      "loss": 0.07447198629379273,
+      "step": 24200
+    },
+    {
+      "epoch": 0.011165081354204534,
+      "grad_norm": 0.5582935810089111,
+      "learning_rate": 7.056517195432847e-05,
+      "loss": 0.054574227333068846,
+      "step": 24210
+    },
+    {
+      "epoch": 0.011457360970806747,
+      "grad_norm": 0.41284844279289246,
+      "learning_rate": 7.048819652517776e-05,
+      "loss": 0.052970665693283084,
+      "step": 24220
+    },
+    {
+      "epoch": 0.011749640587408959,
+      "grad_norm": 0.8441746830940247,
+      "learning_rate": 7.041124024591588e-05,
+      "loss": 0.07383311986923217,
+      "step": 24230
+    },
+    {
+      "epoch": 0.012041920204011172,
+      "grad_norm": 0.5405794978141785,
+      "learning_rate": 7.033430316647898e-05,
+      "loss": 0.05083611607551575,
+      "step": 24240
+    },
+    {
+      "epoch": 0.012334199820613385,
+      "grad_norm": 0.3178989291191101,
+      "learning_rate": 7.025738533679069e-05,
+      "loss": 0.04570271670818329,
+      "step": 24250
+    },
+    {
+      "epoch": 0.012626479437215598,
+      "grad_norm": 0.3733997046947479,
+      "learning_rate": 7.018048680676212e-05,
+      "loss": 0.06502344012260437,
+      "step": 24260
+    },
+    {
+      "epoch": 0.012918759053817811,
+      "grad_norm": 0.5460653305053711,
+      "learning_rate": 7.010360762629194e-05,
+      "loss": 0.04468178153038025,
+      "step": 24270
+    },
+    {
+      "epoch": 0.013211038670420025,
+      "grad_norm": 0.47185811400413513,
+      "learning_rate": 7.002674784526617e-05,
+      "loss": 0.06373377442359925,
+      "step": 24280
+    },
+    {
+      "epoch": 0.013503318287022238,
+      "grad_norm": 0.4594062566757202,
+      "learning_rate": 6.994990751355832e-05,
+      "loss": 0.054783135652542114,
+      "step": 24290
+    },
+    {
+      "epoch": 0.01379559790362445,
+      "grad_norm": 0.312099814414978,
+      "learning_rate": 6.987308668102923e-05,
+      "loss": 0.06075689196586609,
+      "step": 24300
+    },
+    {
+      "epoch": 0.014087877520226662,
+      "grad_norm": 0.42060935497283936,
+      "learning_rate": 6.979628539752711e-05,
+      "loss": 0.05281209349632263,
+      "step": 24310
+    },
+    {
+      "epoch": 0.014380157136828876,
+      "grad_norm": 0.451540470123291,
+      "learning_rate": 6.97195037128875e-05,
+      "loss": 0.05123857259750366,
+      "step": 24320
+    },
+    {
+      "epoch": 0.014672436753431089,
+      "grad_norm": 0.4303482174873352,
+      "learning_rate": 6.964274167693313e-05,
+      "loss": 0.052332305908203126,
+      "step": 24330
+    },
+    {
+      "epoch": 0.014964716370033302,
+      "grad_norm": 0.6100548505783081,
+      "learning_rate": 6.956599933947414e-05,
+      "loss": 0.06352347135543823,
+      "step": 24340
+    },
+    {
+      "epoch": 0.015256995986635515,
+      "grad_norm": 0.5612335801124573,
+      "learning_rate": 6.948927675030771e-05,
+      "loss": 0.06420612931251526,
+      "step": 24350
+    },
+    {
+      "epoch": 0.015549275603237727,
+      "grad_norm": 0.5349525213241577,
+      "learning_rate": 6.941257395921843e-05,
+      "loss": 0.06355353593826293,
+      "step": 24360
+    },
+    {
+      "epoch": 0.01584155521983994,
+      "grad_norm": 0.5295330882072449,
+      "learning_rate": 6.93358910159778e-05,
+      "loss": 0.061083996295928956,
+      "step": 24370
+    },
+    {
+      "epoch": 0.016133834836442155,
+      "grad_norm": 0.340029239654541,
+      "learning_rate": 6.925922797034452e-05,
+      "loss": 0.05503458976745605,
+      "step": 24380
+    },
+    {
+      "epoch": 0.016426114453044365,
+      "grad_norm": 0.5671619772911072,
+      "learning_rate": 6.918258487206453e-05,
+      "loss": 0.07984631061553955,
+      "step": 24390
+    },
+    {
+      "epoch": 0.016718394069646578,
+      "grad_norm": 0.5787197947502136,
+      "learning_rate": 6.910596177087057e-05,
+      "loss": 0.05469493269920349,
+      "step": 24400
+    },
+    {
+      "epoch": 0.01701067368624879,
+      "grad_norm": 0.8051468133926392,
+      "learning_rate": 6.902935871648266e-05,
+      "loss": 0.06979711651802063,
+      "step": 24410
+    },
+    {
+      "epoch": 0.017302953302851004,
+      "grad_norm": 0.4802290201187134,
+      "learning_rate": 6.895277575860765e-05,
+      "loss": 0.07207623720169068,
+      "step": 24420
+    },
+    {
+      "epoch": 0.017595232919453217,
+      "grad_norm": 0.4463903605937958,
+      "learning_rate": 6.887621294693931e-05,
+      "loss": 0.05882283449172974,
+      "step": 24430
+    },
+    {
+      "epoch": 0.01788751253605543,
+      "grad_norm": 0.7048805952072144,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 0.056611716747283936,
+      "step": 24440
+    },
+    {
+      "epoch": 0.018179792152657644,
+      "grad_norm": 0.885055422782898,
+      "learning_rate": 6.872314796093289e-05,
+      "loss": 0.058121681213378906,
+      "step": 24450
+    },
+    {
+      "epoch": 0.018472071769259857,
+      "grad_norm": 0.39340639114379883,
+      "learning_rate": 6.864664588591702e-05,
+      "loss": 0.045943108201026914,
+      "step": 24460
+    },
+    {
+      "epoch": 0.01876435138586207,
+      "grad_norm": 0.5940976142883301,
+      "learning_rate": 6.857016415575224e-05,
+      "loss": 0.061124229431152345,
+      "step": 24470
+    },
+    {
+      "epoch": 0.019056631002464283,
+      "grad_norm": 0.42219656705856323,
+      "learning_rate": 6.849370282006667e-05,
+      "loss": 0.06848087310791015,
+      "step": 24480
+    },
+    {
+      "epoch": 0.019348910619066496,
+      "grad_norm": 0.6950340270996094,
+      "learning_rate": 6.841726192847531e-05,
+      "loss": 0.06713900566101075,
+      "step": 24490
+    },
+    {
+      "epoch": 0.01964119023566871,
+      "grad_norm": 0.40027695894241333,
+      "learning_rate": 6.834084153057976e-05,
+      "loss": 0.05758211612701416,
+      "step": 24500
+    },
+    {
+      "epoch": 0.019933469852270923,
+      "grad_norm": 0.5102626085281372,
+      "learning_rate": 6.826444167596846e-05,
+      "loss": 0.06253283023834229,
+      "step": 24510
+    },
+    {
+      "epoch": 0.020225749468873132,
+      "grad_norm": 0.4730043411254883,
+      "learning_rate": 6.81880624142164e-05,
+      "loss": 0.0585313081741333,
+      "step": 24520
+    },
+    {
+      "epoch": 0.020518029085475346,
+      "grad_norm": 0.35318443179130554,
+      "learning_rate": 6.811170379488526e-05,
+      "loss": 0.03328502178192139,
+      "step": 24530
+    },
+    {
+      "epoch": 0.02081030870207756,
+      "grad_norm": 0.4538300931453705,
+      "learning_rate": 6.803536586752332e-05,
+      "loss": 0.06238994598388672,
+      "step": 24540
+    },
+    {
+      "epoch": 0.021102588318679772,
+      "grad_norm": 0.45233890414237976,
+      "learning_rate": 6.795904868166543e-05,
+      "loss": 0.04827843010425568,
+      "step": 24550
+    },
+    {
+      "epoch": 0.021394867935281985,
+      "grad_norm": 0.5169075727462769,
+      "learning_rate": 6.788275228683298e-05,
+      "loss": 0.04481933116912842,
+      "step": 24560
+    },
+    {
+      "epoch": 0.0216871475518842,
+      "grad_norm": 0.7321431636810303,
+      "learning_rate": 6.780647673253391e-05,
+      "loss": 0.0762115478515625,
+      "step": 24570
+    },
+    {
+      "epoch": 0.02197942716848641,
+      "grad_norm": 0.6765905618667603,
+      "learning_rate": 6.773022206826255e-05,
+      "loss": 0.05706472992897034,
+      "step": 24580
+    },
+    {
+      "epoch": 0.022271706785088625,
+      "grad_norm": 0.42739391326904297,
+      "learning_rate": 6.765398834349973e-05,
+      "loss": 0.048994019627571106,
+      "step": 24590
+    },
+    {
+      "epoch": 0.022563986401690838,
+      "grad_norm": 0.40778204798698425,
+      "learning_rate": 6.757777560771268e-05,
+      "loss": 0.04954122304916382,
+      "step": 24600
+    },
+    {
+      "epoch": 0.02285626601829305,
+      "grad_norm": 0.36776554584503174,
+      "learning_rate": 6.750158391035509e-05,
+      "loss": 0.03918545544147491,
+      "step": 24610
+    },
+    {
+      "epoch": 0.023148545634895264,
+      "grad_norm": 0.5056847333908081,
+      "learning_rate": 6.74254133008668e-05,
+      "loss": 0.06538706421852111,
+      "step": 24620
+    },
+    {
+      "epoch": 0.023440825251497478,
+      "grad_norm": 0.6032892465591431,
+      "learning_rate": 6.734926382867421e-05,
+      "loss": 0.06305201649665833,
+      "step": 24630
+    },
+    {
+      "epoch": 0.02373310486809969,
+      "grad_norm": 0.3110347390174866,
+      "learning_rate": 6.727313554318986e-05,
+      "loss": 0.047031229734420775,
+      "step": 24640
+    },
+    {
+      "epoch": 0.0240253844847019,
+      "grad_norm": 0.39956992864608765,
+      "learning_rate": 6.71970284938125e-05,
+      "loss": 0.0675952136516571,
+      "step": 24650
+    },
+    {
+      "epoch": 0.024317664101304114,
+      "grad_norm": 0.3942258059978485,
+      "learning_rate": 6.712094272992724e-05,
+      "loss": 0.05346781015396118,
+      "step": 24660
+    },
+    {
+      "epoch": 0.024609943717906327,
+      "grad_norm": 0.620644211769104,
+      "learning_rate": 6.704487830090526e-05,
+      "loss": 0.05971034169197083,
+      "step": 24670
+    },
+    {
+      "epoch": 0.02490222333450854,
+      "grad_norm": 0.5135174989700317,
+      "learning_rate": 6.696883525610397e-05,
+      "loss": 0.06254430413246155,
+      "step": 24680
+    },
+    {
+      "epoch": 0.025194502951110753,
+      "grad_norm": 0.335641473531723,
+      "learning_rate": 6.689281364486688e-05,
+      "loss": 0.05586135983467102,
+      "step": 24690
+    },
+    {
+      "epoch": 0.025486782567712966,
+      "grad_norm": 0.5126707553863525,
+      "learning_rate": 6.681681351652356e-05,
+      "loss": 0.05567861795425415,
+      "step": 24700
+    },
+    {
+      "epoch": 0.02577906218431518,
+      "grad_norm": 0.4966842234134674,
+      "learning_rate": 6.674083492038969e-05,
+      "loss": 0.047470131516456605,
+      "step": 24710
+    },
+    {
+      "epoch": 0.026071341800917393,
+      "grad_norm": 0.421207994222641,
+      "learning_rate": 6.666487790576696e-05,
+      "loss": 0.0636415183544159,
+      "step": 24720
+    },
+    {
+      "epoch": 0.026363621417519606,
+      "grad_norm": 0.6463746428489685,
+      "learning_rate": 6.658894252194303e-05,
+      "loss": 0.05859437584877014,
+      "step": 24730
+    },
+    {
+      "epoch": 0.02665590103412182,
+      "grad_norm": 0.5995907187461853,
+      "learning_rate": 6.651302881819159e-05,
+      "loss": 0.04756852388381958,
+      "step": 24740
+    },
+    {
+      "epoch": 0.026948180650724032,
+      "grad_norm": 0.4389137625694275,
+      "learning_rate": 6.643713684377213e-05,
+      "loss": 0.05140033960342407,
+      "step": 24750
+    },
+    {
+      "epoch": 0.027240460267326246,
+      "grad_norm": 0.5610835552215576,
+      "learning_rate": 6.636126664793023e-05,
+      "loss": 0.060775893926620486,
+      "step": 24760
+    },
+    {
+      "epoch": 0.02753273988392846,
+      "grad_norm": 0.6728721857070923,
+      "learning_rate": 6.628541827989716e-05,
+      "loss": 0.056725728511810306,
+      "step": 24770
+    },
+    {
+      "epoch": 0.027825019500530672,
+      "grad_norm": 0.2811421751976013,
+      "learning_rate": 6.620959178889017e-05,
+      "loss": 0.05698311924934387,
+      "step": 24780
+    },
+    {
+      "epoch": 0.02811729911713288,
+      "grad_norm": 0.5115475654602051,
+      "learning_rate": 6.61337872241122e-05,
+      "loss": 0.0604356586933136,
+      "step": 24790
+    },
+    {
+      "epoch": 0.028409578733735095,
+      "grad_norm": 0.31954365968704224,
+      "learning_rate": 6.6058004634752e-05,
+      "loss": 0.055602657794952395,
+      "step": 24800
+    },
+    {
+      "epoch": 0.028701858350337308,
+      "grad_norm": 0.6693370938301086,
+      "learning_rate": 6.59822440699841e-05,
+      "loss": 0.057810574769973755,
+      "step": 24810
+    },
+    {
+      "epoch": 0.02899413796693952,
+      "grad_norm": 0.5260143280029297,
+      "learning_rate": 6.590650557896868e-05,
+      "loss": 0.052703529596328735,
+      "step": 24820
+    },
+    {
+      "epoch": 0.029286417583541734,
+      "grad_norm": 0.4819251596927643,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 0.05563070774078369,
+      "step": 24830
+    },
+    {
+      "epoch": 0.029578697200143948,
+      "grad_norm": 0.6039358973503113,
+      "learning_rate": 6.575509501476455e-05,
+      "loss": 0.08833892345428467,
+      "step": 24840
+    },
+    {
+      "epoch": 0.02987097681674616,
+      "grad_norm": 0.5012726187705994,
+      "learning_rate": 6.56794230398245e-05,
+      "loss": 0.08579878211021423,
+      "step": 24850
+    },
+    {
+      "epoch": 0.030163256433348374,
+      "grad_norm": 0.5411273837089539,
+      "learning_rate": 6.560377333513423e-05,
+      "loss": 0.09530526995658875,
+      "step": 24860
+    },
+    {
+      "epoch": 0.030455536049950587,
+      "grad_norm": 0.5190746188163757,
+      "learning_rate": 6.5528145949782e-05,
+      "loss": 0.10469220876693726,
+      "step": 24870
+    },
+    {
+      "epoch": 0.0307478156665528,
+      "grad_norm": 0.4791228771209717,
+      "learning_rate": 6.54525409328416e-05,
+      "loss": 0.061126315593719484,
+      "step": 24880
+    },
+    {
+      "epoch": 0.031040095283155014,
+      "grad_norm": 0.3211660087108612,
+      "learning_rate": 6.537695833337233e-05,
+      "loss": 0.06827449202537536,
+      "step": 24890
+    },
+    {
+      "epoch": 0.03133237489975722,
+      "grad_norm": 0.42216864228248596,
+      "learning_rate": 6.530139820041889e-05,
+      "loss": 0.09422902464866638,
+      "step": 24900
+    },
+    {
+      "epoch": 0.031624654516359436,
+      "grad_norm": 0.5637394189834595,
+      "learning_rate": 6.522586058301146e-05,
+      "loss": 0.07057323455810546,
+      "step": 24910
+    },
+    {
+      "epoch": 0.03191693413296165,
+      "grad_norm": 0.5219411849975586,
+      "learning_rate": 6.515034553016553e-05,
+      "loss": 0.050888442993164064,
+      "step": 24920
+    },
+    {
+      "epoch": 0.03220921374956386,
+      "grad_norm": 0.4048364758491516,
+      "learning_rate": 6.507485309088206e-05,
+      "loss": 0.06263611912727356,
+      "step": 24930
+    },
+    {
+      "epoch": 0.032501493366166076,
+      "grad_norm": 0.436114102602005,
+      "learning_rate": 6.499938331414719e-05,
+      "loss": 0.0889339029788971,
+      "step": 24940
+    },
+    {
+      "epoch": 0.03279377298276829,
+      "grad_norm": 0.606167197227478,
+      "learning_rate": 6.492393624893254e-05,
+      "loss": 0.06758228540420533,
+      "step": 24950
+    },
+    {
+      "epoch": 0.0330860525993705,
+      "grad_norm": 0.3567550778388977,
+      "learning_rate": 6.484851194419484e-05,
+      "loss": 0.06603872776031494,
+      "step": 24960
+    },
+    {
+      "epoch": 0.033378332215972716,
+      "grad_norm": 0.6353266835212708,
+      "learning_rate": 6.477311044887609e-05,
+      "loss": 0.07201949357986451,
+      "step": 24970
+    },
+    {
+      "epoch": 0.03367061183257493,
+      "grad_norm": 0.4756733477115631,
+      "learning_rate": 6.469773181190352e-05,
+      "loss": 0.08585000038146973,
+      "step": 24980
+    },
+    {
+      "epoch": 0.03396289144917714,
+      "grad_norm": 0.4138638973236084,
+      "learning_rate": 6.462237608218947e-05,
+      "loss": 0.0880472183227539,
+      "step": 24990
+    },
+    {
+      "epoch": 0.034255171065779355,
+      "grad_norm": 0.31874802708625793,
+      "learning_rate": 6.454704330863155e-05,
+      "loss": 0.060063397884368895,
+      "step": 25000
+    },
+    {
+      "epoch": 0.03454745068238157,
+      "grad_norm": 0.5631740093231201,
+      "learning_rate": 6.44717335401123e-05,
+      "loss": 0.08560296893119812,
+      "step": 25010
+    },
+    {
+      "epoch": 0.03483973029898378,
+      "grad_norm": 0.6488601565361023,
+      "learning_rate": 6.439644682549941e-05,
+      "loss": 0.08573988676071168,
+      "step": 25020
+    },
+    {
+      "epoch": 0.035132009915585995,
+      "grad_norm": 0.4319780468940735,
+      "learning_rate": 6.432118321364568e-05,
+      "loss": 0.05632718801498413,
+      "step": 25030
+    },
+    {
+      "epoch": 0.03542428953218821,
+      "grad_norm": 0.5614952445030212,
+      "learning_rate": 6.424594275338873e-05,
+      "loss": 0.07978180050849915,
+      "step": 25040
+    },
+    {
+      "epoch": 0.03571656914879042,
+      "grad_norm": 0.5817257165908813,
+      "learning_rate": 6.417072549355141e-05,
+      "loss": 0.09732577204704285,
+      "step": 25050
+    },
+    {
+      "epoch": 0.036008848765392634,
+      "grad_norm": 0.5329292416572571,
+      "learning_rate": 6.409553148294129e-05,
+      "loss": 0.07865207195281983,
+      "step": 25060
+    },
+    {
+      "epoch": 0.03630112838199485,
+      "grad_norm": 0.5528441071510315,
+      "learning_rate": 6.402036077035101e-05,
+      "loss": 0.06335890293121338,
+      "step": 25070
+    },
+    {
+      "epoch": 0.03659340799859706,
+      "grad_norm": 0.34881412982940674,
+      "learning_rate": 6.394521340455801e-05,
+      "loss": 0.05581914782524109,
+      "step": 25080
+    },
+    {
+      "epoch": 0.036885687615199274,
+      "grad_norm": 0.6952479481697083,
+      "learning_rate": 6.387008943432455e-05,
+      "loss": 0.07483479380607605,
+      "step": 25090
+    },
+    {
+      "epoch": 0.03717796723180149,
+      "grad_norm": 0.34208396077156067,
+      "learning_rate": 6.379498890839783e-05,
+      "loss": 0.09240835309028625,
+      "step": 25100
+    },
+    {
+      "epoch": 0.03747024684840369,
+      "grad_norm": 0.613269031047821,
+      "learning_rate": 6.371991187550972e-05,
+      "loss": 0.09239505529403687,
+      "step": 25110
+    },
+    {
+      "epoch": 0.037762526465005906,
+      "grad_norm": 0.3677777647972107,
+      "learning_rate": 6.364485838437691e-05,
+      "loss": 0.06311186552047729,
+      "step": 25120
+    },
+    {
+      "epoch": 0.03805480608160812,
+      "grad_norm": 0.6809932589530945,
+      "learning_rate": 6.35698284837008e-05,
+      "loss": 0.09031100273132324,
+      "step": 25130
+    },
+    {
+      "epoch": 0.03834708569821033,
+      "grad_norm": 0.6572308540344238,
+      "learning_rate": 6.349482222216747e-05,
+      "loss": 0.06920945048332214,
+      "step": 25140
+    },
+    {
+      "epoch": 0.038639365314812546,
+      "grad_norm": 0.45274817943573,
+      "learning_rate": 6.341983964844766e-05,
+      "loss": 0.08126962780952454,
+      "step": 25150
+    },
+    {
+      "epoch": 0.03893164493141476,
+      "grad_norm": 0.2988074719905853,
+      "learning_rate": 6.334488081119676e-05,
+      "loss": 0.07347806692123413,
+      "step": 25160
+    },
+    {
+      "epoch": 0.03922392454801697,
+      "grad_norm": 0.6382627487182617,
+      "learning_rate": 6.326994575905475e-05,
+      "loss": 0.06357109546661377,
+      "step": 25170
+    },
+    {
+      "epoch": 0.039516204164619186,
+      "grad_norm": 0.44822317361831665,
+      "learning_rate": 6.319503454064618e-05,
+      "loss": 0.09275277853012084,
+      "step": 25180
+    },
+    {
+      "epoch": 0.0398084837812214,
+      "grad_norm": 0.42456361651420593,
+      "learning_rate": 6.312014720458006e-05,
+      "loss": 0.0712854266166687,
+      "step": 25190
+    },
+    {
+      "epoch": 0.04010076339782361,
+      "grad_norm": 0.5751881003379822,
+      "learning_rate": 6.304528379945006e-05,
+      "loss": 0.11173251867294312,
+      "step": 25200
+    },
+    {
+      "epoch": 0.040393043014425825,
+      "grad_norm": 0.6720516681671143,
+      "learning_rate": 6.297044437383415e-05,
+      "loss": 0.06843043565750122,
+      "step": 25210
+    },
+    {
+      "epoch": 0.04068532263102804,
+      "grad_norm": 0.3326621353626251,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 0.07715659737586975,
+      "step": 25220
+    },
+    {
+      "epoch": 0.04097760224763025,
+      "grad_norm": 0.45442450046539307,
+      "learning_rate": 6.282083765537918e-05,
+      "loss": 0.07723469734191894,
+      "step": 25230
+    },
+    {
+      "epoch": 0.041269881864232465,
+      "grad_norm": 0.43177512288093567,
+      "learning_rate": 6.274607045961823e-05,
+      "loss": 0.060142111778259275,
+      "step": 25240
+    },
+    {
+      "epoch": 0.04156216148083468,
+      "grad_norm": 0.5753762722015381,
+      "learning_rate": 6.267132743752773e-05,
+      "loss": 0.11492490768432617,
+      "step": 25250
+    },
+    {
+      "epoch": 0.04185444109743689,
+      "grad_norm": 0.38342735171318054,
+      "learning_rate": 6.259660863760753e-05,
+      "loss": 0.056512337923049924,
+      "step": 25260
+    },
+    {
+      "epoch": 0.042146720714039104,
+      "grad_norm": 0.7967780828475952,
+      "learning_rate": 6.252191410834194e-05,
+      "loss": 0.09975331425666809,
+      "step": 25270
+    },
+    {
+      "epoch": 0.04243900033064132,
+      "grad_norm": 0.525600016117096,
+      "learning_rate": 6.244724389819941e-05,
+      "loss": 0.08848577737808228,
+      "step": 25280
+    },
+    {
+      "epoch": 0.04273127994724353,
+      "grad_norm": 0.39150717854499817,
+      "learning_rate": 6.237259805563256e-05,
+      "loss": 0.10443159341812133,
+      "step": 25290
+    },
+    {
+      "epoch": 0.043023559563845744,
+      "grad_norm": 0.47313767671585083,
+      "learning_rate": 6.229797662907835e-05,
+      "loss": 0.07029808759689331,
+      "step": 25300
+    },
+    {
+      "epoch": 0.04331583918044796,
+      "grad_norm": 0.5901621580123901,
+      "learning_rate": 6.222337966695776e-05,
+      "loss": 0.06134530305862427,
+      "step": 25310
+    },
+    {
+      "epoch": 0.04360811879705017,
+      "grad_norm": 0.5567722916603088,
+      "learning_rate": 6.214880721767599e-05,
+      "loss": 0.07228773236274719,
+      "step": 25320
+    },
+    {
+      "epoch": 0.04390039841365238,
+      "grad_norm": 0.5567444562911987,
+      "learning_rate": 6.207425932962226e-05,
+      "loss": 0.09007618427276612,
+      "step": 25330
+    },
+    {
+      "epoch": 0.0441926780302546,
+      "grad_norm": 0.5103462338447571,
+      "learning_rate": 6.199973605116991e-05,
+      "loss": 0.06993064284324646,
+      "step": 25340
+    },
+    {
+      "epoch": 0.04448495764685681,
+      "grad_norm": 0.5902025103569031,
+      "learning_rate": 6.192523743067626e-05,
+      "loss": 0.08377394080162048,
+      "step": 25350
+    },
+    {
+      "epoch": 0.04477723726345902,
+      "grad_norm": 0.411850243806839,
+      "learning_rate": 6.185076351648266e-05,
+      "loss": 0.0676827847957611,
+      "step": 25360
+    },
+    {
+      "epoch": 0.04506951688006123,
+      "grad_norm": 0.34316718578338623,
+      "learning_rate": 6.177631435691447e-05,
+      "loss": 0.050176918506622314,
+      "step": 25370
+    },
+    {
+      "epoch": 0.04536179649666344,
+      "grad_norm": 0.46456581354141235,
+      "learning_rate": 6.170189000028087e-05,
+      "loss": 0.05937499403953552,
+      "step": 25380
+    },
+    {
+      "epoch": 0.045654076113265656,
+      "grad_norm": 0.5560199022293091,
+      "learning_rate": 6.162749049487509e-05,
+      "loss": 0.058543235063552856,
+      "step": 25390
+    },
+    {
+      "epoch": 0.04594635572986787,
+      "grad_norm": 0.45371419191360474,
+      "learning_rate": 6.15531158889741e-05,
+      "loss": 0.061385565996170045,
+      "step": 25400
+    },
+    {
+      "epoch": 0.04623863534647008,
+      "grad_norm": 0.5840332508087158,
+      "learning_rate": 6.147876623083878e-05,
+      "loss": 0.09458993077278137,
+      "step": 25410
+    },
+    {
+      "epoch": 0.046530914963072295,
+      "grad_norm": 0.34834399819374084,
+      "learning_rate": 6.140444156871386e-05,
+      "loss": 0.08331232070922852,
+      "step": 25420
+    },
+    {
+      "epoch": 0.04682319457967451,
+      "grad_norm": 0.6470980048179626,
+      "learning_rate": 6.133014195082773e-05,
+      "loss": 0.08125271201133728,
+      "step": 25430
+    },
+    {
+      "epoch": 0.04711547419627672,
+      "grad_norm": 0.6033153533935547,
+      "learning_rate": 6.125586742539271e-05,
+      "loss": 0.07806663513183594,
+      "step": 25440
+    },
+    {
+      "epoch": 0.047407753812878935,
+      "grad_norm": 0.5457965731620789,
+      "learning_rate": 6.118161804060469e-05,
+      "loss": 0.07718234658241271,
+      "step": 25450
+    },
+    {
+      "epoch": 0.04770003342948115,
+      "grad_norm": 0.5240962505340576,
+      "learning_rate": 6.110739384464323e-05,
+      "loss": 0.07397805452346802,
+      "step": 25460
+    },
+    {
+      "epoch": 0.04799231304608336,
+      "grad_norm": 0.35100477933883667,
+      "learning_rate": 6.1033194885671684e-05,
+      "loss": 0.09175940155982971,
+      "step": 25470
+    },
+    {
+      "epoch": 0.048284592662685574,
+      "grad_norm": 0.3964548110961914,
+      "learning_rate": 6.09590212118369e-05,
+      "loss": 0.06746885776519776,
+      "step": 25480
+    },
+    {
+      "epoch": 0.04857687227928779,
+      "grad_norm": 0.5447602272033691,
+      "learning_rate": 6.088487287126941e-05,
+      "loss": 0.06217803955078125,
+      "step": 25490
+    },
+    {
+      "epoch": 0.04886915189589,
+      "grad_norm": 0.51618492603302,
+      "learning_rate": 6.081074991208325e-05,
+      "loss": 0.05974909067153931,
+      "step": 25500
+    },
+    {
+      "epoch": 0.049161431512492214,
+      "grad_norm": 0.4656076729297638,
+      "learning_rate": 6.073665238237598e-05,
+      "loss": 0.06172404885292053,
+      "step": 25510
+    },
+    {
+      "epoch": 0.04945371112909443,
+      "grad_norm": 0.5572783946990967,
+      "learning_rate": 6.0662580330228726e-05,
+      "loss": 0.07774239182472228,
+      "step": 25520
+    },
+    {
+      "epoch": 0.04974599074569664,
+      "grad_norm": 0.46146318316459656,
+      "learning_rate": 6.0588533803705994e-05,
+      "loss": 0.061299079656600954,
+      "step": 25530
+    },
+    {
+      "epoch": 0.05003827036229885,
+      "grad_norm": 0.43785396218299866,
+      "learning_rate": 6.051451285085581e-05,
+      "loss": 0.0788904845714569,
+      "step": 25540
+    },
+    {
+      "epoch": 0.05033054997890107,
+      "grad_norm": 0.35123780369758606,
+      "learning_rate": 6.044051751970952e-05,
+      "loss": 0.06761659979820252,
+      "step": 25550
+    },
+    {
+      "epoch": 0.05062282959550328,
+      "grad_norm": 0.5729063153266907,
+      "learning_rate": 6.036654785828195e-05,
+      "loss": 0.09136456251144409,
+      "step": 25560
+    },
+    {
+      "epoch": 0.05091510921210549,
+      "grad_norm": 0.5067956447601318,
+      "learning_rate": 6.029260391457119e-05,
+      "loss": 0.07008501887321472,
+      "step": 25570
+    },
+    {
+      "epoch": 0.051207388828707706,
+      "grad_norm": 0.7191058993339539,
+      "learning_rate": 6.021868573655862e-05,
+      "loss": 0.08219128251075744,
+      "step": 25580
+    },
+    {
+      "epoch": 0.05149966844530992,
+      "grad_norm": 0.48772290349006653,
+      "learning_rate": 6.0144793372209e-05,
+      "loss": 0.039920774102210996,
+      "step": 25590
+    },
+    {
+      "epoch": 0.05179194806191213,
+      "grad_norm": 0.5411823391914368,
+      "learning_rate": 6.007092686947021e-05,
+      "loss": 0.051695460081100465,
+      "step": 25600
+    },
+    {
+      "epoch": 0.052084227678514346,
+      "grad_norm": 0.3366369307041168,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 0.06556240320205689,
+      "step": 25610
+    },
+    {
+      "epoch": 0.05237650729511656,
+      "grad_norm": 0.39403146505355835,
+      "learning_rate": 5.9923271640533264e-05,
+      "loss": 0.06177469491958618,
+      "step": 25620
+    },
+    {
+      "epoch": 0.052668786911718765,
+      "grad_norm": 0.5240152478218079,
+      "learning_rate": 5.9849483010146855e-05,
+      "loss": 0.06305868029594422,
+      "step": 25630
+    },
+    {
+      "epoch": 0.05296106652832098,
+      "grad_norm": 0.486549437046051,
+      "learning_rate": 5.977572043299508e-05,
+      "loss": 0.09278948903083802,
+      "step": 25640
+    },
+    {
+      "epoch": 0.05325334614492319,
+      "grad_norm": 0.5442919135093689,
+      "learning_rate": 5.970198395694156e-05,
+      "loss": 0.058168810606002805,
+      "step": 25650
+    },
+    {
+      "epoch": 0.053545625761525405,
+      "grad_norm": 0.6443012952804565,
+      "learning_rate": 5.962827362983319e-05,
+      "loss": 0.07052817940711975,
+      "step": 25660
+    },
+    {
+      "epoch": 0.05383790537812762,
+      "grad_norm": 0.5261877775192261,
+      "learning_rate": 5.9554589499499744e-05,
+      "loss": 0.07178923487663269,
+      "step": 25670
+    },
+    {
+      "epoch": 0.05413018499472983,
+      "grad_norm": 0.3359242081642151,
+      "learning_rate": 5.948093161375402e-05,
+      "loss": 0.058383262157440184,
+      "step": 25680
+    },
+    {
+      "epoch": 0.054422464611332044,
+      "grad_norm": 0.6166980862617493,
+      "learning_rate": 5.94073000203919e-05,
+      "loss": 0.08712172508239746,
+      "step": 25690
+    },
+    {
+      "epoch": 0.05471474422793426,
+      "grad_norm": 0.5819984078407288,
+      "learning_rate": 5.933369476719206e-05,
+      "loss": 0.08704743385314942,
+      "step": 25700
+    },
+    {
+      "epoch": 0.05500702384453647,
+      "grad_norm": 0.5098153948783875,
+      "learning_rate": 5.92601159019162e-05,
+      "loss": 0.06231500506401062,
+      "step": 25710
+    },
+    {
+      "epoch": 0.055299303461138684,
+      "grad_norm": 0.33018723130226135,
+      "learning_rate": 5.918656347230883e-05,
+      "loss": 0.07798768877983094,
+      "step": 25720
+    },
+    {
+      "epoch": 0.0555915830777409,
+      "grad_norm": 0.4951116144657135,
+      "learning_rate": 5.9113037526097296e-05,
+      "loss": 0.05216352939605713,
+      "step": 25730
+    },
+    {
+      "epoch": 0.05588386269434311,
+      "grad_norm": 0.4441530108451843,
+      "learning_rate": 5.903953811099183e-05,
+      "loss": 0.06253697276115418,
+      "step": 25740
+    },
+    {
+      "epoch": 0.05617614231094532,
+      "grad_norm": 0.4798872470855713,
+      "learning_rate": 5.896606527468535e-05,
+      "loss": 0.07455722093582154,
+      "step": 25750
+    },
+    {
+      "epoch": 0.05646842192754754,
+      "grad_norm": 0.45665329694747925,
+      "learning_rate": 5.889261906485366e-05,
+      "loss": 0.05573180913925171,
+      "step": 25760
+    },
+    {
+      "epoch": 0.05676070154414975,
+      "grad_norm": 0.41299259662628174,
+      "learning_rate": 5.8819199529155134e-05,
+      "loss": 0.061070823669433595,
+      "step": 25770
+    },
+    {
+      "epoch": 0.05705298116075196,
+      "grad_norm": 0.44594302773475647,
+      "learning_rate": 5.874580671523093e-05,
+      "loss": 0.07389472723007202,
+      "step": 25780
+    },
+    {
+      "epoch": 0.057345260777354176,
+      "grad_norm": 0.36203038692474365,
+      "learning_rate": 5.8672440670704886e-05,
+      "loss": 0.06754085421562195,
+      "step": 25790
+    },
+    {
+      "epoch": 0.05763754039395639,
+      "grad_norm": 0.5072854161262512,
+      "learning_rate": 5.859910144318338e-05,
+      "loss": 0.05979921817779541,
+      "step": 25800
+    },
+    {
+      "epoch": 0.0579298200105586,
+      "grad_norm": 0.9090243577957153,
+      "learning_rate": 5.8525789080255475e-05,
+      "loss": 0.07524786591529846,
+      "step": 25810
+    },
+    {
+      "epoch": 0.058222099627160816,
+      "grad_norm": 0.4135330021381378,
+      "learning_rate": 5.8452503629492726e-05,
+      "loss": 0.0595844566822052,
+      "step": 25820
+    },
+    {
+      "epoch": 0.05851437924376303,
+      "grad_norm": 0.48103225231170654,
+      "learning_rate": 5.8379245138449346e-05,
+      "loss": 0.06255961656570434,
+      "step": 25830
+    },
+    {
+      "epoch": 0.05880665886036524,
+      "grad_norm": 0.6845728158950806,
+      "learning_rate": 5.830601365466193e-05,
+      "loss": 0.08557510972023011,
+      "step": 25840
+    },
+    {
+      "epoch": 0.059098938476967455,
+      "grad_norm": 0.4284327030181885,
+      "learning_rate": 5.823280922564955e-05,
+      "loss": 0.0919042944908142,
+      "step": 25850
+    },
+    {
+      "epoch": 0.05939121809356967,
+      "grad_norm": 0.920286238193512,
+      "learning_rate": 5.815963189891381e-05,
+      "loss": 0.06455819606781006,
+      "step": 25860
+    },
+    {
+      "epoch": 0.05968349771017188,
+      "grad_norm": 0.5946494936943054,
+      "learning_rate": 5.808648172193868e-05,
+      "loss": 0.07752955555915833,
+      "step": 25870
+    },
+    {
+      "epoch": 0.059975777326774095,
+      "grad_norm": 0.5104311108589172,
+      "learning_rate": 5.801335874219054e-05,
+      "loss": 0.060300403833389284,
+      "step": 25880
+    },
+    {
+      "epoch": 0.0602680569433763,
+      "grad_norm": 0.7605300545692444,
+      "learning_rate": 5.7940263007118035e-05,
+      "loss": 0.05605647563934326,
+      "step": 25890
+    },
+    {
+      "epoch": 0.060560336559978514,
+      "grad_norm": 0.3852718472480774,
+      "learning_rate": 5.786719456415224e-05,
+      "loss": 0.05162520408630371,
+      "step": 25900
+    },
+    {
+      "epoch": 0.06085261617658073,
+      "grad_norm": 0.5279809236526489,
+      "learning_rate": 5.779415346070649e-05,
+      "loss": 0.07617447972297668,
+      "step": 25910
+    },
+    {
+      "epoch": 0.06114489579318294,
+      "grad_norm": 0.40027275681495667,
+      "learning_rate": 5.772113974417626e-05,
+      "loss": 0.046496248245239256,
+      "step": 25920
+    },
+    {
+      "epoch": 0.061437175409785154,
+      "grad_norm": 0.6157183647155762,
+      "learning_rate": 5.76481534619395e-05,
+      "loss": 0.06601297855377197,
+      "step": 25930
+    },
+    {
+      "epoch": 0.06172945502638737,
+      "grad_norm": 0.4839232563972473,
+      "learning_rate": 5.757519466135617e-05,
+      "loss": 0.0670687735080719,
+      "step": 25940
+    },
+    {
+      "epoch": 0.06202173464298958,
+      "grad_norm": 0.4323160648345947,
+      "learning_rate": 5.750226338976834e-05,
+      "loss": 0.05756822824478149,
+      "step": 25950
+    },
+    {
+      "epoch": 0.06231401425959179,
+      "grad_norm": 0.5168435573577881,
+      "learning_rate": 5.74293596945005e-05,
+      "loss": 0.08524787425994873,
+      "step": 25960
+    },
+    {
+      "epoch": 0.06260629387619401,
+      "grad_norm": 0.4545714259147644,
+      "learning_rate": 5.73564836228589e-05,
+      "loss": 0.09718636870384216,
+      "step": 25970
+    },
+    {
+      "epoch": 0.06289857349279622,
+      "grad_norm": 0.40314778685569763,
+      "learning_rate": 5.7283635222132124e-05,
+      "loss": 0.07916260361671448,
+      "step": 25980
+    },
+    {
+      "epoch": 0.06319085310939844,
+      "grad_norm": 0.554923951625824,
+      "learning_rate": 5.721081453959073e-05,
+      "loss": 0.10156934261322022,
+      "step": 25990
+    },
+    {
+      "epoch": 0.06348313272600065,
+      "grad_norm": 0.27900126576423645,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 0.07747212052345276,
+      "step": 26000
+    },
+    {
+      "epoch": 0.06377541234260287,
+      "grad_norm": 0.4654293656349182,
+      "learning_rate": 5.7065256518056054e-05,
+      "loss": 0.07203555703163148,
+      "step": 26010
+    },
+    {
+      "epoch": 0.06406769195920507,
+      "grad_norm": 0.44500094652175903,
+      "learning_rate": 5.6992519273513836e-05,
+      "loss": 0.05634813904762268,
+      "step": 26020
+    },
+    {
+      "epoch": 0.06435997157580728,
+      "grad_norm": 0.39121586084365845,
+      "learning_rate": 5.6919809936058964e-05,
+      "loss": 0.08185821175575256,
+      "step": 26030
+    },
+    {
+      "epoch": 0.0646522511924095,
+      "grad_norm": 0.2775823771953583,
+      "learning_rate": 5.684712855287169e-05,
+      "loss": 0.08072077035903931,
+      "step": 26040
+    },
+    {
+      "epoch": 0.0649445308090117,
+      "grad_norm": 0.5147310495376587,
+      "learning_rate": 5.677447517111418e-05,
+      "loss": 0.07316410541534424,
+      "step": 26050
+    },
+    {
+      "epoch": 0.06523681042561393,
+      "grad_norm": 0.4894447922706604,
+      "learning_rate": 5.670184983793049e-05,
+      "loss": 0.07036622762680053,
+      "step": 26060
+    },
+    {
+      "epoch": 0.06552909004221613,
+      "grad_norm": 0.4137035012245178,
+      "learning_rate": 5.6629252600446314e-05,
+      "loss": 0.07442597150802613,
+      "step": 26070
+    },
+    {
+      "epoch": 0.06582136965881835,
+      "grad_norm": 0.5181429386138916,
+      "learning_rate": 5.655668350576926e-05,
+      "loss": 0.06565752029418945,
+      "step": 26080
+    },
+    {
+      "epoch": 0.06611364927542056,
+      "grad_norm": 0.4495132863521576,
+      "learning_rate": 5.648414260098861e-05,
+      "loss": 0.08064934611320496,
+      "step": 26090
+    },
+    {
+      "epoch": 0.06640592889202278,
+      "grad_norm": 0.5050700306892395,
+      "learning_rate": 5.641162993317544e-05,
+      "loss": 0.07083800435066223,
+      "step": 26100
+    },
+    {
+      "epoch": 0.06669820850862498,
+      "grad_norm": 0.4610472023487091,
+      "learning_rate": 5.633914554938231e-05,
+      "loss": 0.08283621072769165,
+      "step": 26110
+    },
+    {
+      "epoch": 0.0669904881252272,
+      "grad_norm": 0.35165438055992126,
+      "learning_rate": 5.626668949664362e-05,
+      "loss": 0.07126052975654602,
+      "step": 26120
+    },
+    {
+      "epoch": 0.06728276774182941,
+      "grad_norm": 0.39704883098602295,
+      "learning_rate": 5.619426182197536e-05,
+      "loss": 0.07787105441093445,
+      "step": 26130
+    },
+    {
+      "epoch": 0.06757504735843163,
+      "grad_norm": 0.6571190357208252,
+      "learning_rate": 5.612186257237493e-05,
+      "loss": 0.0765614628791809,
+      "step": 26140
+    },
+    {
+      "epoch": 0.06786732697503384,
+      "grad_norm": 0.4015870690345764,
+      "learning_rate": 5.6049491794821573e-05,
+      "loss": 0.14204491376876832,
+      "step": 26150
+    },
+    {
+      "epoch": 0.06815960659163606,
+      "grad_norm": 0.3941798508167267,
+      "learning_rate": 5.597714953627584e-05,
+      "loss": 0.1126672863960266,
+      "step": 26160
+    },
+    {
+      "epoch": 0.06845188620823826,
+      "grad_norm": 0.49161991477012634,
+      "learning_rate": 5.5904835843679736e-05,
+      "loss": 0.09784611463546752,
+      "step": 26170
+    },
+    {
+      "epoch": 0.06874416582484048,
+      "grad_norm": 0.5977054238319397,
+      "learning_rate": 5.5832550763956994e-05,
+      "loss": 0.14737846851348876,
+      "step": 26180
+    },
+    {
+      "epoch": 0.06903644544144269,
+      "grad_norm": 0.4089168608188629,
+      "learning_rate": 5.576029434401251e-05,
+      "loss": 0.1128377914428711,
+      "step": 26190
+    },
+    {
+      "epoch": 0.06932872505804491,
+      "grad_norm": 0.6502609848976135,
+      "learning_rate": 5.5688066630732716e-05,
+      "loss": 0.13638105392456054,
+      "step": 26200
+    },
+    {
+      "epoch": 0.06962100467464712,
+      "grad_norm": 0.27721264958381653,
+      "learning_rate": 5.561586767098543e-05,
+      "loss": 0.08066203594207763,
+      "step": 26210
+    },
+    {
+      "epoch": 0.06991328429124934,
+      "grad_norm": 0.39576664566993713,
+      "learning_rate": 5.554369751161972e-05,
+      "loss": 0.11493515968322754,
+      "step": 26220
+    },
+    {
+      "epoch": 0.07020556390785154,
+      "grad_norm": 0.39894989132881165,
+      "learning_rate": 5.5471556199466026e-05,
+      "loss": 0.10165752172470092,
+      "step": 26230
+    },
+    {
+      "epoch": 0.07049784352445376,
+      "grad_norm": 0.37745293974876404,
+      "learning_rate": 5.539944378133608e-05,
+      "loss": 0.08957338333129883,
+      "step": 26240
+    },
+    {
+      "epoch": 0.07079012314105597,
+      "grad_norm": 0.47377991676330566,
+      "learning_rate": 5.532736030402288e-05,
+      "loss": 0.09178929924964904,
+      "step": 26250
+    },
+    {
+      "epoch": 0.07108240275765819,
+      "grad_norm": 0.550891637802124,
+      "learning_rate": 5.525530581430054e-05,
+      "loss": 0.08451678156852722,
+      "step": 26260
+    },
+    {
+      "epoch": 0.0713746823742604,
+      "grad_norm": 0.6345214247703552,
+      "learning_rate": 5.518328035892446e-05,
+      "loss": 0.09793843030929565,
+      "step": 26270
+    },
+    {
+      "epoch": 0.0716669619908626,
+      "grad_norm": 0.7177239060401917,
+      "learning_rate": 5.5111283984631236e-05,
+      "loss": 0.11980003118515015,
+      "step": 26280
+    },
+    {
+      "epoch": 0.07195924160746482,
+      "grad_norm": 0.27169811725616455,
+      "learning_rate": 5.503931673813845e-05,
+      "loss": 0.09425142407417297,
+      "step": 26290
+    },
+    {
+      "epoch": 0.07225152122406703,
+      "grad_norm": 0.6814448237419128,
+      "learning_rate": 5.4967378666144885e-05,
+      "loss": 0.10374590158462524,
+      "step": 26300
+    },
+    {
+      "epoch": 0.07254380084066925,
+      "grad_norm": 0.6821067929267883,
+      "learning_rate": 5.48954698153304e-05,
+      "loss": 0.12443258762359619,
+      "step": 26310
+    },
+    {
+      "epoch": 0.07283608045727145,
+      "grad_norm": 0.42043808102607727,
+      "learning_rate": 5.4823590232355904e-05,
+      "loss": 0.10479798316955566,
+      "step": 26320
+    },
+    {
+      "epoch": 0.07312836007387367,
+      "grad_norm": 0.6145655512809753,
+      "learning_rate": 5.475173996386319e-05,
+      "loss": 0.10997918844223023,
+      "step": 26330
+    },
+    {
+      "epoch": 0.07342063969047588,
+      "grad_norm": 0.5663999915122986,
+      "learning_rate": 5.467991905647516e-05,
+      "loss": 0.11678080558776856,
+      "step": 26340
+    },
+    {
+      "epoch": 0.0737129193070781,
+      "grad_norm": 0.740218460559845,
+      "learning_rate": 5.4608127556795655e-05,
+      "loss": 0.12079640626907348,
+      "step": 26350
+    },
+    {
+      "epoch": 0.07400519892368031,
+      "grad_norm": 0.32211267948150635,
+      "learning_rate": 5.45363655114093e-05,
+      "loss": 0.07736040949821472,
+      "step": 26360
+    },
+    {
+      "epoch": 0.07429747854028253,
+      "grad_norm": 0.5076852440834045,
+      "learning_rate": 5.446463296688184e-05,
+      "loss": 0.12274774312973022,
+      "step": 26370
+    },
+    {
+      "epoch": 0.07458975815688473,
+      "grad_norm": 0.43246200680732727,
+      "learning_rate": 5.4392929969759685e-05,
+      "loss": 0.10228569507598877,
+      "step": 26380
+    },
+    {
+      "epoch": 0.07488203777348695,
+      "grad_norm": 0.372988760471344,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 0.10375548601150512,
+      "step": 26390
+    },
+    {
+      "epoch": 0.07517431739008916,
+      "grad_norm": 0.6447717547416687,
+      "learning_rate": 5.424961280382116e-05,
+      "loss": 0.13338868618011473,
+      "step": 26400
+    },
+    {
+      "epoch": 0.07546659700669138,
+      "grad_norm": 0.5936722755432129,
+      "learning_rate": 5.417799872800179e-05,
+      "loss": 0.11825110912322997,
+      "step": 26410
+    },
+    {
+      "epoch": 0.07575887662329359,
+      "grad_norm": 0.3081055283546448,
+      "learning_rate": 5.410641438558156e-05,
+      "loss": 0.08084624409675598,
+      "step": 26420
+    },
+    {
+      "epoch": 0.0760511562398958,
+      "grad_norm": 0.6174929738044739,
+      "learning_rate": 5.4034859823010816e-05,
+      "loss": 0.1702901005744934,
+      "step": 26430
+    },
+    {
+      "epoch": 0.07634343585649801,
+      "grad_norm": 0.4797796607017517,
+      "learning_rate": 5.396333508672047e-05,
+      "loss": 0.15757374763488768,
+      "step": 26440
+    },
+    {
+      "epoch": 0.07663571547310023,
+      "grad_norm": 0.38025957345962524,
+      "learning_rate": 5.389184022312215e-05,
+      "loss": 0.17411439418792723,
+      "step": 26450
+    },
+    {
+      "epoch": 0.07692799508970244,
+      "grad_norm": 0.4931367039680481,
+      "learning_rate": 5.382037527860814e-05,
+      "loss": 0.14486730098724365,
+      "step": 26460
+    },
+    {
+      "epoch": 0.07722027470630466,
+      "grad_norm": 0.42914897203445435,
+      "learning_rate": 5.3748940299551276e-05,
+      "loss": 0.20778086185455322,
+      "step": 26470
+    },
+    {
+      "epoch": 0.07751255432290687,
+      "grad_norm": 0.5748200416564941,
+      "learning_rate": 5.3677535332304874e-05,
+      "loss": 0.17055492401123046,
+      "step": 26480
+    },
+    {
+      "epoch": 0.07780483393950909,
+      "grad_norm": 0.42499110102653503,
+      "learning_rate": 5.360616042320288e-05,
+      "loss": 0.19765783548355104,
+      "step": 26490
+    },
+    {
+      "epoch": 0.07809711355611129,
+      "grad_norm": 0.6965202689170837,
+      "learning_rate": 5.353481561855973e-05,
+      "loss": 0.17228747606277467,
+      "step": 26500
+    },
+    {
+      "epoch": 0.07838939317271351,
+      "grad_norm": 0.4805402457714081,
+      "learning_rate": 5.346350096467023e-05,
+      "loss": 0.19510977268218993,
+      "step": 26510
+    },
+    {
+      "epoch": 0.07868167278931572,
+      "grad_norm": 0.4773094356060028,
+      "learning_rate": 5.3392216507809714e-05,
+      "loss": 0.1962913155555725,
+      "step": 26520
+    },
+    {
+      "epoch": 0.07897395240591794,
+      "grad_norm": 0.4109029173851013,
+      "learning_rate": 5.332096229423389e-05,
+      "loss": 0.13397789001464844,
+      "step": 26530
+    },
+    {
+      "epoch": 0.07926623202252014,
+      "grad_norm": 0.32015061378479004,
+      "learning_rate": 5.324973837017888e-05,
+      "loss": 0.16465796232223512,
+      "step": 26540
+    },
+    {
+      "epoch": 0.07955851163912235,
+      "grad_norm": 0.44110772013664246,
+      "learning_rate": 5.317854478186104e-05,
+      "loss": 0.1621596097946167,
+      "step": 26550
+    },
+    {
+      "epoch": 0.07985079125572457,
+      "grad_norm": 0.5469281673431396,
+      "learning_rate": 5.3107381575477166e-05,
+      "loss": 0.12018738985061646,
+      "step": 26560
+    },
+    {
+      "epoch": 0.08014307087232678,
+      "grad_norm": 0.38715219497680664,
+      "learning_rate": 5.303624879720433e-05,
+      "loss": 0.15567872524261475,
+      "step": 26570
+    },
+    {
+      "epoch": 0.080435350488929,
+      "grad_norm": 0.6251632571220398,
+      "learning_rate": 5.296514649319968e-05,
+      "loss": 0.2060802936553955,
+      "step": 26580
+    },
+    {
+      "epoch": 0.0807276301055312,
+      "grad_norm": 0.5366135835647583,
+      "learning_rate": 5.289407470960089e-05,
+      "loss": 0.17430270910263063,
+      "step": 26590
+    },
+    {
+      "epoch": 0.08101990972213342,
+      "grad_norm": 0.4725992679595947,
+      "learning_rate": 5.282303349252562e-05,
+      "loss": 0.19848705530166627,
+      "step": 26600
+    },
+    {
+      "epoch": 0.08131218933873563,
+      "grad_norm": 0.4279593527317047,
+      "learning_rate": 5.2752022888071625e-05,
+      "loss": 0.2082925796508789,
+      "step": 26610
+    },
+    {
+      "epoch": 0.08160446895533785,
+      "grad_norm": 0.5000889897346497,
+      "learning_rate": 5.268104294231711e-05,
+      "loss": 0.14449467658996581,
+      "step": 26620
+    },
+    {
+      "epoch": 0.08189674857194006,
+      "grad_norm": 0.41607484221458435,
+      "learning_rate": 5.261009370132004e-05,
+      "loss": 0.15890104770660402,
+      "step": 26630
+    },
+    {
+      "epoch": 0.08218902818854228,
+      "grad_norm": 0.44860905408859253,
+      "learning_rate": 5.2539175211118685e-05,
+      "loss": 0.1452787160873413,
+      "step": 26640
+    },
+    {
+      "epoch": 0.08248130780514448,
+      "grad_norm": 0.4956583082675934,
+      "learning_rate": 5.2468287517731276e-05,
+      "loss": 0.18281996250152588,
+      "step": 26650
+    },
+    {
+      "epoch": 0.0827735874217467,
+      "grad_norm": 0.41813892126083374,
+      "learning_rate": 5.239743066715603e-05,
+      "loss": 0.15938519239425658,
+      "step": 26660
+    },
+    {
+      "epoch": 0.08306586703834891,
+      "grad_norm": 0.4784422814846039,
+      "learning_rate": 5.2326604705371206e-05,
+      "loss": 0.1637699007987976,
+      "step": 26670
+    },
+    {
+      "epoch": 0.08335814665495113,
+      "grad_norm": 0.642093300819397,
+      "learning_rate": 5.2255809678335e-05,
+      "loss": 0.19331544637680054,
+      "step": 26680
+    },
+    {
+      "epoch": 0.08365042627155334,
+      "grad_norm": 0.46461957693099976,
+      "learning_rate": 5.2185045631985586e-05,
+      "loss": 0.1477596402168274,
+      "step": 26690
+    },
+    {
+      "epoch": 0.08394270588815556,
+      "grad_norm": 0.47970134019851685,
+      "learning_rate": 5.21143126122409e-05,
+      "loss": 0.17361629009246826,
+      "step": 26700
+    },
+    {
+      "epoch": 0.08423498550475776,
+      "grad_norm": 0.3752261996269226,
+      "learning_rate": 5.204361066499888e-05,
+      "loss": 0.15086857080459595,
+      "step": 26710
+    },
+    {
+      "epoch": 0.08452726512135998,
+      "grad_norm": 0.4635632038116455,
+      "learning_rate": 5.197293983613727e-05,
+      "loss": 0.15668410062789917,
+      "step": 26720
+    },
+    {
+      "epoch": 0.08481954473796219,
+      "grad_norm": 0.3772331476211548,
+      "learning_rate": 5.190230017151355e-05,
+      "loss": 0.17115131616592408,
+      "step": 26730
+    },
+    {
+      "epoch": 0.08511182435456441,
+      "grad_norm": 0.6330028176307678,
+      "learning_rate": 5.183169171696506e-05,
+      "loss": 0.18338911533355712,
+      "step": 26740
+    },
+    {
+      "epoch": 0.08540410397116661,
+      "grad_norm": 0.6345994472503662,
+      "learning_rate": 5.176111451830889e-05,
+      "loss": 0.22149381637573243,
+      "step": 26750
+    },
+    {
+      "epoch": 0.08569638358776883,
+      "grad_norm": 0.5325899720191956,
+      "learning_rate": 5.169056862134176e-05,
+      "loss": 0.14443697929382324,
+      "step": 26760
+    },
+    {
+      "epoch": 0.08598866320437104,
+      "grad_norm": 0.4610198736190796,
+      "learning_rate": 5.162005407184015e-05,
+      "loss": 0.15655972957611083,
+      "step": 26770
+    },
+    {
+      "epoch": 0.08628094282097326,
+      "grad_norm": 0.29812395572662354,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 0.11951093673706055,
+      "step": 26780
+    },
+    {
+      "epoch": 0.08657322243757547,
+      "grad_norm": 0.6232212781906128,
+      "learning_rate": 5.147911919823771e-05,
+      "loss": 0.18175548315048218,
+      "step": 26790
+    },
+    {
+      "epoch": 0.08686550205417769,
+      "grad_norm": 0.5044869184494019,
+      "learning_rate": 5.140869896558791e-05,
+      "loss": 0.14323347806930542,
+      "step": 26800
+    },
+    {
+      "epoch": 0.0871577816707799,
+      "grad_norm": 0.42968565225601196,
+      "learning_rate": 5.133831026330589e-05,
+      "loss": 0.13504555225372314,
+      "step": 26810
+    },
+    {
+      "epoch": 0.0874500612873821,
+      "grad_norm": 0.4756203293800354,
+      "learning_rate": 5.1267953137066013e-05,
+      "loss": 0.14566186666488648,
+      "step": 26820
+    },
+    {
+      "epoch": 0.08774234090398432,
+      "grad_norm": 0.398193359375,
+      "learning_rate": 5.1197627632522205e-05,
+      "loss": 0.16261576414108275,
+      "step": 26830
+    },
+    {
+      "epoch": 0.08803462052058653,
+      "grad_norm": 0.34095948934555054,
+      "learning_rate": 5.112733379530805e-05,
+      "loss": 0.18078892230987548,
+      "step": 26840
+    },
+    {
+      "epoch": 0.08832690013718875,
+      "grad_norm": 0.3014567494392395,
+      "learning_rate": 5.105707167103635e-05,
+      "loss": 0.2191941261291504,
+      "step": 26850
+    },
+    {
+      "epoch": 0.08861917975379095,
+      "grad_norm": 0.5045307278633118,
+      "learning_rate": 5.098684130529949e-05,
+      "loss": 0.17441322803497314,
+      "step": 26860
+    },
+    {
+      "epoch": 0.08891145937039317,
+      "grad_norm": 0.34677854180336,
+      "learning_rate": 5.091664274366922e-05,
+      "loss": 0.156475830078125,
+      "step": 26870
+    },
+    {
+      "epoch": 0.08920373898699538,
+      "grad_norm": 0.7270979881286621,
+      "learning_rate": 5.084647603169653e-05,
+      "loss": 0.19197221994400024,
+      "step": 26880
+    },
+    {
+      "epoch": 0.0894960186035976,
+      "grad_norm": 0.37596720457077026,
+      "learning_rate": 5.0776341214911906e-05,
+      "loss": 0.18734875917434693,
+      "step": 26890
+    },
+    {
+      "epoch": 0.0897882982201998,
+      "grad_norm": 0.549423336982727,
+      "learning_rate": 5.070623833882504e-05,
+      "loss": 0.16589057445526123,
+      "step": 26900
+    },
+    {
+      "epoch": 0.09008057783680203,
+      "grad_norm": 0.36934056878089905,
+      "learning_rate": 5.0636167448924987e-05,
+      "loss": 0.12753794193267823,
+      "step": 26910
+    },
+    {
+      "epoch": 0.09037285745340423,
+      "grad_norm": 0.48045337200164795,
+      "learning_rate": 5.05661285906799e-05,
+      "loss": 0.16103850603103637,
+      "step": 26920
+    },
+    {
+      "epoch": 0.09066513707000645,
+      "grad_norm": 0.5084017515182495,
+      "learning_rate": 5.049612180953729e-05,
+      "loss": 0.14875986576080322,
+      "step": 26930
+    },
+    {
+      "epoch": 0.09095741668660866,
+      "grad_norm": 0.4966805577278137,
+      "learning_rate": 5.04261471509238e-05,
+      "loss": 0.14230921268463134,
+      "step": 26940
+    },
+    {
+      "epoch": 0.09124969630321088,
+      "grad_norm": 0.4646106958389282,
+      "learning_rate": 5.0356204660245176e-05,
+      "loss": 0.15471816062927246,
+      "step": 26950
+    },
+    {
+      "epoch": 0.09154197591981308,
+      "grad_norm": 0.5369713306427002,
+      "learning_rate": 5.028629438288638e-05,
+      "loss": 0.16562976837158203,
+      "step": 26960
+    },
+    {
+      "epoch": 0.0918342555364153,
+      "grad_norm": 0.509137749671936,
+      "learning_rate": 5.021641636421146e-05,
+      "loss": 0.16127376556396483,
+      "step": 26970
+    },
+    {
+      "epoch": 0.09212653515301751,
+      "grad_norm": 0.4069735109806061,
+      "learning_rate": 5.014657064956339e-05,
+      "loss": 0.16296347379684448,
+      "step": 26980
+    },
+    {
+      "epoch": 0.09241881476961973,
+      "grad_norm": 0.3400791585445404,
+      "learning_rate": 5.0076757284264444e-05,
+      "loss": 0.14605835676193238,
+      "step": 26990
+    },
+    {
+      "epoch": 0.09271109438622194,
+      "grad_norm": 0.40890398621559143,
+      "learning_rate": 5.000697631361566e-05,
+      "loss": 0.1404279351234436,
+      "step": 27000
+    },
+    {
+      "epoch": 0.09300337400282416,
+      "grad_norm": 0.548978865146637,
+      "learning_rate": 4.9937227782897165e-05,
+      "loss": 0.15374783277511597,
+      "step": 27010
+    },
+    {
+      "epoch": 0.09329565361942636,
+      "grad_norm": 0.4585045874118805,
+      "learning_rate": 4.986751173736805e-05,
+      "loss": 0.14867924451828002,
+      "step": 27020
+    },
+    {
+      "epoch": 0.09358793323602858,
+      "grad_norm": 0.47758546471595764,
+      "learning_rate": 4.979782822226633e-05,
+      "loss": 0.129884934425354,
+      "step": 27030
+    },
+    {
+      "epoch": 0.09388021285263079,
+      "grad_norm": 0.5782788395881653,
+      "learning_rate": 4.97281772828088e-05,
+      "loss": 0.1697521686553955,
+      "step": 27040
+    },
+    {
+      "epoch": 0.09417249246923301,
+      "grad_norm": 0.4642244577407837,
+      "learning_rate": 4.965855896419123e-05,
+      "loss": 0.1452176332473755,
+      "step": 27050
+    },
+    {
+      "epoch": 0.09446477208583522,
+      "grad_norm": 0.3888194262981415,
+      "learning_rate": 4.958897331158824e-05,
+      "loss": 0.15250990390777588,
+      "step": 27060
+    },
+    {
+      "epoch": 0.09475705170243742,
+      "grad_norm": 0.5959738492965698,
+      "learning_rate": 4.951942037015313e-05,
+      "loss": 0.16284431219100953,
+      "step": 27070
+    },
+    {
+      "epoch": 0.09504933131903964,
+      "grad_norm": 0.6176201105117798,
+      "learning_rate": 4.944990018501807e-05,
+      "loss": 0.21155858039855957,
+      "step": 27080
+    },
+    {
+      "epoch": 0.09534161093564185,
+      "grad_norm": 0.40161213278770447,
+      "learning_rate": 4.938041280129402e-05,
+      "loss": 0.11691936254501342,
+      "step": 27090
+    },
+    {
+      "epoch": 0.09563389055224407,
+      "grad_norm": 0.4430828392505646,
+      "learning_rate": 4.9310958264070485e-05,
+      "loss": 0.1163224458694458,
+      "step": 27100
+    },
+    {
+      "epoch": 0.09592617016884628,
+      "grad_norm": 0.42109787464141846,
+      "learning_rate": 4.9241536618415805e-05,
+      "loss": 0.14318023920059203,
+      "step": 27110
+    },
+    {
+      "epoch": 0.0962184497854485,
+      "grad_norm": 0.7202895879745483,
+      "learning_rate": 4.917214790937694e-05,
+      "loss": 0.16015310287475587,
+      "step": 27120
+    },
+    {
+      "epoch": 0.0965107294020507,
+      "grad_norm": 0.3371022343635559,
+      "learning_rate": 4.910279218197952e-05,
+      "loss": 0.17518346309661864,
+      "step": 27130
+    },
+    {
+      "epoch": 0.09680300901865292,
+      "grad_norm": 0.6276895403862,
+      "learning_rate": 4.903346948122761e-05,
+      "loss": 0.17033191919326782,
+      "step": 27140
+    },
+    {
+      "epoch": 0.09709528863525513,
+      "grad_norm": 0.20702749490737915,
+      "learning_rate": 4.896417985210403e-05,
+      "loss": 0.16485074758529664,
+      "step": 27150
+    },
+    {
+      "epoch": 0.09738756825185735,
+      "grad_norm": 0.43426668643951416,
+      "learning_rate": 4.8894923339570096e-05,
+      "loss": 0.13405003547668456,
+      "step": 27160
+    },
+    {
+      "epoch": 0.09767984786845955,
+      "grad_norm": 0.48570895195007324,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 0.14499945640563966,
+      "step": 27170
+    },
+    {
+      "epoch": 0.09797212748506177,
+      "grad_norm": 0.4123452305793762,
+      "learning_rate": 4.875650984400864e-05,
+      "loss": 0.15157934427261352,
+      "step": 27180
+    },
+    {
+      "epoch": 0.09826440710166398,
+      "grad_norm": 0.46461910009384155,
+      "learning_rate": 4.868735295079622e-05,
+      "loss": 0.16436729431152344,
+      "step": 27190
+    },
+    {
+      "epoch": 0.0985566867182662,
+      "grad_norm": 0.3609086275100708,
+      "learning_rate": 4.861822935380329e-05,
+      "loss": 0.1235634446144104,
+      "step": 27200
+    },
+    {
+      "epoch": 0.09884896633486841,
+      "grad_norm": 0.5201540589332581,
+      "learning_rate": 4.854913909788357e-05,
+      "loss": 0.11965816020965576,
+      "step": 27210
+    },
+    {
+      "epoch": 0.09914124595147063,
+      "grad_norm": 0.617535412311554,
+      "learning_rate": 4.848008222786885e-05,
+      "loss": 0.1726251006126404,
+      "step": 27220
+    },
+    {
+      "epoch": 0.09943352556807283,
+      "grad_norm": 0.5924291610717773,
+      "learning_rate": 4.841105878856943e-05,
+      "loss": 0.15504944324493408,
+      "step": 27230
+    },
+    {
+      "epoch": 0.09972580518467505,
+      "grad_norm": 0.368997186422348,
+      "learning_rate": 4.834206882477391e-05,
+      "loss": 0.13500646352767945,
+      "step": 27240
+    },
+    {
+      "epoch": 0.10001808480127726,
+      "grad_norm": 0.5389465093612671,
+      "learning_rate": 4.827311238124905e-05,
+      "loss": 0.18208673000335693,
+      "step": 27250
+    },
+    {
+      "epoch": 0.10031036441787948,
+      "grad_norm": 0.35124799609184265,
+      "learning_rate": 4.8204189502739985e-05,
+      "loss": 0.13116272687911987,
+      "step": 27260
+    },
+    {
+      "epoch": 0.10060264403448169,
+      "grad_norm": 0.42067965865135193,
+      "learning_rate": 4.813530023397e-05,
+      "loss": 0.14455199241638184,
+      "step": 27270
+    },
+    {
+      "epoch": 0.1008949236510839,
+      "grad_norm": 0.6197178959846497,
+      "learning_rate": 4.806644461964065e-05,
+      "loss": 0.1423550248146057,
+      "step": 27280
+    },
+    {
+      "epoch": 0.10118720326768611,
+      "grad_norm": 0.5224197506904602,
+      "learning_rate": 4.799762270443151e-05,
+      "loss": 0.16640561819076538,
+      "step": 27290
+    },
+    {
+      "epoch": 0.10147948288428833,
+      "grad_norm": 0.3606843650341034,
+      "learning_rate": 4.792883453300042e-05,
+      "loss": 0.12231427431106567,
+      "step": 27300
+    },
+    {
+      "epoch": 0.10177176250089054,
+      "grad_norm": 0.5512232780456543,
+      "learning_rate": 4.78600801499833e-05,
+      "loss": 0.12757351398468017,
+      "step": 27310
+    },
+    {
+      "epoch": 0.10206404211749276,
+      "grad_norm": 0.4447842240333557,
+      "learning_rate": 4.7791359599994066e-05,
+      "loss": 0.17506953477859497,
+      "step": 27320
+    },
+    {
+      "epoch": 0.10235632173409497,
+      "grad_norm": 0.435726523399353,
+      "learning_rate": 4.7722672927624764e-05,
+      "loss": 0.1436169385910034,
+      "step": 27330
+    },
+    {
+      "epoch": 0.10264860135069717,
+      "grad_norm": 0.42706337571144104,
+      "learning_rate": 4.765402017744544e-05,
+      "loss": 0.172484827041626,
+      "step": 27340
+    },
+    {
+      "epoch": 0.10294088096729939,
+      "grad_norm": 0.517407238483429,
+      "learning_rate": 4.758540139400415e-05,
+      "loss": 0.13795779943466185,
+      "step": 27350
+    },
+    {
+      "epoch": 0.1032331605839016,
+      "grad_norm": 0.4375555217266083,
+      "learning_rate": 4.7516816621826784e-05,
+      "loss": 0.15619373321533203,
+      "step": 27360
+    },
+    {
+      "epoch": 0.10352544020050382,
+      "grad_norm": 0.4089914858341217,
+      "learning_rate": 4.7448265905417324e-05,
+      "loss": 0.1537814736366272,
+      "step": 27370
+    },
+    {
+      "epoch": 0.10381771981710602,
+      "grad_norm": 0.4519176781177521,
+      "learning_rate": 4.73797492892576e-05,
+      "loss": 0.15159679651260377,
+      "step": 27380
+    },
+    {
+      "epoch": 0.10410999943370824,
+      "grad_norm": 0.43496692180633545,
+      "learning_rate": 4.731126681780721e-05,
+      "loss": 0.12926777601242065,
+      "step": 27390
+    },
+    {
+      "epoch": 0.10440227905031045,
+      "grad_norm": 0.44301870465278625,
+      "learning_rate": 4.724281853550382e-05,
+      "loss": 0.12694705724716188,
+      "step": 27400
+    },
+    {
+      "epoch": 0.10469455866691267,
+      "grad_norm": 0.4813741147518158,
+      "learning_rate": 4.7174404486762694e-05,
+      "loss": 0.14997857809066772,
+      "step": 27410
+    },
+    {
+      "epoch": 0.10498683828351488,
+      "grad_norm": 0.523106038570404,
+      "learning_rate": 4.710602471597692e-05,
+      "loss": 0.1362652897834778,
+      "step": 27420
+    },
+    {
+      "epoch": 0.1052791179001171,
+      "grad_norm": 0.3514653742313385,
+      "learning_rate": 4.703767926751751e-05,
+      "loss": 0.1448134422302246,
+      "step": 27430
+    },
+    {
+      "epoch": 0.1055713975167193,
+      "grad_norm": 0.5577013492584229,
+      "learning_rate": 4.696936818573298e-05,
+      "loss": 0.18804486989974975,
+      "step": 27440
+    },
+    {
+      "epoch": 0.10586367713332152,
+      "grad_norm": 0.38858306407928467,
+      "learning_rate": 4.690109151494969e-05,
+      "loss": 0.15231740474700928,
+      "step": 27450
+    },
+    {
+      "epoch": 0.10615595674992373,
+      "grad_norm": 0.6015412211418152,
+      "learning_rate": 4.6832849299471674e-05,
+      "loss": 0.09168019294738769,
+      "step": 27460
+    },
+    {
+      "epoch": 0.10644823636652595,
+      "grad_norm": 0.33207976818084717,
+      "learning_rate": 4.676464158358047e-05,
+      "loss": 0.1605085015296936,
+      "step": 27470
+    },
+    {
+      "epoch": 0.10674051598312816,
+      "grad_norm": 0.5263932347297668,
+      "learning_rate": 4.669646841153539e-05,
+      "loss": 0.15719871520996093,
+      "step": 27480
+    },
+    {
+      "epoch": 0.10703279559973038,
+      "grad_norm": 0.3147108554840088,
+      "learning_rate": 4.6628329827573234e-05,
+      "loss": 0.14965027570724487,
+      "step": 27490
+    },
+    {
+      "epoch": 0.10732507521633258,
+      "grad_norm": 0.4358750581741333,
+      "learning_rate": 4.656022587590846e-05,
+      "loss": 0.15058099031448363,
+      "step": 27500
+    },
+    {
+      "epoch": 0.1076173548329348,
+      "grad_norm": 0.38839757442474365,
+      "learning_rate": 4.649215660073287e-05,
+      "loss": 0.1313214421272278,
+      "step": 27510
+    },
+    {
+      "epoch": 0.10790963444953701,
+      "grad_norm": 0.4989486336708069,
+      "learning_rate": 4.6424122046215956e-05,
+      "loss": 0.1751583695411682,
+      "step": 27520
+    },
+    {
+      "epoch": 0.10820191406613923,
+      "grad_norm": 0.5157408118247986,
+      "learning_rate": 4.635612225650461e-05,
+      "loss": 0.20897183418273926,
+      "step": 27530
+    },
+    {
+      "epoch": 0.10849419368274144,
+      "grad_norm": 0.5855669975280762,
+      "learning_rate": 4.6288157275723086e-05,
+      "loss": 0.15495749711990356,
+      "step": 27540
+    },
+    {
+      "epoch": 0.10878647329934366,
+      "grad_norm": 0.401619017124176,
+      "learning_rate": 4.6220227147973175e-05,
+      "loss": 0.12434269189834594,
+      "step": 27550
+    },
+    {
+      "epoch": 0.10907875291594586,
+      "grad_norm": 0.5897713303565979,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 0.1180533528327942,
+      "step": 27560
+    },
+    {
+      "epoch": 0.10937103253254808,
+      "grad_norm": 0.5452660322189331,
+      "learning_rate": 4.608447162786202e-05,
+      "loss": 0.19550626277923583,
+      "step": 27570
+    },
+    {
+      "epoch": 0.10966331214915029,
+      "grad_norm": 0.4595751464366913,
+      "learning_rate": 4.6016646323591036e-05,
+      "loss": 0.16359970569610596,
+      "step": 27580
+    },
+    {
+      "epoch": 0.1099555917657525,
+      "grad_norm": 0.4769741892814636,
+      "learning_rate": 4.594885604853217e-05,
+      "loss": 0.14821945428848265,
+      "step": 27590
+    },
+    {
+      "epoch": 0.11024787138235471,
+      "grad_norm": 0.5596093535423279,
+      "learning_rate": 4.588110084667383e-05,
+      "loss": 0.18486030101776124,
+      "step": 27600
+    },
+    {
+      "epoch": 0.11054015099895692,
+      "grad_norm": 0.44595563411712646,
+      "learning_rate": 4.5813380761981504e-05,
+      "loss": 0.1940765142440796,
+      "step": 27610
+    },
+    {
+      "epoch": 0.11083243061555914,
+      "grad_norm": 0.3472349941730499,
+      "learning_rate": 4.5745695838398196e-05,
+      "loss": 0.14513286352157592,
+      "step": 27620
+    },
+    {
+      "epoch": 0.11112471023216135,
+      "grad_norm": 0.48911619186401367,
+      "learning_rate": 4.567804611984381e-05,
+      "loss": 0.17585965394973754,
+      "step": 27630
+    },
+    {
+      "epoch": 0.11141698984876357,
+      "grad_norm": 0.2544448971748352,
+      "learning_rate": 4.561043165021547e-05,
+      "loss": 0.1395537495613098,
+      "step": 27640
+    },
+    {
+      "epoch": 0.11170926946536577,
+      "grad_norm": 0.33828312158584595,
+      "learning_rate": 4.55428524733876e-05,
+      "loss": 0.1431349039077759,
+      "step": 27650
+    },
+    {
+      "epoch": 0.112001549081968,
+      "grad_norm": 0.3547646105289459,
+      "learning_rate": 4.547530863321149e-05,
+      "loss": 0.14767364263534546,
+      "step": 27660
+    },
+    {
+      "epoch": 0.1122938286985702,
+      "grad_norm": 0.5337129235267639,
+      "learning_rate": 4.5407800173515644e-05,
+      "loss": 0.13381657600402833,
+      "step": 27670
+    },
+    {
+      "epoch": 0.11258610831517242,
+      "grad_norm": 0.4012839198112488,
+      "learning_rate": 4.534032713810561e-05,
+      "loss": 0.19537571668624878,
+      "step": 27680
+    },
+    {
+      "epoch": 0.11287838793177463,
+      "grad_norm": 0.48562943935394287,
+      "learning_rate": 4.527288957076382e-05,
+      "loss": 0.1718903064727783,
+      "step": 27690
+    },
+    {
+      "epoch": 0.11317066754837685,
+      "grad_norm": 0.3646818697452545,
+      "learning_rate": 4.520548751524982e-05,
+      "loss": 0.18174225091934204,
+      "step": 27700
+    },
+    {
+      "epoch": 0.11346294716497905,
+      "grad_norm": 0.6121250987052917,
+      "learning_rate": 4.513812101530008e-05,
+      "loss": 0.1752510190010071,
+      "step": 27710
+    },
+    {
+      "epoch": 0.11375522678158127,
+      "grad_norm": 0.7941213846206665,
+      "learning_rate": 4.507079011462803e-05,
+      "loss": 0.1889721155166626,
+      "step": 27720
+    },
+    {
+      "epoch": 0.11404750639818348,
+      "grad_norm": 0.721476674079895,
+      "learning_rate": 4.500349485692387e-05,
+      "loss": 0.2076568841934204,
+      "step": 27730
+    },
+    {
+      "epoch": 0.1143397860147857,
+      "grad_norm": 0.4743868112564087,
+      "learning_rate": 4.493623528585481e-05,
+      "loss": 0.17123620510101317,
+      "step": 27740
+    },
+    {
+      "epoch": 0.1146320656313879,
+      "grad_norm": 0.5294836163520813,
+      "learning_rate": 4.48690114450649e-05,
+      "loss": 0.27049806118011477,
+      "step": 27750
+    },
+    {
+      "epoch": 0.11492434524799013,
+      "grad_norm": 0.5381839275360107,
+      "learning_rate": 4.4801823378174866e-05,
+      "loss": 0.1601255178451538,
+      "step": 27760
+    },
+    {
+      "epoch": 0.11521662486459233,
+      "grad_norm": 0.3507229685783386,
+      "learning_rate": 4.473467112878238e-05,
+      "loss": 0.17364072799682617,
+      "step": 27770
+    },
+    {
+      "epoch": 0.11550890448119455,
+      "grad_norm": 0.46900781989097595,
+      "learning_rate": 4.466755474046177e-05,
+      "loss": 0.21106083393096925,
+      "step": 27780
+    },
+    {
+      "epoch": 0.11580118409779676,
+      "grad_norm": 0.7412432432174683,
+      "learning_rate": 4.4600474256764204e-05,
+      "loss": 0.17794123888015748,
+      "step": 27790
+    },
+    {
+      "epoch": 0.11609346371439898,
+      "grad_norm": 0.40566137433052063,
+      "learning_rate": 4.453342972121739e-05,
+      "loss": 0.20559492111206054,
+      "step": 27800
+    },
+    {
+      "epoch": 0.11638574333100118,
+      "grad_norm": 0.5269231796264648,
+      "learning_rate": 4.4466421177325835e-05,
+      "loss": 0.20143918991088866,
+      "step": 27810
+    },
+    {
+      "epoch": 0.1166780229476034,
+      "grad_norm": 0.45729300379753113,
+      "learning_rate": 4.43994486685707e-05,
+      "loss": 0.1545391321182251,
+      "step": 27820
+    },
+    {
+      "epoch": 0.11697030256420561,
+      "grad_norm": 0.44985467195510864,
+      "learning_rate": 4.433251223840961e-05,
+      "loss": 0.2043304204940796,
+      "step": 27830
+    },
+    {
+      "epoch": 0.11726258218080783,
+      "grad_norm": 0.44185739755630493,
+      "learning_rate": 4.4265611930277016e-05,
+      "loss": 0.18514822721481322,
+      "step": 27840
+    },
+    {
+      "epoch": 0.11755486179741004,
+      "grad_norm": 0.3882181644439697,
+      "learning_rate": 4.419874778758374e-05,
+      "loss": 0.18056503534317017,
+      "step": 27850
+    },
+    {
+      "epoch": 0.11784714141401224,
+      "grad_norm": 0.49920275807380676,
+      "learning_rate": 4.413191985371713e-05,
+      "loss": 0.17281390428543092,
+      "step": 27860
+    },
+    {
+      "epoch": 0.11813942103061446,
+      "grad_norm": 0.36730948090553284,
+      "learning_rate": 4.406512817204125e-05,
+      "loss": 0.18176440000534058,
+      "step": 27870
+    },
+    {
+      "epoch": 0.11843170064721667,
+      "grad_norm": 0.47389402985572815,
+      "learning_rate": 4.399837278589639e-05,
+      "loss": 0.15645064115524293,
+      "step": 27880
+    },
+    {
+      "epoch": 0.11872398026381889,
+      "grad_norm": 0.46471214294433594,
+      "learning_rate": 4.393165373859942e-05,
+      "loss": 0.21132197380065917,
+      "step": 27890
+    },
+    {
+      "epoch": 0.1190162598804211,
+      "grad_norm": 0.4321572482585907,
+      "learning_rate": 4.386497107344366e-05,
+      "loss": 0.1972594141960144,
+      "step": 27900
+    },
+    {
+      "epoch": 0.11930853949702332,
+      "grad_norm": 0.43716272711753845,
+      "learning_rate": 4.379832483369868e-05,
+      "loss": 0.18363574743270875,
+      "step": 27910
+    },
+    {
+      "epoch": 0.11960081911362552,
+      "grad_norm": 0.4089851379394531,
+      "learning_rate": 4.373171506261054e-05,
+      "loss": 0.2087033987045288,
+      "step": 27920
+    },
+    {
+      "epoch": 0.11989309873022774,
+      "grad_norm": 0.41704821586608887,
+      "learning_rate": 4.36651418034016e-05,
+      "loss": 0.14927574396133422,
+      "step": 27930
+    },
+    {
+      "epoch": 0.12018537834682995,
+      "grad_norm": 0.3968091607093811,
+      "learning_rate": 4.359860509927056e-05,
+      "loss": 0.1861981511116028,
+      "step": 27940
+    },
+    {
+      "epoch": 0.12047765796343217,
+      "grad_norm": 0.33855536580085754,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 0.16548330783843995,
+      "step": 27950
+    },
+    {
+      "epoch": 0.12076993758003438,
+      "grad_norm": 0.4153297245502472,
+      "learning_rate": 4.3465641528918047e-05,
+      "loss": 0.19815868139266968,
+      "step": 27960
+    },
+    {
+      "epoch": 0.1210622171966366,
+      "grad_norm": 0.4304749369621277,
+      "learning_rate": 4.339921474897526e-05,
+      "loss": 0.18542084693908692,
+      "step": 27970
+    },
+    {
+      "epoch": 0.1213544968132388,
+      "grad_norm": 0.41935065388679504,
+      "learning_rate": 4.3332824696667474e-05,
+      "loss": 0.16377384662628175,
+      "step": 27980
+    },
+    {
+      "epoch": 0.12164677642984102,
+      "grad_norm": 0.31701722741127014,
+      "learning_rate": 4.3266471415074514e-05,
+      "loss": 0.1436556339263916,
+      "step": 27990
+    },
+    {
+      "epoch": 0.12193905604644323,
+      "grad_norm": 0.5640013217926025,
+      "learning_rate": 4.32001549472523e-05,
+      "loss": 0.2129084825515747,
+      "step": 28000
+    },
+    {
+      "epoch": 0.12223133566304545,
+      "grad_norm": 0.4181281626224518,
+      "learning_rate": 4.313387533623292e-05,
+      "loss": 0.20640838146209717,
+      "step": 28010
+    },
+    {
+      "epoch": 0.12252361527964765,
+      "grad_norm": 0.41855931282043457,
+      "learning_rate": 4.3067632625024414e-05,
+      "loss": 0.17629060745239258,
+      "step": 28020
+    },
+    {
+      "epoch": 0.12281589489624987,
+      "grad_norm": 0.5060747265815735,
+      "learning_rate": 4.300142685661098e-05,
+      "loss": 0.14987777471542357,
+      "step": 28030
+    },
+    {
+      "epoch": 0.12310817451285208,
+      "grad_norm": 0.4886680841445923,
+      "learning_rate": 4.2935258073952875e-05,
+      "loss": 0.16464685201644896,
+      "step": 28040
+    },
+    {
+      "epoch": 0.1234004541294543,
+      "grad_norm": 0.40735363960266113,
+      "learning_rate": 4.2869126319986195e-05,
+      "loss": 0.20623316764831542,
+      "step": 28050
+    },
+    {
+      "epoch": 0.12369273374605651,
+      "grad_norm": 0.44593724608421326,
+      "learning_rate": 4.2803031637623256e-05,
+      "loss": 0.2364881753921509,
+      "step": 28060
+    },
+    {
+      "epoch": 0.12398501336265873,
+      "grad_norm": 0.35898450016975403,
+      "learning_rate": 4.273697406975211e-05,
+      "loss": 0.22410366535186768,
+      "step": 28070
+    },
+    {
+      "epoch": 0.12427729297926093,
+      "grad_norm": 0.5747163891792297,
+      "learning_rate": 4.267095365923672e-05,
+      "loss": 0.280694317817688,
+      "step": 28080
+    },
+    {
+      "epoch": 0.12456957259586315,
+      "grad_norm": 0.4064870774745941,
+      "learning_rate": 4.2604970448917156e-05,
+      "loss": 0.17308086156845093,
+      "step": 28090
+    },
+    {
+      "epoch": 0.12486185221246536,
+      "grad_norm": 0.5090318918228149,
+      "learning_rate": 4.2539024481609116e-05,
+      "loss": 0.17569255828857422,
+      "step": 28100
+    },
+    {
+      "epoch": 0.12515413182906757,
+      "grad_norm": 0.3582715094089508,
+      "learning_rate": 4.247311580010424e-05,
+      "loss": 0.19092520475387573,
+      "step": 28110
+    },
+    {
+      "epoch": 0.1254464114456698,
+      "grad_norm": 0.3470022678375244,
+      "learning_rate": 4.240724444716998e-05,
+      "loss": 0.21732716560363768,
+      "step": 28120
+    },
+    {
+      "epoch": 0.125738691062272,
+      "grad_norm": 0.4961210787296295,
+      "learning_rate": 4.23414104655495e-05,
+      "loss": 0.25637814998626707,
+      "step": 28130
+    },
+    {
+      "epoch": 0.1260309706788742,
+      "grad_norm": 0.503395140171051,
+      "learning_rate": 4.227561389796179e-05,
+      "loss": 0.2018885612487793,
+      "step": 28140
+    },
+    {
+      "epoch": 0.12632325029547642,
+      "grad_norm": 0.4107346832752228,
+      "learning_rate": 4.22098547871015e-05,
+      "loss": 0.18601430654525758,
+      "step": 28150
+    },
+    {
+      "epoch": 0.12661552991207864,
+      "grad_norm": 0.7840616703033447,
+      "learning_rate": 4.214413317563906e-05,
+      "loss": 0.22185537815093995,
+      "step": 28160
+    },
+    {
+      "epoch": 0.12690780952868086,
+      "grad_norm": 0.5339564085006714,
+      "learning_rate": 4.207844910622043e-05,
+      "loss": 0.2097491979598999,
+      "step": 28170
+    },
+    {
+      "epoch": 0.12720008914528305,
+      "grad_norm": 0.44421061873435974,
+      "learning_rate": 4.201280262146734e-05,
+      "loss": 0.14907686710357665,
+      "step": 28180
+    },
+    {
+      "epoch": 0.12749236876188527,
+      "grad_norm": 0.44252943992614746,
+      "learning_rate": 4.19471937639771e-05,
+      "loss": 0.15797396898269653,
+      "step": 28190
+    },
+    {
+      "epoch": 0.1277846483784875,
+      "grad_norm": 0.5497806668281555,
+      "learning_rate": 4.188162257632252e-05,
+      "loss": 0.23098313808441162,
+      "step": 28200
+    },
+    {
+      "epoch": 0.1280769279950897,
+      "grad_norm": 0.4373284578323364,
+      "learning_rate": 4.181608910105207e-05,
+      "loss": 0.1583493709564209,
+      "step": 28210
+    },
+    {
+      "epoch": 0.1283692076116919,
+      "grad_norm": 0.43907853960990906,
+      "learning_rate": 4.175059338068974e-05,
+      "loss": 0.17469092607498168,
+      "step": 28220
+    },
+    {
+      "epoch": 0.12866148722829412,
+      "grad_norm": 0.686427891254425,
+      "learning_rate": 4.1685135457734905e-05,
+      "loss": 0.14792318344116212,
+      "step": 28230
+    },
+    {
+      "epoch": 0.12895376684489634,
+      "grad_norm": 0.3631090223789215,
+      "learning_rate": 4.161971537466261e-05,
+      "loss": 0.17702265977859497,
+      "step": 28240
+    },
+    {
+      "epoch": 0.12924604646149856,
+      "grad_norm": 0.5830815434455872,
+      "learning_rate": 4.155433317392315e-05,
+      "loss": 0.17246402502059938,
+      "step": 28250
+    },
+    {
+      "epoch": 0.12953832607810076,
+      "grad_norm": 0.524885356426239,
+      "learning_rate": 4.148898889794236e-05,
+      "loss": 0.22747671604156494,
+      "step": 28260
+    },
+    {
+      "epoch": 0.12983060569470298,
+      "grad_norm": 0.5487504005432129,
+      "learning_rate": 4.1423682589121435e-05,
+      "loss": 0.2050922393798828,
+      "step": 28270
+    },
+    {
+      "epoch": 0.1301228853113052,
+      "grad_norm": 0.3661110997200012,
+      "learning_rate": 4.135841428983695e-05,
+      "loss": 0.16184931993484497,
+      "step": 28280
+    },
+    {
+      "epoch": 0.13041516492790742,
+      "grad_norm": 0.48653921484947205,
+      "learning_rate": 4.1293184042440744e-05,
+      "loss": 0.18359391689300536,
+      "step": 28290
+    },
+    {
+      "epoch": 0.1307074445445096,
+      "grad_norm": 0.42407381534576416,
+      "learning_rate": 4.1227991889260034e-05,
+      "loss": 0.1631341338157654,
+      "step": 28300
+    },
+    {
+      "epoch": 0.13099972416111183,
+      "grad_norm": 0.3518393337726593,
+      "learning_rate": 4.116283787259735e-05,
+      "loss": 0.20204582214355468,
+      "step": 28310
+    },
+    {
+      "epoch": 0.13129200377771405,
+      "grad_norm": 0.602037787437439,
+      "learning_rate": 4.109772203473035e-05,
+      "loss": 0.20396575927734376,
+      "step": 28320
+    },
+    {
+      "epoch": 0.13158428339431627,
+      "grad_norm": 0.3457263708114624,
+      "learning_rate": 4.103264441791203e-05,
+      "loss": 0.18486883640289306,
+      "step": 28330
+    },
+    {
+      "epoch": 0.13187656301091846,
+      "grad_norm": 0.3400459289550781,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 0.1795599102973938,
+      "step": 28340
+    },
+    {
+      "epoch": 0.13216884262752068,
+      "grad_norm": 0.5426247715950012,
+      "learning_rate": 4.0902604016309255e-05,
+      "loss": 0.19458307027816774,
+      "step": 28350
+    },
+    {
+      "epoch": 0.1324611222441229,
+      "grad_norm": 0.49179065227508545,
+      "learning_rate": 4.083764131590657e-05,
+      "loss": 0.2034163236618042,
+      "step": 28360
+    },
+    {
+      "epoch": 0.1327534018607251,
+      "grad_norm": 0.33926406502723694,
+      "learning_rate": 4.077271700531613e-05,
+      "loss": 0.23384332656860352,
+      "step": 28370
+    },
+    {
+      "epoch": 0.13304568147732732,
+      "grad_norm": 0.38904204964637756,
+      "learning_rate": 4.0707831126666643e-05,
+      "loss": 0.17011070251464844,
+      "step": 28380
+    },
+    {
+      "epoch": 0.13333796109392954,
+      "grad_norm": 0.5034767389297485,
+      "learning_rate": 4.0642983722061764e-05,
+      "loss": 0.1835199475288391,
+      "step": 28390
+    },
+    {
+      "epoch": 0.13363024071053176,
+      "grad_norm": 0.5227706432342529,
+      "learning_rate": 4.057817483358033e-05,
+      "loss": 0.18817261457443238,
+      "step": 28400
+    },
+    {
+      "epoch": 0.13392252032713395,
+      "grad_norm": 0.4366223216056824,
+      "learning_rate": 4.051340450327614e-05,
+      "loss": 0.20845625400543213,
+      "step": 28410
+    },
+    {
+      "epoch": 0.13421479994373617,
+      "grad_norm": 0.45262467861175537,
+      "learning_rate": 4.044867277317786e-05,
+      "loss": 0.19027016162872315,
+      "step": 28420
+    },
+    {
+      "epoch": 0.1345070795603384,
+      "grad_norm": 0.5794299244880676,
+      "learning_rate": 4.038397968528935e-05,
+      "loss": 0.1755579710006714,
+      "step": 28430
+    },
+    {
+      "epoch": 0.1347993591769406,
+      "grad_norm": 0.49212735891342163,
+      "learning_rate": 4.0319325281589185e-05,
+      "loss": 0.15251225233078003,
+      "step": 28440
+    },
+    {
+      "epoch": 0.1350916387935428,
+      "grad_norm": 0.4970851242542267,
+      "learning_rate": 4.025470960403083e-05,
+      "loss": 0.19374841451644897,
+      "step": 28450
+    },
+    {
+      "epoch": 0.13538391841014502,
+      "grad_norm": 0.6789295673370361,
+      "learning_rate": 4.019013269454284e-05,
+      "loss": 0.21508312225341797,
+      "step": 28460
+    },
+    {
+      "epoch": 0.13567619802674724,
+      "grad_norm": 0.36547836661338806,
+      "learning_rate": 4.012559459502835e-05,
+      "loss": 0.20945651531219484,
+      "step": 28470
+    },
+    {
+      "epoch": 0.13596847764334946,
+      "grad_norm": 0.2961515486240387,
+      "learning_rate": 4.006109534736548e-05,
+      "loss": 0.1742892622947693,
+      "step": 28480
+    },
+    {
+      "epoch": 0.13626075725995165,
+      "grad_norm": 0.5706232190132141,
+      "learning_rate": 3.9996634993407114e-05,
+      "loss": 0.18005183935165406,
+      "step": 28490
+    },
+    {
+      "epoch": 0.13655303687655387,
+      "grad_norm": 0.5240952968597412,
+      "learning_rate": 3.993221357498087e-05,
+      "loss": 0.154144549369812,
+      "step": 28500
+    },
+    {
+      "epoch": 0.1368453164931561,
+      "grad_norm": 0.498589426279068,
+      "learning_rate": 3.9867831133889064e-05,
+      "loss": 0.1730855345726013,
+      "step": 28510
+    },
+    {
+      "epoch": 0.13713759610975831,
+      "grad_norm": 0.5033027529716492,
+      "learning_rate": 3.980348771190878e-05,
+      "loss": 0.22879428863525392,
+      "step": 28520
+    },
+    {
+      "epoch": 0.1374298757263605,
+      "grad_norm": 0.7028383612632751,
+      "learning_rate": 3.973918335079182e-05,
+      "loss": 0.18266732692718507,
+      "step": 28530
+    },
+    {
+      "epoch": 0.13772215534296273,
+      "grad_norm": 0.5395593047142029,
+      "learning_rate": 3.96749180922645e-05,
+      "loss": 0.17976876497268676,
+      "step": 28540
+    },
+    {
+      "epoch": 0.13801443495956495,
+      "grad_norm": 0.49240413308143616,
+      "learning_rate": 3.961069197802789e-05,
+      "loss": 0.1833338975906372,
+      "step": 28550
+    },
+    {
+      "epoch": 0.13830671457616717,
+      "grad_norm": 0.4961566627025604,
+      "learning_rate": 3.954650504975767e-05,
+      "loss": 0.2574249505996704,
+      "step": 28560
+    },
+    {
+      "epoch": 0.13859899419276936,
+      "grad_norm": 0.48584648966789246,
+      "learning_rate": 3.948235734910394e-05,
+      "loss": 0.2239234447479248,
+      "step": 28570
+    },
+    {
+      "epoch": 0.13889127380937158,
+      "grad_norm": 0.4367305040359497,
+      "learning_rate": 3.941824891769149e-05,
+      "loss": 0.14437726736068726,
+      "step": 28580
+    },
+    {
+      "epoch": 0.1391835534259738,
+      "grad_norm": 0.7312074303627014,
+      "learning_rate": 3.935417979711959e-05,
+      "loss": 0.2049402952194214,
+      "step": 28590
+    },
+    {
+      "epoch": 0.13947583304257602,
+      "grad_norm": 0.5530930161476135,
+      "learning_rate": 3.9290150028962044e-05,
+      "loss": 0.19575185775756837,
+      "step": 28600
+    },
+    {
+      "epoch": 0.1397681126591782,
+      "grad_norm": 0.47369861602783203,
+      "learning_rate": 3.9226159654766994e-05,
+      "loss": 0.20326611995697022,
+      "step": 28610
+    },
+    {
+      "epoch": 0.14006039227578043,
+      "grad_norm": 0.6182199120521545,
+      "learning_rate": 3.9162208716057114e-05,
+      "loss": 0.2695906639099121,
+      "step": 28620
+    },
+    {
+      "epoch": 0.14035267189238265,
+      "grad_norm": 0.46619608998298645,
+      "learning_rate": 3.909829725432954e-05,
+      "loss": 0.19612321853637696,
+      "step": 28630
+    },
+    {
+      "epoch": 0.14064495150898484,
+      "grad_norm": 0.3851034939289093,
+      "learning_rate": 3.903442531105561e-05,
+      "loss": 0.19507240056991576,
+      "step": 28640
+    },
+    {
+      "epoch": 0.14093723112558706,
+      "grad_norm": 0.43064650893211365,
+      "learning_rate": 3.897059292768126e-05,
+      "loss": 0.20645384788513182,
+      "step": 28650
+    },
+    {
+      "epoch": 0.14122951074218928,
+      "grad_norm": 0.5684599876403809,
+      "learning_rate": 3.890680014562658e-05,
+      "loss": 0.19083259105682374,
+      "step": 28660
+    },
+    {
+      "epoch": 0.1415217903587915,
+      "grad_norm": 0.433711439371109,
+      "learning_rate": 3.8843047006285936e-05,
+      "loss": 0.21319074630737306,
+      "step": 28670
+    },
+    {
+      "epoch": 0.1418140699753937,
+      "grad_norm": 0.702174186706543,
+      "learning_rate": 3.8779333551028183e-05,
+      "loss": 0.20057661533355714,
+      "step": 28680
+    },
+    {
+      "epoch": 0.14210634959199592,
+      "grad_norm": 0.70973801612854,
+      "learning_rate": 3.871565982119621e-05,
+      "loss": 0.20836129188537597,
+      "step": 28690
+    },
+    {
+      "epoch": 0.14239862920859814,
+      "grad_norm": 0.47371482849121094,
+      "learning_rate": 3.86520258581072e-05,
+      "loss": 0.163743793964386,
+      "step": 28700
+    },
+    {
+      "epoch": 0.14269090882520036,
+      "grad_norm": 0.4650250971317291,
+      "learning_rate": 3.858843170305261e-05,
+      "loss": 0.26329262256622316,
+      "step": 28710
+    },
+    {
+      "epoch": 0.14298318844180255,
+      "grad_norm": 0.3933168947696686,
+      "learning_rate": 3.852487739729792e-05,
+      "loss": 0.19892170429229736,
+      "step": 28720
+    },
+    {
+      "epoch": 0.14327546805840477,
+      "grad_norm": 0.4648030400276184,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 0.1944313168525696,
+      "step": 28730
+    },
+    {
+      "epoch": 0.143567747675007,
+      "grad_norm": 0.30299732089042664,
+      "learning_rate": 3.839788849862124e-05,
+      "loss": 0.24366040229797364,
+      "step": 28740
+    },
+    {
+      "epoch": 0.1438600272916092,
+      "grad_norm": 0.4619271457195282,
+      "learning_rate": 3.833445398810101e-05,
+      "loss": 0.2309725046157837,
+      "step": 28750
+    },
+    {
+      "epoch": 0.1441523069082114,
+      "grad_norm": 0.4481695592403412,
+      "learning_rate": 3.827105949168405e-05,
+      "loss": 0.19631146192550658,
+      "step": 28760
+    },
+    {
+      "epoch": 0.14444458652481362,
+      "grad_norm": 0.6589859127998352,
+      "learning_rate": 3.820770505050642e-05,
+      "loss": 0.23353807926177977,
+      "step": 28770
+    },
+    {
+      "epoch": 0.14473686614141584,
+      "grad_norm": 0.46461009979248047,
+      "learning_rate": 3.8144390705678123e-05,
+      "loss": 0.20361881256103515,
+      "step": 28780
+    },
+    {
+      "epoch": 0.14502914575801806,
+      "grad_norm": 0.4389754831790924,
+      "learning_rate": 3.808111649828313e-05,
+      "loss": 0.20992677211761473,
+      "step": 28790
+    },
+    {
+      "epoch": 0.14532142537462026,
+      "grad_norm": 0.5616406202316284,
+      "learning_rate": 3.801788246937938e-05,
+      "loss": 0.24943602085113525,
+      "step": 28800
+    },
+    {
+      "epoch": 0.14561370499122248,
+      "grad_norm": 0.4695107638835907,
+      "learning_rate": 3.795468865999878e-05,
+      "loss": 0.19450711011886596,
+      "step": 28810
+    },
+    {
+      "epoch": 0.1459059846078247,
+      "grad_norm": 0.40403035283088684,
+      "learning_rate": 3.7891535111147125e-05,
+      "loss": 0.18192031383514404,
+      "step": 28820
+    },
+    {
+      "epoch": 0.14619826422442692,
+      "grad_norm": 0.3371562063694,
+      "learning_rate": 3.7828421863803985e-05,
+      "loss": 0.20170538425445556,
+      "step": 28830
+    },
+    {
+      "epoch": 0.1464905438410291,
+      "grad_norm": 0.37995776534080505,
+      "learning_rate": 3.776534895892293e-05,
+      "loss": 0.1938255786895752,
+      "step": 28840
+    },
+    {
+      "epoch": 0.14678282345763133,
+      "grad_norm": 0.5093180537223816,
+      "learning_rate": 3.7702316437431284e-05,
+      "loss": 0.21081192493438722,
+      "step": 28850
+    },
+    {
+      "epoch": 0.14707510307423355,
+      "grad_norm": 0.49890217185020447,
+      "learning_rate": 3.7639324340230085e-05,
+      "loss": 0.23567750453948974,
+      "step": 28860
+    },
+    {
+      "epoch": 0.14736738269083577,
+      "grad_norm": 0.5042874813079834,
+      "learning_rate": 3.757637270819435e-05,
+      "loss": 0.2177427053451538,
+      "step": 28870
+    },
+    {
+      "epoch": 0.14765966230743796,
+      "grad_norm": 0.5157811641693115,
+      "learning_rate": 3.7513461582172635e-05,
+      "loss": 0.23290653228759767,
+      "step": 28880
+    },
+    {
+      "epoch": 0.14795194192404018,
+      "grad_norm": 0.6270256042480469,
+      "learning_rate": 3.745059100298722e-05,
+      "loss": 0.22153949737548828,
+      "step": 28890
+    },
+    {
+      "epoch": 0.1482442215406424,
+      "grad_norm": 0.49984216690063477,
+      "learning_rate": 3.7387761011434296e-05,
+      "loss": 0.19958875179290772,
+      "step": 28900
+    },
+    {
+      "epoch": 0.1485365011572446,
+      "grad_norm": 0.4756910502910614,
+      "learning_rate": 3.732497164828345e-05,
+      "loss": 0.1413467764854431,
+      "step": 28910
+    },
+    {
+      "epoch": 0.1488287807738468,
+      "grad_norm": 0.5184823870658875,
+      "learning_rate": 3.7262222954278036e-05,
+      "loss": 0.1737914800643921,
+      "step": 28920
+    },
+    {
+      "epoch": 0.14912106039044903,
+      "grad_norm": 0.4724036157131195,
+      "learning_rate": 3.7199514970135044e-05,
+      "loss": 0.18252692222595215,
+      "step": 28930
+    },
+    {
+      "epoch": 0.14941334000705125,
+      "grad_norm": 0.4956326484680176,
+      "learning_rate": 3.713684773654494e-05,
+      "loss": 0.2000032901763916,
+      "step": 28940
+    },
+    {
+      "epoch": 0.14970561962365345,
+      "grad_norm": 0.6846886873245239,
+      "learning_rate": 3.7074221294171855e-05,
+      "loss": 0.18906841278076172,
+      "step": 28950
+    },
+    {
+      "epoch": 0.14999789924025567,
+      "grad_norm": 0.5114481449127197,
+      "learning_rate": 3.701163568365339e-05,
+      "loss": 0.20319364070892335,
+      "step": 28960
+    },
+    {
+      "epoch": 0.1502901788568579,
+      "grad_norm": 0.5294245481491089,
+      "learning_rate": 3.694909094560071e-05,
+      "loss": 0.2252985954284668,
+      "step": 28970
+    },
+    {
+      "epoch": 0.1505824584734601,
+      "grad_norm": 0.3485918343067169,
+      "learning_rate": 3.688658712059835e-05,
+      "loss": 0.1761780261993408,
+      "step": 28980
+    },
+    {
+      "epoch": 0.1508747380900623,
+      "grad_norm": 0.5018284320831299,
+      "learning_rate": 3.682412424920438e-05,
+      "loss": 0.18765780925750733,
+      "step": 28990
+    },
+    {
+      "epoch": 0.15116701770666452,
+      "grad_norm": 0.3857821524143219,
+      "learning_rate": 3.6761702371950344e-05,
+      "loss": 0.14553301334381102,
+      "step": 29000
+    },
+    {
+      "epoch": 0.15145929732326674,
+      "grad_norm": 0.55930095911026,
+      "learning_rate": 3.6699321529341037e-05,
+      "loss": 0.19647326469421386,
+      "step": 29010
+    },
+    {
+      "epoch": 0.15175157693986896,
+      "grad_norm": 0.4785193204879761,
+      "learning_rate": 3.663698176185474e-05,
+      "loss": 0.1489745259284973,
+      "step": 29020
+    },
+    {
+      "epoch": 0.15204385655647115,
+      "grad_norm": 0.6820440292358398,
+      "learning_rate": 3.6574683109943055e-05,
+      "loss": 0.19364893436431885,
+      "step": 29030
+    },
+    {
+      "epoch": 0.15233613617307337,
+      "grad_norm": 0.45923298597335815,
+      "learning_rate": 3.651242561403093e-05,
+      "loss": 0.1679161310195923,
+      "step": 29040
+    },
+    {
+      "epoch": 0.1526284157896756,
+      "grad_norm": 0.4131610691547394,
+      "learning_rate": 3.645020931451652e-05,
+      "loss": 0.15961675643920897,
+      "step": 29050
+    },
+    {
+      "epoch": 0.1529206954062778,
+      "grad_norm": 0.34882524609565735,
+      "learning_rate": 3.638803425177134e-05,
+      "loss": 0.18716013431549072,
+      "step": 29060
+    },
+    {
+      "epoch": 0.15321297502288,
+      "grad_norm": 0.6257960796356201,
+      "learning_rate": 3.632590046614014e-05,
+      "loss": 0.23960530757904053,
+      "step": 29070
+    },
+    {
+      "epoch": 0.15350525463948222,
+      "grad_norm": 0.5228058099746704,
+      "learning_rate": 3.626380799794076e-05,
+      "loss": 0.3398411750793457,
+      "step": 29080
+    },
+    {
+      "epoch": 0.15379753425608444,
+      "grad_norm": 0.5799852013587952,
+      "learning_rate": 3.6201756887464455e-05,
+      "loss": 0.29375035762786866,
+      "step": 29090
+    },
+    {
+      "epoch": 0.15408981387268667,
+      "grad_norm": 0.599976122379303,
+      "learning_rate": 3.613974717497547e-05,
+      "loss": 0.30395660400390623,
+      "step": 29100
+    },
+    {
+      "epoch": 0.15438209348928886,
+      "grad_norm": 0.5709936022758484,
+      "learning_rate": 3.607777890071115e-05,
+      "loss": 0.31661694049835204,
+      "step": 29110
+    },
+    {
+      "epoch": 0.15467437310589108,
+      "grad_norm": 0.563369631767273,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 0.2866456747055054,
+      "step": 29120
+    },
+    {
+      "epoch": 0.1549666527224933,
+      "grad_norm": 0.46088477969169617,
+      "learning_rate": 3.5953966827672066e-05,
+      "loss": 0.290778660774231,
+      "step": 29130
+    },
+    {
+      "epoch": 0.15525893233909552,
+      "grad_norm": 0.3851511478424072,
+      "learning_rate": 3.5892123109237554e-05,
+      "loss": 0.23082396984100342,
+      "step": 29140
+    },
+    {
+      "epoch": 0.1555512119556977,
+      "grad_norm": 0.3564746081829071,
+      "learning_rate": 3.583032098970838e-05,
+      "loss": 0.31605370044708253,
+      "step": 29150
+    },
+    {
+      "epoch": 0.15584349157229993,
+      "grad_norm": 0.4128909111022949,
+      "learning_rate": 3.576856050918719e-05,
+      "loss": 0.27744910717010496,
+      "step": 29160
+    },
+    {
+      "epoch": 0.15613577118890215,
+      "grad_norm": 0.34181347489356995,
+      "learning_rate": 3.570684170774974e-05,
+      "loss": 0.2934889554977417,
+      "step": 29170
+    },
+    {
+      "epoch": 0.15642805080550434,
+      "grad_norm": 0.3412753939628601,
+      "learning_rate": 3.5645164625444694e-05,
+      "loss": 0.27491016387939454,
+      "step": 29180
+    },
+    {
+      "epoch": 0.15672033042210656,
+      "grad_norm": 0.5635940432548523,
+      "learning_rate": 3.558352930229365e-05,
+      "loss": 0.3293509006500244,
+      "step": 29190
+    },
+    {
+      "epoch": 0.15701261003870878,
+      "grad_norm": 0.4219683110713959,
+      "learning_rate": 3.552193577829106e-05,
+      "loss": 0.2884700298309326,
+      "step": 29200
+    },
+    {
+      "epoch": 0.157304889655311,
+      "grad_norm": 0.42664945125579834,
+      "learning_rate": 3.546038409340432e-05,
+      "loss": 0.30024359226226804,
+      "step": 29210
+    },
+    {
+      "epoch": 0.1575971692719132,
+      "grad_norm": 0.4336451590061188,
+      "learning_rate": 3.539887428757367e-05,
+      "loss": 0.31950252056121825,
+      "step": 29220
+    },
+    {
+      "epoch": 0.15788944888851542,
+      "grad_norm": 0.4691856801509857,
+      "learning_rate": 3.533740640071209e-05,
+      "loss": 0.25833241939544677,
+      "step": 29230
+    },
+    {
+      "epoch": 0.15818172850511764,
+      "grad_norm": 0.3956485688686371,
+      "learning_rate": 3.5275980472705484e-05,
+      "loss": 0.3047396183013916,
+      "step": 29240
+    },
+    {
+      "epoch": 0.15847400812171986,
+      "grad_norm": 0.5770565271377563,
+      "learning_rate": 3.521459654341244e-05,
+      "loss": 0.2951256275177002,
+      "step": 29250
+    },
+    {
+      "epoch": 0.15876628773832205,
+      "grad_norm": 0.6213288307189941,
+      "learning_rate": 3.515325465266436e-05,
+      "loss": 0.2792485237121582,
+      "step": 29260
+    },
+    {
+      "epoch": 0.15905856735492427,
+      "grad_norm": 0.413501113653183,
+      "learning_rate": 3.5091954840265286e-05,
+      "loss": 0.2694176435470581,
+      "step": 29270
+    },
+    {
+      "epoch": 0.1593508469715265,
+      "grad_norm": 0.49126946926116943,
+      "learning_rate": 3.503069714599201e-05,
+      "loss": 0.3735769033432007,
+      "step": 29280
+    },
+    {
+      "epoch": 0.1596431265881287,
+      "grad_norm": 0.4455559551715851,
+      "learning_rate": 3.4969481609594026e-05,
+      "loss": 0.3086733341217041,
+      "step": 29290
+    },
+    {
+      "epoch": 0.1599354062047309,
+      "grad_norm": 0.3725144863128662,
+      "learning_rate": 3.4908308270793344e-05,
+      "loss": 0.24934673309326172,
+      "step": 29300
+    },
+    {
+      "epoch": 0.16022768582133312,
+      "grad_norm": 0.5326842665672302,
+      "learning_rate": 3.4847177169284784e-05,
+      "loss": 0.27558863162994385,
+      "step": 29310
+    },
+    {
+      "epoch": 0.16051996543793534,
+      "grad_norm": 0.5251521468162537,
+      "learning_rate": 3.478608834473562e-05,
+      "loss": 0.2876440525054932,
+      "step": 29320
+    },
+    {
+      "epoch": 0.16081224505453756,
+      "grad_norm": 0.47830405831336975,
+      "learning_rate": 3.472504183678564e-05,
+      "loss": 0.3317052125930786,
+      "step": 29330
+    },
+    {
+      "epoch": 0.16110452467113975,
+      "grad_norm": 0.41410595178604126,
+      "learning_rate": 3.46640376850474e-05,
+      "loss": 0.2519268035888672,
+      "step": 29340
+    },
+    {
+      "epoch": 0.16139680428774197,
+      "grad_norm": 0.4861215054988861,
+      "learning_rate": 3.460307592910573e-05,
+      "loss": 0.3285313606262207,
+      "step": 29350
+    },
+    {
+      "epoch": 0.1616890839043442,
+      "grad_norm": 0.545689582824707,
+      "learning_rate": 3.454215660851808e-05,
+      "loss": 0.3384054660797119,
+      "step": 29360
+    },
+    {
+      "epoch": 0.16198136352094641,
+      "grad_norm": 0.46531838178634644,
+      "learning_rate": 3.4481279762814386e-05,
+      "loss": 0.2807896137237549,
+      "step": 29370
+    },
+    {
+      "epoch": 0.1622736431375486,
+      "grad_norm": 0.44436612725257874,
+      "learning_rate": 3.442044543149688e-05,
+      "loss": 0.29774482250213624,
+      "step": 29380
+    },
+    {
+      "epoch": 0.16256592275415083,
+      "grad_norm": 0.39731815457344055,
+      "learning_rate": 3.435965365404036e-05,
+      "loss": 0.27618086338043213,
+      "step": 29390
+    },
+    {
+      "epoch": 0.16285820237075305,
+      "grad_norm": 0.40577206015586853,
+      "learning_rate": 3.4298904469891936e-05,
+      "loss": 0.30148136615753174,
+      "step": 29400
+    },
+    {
+      "epoch": 0.16315048198735524,
+      "grad_norm": 0.39966824650764465,
+      "learning_rate": 3.4238197918471115e-05,
+      "loss": 0.3293858528137207,
+      "step": 29410
+    },
+    {
+      "epoch": 0.16344276160395746,
+      "grad_norm": 0.3944089412689209,
+      "learning_rate": 3.417753403916968e-05,
+      "loss": 0.297630763053894,
+      "step": 29420
+    },
+    {
+      "epoch": 0.16373504122055968,
+      "grad_norm": 0.44451671838760376,
+      "learning_rate": 3.411691287135177e-05,
+      "loss": 0.29928138256073,
+      "step": 29430
+    },
+    {
+      "epoch": 0.1640273208371619,
+      "grad_norm": 0.45176878571510315,
+      "learning_rate": 3.4056334454353854e-05,
+      "loss": 0.28041954040527345,
+      "step": 29440
+    },
+    {
+      "epoch": 0.1643196004537641,
+      "grad_norm": 0.42176857590675354,
+      "learning_rate": 3.39957988274845e-05,
+      "loss": 0.273341703414917,
+      "step": 29450
+    },
+    {
+      "epoch": 0.1646118800703663,
+      "grad_norm": 0.4102049171924591,
+      "learning_rate": 3.3935306030024773e-05,
+      "loss": 0.2911419630050659,
+      "step": 29460
+    },
+    {
+      "epoch": 0.16490415968696853,
+      "grad_norm": 0.4470144808292389,
+      "learning_rate": 3.387485610122766e-05,
+      "loss": 0.2940429449081421,
+      "step": 29470
+    },
+    {
+      "epoch": 0.16519643930357075,
+      "grad_norm": 0.34494346380233765,
+      "learning_rate": 3.3814449080318524e-05,
+      "loss": 0.2626910448074341,
+      "step": 29480
+    },
+    {
+      "epoch": 0.16548871892017294,
+      "grad_norm": 0.5578533411026001,
+      "learning_rate": 3.375408500649483e-05,
+      "loss": 0.28085010051727294,
+      "step": 29490
+    },
+    {
+      "epoch": 0.16578099853677516,
+      "grad_norm": 0.377634733915329,
+      "learning_rate": 3.3693763918926134e-05,
+      "loss": 0.32071495056152344,
+      "step": 29500
+    },
+    {
+      "epoch": 0.16607327815337738,
+      "grad_norm": 0.4433046579360962,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 0.34470200538635254,
+      "step": 29510
+    },
+    {
+      "epoch": 0.1663655577699796,
+      "grad_norm": 0.47977587580680847,
+      "learning_rate": 3.3573250859092666e-05,
+      "loss": 0.23838646411895753,
+      "step": 29520
+    },
+    {
+      "epoch": 0.1666578373865818,
+      "grad_norm": 0.42079079151153564,
+      "learning_rate": 3.3513058965027545e-05,
+      "loss": 0.2690751314163208,
+      "step": 29530
+    },
+    {
+      "epoch": 0.16695011700318402,
+      "grad_norm": 0.46342065930366516,
+      "learning_rate": 3.345291021361659e-05,
+      "loss": 0.30729141235351565,
+      "step": 29540
+    },
+    {
+      "epoch": 0.16724239661978624,
+      "grad_norm": 0.34235286712646484,
+      "learning_rate": 3.3392804643889694e-05,
+      "loss": 0.2764017105102539,
+      "step": 29550
+    },
+    {
+      "epoch": 0.16753467623638846,
+      "grad_norm": 0.3624505400657654,
+      "learning_rate": 3.3332742294848776e-05,
+      "loss": 0.245084547996521,
+      "step": 29560
+    },
+    {
+      "epoch": 0.16782695585299065,
+      "grad_norm": 0.36505126953125,
+      "learning_rate": 3.327272320546754e-05,
+      "loss": 0.2980503559112549,
+      "step": 29570
+    },
+    {
+      "epoch": 0.16811923546959287,
+      "grad_norm": 0.46576160192489624,
+      "learning_rate": 3.321274741469177e-05,
+      "loss": 0.34157936573028563,
+      "step": 29580
+    },
+    {
+      "epoch": 0.1684115150861951,
+      "grad_norm": 0.36173564195632935,
+      "learning_rate": 3.3152814961439137e-05,
+      "loss": 0.259122633934021,
+      "step": 29590
+    },
+    {
+      "epoch": 0.1687037947027973,
+      "grad_norm": 0.4300622045993805,
+      "learning_rate": 3.3092925884599094e-05,
+      "loss": 0.334716272354126,
+      "step": 29600
+    },
+    {
+      "epoch": 0.1689960743193995,
+      "grad_norm": 0.41738736629486084,
+      "learning_rate": 3.3033080223033044e-05,
+      "loss": 0.29223344326019285,
+      "step": 29610
+    },
+    {
+      "epoch": 0.16928835393600172,
+      "grad_norm": 0.5646920800209045,
+      "learning_rate": 3.2973278015574196e-05,
+      "loss": 0.28518683910369874,
+      "step": 29620
+    },
+    {
+      "epoch": 0.16958063355260394,
+      "grad_norm": 0.5043878555297852,
+      "learning_rate": 3.2913519301027576e-05,
+      "loss": 0.23273320198059083,
+      "step": 29630
+    },
+    {
+      "epoch": 0.16987291316920616,
+      "grad_norm": 0.40135180950164795,
+      "learning_rate": 3.285380411816988e-05,
+      "loss": 0.2461369276046753,
+      "step": 29640
+    },
+    {
+      "epoch": 0.17016519278580836,
+      "grad_norm": 0.42611992359161377,
+      "learning_rate": 3.279413250574972e-05,
+      "loss": 0.2348939895629883,
+      "step": 29650
+    },
+    {
+      "epoch": 0.17045747240241058,
+      "grad_norm": 0.4052818715572357,
+      "learning_rate": 3.273450450248736e-05,
+      "loss": 0.28273630142211914,
+      "step": 29660
+    },
+    {
+      "epoch": 0.1707497520190128,
+      "grad_norm": 0.3995920419692993,
+      "learning_rate": 3.267492014707467e-05,
+      "loss": 0.2929879903793335,
+      "step": 29670
+    },
+    {
+      "epoch": 0.171042031635615,
+      "grad_norm": 0.32996323704719543,
+      "learning_rate": 3.2615379478175424e-05,
+      "loss": 0.2315880537033081,
+      "step": 29680
+    },
+    {
+      "epoch": 0.1713343112522172,
+      "grad_norm": 0.301029771566391,
+      "learning_rate": 3.2555882534424855e-05,
+      "loss": 0.27566049098968504,
+      "step": 29690
+    },
+    {
+      "epoch": 0.17162659086881943,
+      "grad_norm": 0.3210665285587311,
+      "learning_rate": 3.2496429354429806e-05,
+      "loss": 0.2651090383529663,
+      "step": 29700
+    },
+    {
+      "epoch": 0.17191887048542165,
+      "grad_norm": 0.4376218616962433,
+      "learning_rate": 3.243701997676893e-05,
+      "loss": 0.26818721294403075,
+      "step": 29710
+    },
+    {
+      "epoch": 0.17221115010202384,
+      "grad_norm": 0.8171451687812805,
+      "learning_rate": 3.2377654439992255e-05,
+      "loss": 0.3270720481872559,
+      "step": 29720
+    },
+    {
+      "epoch": 0.17250342971862606,
+      "grad_norm": 0.4278380870819092,
+      "learning_rate": 3.231833278262145e-05,
+      "loss": 0.276395320892334,
+      "step": 29730
+    },
+    {
+      "epoch": 0.17279570933522828,
+      "grad_norm": 0.4471166729927063,
+      "learning_rate": 3.225905504314969e-05,
+      "loss": 0.24598653316497804,
+      "step": 29740
+    },
+    {
+      "epoch": 0.1730879889518305,
+      "grad_norm": 0.3617573082447052,
+      "learning_rate": 3.2199821260041683e-05,
+      "loss": 0.2993766784667969,
+      "step": 29750
+    },
+    {
+      "epoch": 0.1733802685684327,
+      "grad_norm": 0.4057319462299347,
+      "learning_rate": 3.214063147173354e-05,
+      "loss": 0.250543999671936,
+      "step": 29760
+    },
+    {
+      "epoch": 0.17367254818503491,
+      "grad_norm": 0.40494492650032043,
+      "learning_rate": 3.208148571663289e-05,
+      "loss": 0.3049450397491455,
+      "step": 29770
+    },
+    {
+      "epoch": 0.17396482780163713,
+      "grad_norm": 0.3548034131526947,
+      "learning_rate": 3.2022384033118815e-05,
+      "loss": 0.30169928073883057,
+      "step": 29780
+    },
+    {
+      "epoch": 0.17425710741823935,
+      "grad_norm": 0.5113735795021057,
+      "learning_rate": 3.196332645954169e-05,
+      "loss": 0.3650834560394287,
+      "step": 29790
+    },
+    {
+      "epoch": 0.17454938703484155,
+      "grad_norm": 0.4839605689048767,
+      "learning_rate": 3.190431303422336e-05,
+      "loss": 0.3876906156539917,
+      "step": 29800
+    },
+    {
+      "epoch": 0.17484166665144377,
+      "grad_norm": 0.477368026971817,
+      "learning_rate": 3.184534379545702e-05,
+      "loss": 0.37152698040008547,
+      "step": 29810
+    },
+    {
+      "epoch": 0.175133946268046,
+      "grad_norm": 0.3773256540298462,
+      "learning_rate": 3.178641878150712e-05,
+      "loss": 0.3911390781402588,
+      "step": 29820
+    },
+    {
+      "epoch": 0.1754262258846482,
+      "grad_norm": 0.295611172914505,
+      "learning_rate": 3.1727538030609495e-05,
+      "loss": 0.34877662658691405,
+      "step": 29830
+    },
+    {
+      "epoch": 0.1757185055012504,
+      "grad_norm": 0.34554120898246765,
+      "learning_rate": 3.1668701580971205e-05,
+      "loss": 0.39424092769622804,
+      "step": 29840
+    },
+    {
+      "epoch": 0.17601078511785262,
+      "grad_norm": 0.42011308670043945,
+      "learning_rate": 3.160990947077064e-05,
+      "loss": 0.3001796960830688,
+      "step": 29850
+    },
+    {
+      "epoch": 0.17630306473445484,
+      "grad_norm": 0.447304904460907,
+      "learning_rate": 3.155116173815729e-05,
+      "loss": 0.35801091194152834,
+      "step": 29860
+    },
+    {
+      "epoch": 0.17659534435105706,
+      "grad_norm": 0.5688051581382751,
+      "learning_rate": 3.149245842125194e-05,
+      "loss": 0.39142580032348634,
+      "step": 29870
+    },
+    {
+      "epoch": 0.17688762396765925,
+      "grad_norm": 0.3813949227333069,
+      "learning_rate": 3.143379955814658e-05,
+      "loss": 0.3552913427352905,
+      "step": 29880
+    },
+    {
+      "epoch": 0.17717990358426147,
+      "grad_norm": 0.547062337398529,
+      "learning_rate": 3.137518518690421e-05,
+      "loss": 0.33888659477233884,
+      "step": 29890
+    },
+    {
+      "epoch": 0.1774721832008637,
+      "grad_norm": 0.45928165316581726,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 0.36072754859924316,
+      "step": 29900
+    },
+    {
+      "epoch": 0.1777644628174659,
+      "grad_norm": 0.5510131120681763,
+      "learning_rate": 3.125809007211679e-05,
+      "loss": 0.3214040517807007,
+      "step": 29910
+    },
+    {
+      "epoch": 0.1780567424340681,
+      "grad_norm": 0.39881381392478943,
+      "learning_rate": 3.119960940455334e-05,
+      "loss": 0.38784019947052,
+      "step": 29920
+    },
+    {
+      "epoch": 0.17834902205067032,
+      "grad_norm": 0.36602431535720825,
+      "learning_rate": 3.114117338081648e-05,
+      "loss": 0.35827977657318116,
+      "step": 29930
+    },
+    {
+      "epoch": 0.17864130166727255,
+      "grad_norm": 0.45217660069465637,
+      "learning_rate": 3.108278203882458e-05,
+      "loss": 0.37975201606750486,
+      "step": 29940
+    },
+    {
+      "epoch": 0.17893358128387474,
+      "grad_norm": 0.35449516773223877,
+      "learning_rate": 3.102443541646719e-05,
+      "loss": 0.42003498077392576,
+      "step": 29950
+    },
+    {
+      "epoch": 0.17922586090047696,
+      "grad_norm": 0.39860430359840393,
+      "learning_rate": 3.09661335516048e-05,
+      "loss": 0.3907869100570679,
+      "step": 29960
+    },
+    {
+      "epoch": 0.17951814051707918,
+      "grad_norm": 0.5728691220283508,
+      "learning_rate": 3.09078764820689e-05,
+      "loss": 0.29853367805480957,
+      "step": 29970
+    },
+    {
+      "epoch": 0.1798104201336814,
+      "grad_norm": 0.40936318039894104,
+      "learning_rate": 3.084966424566179e-05,
+      "loss": 0.32233223915100095,
+      "step": 29980
+    },
+    {
+      "epoch": 0.1801026997502836,
+      "grad_norm": 0.3031325042247772,
+      "learning_rate": 3.079149688015679e-05,
+      "loss": 0.34882681369781493,
+      "step": 29990
+    },
+    {
+      "epoch": 0.1803949793668858,
+      "grad_norm": 0.40131714940071106,
+      "learning_rate": 3.0733374423298124e-05,
+      "loss": 0.33974432945251465,
+      "step": 30000
+    },
+    {
+      "epoch": 0.18068725898348803,
+      "grad_norm": 0.39961910247802734,
+      "learning_rate": 3.0675296912800736e-05,
+      "loss": 0.39501338005065917,
+      "step": 30010
+    },
+    {
+      "epoch": 0.18097953860009025,
+      "grad_norm": 0.3837500810623169,
+      "learning_rate": 3.061726438635054e-05,
+      "loss": 0.3640526294708252,
+      "step": 30020
+    },
+    {
+      "epoch": 0.18127181821669244,
+      "grad_norm": 0.38006001710891724,
+      "learning_rate": 3.0559276881604236e-05,
+      "loss": 0.3666640281677246,
+      "step": 30030
+    },
+    {
+      "epoch": 0.18156409783329466,
+      "grad_norm": 0.3662281036376953,
+      "learning_rate": 3.050133443618922e-05,
+      "loss": 0.3733483552932739,
+      "step": 30040
+    },
+    {
+      "epoch": 0.18185637744989688,
+      "grad_norm": 0.3845963776111603,
+      "learning_rate": 3.0443437087703762e-05,
+      "loss": 0.316637921333313,
+      "step": 30050
+    },
+    {
+      "epoch": 0.1821486570664991,
+      "grad_norm": 0.3406497836112976,
+      "learning_rate": 3.038558487371681e-05,
+      "loss": 0.3951242208480835,
+      "step": 30060
+    },
+    {
+      "epoch": 0.1824409366831013,
+      "grad_norm": 0.4281523525714874,
+      "learning_rate": 3.0327777831768067e-05,
+      "loss": 0.4334969997406006,
+      "step": 30070
+    },
+    {
+      "epoch": 0.18273321629970352,
+      "grad_norm": 0.3316340148448944,
+      "learning_rate": 3.027001599936784e-05,
+      "loss": 0.3702265977859497,
+      "step": 30080
+    },
+    {
+      "epoch": 0.18302549591630574,
+      "grad_norm": 0.48244062066078186,
+      "learning_rate": 3.021229941399718e-05,
+      "loss": 0.3473745584487915,
+      "step": 30090
+    },
+    {
+      "epoch": 0.18331777553290796,
+      "grad_norm": 0.3406129479408264,
+      "learning_rate": 3.0154628113107787e-05,
+      "loss": 0.29930007457733154,
+      "step": 30100
+    },
+    {
+      "epoch": 0.18361005514951015,
+      "grad_norm": 0.4943349063396454,
+      "learning_rate": 3.0097002134121855e-05,
+      "loss": 0.4322549819946289,
+      "step": 30110
+    },
+    {
+      "epoch": 0.18390233476611237,
+      "grad_norm": 0.4638117551803589,
+      "learning_rate": 3.0039421514432375e-05,
+      "loss": 0.33375422954559325,
+      "step": 30120
+    },
+    {
+      "epoch": 0.1841946143827146,
+      "grad_norm": 0.3749340772628784,
+      "learning_rate": 2.9981886291402715e-05,
+      "loss": 0.37318236827850343,
+      "step": 30130
+    },
+    {
+      "epoch": 0.1844868939993168,
+      "grad_norm": 0.40442243218421936,
+      "learning_rate": 2.99243965023668e-05,
+      "loss": 0.3808151721954346,
+      "step": 30140
+    },
+    {
+      "epoch": 0.184779173615919,
+      "grad_norm": 0.47701358795166016,
+      "learning_rate": 2.986695218462927e-05,
+      "loss": 0.3489521980285645,
+      "step": 30150
+    },
+    {
+      "epoch": 0.18507145323252122,
+      "grad_norm": 0.3509625792503357,
+      "learning_rate": 2.9809553375465004e-05,
+      "loss": 0.3411436080932617,
+      "step": 30160
+    },
+    {
+      "epoch": 0.18536373284912344,
+      "grad_norm": 0.31631937623023987,
+      "learning_rate": 2.975220011211951e-05,
+      "loss": 0.38819930553436277,
+      "step": 30170
+    },
+    {
+      "epoch": 0.18565601246572566,
+      "grad_norm": 0.40450090169906616,
+      "learning_rate": 2.9694892431808696e-05,
+      "loss": 0.33304221630096437,
+      "step": 30180
+    },
+    {
+      "epoch": 0.18594829208232785,
+      "grad_norm": 0.39138996601104736,
+      "learning_rate": 2.9637630371718916e-05,
+      "loss": 0.3920623779296875,
+      "step": 30190
+    },
+    {
+      "epoch": 0.18624057169893007,
+      "grad_norm": 0.3502466082572937,
+      "learning_rate": 2.9580413969006848e-05,
+      "loss": 0.42269234657287597,
+      "step": 30200
+    },
+    {
+      "epoch": 0.1865328513155323,
+      "grad_norm": 0.5317121148109436,
+      "learning_rate": 2.9523243260799615e-05,
+      "loss": 0.33345770835876465,
+      "step": 30210
+    },
+    {
+      "epoch": 0.1868251309321345,
+      "grad_norm": 0.4839714467525482,
+      "learning_rate": 2.946611828419472e-05,
+      "loss": 0.37590670585632324,
+      "step": 30220
+    },
+    {
+      "epoch": 0.1871174105487367,
+      "grad_norm": 0.44953539967536926,
+      "learning_rate": 2.940903907625985e-05,
+      "loss": 0.3582432746887207,
+      "step": 30230
+    },
+    {
+      "epoch": 0.18740969016533893,
+      "grad_norm": 0.41432836651802063,
+      "learning_rate": 2.9352005674033135e-05,
+      "loss": 0.3124553918838501,
+      "step": 30240
+    },
+    {
+      "epoch": 0.18770196978194115,
+      "grad_norm": 0.3317153751850128,
+      "learning_rate": 2.9295018114522953e-05,
+      "loss": 0.3911390781402588,
+      "step": 30250
+    },
+    {
+      "epoch": 0.18799424939854334,
+      "grad_norm": 0.31571638584136963,
+      "learning_rate": 2.923807643470785e-05,
+      "loss": 0.3495438575744629,
+      "step": 30260
+    },
+    {
+      "epoch": 0.18828652901514556,
+      "grad_norm": 0.4265151619911194,
+      "learning_rate": 2.9181180671536702e-05,
+      "loss": 0.324018669128418,
+      "step": 30270
+    },
+    {
+      "epoch": 0.18857880863174778,
+      "grad_norm": 0.3939460217952728,
+      "learning_rate": 2.912433086192854e-05,
+      "loss": 0.3496082305908203,
+      "step": 30280
+    },
+    {
+      "epoch": 0.18887108824835,
+      "grad_norm": 0.36875006556510925,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 0.3979907274246216,
+      "step": 30290
+    },
+    {
+      "epoch": 0.1891633678649522,
+      "grad_norm": 0.28948017954826355,
+      "learning_rate": 2.9010769250928295e-05,
+      "loss": 0.3815887928009033,
+      "step": 30300
+    },
+    {
+      "epoch": 0.1894556474815544,
+      "grad_norm": 0.4039372503757477,
+      "learning_rate": 2.8954057523225066e-05,
+      "loss": 0.4214956760406494,
+      "step": 30310
+    },
+    {
+      "epoch": 0.18974792709815663,
+      "grad_norm": 0.44159185886383057,
+      "learning_rate": 2.889739189646261e-05,
+      "loss": 0.3507035732269287,
+      "step": 30320
+    },
+    {
+      "epoch": 0.19004020671475885,
+      "grad_norm": 0.30350378155708313,
+      "learning_rate": 2.8840772407410542e-05,
+      "loss": 0.36494951248168944,
+      "step": 30330
+    },
+    {
+      "epoch": 0.19033248633136104,
+      "grad_norm": 0.3325420022010803,
+      "learning_rate": 2.878419909280875e-05,
+      "loss": 0.3464409112930298,
+      "step": 30340
+    },
+    {
+      "epoch": 0.19062476594796327,
+      "grad_norm": 0.44692251086235046,
+      "learning_rate": 2.8727671989366988e-05,
+      "loss": 0.42154641151428224,
+      "step": 30350
+    },
+    {
+      "epoch": 0.19091704556456549,
+      "grad_norm": 0.37338414788246155,
+      "learning_rate": 2.8671191133765018e-05,
+      "loss": 0.3785668611526489,
+      "step": 30360
+    },
+    {
+      "epoch": 0.1912093251811677,
+      "grad_norm": 0.40996572375297546,
+      "learning_rate": 2.8614756562652788e-05,
+      "loss": 0.3122179746627808,
+      "step": 30370
+    },
+    {
+      "epoch": 0.1915016047977699,
+      "grad_norm": 0.33593571186065674,
+      "learning_rate": 2.8558368312649964e-05,
+      "loss": 0.33015730381011965,
+      "step": 30380
+    },
+    {
+      "epoch": 0.19179388441437212,
+      "grad_norm": 0.5369383096694946,
+      "learning_rate": 2.8502026420346318e-05,
+      "loss": 0.33395540714263916,
+      "step": 30390
+    },
+    {
+      "epoch": 0.19208616403097434,
+      "grad_norm": 0.5107163786888123,
+      "learning_rate": 2.8445730922301518e-05,
+      "loss": 0.37331039905548097,
+      "step": 30400
+    },
+    {
+      "epoch": 0.19237844364757656,
+      "grad_norm": 0.38285601139068604,
+      "learning_rate": 2.838948185504503e-05,
+      "loss": 0.38164947032928465,
+      "step": 30410
+    },
+    {
+      "epoch": 0.19267072326417875,
+      "grad_norm": 0.4003716707229614,
+      "learning_rate": 2.8333279255076306e-05,
+      "loss": 0.3435195446014404,
+      "step": 30420
+    },
+    {
+      "epoch": 0.19296300288078097,
+      "grad_norm": 0.38201335072517395,
+      "learning_rate": 2.8277123158864594e-05,
+      "loss": 0.40462265014648435,
+      "step": 30430
+    },
+    {
+      "epoch": 0.1932552824973832,
+      "grad_norm": 0.35732266306877136,
+      "learning_rate": 2.822101360284901e-05,
+      "loss": 0.39876952171325686,
+      "step": 30440
+    },
+    {
+      "epoch": 0.1935475621139854,
+      "grad_norm": 0.40936318039894104,
+      "learning_rate": 2.816495062343839e-05,
+      "loss": 0.3432680368423462,
+      "step": 30450
+    },
+    {
+      "epoch": 0.1938398417305876,
+      "grad_norm": 0.5065858960151672,
+      "learning_rate": 2.81089342570114e-05,
+      "loss": 0.3459235906600952,
+      "step": 30460
+    },
+    {
+      "epoch": 0.19413212134718982,
+      "grad_norm": 0.39030635356903076,
+      "learning_rate": 2.8052964539916482e-05,
+      "loss": 0.3741995573043823,
+      "step": 30470
+    },
+    {
+      "epoch": 0.19442440096379204,
+      "grad_norm": 0.40187081694602966,
+      "learning_rate": 2.7997041508471746e-05,
+      "loss": 0.339365291595459,
+      "step": 30480
+    },
+    {
+      "epoch": 0.19471668058039424,
+      "grad_norm": 0.32880908250808716,
+      "learning_rate": 2.7941165198965034e-05,
+      "loss": 0.36291794776916503,
+      "step": 30490
+    },
+    {
+      "epoch": 0.19500896019699646,
+      "grad_norm": 0.4341806173324585,
+      "learning_rate": 2.7885335647653888e-05,
+      "loss": 0.3181918621063232,
+      "step": 30500
+    },
+    {
+      "epoch": 0.19530123981359868,
+      "grad_norm": 0.45084238052368164,
+      "learning_rate": 2.7829552890765544e-05,
+      "loss": 0.41957736015319824,
+      "step": 30510
+    },
+    {
+      "epoch": 0.1955935194302009,
+      "grad_norm": 0.4353318512439728,
+      "learning_rate": 2.7773816964496758e-05,
+      "loss": 0.3733999013900757,
+      "step": 30520
+    },
+    {
+      "epoch": 0.1958857990468031,
+      "grad_norm": 0.34275874495506287,
+      "learning_rate": 2.771812790501399e-05,
+      "loss": 0.3667341947555542,
+      "step": 30530
+    },
+    {
+      "epoch": 0.1961780786634053,
+      "grad_norm": 0.3720132112503052,
+      "learning_rate": 2.7662485748453305e-05,
+      "loss": 0.3355003118515015,
+      "step": 30540
+    },
+    {
+      "epoch": 0.19647035828000753,
+      "grad_norm": 0.38378554582595825,
+      "learning_rate": 2.7606890530920195e-05,
+      "loss": 0.32172939777374265,
+      "step": 30550
+    },
+    {
+      "epoch": 0.19676263789660975,
+      "grad_norm": 0.330384224653244,
+      "learning_rate": 2.755134228848992e-05,
+      "loss": 0.29499542713165283,
+      "step": 30560
+    },
+    {
+      "epoch": 0.19705491751321194,
+      "grad_norm": 0.40749484300613403,
+      "learning_rate": 2.749584105720707e-05,
+      "loss": 0.3853444576263428,
+      "step": 30570
+    },
+    {
+      "epoch": 0.19734719712981416,
+      "grad_norm": 0.3921387791633606,
+      "learning_rate": 2.7440386873085733e-05,
+      "loss": 0.33362269401550293,
+      "step": 30580
+    },
+    {
+      "epoch": 0.19763947674641638,
+      "grad_norm": 0.3459542393684387,
+      "learning_rate": 2.7384979772109643e-05,
+      "loss": 0.36746368408203123,
+      "step": 30590
+    },
+    {
+      "epoch": 0.1979317563630186,
+      "grad_norm": 0.467636376619339,
+      "learning_rate": 2.7329619790231787e-05,
+      "loss": 0.3533440351486206,
+      "step": 30600
+    },
+    {
+      "epoch": 0.1982240359796208,
+      "grad_norm": 0.3198123276233673,
+      "learning_rate": 2.7274306963374675e-05,
+      "loss": 0.29893224239349364,
+      "step": 30610
+    },
+    {
+      "epoch": 0.19851631559622301,
+      "grad_norm": 0.5530977249145508,
+      "learning_rate": 2.721904132743023e-05,
+      "loss": 0.3656468868255615,
+      "step": 30620
+    },
+    {
+      "epoch": 0.19880859521282523,
+      "grad_norm": 0.48909991979599,
+      "learning_rate": 2.7163822918259662e-05,
+      "loss": 0.39270691871643065,
+      "step": 30630
+    },
+    {
+      "epoch": 0.19910087482942745,
+      "grad_norm": 0.44308730959892273,
+      "learning_rate": 2.710865177169364e-05,
+      "loss": 0.34100518226623533,
+      "step": 30640
+    },
+    {
+      "epoch": 0.19939315444602965,
+      "grad_norm": 0.3659192621707916,
+      "learning_rate": 2.7053527923532106e-05,
+      "loss": 0.35852746963500975,
+      "step": 30650
+    },
+    {
+      "epoch": 0.19968543406263187,
+      "grad_norm": 0.35255858302116394,
+      "learning_rate": 2.699845140954438e-05,
+      "loss": 0.3300185680389404,
+      "step": 30660
+    },
+    {
+      "epoch": 0.1999777136792341,
+      "grad_norm": 0.5250425934791565,
+      "learning_rate": 2.6943422265468888e-05,
+      "loss": 0.35623347759246826,
+      "step": 30670
+    },
+    {
+      "epoch": 0.2002699932958363,
+      "grad_norm": 0.4660249650478363,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 0.3234785795211792,
+      "step": 30680
+    },
+    {
+      "epoch": 0.2005622729124385,
+      "grad_norm": 0.4130048453807831,
+      "learning_rate": 2.6833506229855475e-05,
+      "loss": 0.3791043758392334,
+      "step": 30690
+    },
+    {
+      "epoch": 0.20085455252904072,
+      "grad_norm": 0.30664271116256714,
+      "learning_rate": 2.677861940964075e-05,
+      "loss": 0.37475242614746096,
+      "step": 30700
+    },
+    {
+      "epoch": 0.20114683214564294,
+      "grad_norm": 0.32288944721221924,
+      "learning_rate": 2.6723780101984998e-05,
+      "loss": 0.29735162258148196,
+      "step": 30710
+    },
+    {
+      "epoch": 0.20143911176224513,
+      "grad_norm": 0.27546101808547974,
+      "learning_rate": 2.6668988342472756e-05,
+      "loss": 0.3538494348526001,
+      "step": 30720
+    },
+    {
+      "epoch": 0.20173139137884735,
+      "grad_norm": 0.30604058504104614,
+      "learning_rate": 2.6614244166657843e-05,
+      "loss": 0.3427356004714966,
+      "step": 30730
+    },
+    {
+      "epoch": 0.20202367099544957,
+      "grad_norm": 0.45071831345558167,
+      "learning_rate": 2.6559547610063172e-05,
+      "loss": 0.3636742115020752,
+      "step": 30740
+    },
+    {
+      "epoch": 0.2023159506120518,
+      "grad_norm": 0.4281063973903656,
+      "learning_rate": 2.6504898708180702e-05,
+      "loss": 0.41559510231018065,
+      "step": 30750
+    },
+    {
+      "epoch": 0.20260823022865398,
+      "grad_norm": 0.479788213968277,
+      "learning_rate": 2.645029749647153e-05,
+      "loss": 0.3173357009887695,
+      "step": 30760
+    },
+    {
+      "epoch": 0.2029005098452562,
+      "grad_norm": 0.4618152976036072,
+      "learning_rate": 2.63957440103658e-05,
+      "loss": 0.37471106052398684,
+      "step": 30770
+    },
+    {
+      "epoch": 0.20319278946185843,
+      "grad_norm": 0.3674679398536682,
+      "learning_rate": 2.6341238285262704e-05,
+      "loss": 0.34237487316131593,
+      "step": 30780
+    },
+    {
+      "epoch": 0.00017536776996132774,
+      "grad_norm": 0.2707560658454895,
+      "learning_rate": 2.628678035653035e-05,
+      "loss": 0.13589210311571756,
+      "step": 30790
+    },
+    {
+      "epoch": 0.0004676473865635407,
+      "grad_norm": 0.2517087459564209,
+      "learning_rate": 2.6232370259505922e-05,
+      "loss": 0.11105576753616334,
+      "step": 30800
+    },
+    {
+      "epoch": 0.0007599270031657536,
+      "grad_norm": 0.30016499757766724,
+      "learning_rate": 2.6178008029495592e-05,
+      "loss": 0.07748374342918396,
+      "step": 30810
+    },
+    {
+      "epoch": 0.0010522066197679665,
+      "grad_norm": 0.5185737609863281,
+      "learning_rate": 2.612369370177434e-05,
+      "loss": 0.05731107592582703,
+      "step": 30820
+    },
+    {
+      "epoch": 0.0013444862363701795,
+      "grad_norm": 0.25890663266181946,
+      "learning_rate": 2.606942731158619e-05,
+      "loss": 0.04520007073879242,
+      "step": 30830
+    },
+    {
+      "epoch": 0.0016367658529723925,
+      "grad_norm": 0.4217708110809326,
+      "learning_rate": 2.601520889414403e-05,
+      "loss": 0.048282259702682497,
+      "step": 30840
+    },
+    {
+      "epoch": 0.0019290454695746052,
+      "grad_norm": 0.44400447607040405,
+      "learning_rate": 2.596103848462954e-05,
+      "loss": 0.05709788203239441,
+      "step": 30850
+    },
+    {
+      "epoch": 0.0022213250861768184,
+      "grad_norm": 0.3610702157020569,
+      "learning_rate": 2.5906916118193346e-05,
+      "loss": 0.04304202795028687,
+      "step": 30860
+    },
+    {
+      "epoch": 0.002513604702779031,
+      "grad_norm": 0.3578943908214569,
+      "learning_rate": 2.5852841829954876e-05,
+      "loss": 0.05245540142059326,
+      "step": 30870
+    },
+    {
+      "epoch": 0.002805884319381244,
+      "grad_norm": 0.46720781922340393,
+      "learning_rate": 2.579881565500236e-05,
+      "loss": 0.03525699377059936,
+      "step": 30880
+    },
+    {
+      "epoch": 0.003098163935983457,
+      "grad_norm": 0.32479098439216614,
+      "learning_rate": 2.574483762839276e-05,
+      "loss": 0.044724881649017334,
+      "step": 30890
+    },
+    {
+      "epoch": 0.00339044355258567,
+      "grad_norm": 0.6349447965621948,
+      "learning_rate": 2.569090778515183e-05,
+      "loss": 0.05059219002723694,
+      "step": 30900
+    },
+    {
+      "epoch": 0.0036827231691878826,
+      "grad_norm": 0.3748016357421875,
+      "learning_rate": 2.5637026160274114e-05,
+      "loss": 0.041036474704742434,
+      "step": 30910
+    },
+    {
+      "epoch": 0.003975002785790095,
+      "grad_norm": 0.43538960814476013,
+      "learning_rate": 2.5583192788722722e-05,
+      "loss": 0.05090898871421814,
+      "step": 30920
+    },
+    {
+      "epoch": 0.0042672824023923086,
+      "grad_norm": 0.45018938183784485,
+      "learning_rate": 2.552940770542964e-05,
+      "loss": 0.04495919644832611,
+      "step": 30930
+    },
+    {
+      "epoch": 0.004559562018994522,
+      "grad_norm": 0.46432650089263916,
+      "learning_rate": 2.547567094529537e-05,
+      "loss": 0.031420457363128665,
+      "step": 30940
+    },
+    {
+      "epoch": 0.004851841635596735,
+      "grad_norm": 0.4390200078487396,
+      "learning_rate": 2.5421982543189116e-05,
+      "loss": 0.03616984486579895,
+      "step": 30950
+    },
+    {
+      "epoch": 0.005144121252198947,
+      "grad_norm": 0.40570294857025146,
+      "learning_rate": 2.536834253394874e-05,
+      "loss": 0.030301403999328614,
+      "step": 30960
+    },
+    {
+      "epoch": 0.0054364008688011604,
+      "grad_norm": 0.5110713839530945,
+      "learning_rate": 2.5314750952380593e-05,
+      "loss": 0.04134462773799896,
+      "step": 30970
+    },
+    {
+      "epoch": 0.005728680485403374,
+      "grad_norm": 0.5130481719970703,
+      "learning_rate": 2.5261207833259714e-05,
+      "loss": 0.03148481249809265,
+      "step": 30980
+    },
+    {
+      "epoch": 0.006020960102005586,
+      "grad_norm": 0.2104519158601761,
+      "learning_rate": 2.5207713211329643e-05,
+      "loss": 0.04479070901870728,
+      "step": 30990
+    },
+    {
+      "epoch": 0.006313239718607799,
+      "grad_norm": 0.5017517805099487,
+      "learning_rate": 2.5154267121302488e-05,
+      "loss": 0.033181813359260556,
+      "step": 31000
+    },
+    {
+      "epoch": 0.006605519335210012,
+      "grad_norm": 0.44545426964759827,
+      "learning_rate": 2.510086959785878e-05,
+      "loss": 0.04562610387802124,
+      "step": 31010
+    },
+    {
+      "epoch": 0.006897798951812225,
+      "grad_norm": 0.4691000282764435,
+      "learning_rate": 2.504752067564762e-05,
+      "loss": 0.046014347672462465,
+      "step": 31020
+    },
+    {
+      "epoch": 0.007190078568414438,
+      "grad_norm": 0.7323338985443115,
+      "learning_rate": 2.499422038928658e-05,
+      "loss": 0.042394807934761046,
+      "step": 31030
+    },
+    {
+      "epoch": 0.007482358185016651,
+      "grad_norm": 0.23409044742584229,
+      "learning_rate": 2.4940968773361572e-05,
+      "loss": 0.02687516510486603,
+      "step": 31040
+    },
+    {
+      "epoch": 0.007774637801618863,
+      "grad_norm": 0.5946905612945557,
+      "learning_rate": 2.4887765862427026e-05,
+      "loss": 0.03483427762985229,
+      "step": 31050
+    },
+    {
+      "epoch": 0.008066917418221077,
+      "grad_norm": 0.31266510486602783,
+      "learning_rate": 2.483461169100576e-05,
+      "loss": 0.03945198655128479,
+      "step": 31060
+    },
+    {
+      "epoch": 0.008359197034823289,
+      "grad_norm": 0.3112914562225342,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 0.034811544418334964,
+      "step": 31070
+    },
+    {
+      "epoch": 0.008651476651425502,
+      "grad_norm": 0.41006046533584595,
+      "learning_rate": 2.4728449704635914e-05,
+      "loss": 0.03897059261798859,
+      "step": 31080
+    },
+    {
+      "epoch": 0.008943756268027715,
+      "grad_norm": 0.5569542050361633,
+      "learning_rate": 2.467544195857474e-05,
+      "loss": 0.03799519240856171,
+      "step": 31090
+    },
+    {
+      "epoch": 0.009236035884629928,
+      "grad_norm": 0.2839227616786957,
+      "learning_rate": 2.462248308980152e-05,
+      "loss": 0.03459154665470123,
+      "step": 31100
+    },
+    {
+      "epoch": 0.009528315501232142,
+      "grad_norm": 0.47007253766059875,
+      "learning_rate": 2.4569573132680622e-05,
+      "loss": 0.03303917646408081,
+      "step": 31110
+    },
+    {
+      "epoch": 0.009820595117834355,
+      "grad_norm": 0.4322066009044647,
+      "learning_rate": 2.4516712121544794e-05,
+      "loss": 0.039104807376861575,
+      "step": 31120
+    },
+    {
+      "epoch": 0.010112874734436566,
+      "grad_norm": 0.4930959939956665,
+      "learning_rate": 2.446390009069499e-05,
+      "loss": 0.040959838032722476,
+      "step": 31130
+    },
+    {
+      "epoch": 0.01040515435103878,
+      "grad_norm": 0.2450665980577469,
+      "learning_rate": 2.441113707440027e-05,
+      "loss": 0.02899554967880249,
+      "step": 31140
+    },
+    {
+      "epoch": 0.010697433967640993,
+      "grad_norm": 0.3187698721885681,
+      "learning_rate": 2.43584231068981e-05,
+      "loss": 0.041855734586715695,
+      "step": 31150
+    },
+    {
+      "epoch": 0.010989713584243206,
+      "grad_norm": 0.2259834259748459,
+      "learning_rate": 2.4305758222393927e-05,
+      "loss": 0.038677138090133664,
+      "step": 31160
+    },
+    {
+      "epoch": 0.011281993200845419,
+      "grad_norm": 0.6648458242416382,
+      "learning_rate": 2.425314245506144e-05,
+      "loss": 0.03533812761306763,
+      "step": 31170
+    },
+    {
+      "epoch": 0.011574272817447632,
+      "grad_norm": 0.35770630836486816,
+      "learning_rate": 2.420057583904246e-05,
+      "loss": 0.04079995453357697,
+      "step": 31180
+    },
+    {
+      "epoch": 0.011866552434049845,
+      "grad_norm": 0.41780030727386475,
+      "learning_rate": 2.414805840844686e-05,
+      "loss": 0.03529571890830994,
+      "step": 31190
+    },
+    {
+      "epoch": 0.012158832050652057,
+      "grad_norm": 0.4948868453502655,
+      "learning_rate": 2.4095590197352645e-05,
+      "loss": 0.030538800358772277,
+      "step": 31200
+    },
+    {
+      "epoch": 0.01245111166725427,
+      "grad_norm": 0.4128345549106598,
+      "learning_rate": 2.404317123980586e-05,
+      "loss": 0.037554800510406494,
+      "step": 31210
+    },
+    {
+      "epoch": 0.012743391283856483,
+      "grad_norm": 0.45742449164390564,
+      "learning_rate": 2.3990801569820643e-05,
+      "loss": 0.0345581442117691,
+      "step": 31220
+    },
+    {
+      "epoch": 0.013035670900458696,
+      "grad_norm": 0.5399410724639893,
+      "learning_rate": 2.3938481221379026e-05,
+      "loss": 0.026771077513694765,
+      "step": 31230
+    },
+    {
+      "epoch": 0.01332795051706091,
+      "grad_norm": 0.5230187177658081,
+      "learning_rate": 2.388621022843117e-05,
+      "loss": 0.04226500391960144,
+      "step": 31240
+    },
+    {
+      "epoch": 0.013620230133663123,
+      "grad_norm": 0.47975000739097595,
+      "learning_rate": 2.3833988624895165e-05,
+      "loss": 0.046533650159835814,
+      "step": 31250
+    },
+    {
+      "epoch": 0.013912509750265336,
+      "grad_norm": 0.4004831910133362,
+      "learning_rate": 2.378181644465699e-05,
+      "loss": 0.029512953758239747,
+      "step": 31260
+    },
+    {
+      "epoch": 0.014204789366867547,
+      "grad_norm": 0.5633212327957153,
+      "learning_rate": 2.3729693721570635e-05,
+      "loss": 0.029715001583099365,
+      "step": 31270
+    },
+    {
+      "epoch": 0.01449706898346976,
+      "grad_norm": 0.32318446040153503,
+      "learning_rate": 2.367762048945801e-05,
+      "loss": 0.03455106317996979,
+      "step": 31280
+    },
+    {
+      "epoch": 0.014789348600071974,
+      "grad_norm": 0.5322811603546143,
+      "learning_rate": 2.3625596782108782e-05,
+      "loss": 0.03868768215179443,
+      "step": 31290
+    },
+    {
+      "epoch": 0.015081628216674187,
+      "grad_norm": 0.40740758180618286,
+      "learning_rate": 2.3573622633280623e-05,
+      "loss": 0.03716544210910797,
+      "step": 31300
+    },
+    {
+      "epoch": 0.0153739078332764,
+      "grad_norm": 0.6749125719070435,
+      "learning_rate": 2.3521698076698996e-05,
+      "loss": 0.04570290744304657,
+      "step": 31310
+    },
+    {
+      "epoch": 0.01566618744987861,
+      "grad_norm": 0.4723735749721527,
+      "learning_rate": 2.346982314605719e-05,
+      "loss": 0.03807953596115112,
+      "step": 31320
+    },
+    {
+      "epoch": 0.015958467066480825,
+      "grad_norm": 0.611591100692749,
+      "learning_rate": 2.341799787501625e-05,
+      "loss": 0.04045450687408447,
+      "step": 31330
+    },
+    {
+      "epoch": 0.016250746683083038,
+      "grad_norm": 0.4128269851207733,
+      "learning_rate": 2.3366222297205044e-05,
+      "loss": 0.0468936562538147,
+      "step": 31340
+    },
+    {
+      "epoch": 0.01654302629968525,
+      "grad_norm": 0.5142440795898438,
+      "learning_rate": 2.3314496446220213e-05,
+      "loss": 0.04724392294883728,
+      "step": 31350
+    },
+    {
+      "epoch": 0.016835305916287464,
+      "grad_norm": 0.48125791549682617,
+      "learning_rate": 2.3262820355626026e-05,
+      "loss": 0.038220140337944034,
+      "step": 31360
+    },
+    {
+      "epoch": 0.017127585532889678,
+      "grad_norm": 0.55055832862854,
+      "learning_rate": 2.3211194058954644e-05,
+      "loss": 0.04546545743942261,
+      "step": 31370
+    },
+    {
+      "epoch": 0.01741986514949189,
+      "grad_norm": 0.3137090802192688,
+      "learning_rate": 2.315961758970575e-05,
+      "loss": 0.04613784551620483,
+      "step": 31380
+    },
+    {
+      "epoch": 0.017712144766094104,
+      "grad_norm": 0.7191341519355774,
+      "learning_rate": 2.3108090981346697e-05,
+      "loss": 0.032217130064964294,
+      "step": 31390
+    },
+    {
+      "epoch": 0.018004424382696317,
+      "grad_norm": 0.20661820471286774,
+      "learning_rate": 2.3056614267312672e-05,
+      "loss": 0.038537850975990294,
+      "step": 31400
+    },
+    {
+      "epoch": 0.01829670399929853,
+      "grad_norm": 0.5620737075805664,
+      "learning_rate": 2.3005187481006252e-05,
+      "loss": 0.03490525782108307,
+      "step": 31410
+    },
+    {
+      "epoch": 0.018588983615900743,
+      "grad_norm": 0.5708649158477783,
+      "learning_rate": 2.2953810655797747e-05,
+      "loss": 0.04009975790977478,
+      "step": 31420
+    },
+    {
+      "epoch": 0.018881263232502953,
+      "grad_norm": 0.4738178849220276,
+      "learning_rate": 2.2902483825025046e-05,
+      "loss": 0.038276398181915285,
+      "step": 31430
+    },
+    {
+      "epoch": 0.019173542849105166,
+      "grad_norm": 0.475565105676651,
+      "learning_rate": 2.285120702199357e-05,
+      "loss": 0.04238824844360352,
+      "step": 31440
+    },
+    {
+      "epoch": 0.01946582246570738,
+      "grad_norm": 0.33945712447166443,
+      "learning_rate": 2.279998027997624e-05,
+      "loss": 0.04524965882301331,
+      "step": 31450
+    },
+    {
+      "epoch": 0.019758102082309593,
+      "grad_norm": 0.5164212584495544,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 0.042377158999443054,
+      "step": 31460
+    },
+    {
+      "epoch": 0.020050381698911806,
+      "grad_norm": 0.5562757849693298,
+      "learning_rate": 2.269767711191352e-05,
+      "loss": 0.03618437051773071,
+      "step": 31470
+    },
+    {
+      "epoch": 0.02034266131551402,
+      "grad_norm": 0.29933780431747437,
+      "learning_rate": 2.2646600752251513e-05,
+      "loss": 0.037383508682250974,
+      "step": 31480
+    },
+    {
+      "epoch": 0.020634940932116232,
+      "grad_norm": 0.5189642906188965,
+      "learning_rate": 2.259557458637047e-05,
+      "loss": 0.028320759534835815,
+      "step": 31490
+    },
+    {
+      "epoch": 0.020927220548718446,
+      "grad_norm": 0.3747444748878479,
+      "learning_rate": 2.2544598647380745e-05,
+      "loss": 0.03934798538684845,
+      "step": 31500
+    },
+    {
+      "epoch": 0.02121950016532066,
+      "grad_norm": 0.4220009744167328,
+      "learning_rate": 2.2493672968360003e-05,
+      "loss": 0.029090550541877747,
+      "step": 31510
+    },
+    {
+      "epoch": 0.021511779781922872,
+      "grad_norm": 0.5097419023513794,
+      "learning_rate": 2.2442797582353393e-05,
+      "loss": 0.04183672368526459,
+      "step": 31520
+    },
+    {
+      "epoch": 0.021804059398525085,
+      "grad_norm": 0.5790647864341736,
+      "learning_rate": 2.2391972522373406e-05,
+      "loss": 0.04003177583217621,
+      "step": 31530
+    },
+    {
+      "epoch": 0.0220963390151273,
+      "grad_norm": 0.5781935453414917,
+      "learning_rate": 2.2341197821399905e-05,
+      "loss": 0.0413451611995697,
+      "step": 31540
+    },
+    {
+      "epoch": 0.02238861863172951,
+      "grad_norm": 0.48431098461151123,
+      "learning_rate": 2.2290473512379973e-05,
+      "loss": 0.0268336147069931,
+      "step": 31550
+    },
+    {
+      "epoch": 0.02268089824833172,
+      "grad_norm": 0.5021381974220276,
+      "learning_rate": 2.223979962822811e-05,
+      "loss": 0.027185022830963135,
+      "step": 31560
+    },
+    {
+      "epoch": 0.022973177864933934,
+      "grad_norm": 0.3669155538082123,
+      "learning_rate": 2.2189176201826056e-05,
+      "loss": 0.033669310808181765,
+      "step": 31570
+    },
+    {
+      "epoch": 0.023265457481536148,
+      "grad_norm": 0.469275563955307,
+      "learning_rate": 2.213860326602273e-05,
+      "loss": 0.04097016751766205,
+      "step": 31580
+    },
+    {
+      "epoch": 0.02355773709813836,
+      "grad_norm": 0.4300369918346405,
+      "learning_rate": 2.2088080853634473e-05,
+      "loss": 0.03791420757770538,
+      "step": 31590
+    },
+    {
+      "epoch": 0.023850016714740574,
+      "grad_norm": 0.28294146060943604,
+      "learning_rate": 2.203760899744467e-05,
+      "loss": 0.037059289216995236,
+      "step": 31600
+    },
+    {
+      "epoch": 0.024142296331342787,
+      "grad_norm": 0.3775625228881836,
+      "learning_rate": 2.198718773020392e-05,
+      "loss": 0.04303731024265289,
+      "step": 31610
+    },
+    {
+      "epoch": 0.024434575947945,
+      "grad_norm": 0.33136215806007385,
+      "learning_rate": 2.1936817084630114e-05,
+      "loss": 0.03654118478298187,
+      "step": 31620
+    },
+    {
+      "epoch": 0.024726855564547214,
+      "grad_norm": 0.3154297471046448,
+      "learning_rate": 2.1886497093408164e-05,
+      "loss": 0.031285032629966736,
+      "step": 31630
+    },
+    {
+      "epoch": 0.025019135181149427,
+      "grad_norm": 0.554410457611084,
+      "learning_rate": 2.183622778919019e-05,
+      "loss": 0.048117217421531674,
+      "step": 31640
+    },
+    {
+      "epoch": 0.02531141479775164,
+      "grad_norm": 0.31969505548477173,
+      "learning_rate": 2.178600920459538e-05,
+      "loss": 0.03263488113880157,
+      "step": 31650
+    },
+    {
+      "epoch": 0.025603694414353853,
+      "grad_norm": 0.3681676387786865,
+      "learning_rate": 2.1735841372210074e-05,
+      "loss": 0.03770647346973419,
+      "step": 31660
+    },
+    {
+      "epoch": 0.025895974030956066,
+      "grad_norm": 0.6038069128990173,
+      "learning_rate": 2.1685724324587576e-05,
+      "loss": 0.031424093246459964,
+      "step": 31670
+    },
+    {
+      "epoch": 0.02618825364755828,
+      "grad_norm": 0.735026478767395,
+      "learning_rate": 2.163565809424831e-05,
+      "loss": 0.04351509511470795,
+      "step": 31680
+    },
+    {
+      "epoch": 0.02648053326416049,
+      "grad_norm": 0.24993659555912018,
+      "learning_rate": 2.1585642713679753e-05,
+      "loss": 0.03310315608978272,
+      "step": 31690
+    },
+    {
+      "epoch": 0.026772812880762702,
+      "grad_norm": 0.4668404757976532,
+      "learning_rate": 2.1535678215336274e-05,
+      "loss": 0.0275485098361969,
+      "step": 31700
+    },
+    {
+      "epoch": 0.027065092497364916,
+      "grad_norm": 0.45025038719177246,
+      "learning_rate": 2.1485764631639328e-05,
+      "loss": 0.03712981939315796,
+      "step": 31710
+    },
+    {
+      "epoch": 0.02735737211396713,
+      "grad_norm": 0.5361169576644897,
+      "learning_rate": 2.1435901994977326e-05,
+      "loss": 0.03676751852035522,
+      "step": 31720
+    },
+    {
+      "epoch": 0.027649651730569342,
+      "grad_norm": 0.2743486762046814,
+      "learning_rate": 2.1386090337705522e-05,
+      "loss": 0.05056242942810059,
+      "step": 31730
+    },
+    {
+      "epoch": 0.027941931347171555,
+      "grad_norm": 0.246163010597229,
+      "learning_rate": 2.1336329692146207e-05,
+      "loss": 0.03135035634040832,
+      "step": 31740
+    },
+    {
+      "epoch": 0.02823421096377377,
+      "grad_norm": 0.4383198618888855,
+      "learning_rate": 2.1286620090588515e-05,
+      "loss": 0.04032829999923706,
+      "step": 31750
+    },
+    {
+      "epoch": 0.02852649058037598,
+      "grad_norm": 0.44198882579803467,
+      "learning_rate": 2.1236961565288495e-05,
+      "loss": 0.04002164304256439,
+      "step": 31760
+    },
+    {
+      "epoch": 0.028818770196978195,
+      "grad_norm": 0.442519873380661,
+      "learning_rate": 2.1187354148468976e-05,
+      "loss": 0.038885998725891116,
+      "step": 31770
+    },
+    {
+      "epoch": 0.029111049813580408,
+      "grad_norm": 0.46529823541641235,
+      "learning_rate": 2.1137797872319687e-05,
+      "loss": 0.03308583199977875,
+      "step": 31780
+    },
+    {
+      "epoch": 0.02940332943018262,
+      "grad_norm": 0.4279099404811859,
+      "learning_rate": 2.1088292768997185e-05,
+      "loss": 0.04584682881832123,
+      "step": 31790
+    },
+    {
+      "epoch": 0.029695609046784834,
+      "grad_norm": 0.29172927141189575,
+      "learning_rate": 2.1038838870624732e-05,
+      "loss": 0.05990554094314575,
+      "step": 31800
+    },
+    {
+      "epoch": 0.029987888663387047,
+      "grad_norm": 0.3993830382823944,
+      "learning_rate": 2.098943620929251e-05,
+      "loss": 0.056998652219772336,
+      "step": 31810
+    },
+    {
+      "epoch": 0.030280168279989257,
+      "grad_norm": 0.6501842737197876,
+      "learning_rate": 2.0940084817057304e-05,
+      "loss": 0.06923367977142333,
+      "step": 31820
+    },
+    {
+      "epoch": 0.03057244789659147,
+      "grad_norm": 0.3089783787727356,
+      "learning_rate": 2.0890784725942726e-05,
+      "loss": 0.04636120200157166,
+      "step": 31830
+    },
+    {
+      "epoch": 0.030864727513193684,
+      "grad_norm": 0.31864693760871887,
+      "learning_rate": 2.0841535967939073e-05,
+      "loss": 0.04129012227058411,
+      "step": 31840
+    },
+    {
+      "epoch": 0.031157007129795897,
+      "grad_norm": 0.5650334358215332,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 0.05147960186004639,
+      "step": 31850
+    },
+    {
+      "epoch": 0.03144928674639811,
+      "grad_norm": 0.4118666350841522,
+      "learning_rate": 2.074319257905909e-05,
+      "loss": 0.04803649485111237,
+      "step": 31860
+    },
+    {
+      "epoch": 0.03174156636300032,
+      "grad_norm": 0.38670656085014343,
+      "learning_rate": 2.0694098011996755e-05,
+      "loss": 0.048144778609275816,
+      "step": 31870
+    },
+    {
+      "epoch": 0.032033845979602536,
+      "grad_norm": 0.387030690908432,
+      "learning_rate": 2.0645054905673178e-05,
+      "loss": 0.03306574821472168,
+      "step": 31880
+    },
+    {
+      "epoch": 0.03232612559620475,
+      "grad_norm": 0.6669332385063171,
+      "learning_rate": 2.0596063291911938e-05,
+      "loss": 0.04260598123073578,
+      "step": 31890
+    },
+    {
+      "epoch": 0.03261840521280696,
+      "grad_norm": 0.48304450511932373,
+      "learning_rate": 2.054712320250314e-05,
+      "loss": 0.06114354729652405,
+      "step": 31900
+    },
+    {
+      "epoch": 0.032910684829409176,
+      "grad_norm": 0.47821900248527527,
+      "learning_rate": 2.049823466920352e-05,
+      "loss": 0.04245083630084991,
+      "step": 31910
+    },
+    {
+      "epoch": 0.03320296444601139,
+      "grad_norm": 0.3573174476623535,
+      "learning_rate": 2.0449397723736242e-05,
+      "loss": 0.044038066267967226,
+      "step": 31920
+    },
+    {
+      "epoch": 0.0334952440626136,
+      "grad_norm": 0.5889163613319397,
+      "learning_rate": 2.0400612397791153e-05,
+      "loss": 0.0486553281545639,
+      "step": 31930
+    },
+    {
+      "epoch": 0.033787523679215815,
+      "grad_norm": 0.5002954602241516,
+      "learning_rate": 2.0351878723024485e-05,
+      "loss": 0.061679935455322264,
+      "step": 31940
+    },
+    {
+      "epoch": 0.03407980329581803,
+      "grad_norm": 0.4039132595062256,
+      "learning_rate": 2.0303196731058925e-05,
+      "loss": 0.05008590221405029,
+      "step": 31950
+    },
+    {
+      "epoch": 0.03437208291242024,
+      "grad_norm": 0.4943726360797882,
+      "learning_rate": 2.0254566453483813e-05,
+      "loss": 0.03969822525978088,
+      "step": 31960
+    },
+    {
+      "epoch": 0.034664362529022455,
+      "grad_norm": 0.44370079040527344,
+      "learning_rate": 2.020598792185472e-05,
+      "loss": 0.055040568113327026,
+      "step": 31970
+    },
+    {
+      "epoch": 0.03495664214562467,
+      "grad_norm": 0.5418318510055542,
+      "learning_rate": 2.0157461167693748e-05,
+      "loss": 0.05637938380241394,
+      "step": 31980
+    },
+    {
+      "epoch": 0.03524892176222688,
+      "grad_norm": 0.34446701407432556,
+      "learning_rate": 2.0108986222489424e-05,
+      "loss": 0.03586422204971314,
+      "step": 31990
+    },
+    {
+      "epoch": 0.035541201378829095,
+      "grad_norm": 0.3788282573223114,
+      "learning_rate": 2.006056311769654e-05,
+      "loss": 0.060768669843673705,
+      "step": 32000
+    },
+    {
+      "epoch": 0.00026305165494199163,
+      "grad_norm": 0.4973739683628082,
+      "learning_rate": 2.0012191884736366e-05,
+      "loss": 0.05248960521486071,
+      "step": 32010
+    },
+    {
+      "epoch": 0.0005553312715442046,
+      "grad_norm": 0.26003333926200867,
+      "learning_rate": 1.9963872554996445e-05,
+      "loss": 0.03776390850543976,
+      "step": 32020
+    },
+    {
+      "epoch": 0.0008476108881464175,
+      "grad_norm": 0.17580893635749817,
+      "learning_rate": 1.991560515983073e-05,
+      "loss": 0.02353486865758896,
+      "step": 32030
+    },
+    {
+      "epoch": 0.0011398905047486304,
+      "grad_norm": 0.24444793164730072,
+      "learning_rate": 1.9867389730559326e-05,
+      "loss": 0.022889670729637147,
+      "step": 32040
+    },
+    {
+      "epoch": 0.0014321701213508434,
+      "grad_norm": 0.24159517884254456,
+      "learning_rate": 1.981922629846874e-05,
+      "loss": 0.018500922620296477,
+      "step": 32050
+    },
+    {
+      "epoch": 0.0017244497379530562,
+      "grad_norm": 0.3611849844455719,
+      "learning_rate": 1.9771114894811727e-05,
+      "loss": 0.026837033033370972,
+      "step": 32060
+    },
+    {
+      "epoch": 0.0020167293545552693,
+      "grad_norm": 0.23240919411182404,
+      "learning_rate": 1.97230555508072e-05,
+      "loss": 0.02514037787914276,
+      "step": 32070
+    },
+    {
+      "epoch": 0.002309008971157482,
+      "grad_norm": 0.22325488924980164,
+      "learning_rate": 1.9675048297640363e-05,
+      "loss": 0.016411012411117552,
+      "step": 32080
+    },
+    {
+      "epoch": 0.002601288587759695,
+      "grad_norm": 0.29972967505455017,
+      "learning_rate": 1.962709316646265e-05,
+      "loss": 0.025227734446525575,
+      "step": 32090
+    },
+    {
+      "epoch": 0.002893568204361908,
+      "grad_norm": 0.3807256519794464,
+      "learning_rate": 1.957919018839155e-05,
+      "loss": 0.01727556884288788,
+      "step": 32100
+    },
+    {
+      "epoch": 0.003185847820964121,
+      "grad_norm": 0.4461345076560974,
+      "learning_rate": 1.9531339394510827e-05,
+      "loss": 0.02147582471370697,
+      "step": 32110
+    },
+    {
+      "epoch": 0.003478127437566334,
+      "grad_norm": 0.27030149102211,
+      "learning_rate": 1.9483540815870327e-05,
+      "loss": 0.02091405391693115,
+      "step": 32120
+    },
+    {
+      "epoch": 0.0037704070541685467,
+      "grad_norm": 0.30214038491249084,
+      "learning_rate": 1.943579448348607e-05,
+      "loss": 0.020106695592403412,
+      "step": 32130
+    },
+    {
+      "epoch": 0.0040626866707707595,
+      "grad_norm": 0.4095200002193451,
+      "learning_rate": 1.938810042834005e-05,
+      "loss": 0.02843693196773529,
+      "step": 32140
+    },
+    {
+      "epoch": 0.004354966287372973,
+      "grad_norm": 0.2516385316848755,
+      "learning_rate": 1.9340458681380514e-05,
+      "loss": 0.01896345764398575,
+      "step": 32150
+    },
+    {
+      "epoch": 0.004647245903975186,
+      "grad_norm": 0.22221162915229797,
+      "learning_rate": 1.929286927352164e-05,
+      "loss": 0.014323014020919799,
+      "step": 32160
+    },
+    {
+      "epoch": 0.004939525520577398,
+      "grad_norm": 0.19755686819553375,
+      "learning_rate": 1.9245332235643608e-05,
+      "loss": 0.014702646434307099,
+      "step": 32170
+    },
+    {
+      "epoch": 0.005231805137179611,
+      "grad_norm": 0.20292168855667114,
+      "learning_rate": 1.919784759859281e-05,
+      "loss": 0.014039400219917297,
+      "step": 32180
+    },
+    {
+      "epoch": 0.005524084753781825,
+      "grad_norm": 0.11288988590240479,
+      "learning_rate": 1.915041539318142e-05,
+      "loss": 0.019446593523025513,
+      "step": 32190
+    },
+    {
+      "epoch": 0.005816364370384037,
+      "grad_norm": 0.38899025321006775,
+      "learning_rate": 1.9103035650187707e-05,
+      "loss": 0.016867619752883912,
+      "step": 32200
+    },
+    {
+      "epoch": 0.00610864398698625,
+      "grad_norm": 0.13710832595825195,
+      "learning_rate": 1.90557084003559e-05,
+      "loss": 0.018804381787776946,
+      "step": 32210
+    },
+    {
+      "epoch": 0.006400923603588463,
+      "grad_norm": 0.4738372266292572,
+      "learning_rate": 1.9008433674396085e-05,
+      "loss": 0.019061791896820068,
+      "step": 32220
+    },
+    {
+      "epoch": 0.006693203220190676,
+      "grad_norm": 0.18087215721607208,
+      "learning_rate": 1.8961211502984357e-05,
+      "loss": 0.016344700753688813,
+      "step": 32230
+    },
+    {
+      "epoch": 0.006985482836792889,
+      "grad_norm": 0.36321917176246643,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 0.01985886096954346,
+      "step": 32240
+    },
+    {
+      "epoch": 0.007277762453395102,
+      "grad_norm": 0.2187732458114624,
+      "learning_rate": 1.8866924946338825e-05,
+      "loss": 0.01938354820013046,
+      "step": 32250
+    },
+    {
+      "epoch": 0.007570042069997314,
+      "grad_norm": 0.10396921634674072,
+      "learning_rate": 1.8819860622286534e-05,
+      "loss": 0.011649166792631149,
+      "step": 32260
+    },
+    {
+      "epoch": 0.007862321686599527,
+      "grad_norm": 0.4477599859237671,
+      "learning_rate": 1.877284897514533e-05,
+      "loss": 0.016419678926467896,
+      "step": 32270
+    },
+    {
+      "epoch": 0.00815460130320174,
+      "grad_norm": 0.22246281802654266,
+      "learning_rate": 1.872589003542058e-05,
+      "loss": 0.014308787882328033,
+      "step": 32280
+    },
+    {
+      "epoch": 0.008446880919803954,
+      "grad_norm": 0.3079025149345398,
+      "learning_rate": 1.8678983833583374e-05,
+      "loss": 0.01560411900281906,
+      "step": 32290
+    },
+    {
+      "epoch": 0.008739160536406167,
+      "grad_norm": 0.1909436583518982,
+      "learning_rate": 1.863213040007067e-05,
+      "loss": 0.016433225572109224,
+      "step": 32300
+    },
+    {
+      "epoch": 0.009031440153008378,
+      "grad_norm": 0.3250446319580078,
+      "learning_rate": 1.8585329765285187e-05,
+      "loss": 0.01707385927438736,
+      "step": 32310
+    },
+    {
+      "epoch": 0.009323719769610592,
+      "grad_norm": 0.32244673371315,
+      "learning_rate": 1.853858195959528e-05,
+      "loss": 0.015416412055492401,
+      "step": 32320
+    },
+    {
+      "epoch": 0.009615999386212805,
+      "grad_norm": 0.24997259676456451,
+      "learning_rate": 1.8491887013335117e-05,
+      "loss": 0.010679328441619873,
+      "step": 32330
+    },
+    {
+      "epoch": 0.009908279002815018,
+      "grad_norm": 0.22225984930992126,
+      "learning_rate": 1.8445244956804563e-05,
+      "loss": 0.017385059595108034,
+      "step": 32340
+    },
+    {
+      "epoch": 0.010200558619417231,
+      "grad_norm": 0.27568697929382324,
+      "learning_rate": 1.839865582026915e-05,
+      "loss": 0.016467107832431792,
+      "step": 32350
+    },
+    {
+      "epoch": 0.010492838236019444,
+      "grad_norm": 0.2669641375541687,
+      "learning_rate": 1.835211963396002e-05,
+      "loss": 0.011760076135396957,
+      "step": 32360
+    },
+    {
+      "epoch": 0.010785117852621658,
+      "grad_norm": 0.2670864760875702,
+      "learning_rate": 1.8305636428074014e-05,
+      "loss": 0.018715476989746092,
+      "step": 32370
+    },
+    {
+      "epoch": 0.011077397469223869,
+      "grad_norm": 0.17398963868618011,
+      "learning_rate": 1.8259206232773618e-05,
+      "loss": 0.020170696079730988,
+      "step": 32380
+    },
+    {
+      "epoch": 0.011369677085826082,
+      "grad_norm": 0.48624247312545776,
+      "learning_rate": 1.8212829078186798e-05,
+      "loss": 0.018696312606334687,
+      "step": 32390
+    },
+    {
+      "epoch": 0.011661956702428295,
+      "grad_norm": 0.4576359987258911,
+      "learning_rate": 1.8166504994407285e-05,
+      "loss": 0.020144809782505036,
+      "step": 32400
+    },
+    {
+      "epoch": 0.011954236319030509,
+      "grad_norm": 0.2963448464870453,
+      "learning_rate": 1.812023401149421e-05,
+      "loss": 0.011683215200901032,
+      "step": 32410
+    },
+    {
+      "epoch": 0.012246515935632722,
+      "grad_norm": 0.19167789816856384,
+      "learning_rate": 1.8074016159472317e-05,
+      "loss": 0.01231534332036972,
+      "step": 32420
+    },
+    {
+      "epoch": 0.012538795552234935,
+      "grad_norm": 0.21664723753929138,
+      "learning_rate": 1.8027851468331923e-05,
+      "loss": 0.014816097915172577,
+      "step": 32430
+    },
+    {
+      "epoch": 0.012831075168837146,
+      "grad_norm": 0.18707649409770966,
+      "learning_rate": 1.798173996802872e-05,
+      "loss": 0.014465142786502839,
+      "step": 32440
+    },
+    {
+      "epoch": 0.01312335478543936,
+      "grad_norm": 0.4082675278186798,
+      "learning_rate": 1.7935681688483985e-05,
+      "loss": 0.01341407150030136,
+      "step": 32450
+    },
+    {
+      "epoch": 0.013415634402041573,
+      "grad_norm": 0.15097574889659882,
+      "learning_rate": 1.7889676659584455e-05,
+      "loss": 0.016197820007801057,
+      "step": 32460
+    },
+    {
+      "epoch": 0.013707914018643786,
+      "grad_norm": 0.22646082937717438,
+      "learning_rate": 1.7843724911182292e-05,
+      "loss": 0.020954221487045288,
+      "step": 32470
+    },
+    {
+      "epoch": 0.014000193635246,
+      "grad_norm": 0.33519256114959717,
+      "learning_rate": 1.779782647309506e-05,
+      "loss": 0.014750099182128907,
+      "step": 32480
+    },
+    {
+      "epoch": 0.014292473251848212,
+      "grad_norm": 0.15925352275371552,
+      "learning_rate": 1.7751981375105776e-05,
+      "loss": 0.010437330603599549,
+      "step": 32490
+    },
+    {
+      "epoch": 0.014584752868450426,
+      "grad_norm": 0.22180332243442535,
+      "learning_rate": 1.7706189646962847e-05,
+      "loss": 0.01209150329232216,
+      "step": 32500
+    },
+    {
+      "epoch": 0.014877032485052637,
+      "grad_norm": 0.20685526728630066,
+      "learning_rate": 1.7660451318379977e-05,
+      "loss": 0.01703527420759201,
+      "step": 32510
+    },
+    {
+      "epoch": 0.01516931210165485,
+      "grad_norm": 0.39896267652511597,
+      "learning_rate": 1.76147664190363e-05,
+      "loss": 0.020233039557933808,
+      "step": 32520
+    },
+    {
+      "epoch": 0.015461591718257063,
+      "grad_norm": 0.1432127207517624,
+      "learning_rate": 1.7569134978576273e-05,
+      "loss": 0.018171924352645873,
+      "step": 32530
+    },
+    {
+      "epoch": 0.015753871334859275,
+      "grad_norm": 0.28434422612190247,
+      "learning_rate": 1.7523557026609593e-05,
+      "loss": 0.013687740266323089,
+      "step": 32540
+    },
+    {
+      "epoch": 0.016046150951461488,
+      "grad_norm": 0.36812835931777954,
+      "learning_rate": 1.7478032592711334e-05,
+      "loss": 0.020922017097473145,
+      "step": 32550
+    },
+    {
+      "epoch": 0.0163384305680637,
+      "grad_norm": 0.23475459218025208,
+      "learning_rate": 1.743256170642179e-05,
+      "loss": 0.021994376182556154,
+      "step": 32560
+    },
+    {
+      "epoch": 0.016630710184665914,
+      "grad_norm": 0.31185635924339294,
+      "learning_rate": 1.738714439724656e-05,
+      "loss": 0.01730121225118637,
+      "step": 32570
+    },
+    {
+      "epoch": 0.016922989801268128,
+      "grad_norm": 0.4498734772205353,
+      "learning_rate": 1.7341780694656395e-05,
+      "loss": 0.017756085097789764,
+      "step": 32580
+    },
+    {
+      "epoch": 0.01721526941787034,
+      "grad_norm": 0.21492226421833038,
+      "learning_rate": 1.7296470628087315e-05,
+      "loss": 0.019565393030643464,
+      "step": 32590
+    },
+    {
+      "epoch": 0.017507549034472554,
+      "grad_norm": 0.17935171723365784,
+      "learning_rate": 1.725121422694058e-05,
+      "loss": 0.018504330515861513,
+      "step": 32600
+    },
+    {
+      "epoch": 0.017799828651074767,
+      "grad_norm": 0.41158759593963623,
+      "learning_rate": 1.72060115205825e-05,
+      "loss": 0.013376963138580323,
+      "step": 32610
+    },
+    {
+      "epoch": 0.01809210826767698,
+      "grad_norm": 0.3655569851398468,
+      "learning_rate": 1.71608625383447e-05,
+      "loss": 0.01625903844833374,
+      "step": 32620
+    },
+    {
+      "epoch": 0.018384387884279194,
+      "grad_norm": 0.38022857904434204,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 0.013181254267692566,
+      "step": 32630
+    },
+    {
+      "epoch": 0.018676667500881407,
+      "grad_norm": 0.4689427614212036,
+      "learning_rate": 1.7070725863381653e-05,
+      "loss": 0.018557126820087432,
+      "step": 32640
+    },
+    {
+      "epoch": 0.01896894711748362,
+      "grad_norm": 0.12309762835502625,
+      "learning_rate": 1.702573822914515e-05,
+      "loss": 0.016794680058956145,
+      "step": 32650
+    },
+    {
+      "epoch": 0.019261226734085833,
+      "grad_norm": 0.34093308448791504,
+      "learning_rate": 1.6980804436006248e-05,
+      "loss": 0.016108806431293487,
+      "step": 32660
+    },
+    {
+      "epoch": 0.019553506350688043,
+      "grad_norm": 0.3414905369281769,
+      "learning_rate": 1.6935924513122026e-05,
+      "loss": 0.021218304336071015,
+      "step": 32670
+    },
+    {
+      "epoch": 0.019845785967290256,
+      "grad_norm": 0.3480222821235657,
+      "learning_rate": 1.6891098489614577e-05,
+      "loss": 0.01595773845911026,
+      "step": 32680
+    },
+    {
+      "epoch": 0.02013806558389247,
+      "grad_norm": 0.3328452408313751,
+      "learning_rate": 1.6846326394571033e-05,
+      "loss": 0.01621834933757782,
+      "step": 32690
+    },
+    {
+      "epoch": 0.020430345200494682,
+      "grad_norm": 0.26909396052360535,
+      "learning_rate": 1.680160825704349e-05,
+      "loss": 0.011456964164972305,
+      "step": 32700
+    },
+    {
+      "epoch": 0.020722624817096896,
+      "grad_norm": 0.3564660847187042,
+      "learning_rate": 1.6756944106049066e-05,
+      "loss": 0.016938157379627228,
+      "step": 32710
+    },
+    {
+      "epoch": 0.02101490443369911,
+      "grad_norm": 0.16497020423412323,
+      "learning_rate": 1.6712333970569883e-05,
+      "loss": 0.013434399664402009,
+      "step": 32720
+    },
+    {
+      "epoch": 0.021307184050301322,
+      "grad_norm": 0.28860923647880554,
+      "learning_rate": 1.6667777879552928e-05,
+      "loss": 0.012481187283992768,
+      "step": 32730
+    },
+    {
+      "epoch": 0.021599463666903535,
+      "grad_norm": 0.4379538893699646,
+      "learning_rate": 1.6623275861910172e-05,
+      "loss": 0.0222562775015831,
+      "step": 32740
+    },
+    {
+      "epoch": 0.02189174328350575,
+      "grad_norm": 0.3677423298358917,
+      "learning_rate": 1.6578827946518527e-05,
+      "loss": 0.01772304028272629,
+      "step": 32750
+    },
+    {
+      "epoch": 0.02218402290010796,
+      "grad_norm": 0.35193854570388794,
+      "learning_rate": 1.6534434162219727e-05,
+      "loss": 0.01352870613336563,
+      "step": 32760
+    },
+    {
+      "epoch": 0.022476302516710175,
+      "grad_norm": 0.259731650352478,
+      "learning_rate": 1.6490094537820412e-05,
+      "loss": 0.012437423318624496,
+      "step": 32770
+    },
+    {
+      "epoch": 0.022768582133312388,
+      "grad_norm": 0.25573334097862244,
+      "learning_rate": 1.644580910209209e-05,
+      "loss": 0.010144352912902832,
+      "step": 32780
+    },
+    {
+      "epoch": 0.0230608617499146,
+      "grad_norm": 0.2369808554649353,
+      "learning_rate": 1.6401577883771147e-05,
+      "loss": 0.01704389750957489,
+      "step": 32790
+    },
+    {
+      "epoch": 0.02335314136651681,
+      "grad_norm": 0.26524466276168823,
+      "learning_rate": 1.6357400911558674e-05,
+      "loss": 0.02074643075466156,
+      "step": 32800
+    },
+    {
+      "epoch": 0.023645420983119024,
+      "grad_norm": 0.42297810316085815,
+      "learning_rate": 1.6313278214120654e-05,
+      "loss": 0.012604570388793946,
+      "step": 32810
+    },
+    {
+      "epoch": 0.023937700599721237,
+      "grad_norm": 0.2695879638195038,
+      "learning_rate": 1.6269209820087848e-05,
+      "loss": 0.026061999797821044,
+      "step": 32820
+    },
+    {
+      "epoch": 0.02422998021632345,
+      "grad_norm": 0.30303072929382324,
+      "learning_rate": 1.622519575805569e-05,
+      "loss": 0.013605017960071564,
+      "step": 32830
+    },
+    {
+      "epoch": 0.024522259832925664,
+      "grad_norm": 0.27123382687568665,
+      "learning_rate": 1.6181236056584513e-05,
+      "loss": 0.01717102825641632,
+      "step": 32840
+    },
+    {
+      "epoch": 0.024814539449527877,
+      "grad_norm": 0.29478198289871216,
+      "learning_rate": 1.6137330744199262e-05,
+      "loss": 0.01502360701560974,
+      "step": 32850
+    },
+    {
+      "epoch": 0.02510681906613009,
+      "grad_norm": 0.26690873503685,
+      "learning_rate": 1.609347984938956e-05,
+      "loss": 0.0191180020570755,
+      "step": 32860
+    },
+    {
+      "epoch": 0.025399098682732303,
+      "grad_norm": 0.40242666006088257,
+      "learning_rate": 1.604968340060987e-05,
+      "loss": 0.014412066340446473,
+      "step": 32870
+    },
+    {
+      "epoch": 0.025691378299334516,
+      "grad_norm": 0.1930418461561203,
+      "learning_rate": 1.600594142627917e-05,
+      "loss": 0.01486406773328781,
+      "step": 32880
+    },
+    {
+      "epoch": 0.02598365791593673,
+      "grad_norm": 0.2669159770011902,
+      "learning_rate": 1.596225395478116e-05,
+      "loss": 0.015145650506019593,
+      "step": 32890
+    },
+    {
+      "epoch": 0.026275937532538943,
+      "grad_norm": 0.344014048576355,
+      "learning_rate": 1.5918621014464173e-05,
+      "loss": 0.018821634352207184,
+      "step": 32900
+    },
+    {
+      "epoch": 0.026568217149141156,
+      "grad_norm": 0.2288040816783905,
+      "learning_rate": 1.5875042633641168e-05,
+      "loss": 0.013499613106250762,
+      "step": 32910
+    },
+    {
+      "epoch": 0.02686049676574337,
+      "grad_norm": 0.1799820065498352,
+      "learning_rate": 1.5831518840589654e-05,
+      "loss": 0.014769545197486878,
+      "step": 32920
+    },
+    {
+      "epoch": 0.02715277638234558,
+      "grad_norm": 0.4189588129520416,
+      "learning_rate": 1.5788049663551763e-05,
+      "loss": 0.01830632835626602,
+      "step": 32930
+    },
+    {
+      "epoch": 0.027445055998947792,
+      "grad_norm": 0.10951752960681915,
+      "learning_rate": 1.5744635130734186e-05,
+      "loss": 0.014850544929504394,
+      "step": 32940
+    },
+    {
+      "epoch": 0.027737335615550005,
+      "grad_norm": 0.25557202100753784,
+      "learning_rate": 1.57012752703081e-05,
+      "loss": 0.02435782551765442,
+      "step": 32950
+    },
+    {
+      "epoch": 0.02802961523215222,
+      "grad_norm": 0.38698697090148926,
+      "learning_rate": 1.565797011040927e-05,
+      "loss": 0.017513154447078703,
+      "step": 32960
+    },
+    {
+      "epoch": 0.02832189484875443,
+      "grad_norm": 0.26624923944473267,
+      "learning_rate": 1.561471967913798e-05,
+      "loss": 0.016468772292137147,
+      "step": 32970
+    },
+    {
+      "epoch": 0.028614174465356645,
+      "grad_norm": 0.38381075859069824,
+      "learning_rate": 1.55715240045589e-05,
+      "loss": 0.01955130845308304,
+      "step": 32980
+    },
+    {
+      "epoch": 0.028906454081958858,
+      "grad_norm": 0.1439058780670166,
+      "learning_rate": 1.5528383114701284e-05,
+      "loss": 0.015050794184207916,
+      "step": 32990
+    },
+    {
+      "epoch": 0.02919873369856107,
+      "grad_norm": 0.2587265968322754,
+      "learning_rate": 1.5485297037558756e-05,
+      "loss": 0.013450495898723602,
+      "step": 33000
+    },
+    {
+      "epoch": 0.029491013315163284,
+      "grad_norm": 0.23926150798797607,
+      "learning_rate": 1.544226580108944e-05,
+      "loss": 0.02070021480321884,
+      "step": 33010
+    },
+    {
+      "epoch": 0.029783292931765497,
+      "grad_norm": 0.5914262533187866,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 0.0390710324048996,
+      "step": 33020
+    },
+    {
+      "epoch": 0.03007557254836771,
+      "grad_norm": 0.5204606056213379,
+      "learning_rate": 1.5356367961824737e-05,
+      "loss": 0.02930327653884888,
+      "step": 33030
+    },
+    {
+      "epoch": 0.030367852164969924,
+      "grad_norm": 0.29071876406669617,
+      "learning_rate": 1.531350141476755e-05,
+      "loss": 0.031212207674980164,
+      "step": 33040
+    },
+    {
+      "epoch": 0.030660131781572137,
+      "grad_norm": 0.26142480969429016,
+      "learning_rate": 1.5270689819859862e-05,
+      "loss": 0.02368135154247284,
+      "step": 33050
+    },
+    {
+      "epoch": 0.030952411398174347,
+      "grad_norm": 0.1463613659143448,
+      "learning_rate": 1.5227933204881683e-05,
+      "loss": 0.023684953153133393,
+      "step": 33060
+    },
+    {
+      "epoch": 0.03124469101477656,
+      "grad_norm": 0.3913112282752991,
+      "learning_rate": 1.5185231597577277e-05,
+      "loss": 0.033552950620651244,
+      "step": 33070
+    },
+    {
+      "epoch": 0.03153697063137877,
+      "grad_norm": 0.48647817969322205,
+      "learning_rate": 1.5142585025655254e-05,
+      "loss": 0.026082366704940796,
+      "step": 33080
+    },
+    {
+      "epoch": 0.031829250247980986,
+      "grad_norm": 0.22162260115146637,
+      "learning_rate": 1.5099993516788547e-05,
+      "loss": 0.015065094828605652,
+      "step": 33090
+    },
+    {
+      "epoch": 0.0321215298645832,
+      "grad_norm": 0.27202144265174866,
+      "learning_rate": 1.5057457098614281e-05,
+      "loss": 0.019830600917339326,
+      "step": 33100
+    },
+    {
+      "epoch": 0.03241380948118541,
+      "grad_norm": 0.539672315120697,
+      "learning_rate": 1.50149757987339e-05,
+      "loss": 0.02515886425971985,
+      "step": 33110
+    },
+    {
+      "epoch": 0.032706089097787626,
+      "grad_norm": 0.18643370270729065,
+      "learning_rate": 1.4972549644713041e-05,
+      "loss": 0.025296702980995178,
+      "step": 33120
+    },
+    {
+      "epoch": 0.03299836871438984,
+      "grad_norm": 0.1546822041273117,
+      "learning_rate": 1.4930178664081618e-05,
+      "loss": 0.020738226175308228,
+      "step": 33130
+    },
+    {
+      "epoch": 0.03329064833099205,
+      "grad_norm": 0.4232648015022278,
+      "learning_rate": 1.4887862884333659e-05,
+      "loss": 0.022346612811088563,
+      "step": 33140
+    },
+    {
+      "epoch": 0.033582927947594265,
+      "grad_norm": 0.3023698627948761,
+      "learning_rate": 1.484560233292741e-05,
+      "loss": 0.025805747509002684,
+      "step": 33150
+    },
+    {
+      "epoch": 0.03387520756419648,
+      "grad_norm": 0.15867879986763,
+      "learning_rate": 1.4803397037285326e-05,
+      "loss": 0.038519614934921266,
+      "step": 33160
+    },
+    {
+      "epoch": 0.03416748718079869,
+      "grad_norm": 0.3505624532699585,
+      "learning_rate": 1.476124702479389e-05,
+      "loss": 0.024206794798374176,
+      "step": 33170
+    },
+    {
+      "epoch": 0.034459766797400905,
+      "grad_norm": 0.26915377378463745,
+      "learning_rate": 1.4719152322803887e-05,
+      "loss": 0.023101066052913666,
+      "step": 33180
+    },
+    {
+      "epoch": 0.03475204641400312,
+      "grad_norm": 0.1463579535484314,
+      "learning_rate": 1.4677112958630056e-05,
+      "loss": 0.027117082476615907,
+      "step": 33190
+    },
+    {
+      "epoch": 0.03504432603060533,
+      "grad_norm": 0.15641847252845764,
+      "learning_rate": 1.4635128959551247e-05,
+      "loss": 0.026947489380836485,
+      "step": 33200
+    },
+    {
+      "epoch": 0.035336605647207545,
+      "grad_norm": 0.4452631175518036,
+      "learning_rate": 1.459320035281051e-05,
+      "loss": 0.02283456325531006,
+      "step": 33210
+    },
+    {
+      "epoch": 0.03562888526380976,
+      "grad_norm": 0.6916595697402954,
+      "learning_rate": 1.4551327165614792e-05,
+      "loss": 0.046969106793403624,
+      "step": 33220
+    },
+    {
+      "epoch": 0.03592116488041197,
+      "grad_norm": 0.49774789810180664,
+      "learning_rate": 1.450950942513517e-05,
+      "loss": 0.05662800669670105,
+      "step": 33230
+    },
+    {
+      "epoch": 0.036213444497014184,
+      "grad_norm": 0.5510572195053101,
+      "learning_rate": 1.446774715850675e-05,
+      "loss": 0.052986526489257814,
+      "step": 33240
+    },
+    {
+      "epoch": 0.0365057241136164,
+      "grad_norm": 0.41170015931129456,
+      "learning_rate": 1.4426040392828566e-05,
+      "loss": 0.033988097310066225,
+      "step": 33250
+    },
+    {
+      "epoch": 0.036798003730218604,
+      "grad_norm": 0.5689597725868225,
+      "learning_rate": 1.4384389155163714e-05,
+      "loss": 0.04659561812877655,
+      "step": 33260
+    },
+    {
+      "epoch": 0.03709028334682082,
+      "grad_norm": 0.5543137788772583,
+      "learning_rate": 1.4342793472539207e-05,
+      "loss": 0.0560888409614563,
+      "step": 33270
+    },
+    {
+      "epoch": 0.03738256296342303,
+      "grad_norm": 0.560558557510376,
+      "learning_rate": 1.4301253371946089e-05,
+      "loss": 0.05629299879074097,
+      "step": 33280
+    },
+    {
+      "epoch": 0.03767484258002524,
+      "grad_norm": 0.3808022141456604,
+      "learning_rate": 1.4259768880339209e-05,
+      "loss": 0.06510691046714782,
+      "step": 33290
+    },
+    {
+      "epoch": 0.037967122196627456,
+      "grad_norm": 0.6690894365310669,
+      "learning_rate": 1.4218340024637434e-05,
+      "loss": 0.052744680643081666,
+      "step": 33300
+    },
+    {
+      "epoch": 0.03825940181322967,
+      "grad_norm": 0.42849475145339966,
+      "learning_rate": 1.4176966831723527e-05,
+      "loss": 0.04884353280067444,
+      "step": 33310
+    },
+    {
+      "epoch": 0.03855168142983188,
+      "grad_norm": 0.4896790087223053,
+      "learning_rate": 1.4135649328444045e-05,
+      "loss": 0.054973548650741576,
+      "step": 33320
+    },
+    {
+      "epoch": 0.038843961046434096,
+      "grad_norm": 0.5581375956535339,
+      "learning_rate": 1.40943875416095e-05,
+      "loss": 0.06012827157974243,
+      "step": 33330
+    },
+    {
+      "epoch": 0.03913624066303631,
+      "grad_norm": 0.6718844771385193,
+      "learning_rate": 1.4053181497994238e-05,
+      "loss": 0.033290442824363706,
+      "step": 33340
+    },
+    {
+      "epoch": 0.03942852027963852,
+      "grad_norm": 0.44566166400909424,
+      "learning_rate": 1.4012031224336375e-05,
+      "loss": 0.06867111325263978,
+      "step": 33350
+    },
+    {
+      "epoch": 0.039720799896240735,
+      "grad_norm": 0.47675809264183044,
+      "learning_rate": 1.3970936747337881e-05,
+      "loss": 0.050774389505386354,
+      "step": 33360
+    },
+    {
+      "epoch": 0.04001307951284295,
+      "grad_norm": 0.49488356709480286,
+      "learning_rate": 1.3929898093664539e-05,
+      "loss": 0.06320577263832092,
+      "step": 33370
+    },
+    {
+      "epoch": 0.04030535912944516,
+      "grad_norm": 0.421577125787735,
+      "learning_rate": 1.3888915289945893e-05,
+      "loss": 0.05753251314163208,
+      "step": 33380
+    },
+    {
+      "epoch": 0.040597638746047375,
+      "grad_norm": 0.2298666536808014,
+      "learning_rate": 1.3847988362775166e-05,
+      "loss": 0.05316421389579773,
+      "step": 33390
+    },
+    {
+      "epoch": 0.04088991836264959,
+      "grad_norm": 0.6251990795135498,
+      "learning_rate": 1.3807117338709507e-05,
+      "loss": 0.05086994767189026,
+      "step": 33400
+    },
+    {
+      "epoch": 0.0411821979792518,
+      "grad_norm": 0.33022910356521606,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 0.04281286895275116,
+      "step": 33410
+    },
+    {
+      "epoch": 0.041474477595854015,
+      "grad_norm": 0.6224421858787537,
+      "learning_rate": 1.3725543105939942e-05,
+      "loss": 0.0743304431438446,
+      "step": 33420
+    },
+    {
+      "epoch": 0.04176675721245623,
+      "grad_norm": 0.36470380425453186,
+      "learning_rate": 1.3684839950168716e-05,
+      "loss": 0.04776545464992523,
+      "step": 33430
+    },
+    {
+      "epoch": 0.04205903682905844,
+      "grad_norm": 0.43863826990127563,
+      "learning_rate": 1.3644192803367706e-05,
+      "loss": 0.048468881845474245,
+      "step": 33440
+    },
+    {
+      "epoch": 0.042351316445660654,
+      "grad_norm": 0.44012609124183655,
+      "learning_rate": 1.3603601691912448e-05,
+      "loss": 0.06543636322021484,
+      "step": 33450
+    },
+    {
+      "epoch": 0.04264359606226287,
+      "grad_norm": 0.4654460847377777,
+      "learning_rate": 1.3563066642142075e-05,
+      "loss": 0.07518500685691834,
+      "step": 33460
+    },
+    {
+      "epoch": 0.04293587567886508,
+      "grad_norm": 0.3413224518299103,
+      "learning_rate": 1.35225876803593e-05,
+      "loss": 0.053560078144073486,
+      "step": 33470
+    },
+    {
+      "epoch": 0.043228155295467294,
+      "grad_norm": 0.29672011733055115,
+      "learning_rate": 1.3482164832830513e-05,
+      "loss": 0.039417999982833865,
+      "step": 33480
+    },
+    {
+      "epoch": 0.04352043491206951,
+      "grad_norm": 0.6169062852859497,
+      "learning_rate": 1.3441798125785653e-05,
+      "loss": 0.052520865201950075,
+      "step": 33490
+    },
+    {
+      "epoch": 0.04381271452867172,
+      "grad_norm": 0.5814476013183594,
+      "learning_rate": 1.340148758541826e-05,
+      "loss": 0.057974493503570555,
+      "step": 33500
+    },
+    {
+      "epoch": 0.04410499414527393,
+      "grad_norm": 0.48374658823013306,
+      "learning_rate": 1.3361233237885373e-05,
+      "loss": 0.05140966176986694,
+      "step": 33510
+    },
+    {
+      "epoch": 0.04439727376187614,
+      "grad_norm": 0.4957360327243805,
+      "learning_rate": 1.33210351093076e-05,
+      "loss": 0.06187041997909546,
+      "step": 33520
+    },
+    {
+      "epoch": 0.04468955337847835,
+      "grad_norm": 0.6538642048835754,
+      "learning_rate": 1.328089322576912e-05,
+      "loss": 0.04752950668334961,
+      "step": 33530
+    },
+    {
+      "epoch": 0.044981832995080566,
+      "grad_norm": 0.3478364646434784,
+      "learning_rate": 1.3240807613317507e-05,
+      "loss": 0.03778061866760254,
+      "step": 33540
+    },
+    {
+      "epoch": 0.04527411261168278,
+      "grad_norm": 0.29020532965660095,
+      "learning_rate": 1.3200778297963889e-05,
+      "loss": 0.03756671845912933,
+      "step": 33550
+    },
+    {
+      "epoch": 0.04556639222828499,
+      "grad_norm": 0.706080436706543,
+      "learning_rate": 1.31608053056829e-05,
+      "loss": 0.036304301023483275,
+      "step": 33560
+    },
+    {
+      "epoch": 0.045858671844887205,
+      "grad_norm": 0.4034317135810852,
+      "learning_rate": 1.3120888662412534e-05,
+      "loss": 0.04364868104457855,
+      "step": 33570
+    },
+    {
+      "epoch": 0.04615095146148942,
+      "grad_norm": 0.7568108439445496,
+      "learning_rate": 1.3081028394054273e-05,
+      "loss": 0.046861252188682555,
+      "step": 33580
+    },
+    {
+      "epoch": 0.04644323107809163,
+      "grad_norm": 0.89830082654953,
+      "learning_rate": 1.3041224526473017e-05,
+      "loss": 0.08061473369598389,
+      "step": 33590
+    },
+    {
+      "epoch": 0.046735510694693845,
+      "grad_norm": 0.7171100378036499,
+      "learning_rate": 1.3001477085497093e-05,
+      "loss": 0.04803357720375061,
+      "step": 33600
+    },
+    {
+      "epoch": 0.04702779031129606,
+      "grad_norm": 0.4562824070453644,
+      "learning_rate": 1.2961786096918127e-05,
+      "loss": 0.0535660982131958,
+      "step": 33610
+    },
+    {
+      "epoch": 0.04732006992789827,
+      "grad_norm": 0.6582203507423401,
+      "learning_rate": 1.2922151586491227e-05,
+      "loss": 0.04878539443016052,
+      "step": 33620
+    },
+    {
+      "epoch": 0.047612349544500485,
+      "grad_norm": 0.5199447274208069,
+      "learning_rate": 1.2882573579934787e-05,
+      "loss": 0.05310463309288025,
+      "step": 33630
+    },
+    {
+      "epoch": 0.0479046291611027,
+      "grad_norm": 0.8015410900115967,
+      "learning_rate": 1.2843052102930476e-05,
+      "loss": 0.06483718156814575,
+      "step": 33640
+    },
+    {
+      "epoch": 0.04819690877770491,
+      "grad_norm": 0.5843065977096558,
+      "learning_rate": 1.2803587181123456e-05,
+      "loss": 0.056408822536468506,
+      "step": 33650
+    },
+    {
+      "epoch": 0.048489188394307124,
+      "grad_norm": 0.6058087348937988,
+      "learning_rate": 1.276417884012201e-05,
+      "loss": 0.036106881499290464,
+      "step": 33660
+    },
+    {
+      "epoch": 0.04878146801090934,
+      "grad_norm": 0.4453861713409424,
+      "learning_rate": 1.2724827105497816e-05,
+      "loss": 0.03493104577064514,
+      "step": 33670
+    },
+    {
+      "epoch": 0.04907374762751155,
+      "grad_norm": 0.21361681818962097,
+      "learning_rate": 1.2685532002785793e-05,
+      "loss": 0.041265290975570676,
+      "step": 33680
+    },
+    {
+      "epoch": 0.049366027244113764,
+      "grad_norm": 0.6658977270126343,
+      "learning_rate": 1.2646293557484079e-05,
+      "loss": 0.0627876102924347,
+      "step": 33690
+    },
+    {
+      "epoch": 0.04965830686071598,
+      "grad_norm": 0.5152688026428223,
+      "learning_rate": 1.2607111795054095e-05,
+      "loss": 0.031376755237579344,
+      "step": 33700
+    },
+    {
+      "epoch": 0.04995058647731819,
+      "grad_norm": 0.5662060976028442,
+      "learning_rate": 1.2567986740920445e-05,
+      "loss": 0.06039530634880066,
+      "step": 33710
+    },
+    {
+      "epoch": 0.0502428660939204,
+      "grad_norm": 0.5007244944572449,
+      "learning_rate": 1.2528918420470993e-05,
+      "loss": 0.04735149741172791,
+      "step": 33720
+    },
+    {
+      "epoch": 0.050535145710522617,
+      "grad_norm": 0.5353628993034363,
+      "learning_rate": 1.2489906859056689e-05,
+      "loss": 0.049626138806343076,
+      "step": 33730
+    },
+    {
+      "epoch": 0.05082742532712483,
+      "grad_norm": 0.6893242597579956,
+      "learning_rate": 1.2450952081991741e-05,
+      "loss": 0.06103054881095886,
+      "step": 33740
+    },
+    {
+      "epoch": 0.05111970494372704,
+      "grad_norm": 0.46427005529403687,
+      "learning_rate": 1.2412054114553507e-05,
+      "loss": 0.04439499080181122,
+      "step": 33750
+    },
+    {
+      "epoch": 0.051411984560329256,
+      "grad_norm": 0.49867910146713257,
+      "learning_rate": 1.2373212981982407e-05,
+      "loss": 0.040210551023483275,
+      "step": 33760
+    },
+    {
+      "epoch": 0.05170426417693147,
+      "grad_norm": 0.2641783654689789,
+      "learning_rate": 1.2334428709482049e-05,
+      "loss": 0.033705466985702516,
+      "step": 33770
+    },
+    {
+      "epoch": 0.051996543793533676,
+      "grad_norm": 0.527046799659729,
+      "learning_rate": 1.2295701322219156e-05,
+      "loss": 0.045269140601158143,
+      "step": 33780
+    },
+    {
+      "epoch": 0.05228882341013589,
+      "grad_norm": 0.5191704034805298,
+      "learning_rate": 1.2257030845323459e-05,
+      "loss": 0.04536208808422089,
+      "step": 33790
+    },
+    {
+      "epoch": 0.0525811030267381,
+      "grad_norm": 0.33145129680633545,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 0.03620860874652863,
+      "step": 33800
+    },
+    {
+      "epoch": 0.052873382643340315,
+      "grad_norm": 0.4697130620479584,
+      "learning_rate": 1.2179860722968195e-05,
+      "loss": 0.06569225192070008,
+      "step": 33810
+    },
+    {
+      "epoch": 0.05316566225994253,
+      "grad_norm": 0.5600602030754089,
+      "learning_rate": 1.2141361127583506e-05,
+      "loss": 0.04838074743747711,
+      "step": 33820
+    },
+    {
+      "epoch": 0.05345794187654474,
+      "grad_norm": 0.5796632170677185,
+      "learning_rate": 1.210291854271568e-05,
+      "loss": 0.03654422760009766,
+      "step": 33830
+    },
+    {
+      "epoch": 0.053750221493146955,
+      "grad_norm": 0.4382896423339844,
+      "learning_rate": 1.206453299330974e-05,
+      "loss": 0.050191092491149905,
+      "step": 33840
+    },
+    {
+      "epoch": 0.05404250110974917,
+      "grad_norm": 0.49799174070358276,
+      "learning_rate": 1.2026204504273653e-05,
+      "loss": 0.04578710496425629,
+      "step": 33850
+    },
+    {
+      "epoch": 0.05433478072635138,
+      "grad_norm": 0.4525645077228546,
+      "learning_rate": 1.1987933100478299e-05,
+      "loss": 0.037131467461586,
+      "step": 33860
+    },
+    {
+      "epoch": 0.054627060342953594,
+      "grad_norm": 0.5742552280426025,
+      "learning_rate": 1.194971880675767e-05,
+      "loss": 0.08264287710189819,
+      "step": 33870
+    },
+    {
+      "epoch": 0.05491933995955581,
+      "grad_norm": 0.4460553526878357,
+      "learning_rate": 1.1911561647908542e-05,
+      "loss": 0.04796709716320038,
+      "step": 33880
+    },
+    {
+      "epoch": 0.05521161957615802,
+      "grad_norm": 0.5096948146820068,
+      "learning_rate": 1.1873461648690698e-05,
+      "loss": 0.04745099544525146,
+      "step": 33890
+    },
+    {
+      "epoch": 0.055503899192760234,
+      "grad_norm": 0.49294182658195496,
+      "learning_rate": 1.1835418833826827e-05,
+      "loss": 0.042798712849617004,
+      "step": 33900
+    },
+    {
+      "epoch": 0.05579617880936245,
+      "grad_norm": 0.5252227783203125,
+      "learning_rate": 1.1797433228002474e-05,
+      "loss": 0.039584997296333316,
+      "step": 33910
+    },
+    {
+      "epoch": 0.05608845842596466,
+      "grad_norm": 0.36995187401771545,
+      "learning_rate": 1.1759504855866089e-05,
+      "loss": 0.04886587262153626,
+      "step": 33920
+    },
+    {
+      "epoch": 0.05638073804256687,
+      "grad_norm": 0.5218469500541687,
+      "learning_rate": 1.1721633742028992e-05,
+      "loss": 0.042457515001297,
+      "step": 33930
+    },
+    {
+      "epoch": 0.056673017659169087,
+      "grad_norm": 0.3299925625324249,
+      "learning_rate": 1.1683819911065364e-05,
+      "loss": 0.0325689822435379,
+      "step": 33940
+    },
+    {
+      "epoch": 0.0569652972757713,
+      "grad_norm": 0.26507556438446045,
+      "learning_rate": 1.164606338751213e-05,
+      "loss": 0.05239583849906922,
+      "step": 33950
+    },
+    {
+      "epoch": 0.05725757689237351,
+      "grad_norm": 0.5051961541175842,
+      "learning_rate": 1.1608364195869126e-05,
+      "loss": 0.04699195623397827,
+      "step": 33960
+    },
+    {
+      "epoch": 0.057549856508975726,
+      "grad_norm": 0.5581249594688416,
+      "learning_rate": 1.157072236059895e-05,
+      "loss": 0.049959960579872134,
+      "step": 33970
+    },
+    {
+      "epoch": 0.05784213612557794,
+      "grad_norm": 0.48835989832878113,
+      "learning_rate": 1.1533137906126968e-05,
+      "loss": 0.05291617512702942,
+      "step": 33980
+    },
+    {
+      "epoch": 0.05813441574218015,
+      "grad_norm": 0.37549278140068054,
+      "learning_rate": 1.149561085684131e-05,
+      "loss": 0.044364461302757265,
+      "step": 33990
+    },
+    {
+      "epoch": 0.058426695358782366,
+      "grad_norm": 0.4895680844783783,
+      "learning_rate": 1.1458141237092923e-05,
+      "loss": 0.04005611538887024,
+      "step": 34000
+    },
+    {
+      "epoch": 0.05871897497538458,
+      "grad_norm": 0.4668130874633789,
+      "learning_rate": 1.1420729071195379e-05,
+      "loss": 0.048484644293785094,
+      "step": 34010
+    },
+    {
+      "epoch": 0.05901125459198679,
+      "grad_norm": 0.827120840549469,
+      "learning_rate": 1.138337438342505e-05,
+      "loss": 0.0665062665939331,
+      "step": 34020
+    },
+    {
+      "epoch": 0.059303534208589005,
+      "grad_norm": 0.5746980905532837,
+      "learning_rate": 1.1346077198020977e-05,
+      "loss": 0.04909449517726898,
+      "step": 34030
+    },
+    {
+      "epoch": 0.05959581382519121,
+      "grad_norm": 0.7652808427810669,
+      "learning_rate": 1.1308837539184935e-05,
+      "loss": 0.06757973432540894,
+      "step": 34040
+    },
+    {
+      "epoch": 0.059888093441793425,
+      "grad_norm": 0.6857786774635315,
+      "learning_rate": 1.1271655431081285e-05,
+      "loss": 0.029812034964561463,
+      "step": 34050
+    },
+    {
+      "epoch": 0.06018037305839564,
+      "grad_norm": 0.42219269275665283,
+      "learning_rate": 1.1234530897837126e-05,
+      "loss": 0.04557786285877228,
+      "step": 34060
+    },
+    {
+      "epoch": 0.06047265267499785,
+      "grad_norm": 0.27516037225723267,
+      "learning_rate": 1.1197463963542164e-05,
+      "loss": 0.04339747726917267,
+      "step": 34070
+    },
+    {
+      "epoch": 0.060764932291600064,
+      "grad_norm": 0.4118559956550598,
+      "learning_rate": 1.1160454652248687e-05,
+      "loss": 0.03397228717803955,
+      "step": 34080
+    },
+    {
+      "epoch": 0.06105721190820228,
+      "grad_norm": 0.3341304659843445,
+      "learning_rate": 1.1123502987971701e-05,
+      "loss": 0.050447458028793336,
+      "step": 34090
+    },
+    {
+      "epoch": 0.06134949152480449,
+      "grad_norm": 0.42154330015182495,
+      "learning_rate": 1.1086608994688696e-05,
+      "loss": 0.039873361587524414,
+      "step": 34100
+    },
+    {
+      "epoch": 0.061641771141406704,
+      "grad_norm": 0.5591480135917664,
+      "learning_rate": 1.1049772696339788e-05,
+      "loss": 0.04708268344402313,
+      "step": 34110
+    },
+    {
+      "epoch": 0.06193405075800892,
+      "grad_norm": 0.29679563641548157,
+      "learning_rate": 1.1012994116827691e-05,
+      "loss": 0.04279426634311676,
+      "step": 34120
+    },
+    {
+      "epoch": 0.06222633037461113,
+      "grad_norm": 0.5576807260513306,
+      "learning_rate": 1.0976273280017569e-05,
+      "loss": 0.045202425122261046,
+      "step": 34130
+    },
+    {
+      "epoch": 0.06251860999121335,
+      "grad_norm": 0.5076895356178284,
+      "learning_rate": 1.0939610209737194e-05,
+      "loss": 0.08589447736740112,
+      "step": 34140
+    },
+    {
+      "epoch": 0.06281088960781556,
+      "grad_norm": 0.5857372283935547,
+      "learning_rate": 1.090300492977684e-05,
+      "loss": 0.05211997032165527,
+      "step": 34150
+    },
+    {
+      "epoch": 0.06310316922441778,
+      "grad_norm": 0.2982357442378998,
+      "learning_rate": 1.086645746388929e-05,
+      "loss": 0.07178020477294922,
+      "step": 34160
+    },
+    {
+      "epoch": 0.06339544884101998,
+      "grad_norm": 0.5085159540176392,
+      "learning_rate": 1.0829967835789767e-05,
+      "loss": 0.0515937864780426,
+      "step": 34170
+    },
+    {
+      "epoch": 0.06368772845762219,
+      "grad_norm": 0.44178149104118347,
+      "learning_rate": 1.0793536069156018e-05,
+      "loss": 0.055595749616622926,
+      "step": 34180
+    },
+    {
+      "epoch": 0.06398000807422441,
+      "grad_norm": 0.4491787850856781,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 0.04353874623775482,
+      "step": 34190
+    },
+    {
+      "epoch": 0.06427228769082662,
+      "grad_norm": 0.4261537492275238,
+      "learning_rate": 1.0720846214808977e-05,
+      "loss": 0.05219353437423706,
+      "step": 34200
+    },
+    {
+      "epoch": 0.06456456730742884,
+      "grad_norm": 0.40476951003074646,
+      "learning_rate": 1.0684588174263322e-05,
+      "loss": 0.03965598046779632,
+      "step": 34210
+    },
+    {
+      "epoch": 0.06485684692403104,
+      "grad_norm": 0.2841920554637909,
+      "learning_rate": 1.0648388089518747e-05,
+      "loss": 0.0720330834388733,
+      "step": 34220
+    },
+    {
+      "epoch": 0.06514912654063326,
+      "grad_norm": 0.40480294823646545,
+      "learning_rate": 1.061224598406504e-05,
+      "loss": 0.049324429035186766,
+      "step": 34230
+    },
+    {
+      "epoch": 0.06544140615723547,
+      "grad_norm": 0.7512708306312561,
+      "learning_rate": 1.057616188135444e-05,
+      "loss": 0.0534781813621521,
+      "step": 34240
+    },
+    {
+      "epoch": 0.06573368577383769,
+      "grad_norm": 0.48762747645378113,
+      "learning_rate": 1.0540135804801543e-05,
+      "loss": 0.044232887029647824,
+      "step": 34250
+    },
+    {
+      "epoch": 0.0660259653904399,
+      "grad_norm": 0.4850360155105591,
+      "learning_rate": 1.050416777778329e-05,
+      "loss": 0.056512922048568726,
+      "step": 34260
+    },
+    {
+      "epoch": 0.06631824500704211,
+      "grad_norm": 0.5621681809425354,
+      "learning_rate": 1.0468257823638893e-05,
+      "loss": 0.04275169372558594,
+      "step": 34270
+    },
+    {
+      "epoch": 0.06661052462364432,
+      "grad_norm": 0.24818146228790283,
+      "learning_rate": 1.0432405965669978e-05,
+      "loss": 0.061465829610824585,
+      "step": 34280
+    },
+    {
+      "epoch": 0.06690280424024654,
+      "grad_norm": 0.3288441300392151,
+      "learning_rate": 1.0396612227140402e-05,
+      "loss": 0.04255369603633881,
+      "step": 34290
+    },
+    {
+      "epoch": 0.06719508385684875,
+      "grad_norm": 0.3281193971633911,
+      "learning_rate": 1.036087663127634e-05,
+      "loss": 0.053208458423614505,
+      "step": 34300
+    },
+    {
+      "epoch": 0.06748736347345097,
+      "grad_norm": 0.6916245222091675,
+      "learning_rate": 1.0325199201266256e-05,
+      "loss": 0.048221930861473083,
+      "step": 34310
+    },
+    {
+      "epoch": 0.06777964309005317,
+      "grad_norm": 0.4801657199859619,
+      "learning_rate": 1.0289579960260798e-05,
+      "loss": 0.09362080097198486,
+      "step": 34320
+    },
+    {
+      "epoch": 0.0680719227066554,
+      "grad_norm": 0.6675949096679688,
+      "learning_rate": 1.0254018931372944e-05,
+      "loss": 0.07682848572731019,
+      "step": 34330
+    },
+    {
+      "epoch": 0.0683642023232576,
+      "grad_norm": 0.4661373198032379,
+      "learning_rate": 1.0218516137677858e-05,
+      "loss": 0.06984973549842835,
+      "step": 34340
+    },
+    {
+      "epoch": 0.06865648193985982,
+      "grad_norm": 0.6889960169792175,
+      "learning_rate": 1.018307160221289e-05,
+      "loss": 0.08388848304748535,
+      "step": 34350
+    },
+    {
+      "epoch": 0.06894876155646203,
+      "grad_norm": 0.6033889651298523,
+      "learning_rate": 1.0147685347977632e-05,
+      "loss": 0.09418482780456543,
+      "step": 34360
+    },
+    {
+      "epoch": 0.06924104117306425,
+      "grad_norm": 0.7937296628952026,
+      "learning_rate": 1.0112357397933847e-05,
+      "loss": 0.08010193109512329,
+      "step": 34370
+    },
+    {
+      "epoch": 0.06953332078966645,
+      "grad_norm": 0.5029621124267578,
+      "learning_rate": 1.0077087775005456e-05,
+      "loss": 0.06954508423805236,
+      "step": 34380
+    },
+    {
+      "epoch": 0.06982560040626867,
+      "grad_norm": 0.3690824806690216,
+      "learning_rate": 1.0041876502078517e-05,
+      "loss": 0.051458936929702756,
+      "step": 34390
+    },
+    {
+      "epoch": 0.07011788002287088,
+      "grad_norm": 0.5364463925361633,
+      "learning_rate": 1.0006723602001255e-05,
+      "loss": 0.09141680002212524,
+      "step": 34400
+    },
+    {
+      "epoch": 0.0704101596394731,
+      "grad_norm": 0.7409127950668335,
+      "learning_rate": 9.971629097584024e-06,
+      "loss": 0.05852977633476257,
+      "step": 34410
+    },
+    {
+      "epoch": 0.0707024392560753,
+      "grad_norm": 0.2365359365940094,
+      "learning_rate": 9.936593011599205e-06,
+      "loss": 0.04905727505683899,
+      "step": 34420
+    },
+    {
+      "epoch": 0.07099471887267751,
+      "grad_norm": 0.43835005164146423,
+      "learning_rate": 9.901615366781414e-06,
+      "loss": 0.05805126428604126,
+      "step": 34430
+    },
+    {
+      "epoch": 0.07128699848927973,
+      "grad_norm": 0.5249824523925781,
+      "learning_rate": 9.866696185827228e-06,
+      "loss": 0.043165850639343264,
+      "step": 34440
+    },
+    {
+      "epoch": 0.07157927810588194,
+      "grad_norm": 0.3985100984573364,
+      "learning_rate": 9.831835491395292e-06,
+      "loss": 0.0790860652923584,
+      "step": 34450
+    },
+    {
+      "epoch": 0.07187155772248416,
+      "grad_norm": 0.34506112337112427,
+      "learning_rate": 9.797033306106396e-06,
+      "loss": 0.06931524872779846,
+      "step": 34460
+    },
+    {
+      "epoch": 0.07216383733908636,
+      "grad_norm": 0.4406939148902893,
+      "learning_rate": 9.762289652543244e-06,
+      "loss": 0.07681872844696044,
+      "step": 34470
+    },
+    {
+      "epoch": 0.07245611695568858,
+      "grad_norm": 0.4777144491672516,
+      "learning_rate": 9.727604553250646e-06,
+      "loss": 0.07235546112060547,
+      "step": 34480
+    },
+    {
+      "epoch": 0.07274839657229079,
+      "grad_norm": 0.2835845649242401,
+      "learning_rate": 9.6929780307354e-06,
+      "loss": 0.07871950268745423,
+      "step": 34490
+    },
+    {
+      "epoch": 0.07304067618889301,
+      "grad_norm": 0.617375910282135,
+      "learning_rate": 9.658410107466243e-06,
+      "loss": 0.06361361742019653,
+      "step": 34500
+    },
+    {
+      "epoch": 0.07333295580549522,
+      "grad_norm": 0.5350035429000854,
+      "learning_rate": 9.623900805873942e-06,
+      "loss": 0.08758628964424134,
+      "step": 34510
+    },
+    {
+      "epoch": 0.07362523542209744,
+      "grad_norm": 0.44442906975746155,
+      "learning_rate": 9.589450148351209e-06,
+      "loss": 0.07358690500259399,
+      "step": 34520
+    },
+    {
+      "epoch": 0.07391751503869964,
+      "grad_norm": 0.2947746813297272,
+      "learning_rate": 9.555058157252716e-06,
+      "loss": 0.059431082010269164,
+      "step": 34530
+    },
+    {
+      "epoch": 0.07420979465530186,
+      "grad_norm": 0.7169542908668518,
+      "learning_rate": 9.520724854895025e-06,
+      "loss": 0.08271076083183289,
+      "step": 34540
+    },
+    {
+      "epoch": 0.07450207427190407,
+      "grad_norm": 0.6678522825241089,
+      "learning_rate": 9.486450263556656e-06,
+      "loss": 0.06905792355537414,
+      "step": 34550
+    },
+    {
+      "epoch": 0.07479435388850629,
+      "grad_norm": 0.5074385404586792,
+      "learning_rate": 9.452234405478033e-06,
+      "loss": 0.05800063610076904,
+      "step": 34560
+    },
+    {
+      "epoch": 0.0750866335051085,
+      "grad_norm": 0.4099612534046173,
+      "learning_rate": 9.418077302861427e-06,
+      "loss": 0.0714118242263794,
+      "step": 34570
+    },
+    {
+      "epoch": 0.07537891312171072,
+      "grad_norm": 0.4508506655693054,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 0.06984037756919861,
+      "step": 34580
+    },
+    {
+      "epoch": 0.07567119273831292,
+      "grad_norm": 0.49652016162872314,
+      "learning_rate": 9.34993945263285e-06,
+      "loss": 0.09111856818199157,
+      "step": 34590
+    },
+    {
+      "epoch": 0.07596347235491514,
+      "grad_norm": 0.5275205969810486,
+      "learning_rate": 9.315958749234822e-06,
+      "loss": 0.07273449897766113,
+      "step": 34600
+    },
+    {
+      "epoch": 0.07625575197151735,
+      "grad_norm": 0.5305405259132385,
+      "learning_rate": 9.282036889726598e-06,
+      "loss": 0.11469094753265381,
+      "step": 34610
+    },
+    {
+      "epoch": 0.07654803158811957,
+      "grad_norm": 0.40410712361335754,
+      "learning_rate": 9.248173896119716e-06,
+      "loss": 0.11463719606399536,
+      "step": 34620
+    },
+    {
+      "epoch": 0.07684031120472178,
+      "grad_norm": 0.5100727081298828,
+      "learning_rate": 9.214369790387544e-06,
+      "loss": 0.08346616625785827,
+      "step": 34630
+    },
+    {
+      "epoch": 0.077132590821324,
+      "grad_norm": 0.8243907690048218,
+      "learning_rate": 9.180624594465126e-06,
+      "loss": 0.13893865346908568,
+      "step": 34640
+    },
+    {
+      "epoch": 0.0774248704379262,
+      "grad_norm": 0.37129876017570496,
+      "learning_rate": 9.146938330249443e-06,
+      "loss": 0.1054531216621399,
+      "step": 34650
+    },
+    {
+      "epoch": 0.07771715005452842,
+      "grad_norm": 0.5257818102836609,
+      "learning_rate": 9.113311019599092e-06,
+      "loss": 0.1461156964302063,
+      "step": 34660
+    },
+    {
+      "epoch": 0.07800942967113063,
+      "grad_norm": 0.559272289276123,
+      "learning_rate": 9.079742684334447e-06,
+      "loss": 0.09759767055511474,
+      "step": 34670
+    },
+    {
+      "epoch": 0.07830170928773285,
+      "grad_norm": 0.7128331065177917,
+      "learning_rate": 9.046233346237697e-06,
+      "loss": 0.12045934200286865,
+      "step": 34680
+    },
+    {
+      "epoch": 0.07859398890433505,
+      "grad_norm": 0.4552355110645294,
+      "learning_rate": 9.01278302705264e-06,
+      "loss": 0.14952787160873413,
+      "step": 34690
+    },
+    {
+      "epoch": 0.07888626852093726,
+      "grad_norm": 0.5281159281730652,
+      "learning_rate": 8.979391748484844e-06,
+      "loss": 0.10566450357437134,
+      "step": 34700
+    },
+    {
+      "epoch": 0.07917854813753948,
+      "grad_norm": 0.61601722240448,
+      "learning_rate": 8.946059532201567e-06,
+      "loss": 0.10523654222488403,
+      "step": 34710
+    },
+    {
+      "epoch": 0.07947082775414169,
+      "grad_norm": 0.5282240509986877,
+      "learning_rate": 8.912786399831685e-06,
+      "loss": 0.08104625344276428,
+      "step": 34720
+    },
+    {
+      "epoch": 0.07976310737074391,
+      "grad_norm": 0.2776995003223419,
+      "learning_rate": 8.879572372965783e-06,
+      "loss": 0.10662907361984253,
+      "step": 34730
+    },
+    {
+      "epoch": 0.08005538698734611,
+      "grad_norm": 0.44288697838783264,
+      "learning_rate": 8.846417473156088e-06,
+      "loss": 0.09985483288764954,
+      "step": 34740
+    },
+    {
+      "epoch": 0.08034766660394833,
+      "grad_norm": 0.5691683292388916,
+      "learning_rate": 8.813321721916467e-06,
+      "loss": 0.10757981538772583,
+      "step": 34750
+    },
+    {
+      "epoch": 0.08063994622055054,
+      "grad_norm": 0.3976956605911255,
+      "learning_rate": 8.780285140722378e-06,
+      "loss": 0.13828656673431397,
+      "step": 34760
+    },
+    {
+      "epoch": 0.08093222583715276,
+      "grad_norm": 0.594179093837738,
+      "learning_rate": 8.747307751010891e-06,
+      "loss": 0.13858397006988527,
+      "step": 34770
+    },
+    {
+      "epoch": 0.08122450545375497,
+      "grad_norm": 0.6442998051643372,
+      "learning_rate": 8.714389574180715e-06,
+      "loss": 0.13031725883483886,
+      "step": 34780
+    },
+    {
+      "epoch": 0.08151678507035719,
+      "grad_norm": 0.5470430850982666,
+      "learning_rate": 8.681530631592038e-06,
+      "loss": 0.09936890602111817,
+      "step": 34790
+    },
+    {
+      "epoch": 0.08180906468695939,
+      "grad_norm": 0.45249462127685547,
+      "learning_rate": 8.648730944566719e-06,
+      "loss": 0.09623882174491882,
+      "step": 34800
+    },
+    {
+      "epoch": 0.08210134430356161,
+      "grad_norm": 0.6310141682624817,
+      "learning_rate": 8.615990534388097e-06,
+      "loss": 0.09833247065544129,
+      "step": 34810
+    },
+    {
+      "epoch": 0.08239362392016382,
+      "grad_norm": 0.5009254217147827,
+      "learning_rate": 8.583309422301101e-06,
+      "loss": 0.11918139457702637,
+      "step": 34820
+    },
+    {
+      "epoch": 0.08268590353676604,
+      "grad_norm": 0.6045822501182556,
+      "learning_rate": 8.550687629512088e-06,
+      "loss": 0.1061674952507019,
+      "step": 34830
+    },
+    {
+      "epoch": 0.08297818315336825,
+      "grad_norm": 0.45637860894203186,
+      "learning_rate": 8.51812517718904e-06,
+      "loss": 0.123035728931427,
+      "step": 34840
+    },
+    {
+      "epoch": 0.08327046276997047,
+      "grad_norm": 0.5218595266342163,
+      "learning_rate": 8.485622086461365e-06,
+      "loss": 0.0925482153892517,
+      "step": 34850
+    },
+    {
+      "epoch": 0.08356274238657267,
+      "grad_norm": 0.5282291173934937,
+      "learning_rate": 8.453178378419935e-06,
+      "loss": 0.12109495401382446,
+      "step": 34860
+    },
+    {
+      "epoch": 0.08385502200317489,
+      "grad_norm": 0.614937424659729,
+      "learning_rate": 8.420794074117178e-06,
+      "loss": 0.0839550256729126,
+      "step": 34870
+    },
+    {
+      "epoch": 0.0841473016197771,
+      "grad_norm": 0.4173479676246643,
+      "learning_rate": 8.388469194566883e-06,
+      "loss": 0.11765177249908447,
+      "step": 34880
+    },
+    {
+      "epoch": 0.08443958123637932,
+      "grad_norm": 0.4556170403957367,
+      "learning_rate": 8.356203760744274e-06,
+      "loss": 0.08123626708984374,
+      "step": 34890
+    },
+    {
+      "epoch": 0.08473186085298152,
+      "grad_norm": 0.3977719247341156,
+      "learning_rate": 8.323997793586124e-06,
+      "loss": 0.11462039947509765,
+      "step": 34900
+    },
+    {
+      "epoch": 0.08502414046958374,
+      "grad_norm": 0.5335431098937988,
+      "learning_rate": 8.291851313990484e-06,
+      "loss": 0.12480266094207763,
+      "step": 34910
+    },
+    {
+      "epoch": 0.08531642008618595,
+      "grad_norm": 0.3573301136493683,
+      "learning_rate": 8.259764342816856e-06,
+      "loss": 0.13992141485214232,
+      "step": 34920
+    },
+    {
+      "epoch": 0.08560869970278817,
+      "grad_norm": 0.49426358938217163,
+      "learning_rate": 8.227736900886151e-06,
+      "loss": 0.11405361890792846,
+      "step": 34930
+    },
+    {
+      "epoch": 0.08590097931939038,
+      "grad_norm": 0.4588361084461212,
+      "learning_rate": 8.195769008980603e-06,
+      "loss": 0.10775070190429688,
+      "step": 34940
+    },
+    {
+      "epoch": 0.08619325893599258,
+      "grad_norm": 0.3250270187854767,
+      "learning_rate": 8.163860687843839e-06,
+      "loss": 0.07414093017578124,
+      "step": 34950
+    },
+    {
+      "epoch": 0.0864855385525948,
+      "grad_norm": 0.38330334424972534,
+      "learning_rate": 8.13201195818083e-06,
+      "loss": 0.09026837944984437,
+      "step": 34960
+    },
+    {
+      "epoch": 0.08677781816919701,
+      "grad_norm": 0.6113490462303162,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 0.09416397809982299,
+      "step": 34970
+    },
+    {
+      "epoch": 0.08707009778579923,
+      "grad_norm": 0.5501431226730347,
+      "learning_rate": 8.068493355902562e-06,
+      "loss": 0.09925037026405334,
+      "step": 34980
+    },
+    {
+      "epoch": 0.08736237740240144,
+      "grad_norm": 0.26558834314346313,
+      "learning_rate": 8.036823524503823e-06,
+      "loss": 0.10210937261581421,
+      "step": 34990
+    },
+    {
+      "epoch": 0.08765465701900366,
+      "grad_norm": 0.70122230052948,
+      "learning_rate": 8.005213367011888e-06,
+      "loss": 0.10449042320251464,
+      "step": 35000
+    },
+    {
+      "epoch": 0.08794693663560586,
+      "grad_norm": 0.5046303272247314,
+      "learning_rate": 7.973662903938205e-06,
+      "loss": 0.0919480562210083,
+      "step": 35010
+    },
+    {
+      "epoch": 0.08823921625220808,
+      "grad_norm": 0.43984949588775635,
+      "learning_rate": 7.942172155755546e-06,
+      "loss": 0.1620811939239502,
+      "step": 35020
+    },
+    {
+      "epoch": 0.08853149586881029,
+      "grad_norm": 0.6931641101837158,
+      "learning_rate": 7.910741142897949e-06,
+      "loss": 0.12950831651687622,
+      "step": 35030
+    },
+    {
+      "epoch": 0.08882377548541251,
+      "grad_norm": 0.517694890499115,
+      "learning_rate": 7.879369885760591e-06,
+      "loss": 0.09344902634620667,
+      "step": 35040
+    },
+    {
+      "epoch": 0.08911605510201472,
+      "grad_norm": 0.455465167760849,
+      "learning_rate": 7.848058404699988e-06,
+      "loss": 0.1143727421760559,
+      "step": 35050
+    },
+    {
+      "epoch": 0.08940833471861694,
+      "grad_norm": 0.44477835297584534,
+      "learning_rate": 7.816806720033798e-06,
+      "loss": 0.14658366441726683,
+      "step": 35060
+    },
+    {
+      "epoch": 0.08970061433521914,
+      "grad_norm": 0.4985464811325073,
+      "learning_rate": 7.785614852040935e-06,
+      "loss": 0.11717532873153687,
+      "step": 35070
+    },
+    {
+      "epoch": 0.08999289395182136,
+      "grad_norm": 0.5502634644508362,
+      "learning_rate": 7.754482820961396e-06,
+      "loss": 0.08342461585998535,
+      "step": 35080
+    },
+    {
+      "epoch": 0.09028517356842357,
+      "grad_norm": 0.5386865735054016,
+      "learning_rate": 7.723410646996498e-06,
+      "loss": 0.08006536364555358,
+      "step": 35090
+    },
+    {
+      "epoch": 0.09057745318502579,
+      "grad_norm": 0.7822738885879517,
+      "learning_rate": 7.692398350308594e-06,
+      "loss": 0.11152905225753784,
+      "step": 35100
+    },
+    {
+      "epoch": 0.090869732801628,
+      "grad_norm": 0.43481606245040894,
+      "learning_rate": 7.661445951021196e-06,
+      "loss": 0.08059299588203431,
+      "step": 35110
+    },
+    {
+      "epoch": 0.09116201241823021,
+      "grad_norm": 0.5356187224388123,
+      "learning_rate": 7.630553469219049e-06,
+      "loss": 0.09959313869476319,
+      "step": 35120
+    },
+    {
+      "epoch": 0.09145429203483242,
+      "grad_norm": 0.6277801990509033,
+      "learning_rate": 7.5997209249479065e-06,
+      "loss": 0.11210767030715943,
+      "step": 35130
+    },
+    {
+      "epoch": 0.09174657165143464,
+      "grad_norm": 0.46041789650917053,
+      "learning_rate": 7.5689483382146766e-06,
+      "loss": 0.114070725440979,
+      "step": 35140
+    },
+    {
+      "epoch": 0.09203885126803685,
+      "grad_norm": 0.49991375207901,
+      "learning_rate": 7.538235728987375e-06,
+      "loss": 0.09788026213645935,
+      "step": 35150
+    },
+    {
+      "epoch": 0.09233113088463907,
+      "grad_norm": 0.5267115235328674,
+      "learning_rate": 7.507583117195049e-06,
+      "loss": 0.12069894075393676,
+      "step": 35160
+    },
+    {
+      "epoch": 0.09262341050124127,
+      "grad_norm": 0.432651162147522,
+      "learning_rate": 7.476990522727867e-06,
+      "loss": 0.07310009002685547,
+      "step": 35170
+    },
+    {
+      "epoch": 0.0929156901178435,
+      "grad_norm": 0.44736069440841675,
+      "learning_rate": 7.4464579654370144e-06,
+      "loss": 0.0965876042842865,
+      "step": 35180
+    },
+    {
+      "epoch": 0.0932079697344457,
+      "grad_norm": 0.42934346199035645,
+      "learning_rate": 7.415985465134778e-06,
+      "loss": 0.10947409868240357,
+      "step": 35190
+    },
+    {
+      "epoch": 0.09350024935104792,
+      "grad_norm": 0.524398922920227,
+      "learning_rate": 7.385573041594385e-06,
+      "loss": 0.06933254599571229,
+      "step": 35200
+    },
+    {
+      "epoch": 0.09379252896765013,
+      "grad_norm": 0.4171840250492096,
+      "learning_rate": 7.355220714550138e-06,
+      "loss": 0.10686259269714356,
+      "step": 35210
+    },
+    {
+      "epoch": 0.09408480858425233,
+      "grad_norm": 0.3155970275402069,
+      "learning_rate": 7.324928503697359e-06,
+      "loss": 0.1111186385154724,
+      "step": 35220
+    },
+    {
+      "epoch": 0.09437708820085455,
+      "grad_norm": 0.34378135204315186,
+      "learning_rate": 7.2946964286923046e-06,
+      "loss": 0.08119457364082336,
+      "step": 35230
+    },
+    {
+      "epoch": 0.09466936781745676,
+      "grad_norm": 0.5157742500305176,
+      "learning_rate": 7.264524509152249e-06,
+      "loss": 0.08805540800094605,
+      "step": 35240
+    },
+    {
+      "epoch": 0.09496164743405898,
+      "grad_norm": 0.556401789188385,
+      "learning_rate": 7.234412764655441e-06,
+      "loss": 0.15788320302963257,
+      "step": 35250
+    },
+    {
+      "epoch": 0.09525392705066119,
+      "grad_norm": 0.3666079342365265,
+      "learning_rate": 7.204361214741018e-06,
+      "loss": 0.08203710317611694,
+      "step": 35260
+    },
+    {
+      "epoch": 0.0955462066672634,
+      "grad_norm": 0.4647900462150574,
+      "learning_rate": 7.174369878909148e-06,
+      "loss": 0.0695656955242157,
+      "step": 35270
+    },
+    {
+      "epoch": 0.09583848628386561,
+      "grad_norm": 0.20974373817443848,
+      "learning_rate": 7.144438776620877e-06,
+      "loss": 0.09542552828788757,
+      "step": 35280
+    },
+    {
+      "epoch": 0.09613076590046783,
+      "grad_norm": 0.4728606045246124,
+      "learning_rate": 7.114567927298166e-06,
+      "loss": 0.08774740695953369,
+      "step": 35290
+    },
+    {
+      "epoch": 0.09642304551707004,
+      "grad_norm": 0.44567981362342834,
+      "learning_rate": 7.084757350323856e-06,
+      "loss": 0.13875144720077515,
+      "step": 35300
+    },
+    {
+      "epoch": 0.09671532513367226,
+      "grad_norm": 0.6297561526298523,
+      "learning_rate": 7.055007065041775e-06,
+      "loss": 0.10262517929077149,
+      "step": 35310
+    },
+    {
+      "epoch": 0.09700760475027446,
+      "grad_norm": 0.48533350229263306,
+      "learning_rate": 7.025317090756523e-06,
+      "loss": 0.1336029887199402,
+      "step": 35320
+    },
+    {
+      "epoch": 0.09729988436687668,
+      "grad_norm": 0.5486268401145935,
+      "learning_rate": 6.995687446733556e-06,
+      "loss": 0.06903930306434632,
+      "step": 35330
+    },
+    {
+      "epoch": 0.09759216398347889,
+      "grad_norm": 0.6231697797775269,
+      "learning_rate": 6.966118152199308e-06,
+      "loss": 0.09906846880912781,
+      "step": 35340
+    },
+    {
+      "epoch": 0.09788444360008111,
+      "grad_norm": 0.42271775007247925,
+      "learning_rate": 6.936609226340918e-06,
+      "loss": 0.08609395027160645,
+      "step": 35350
+    },
+    {
+      "epoch": 0.09817672321668332,
+      "grad_norm": 0.39531034231185913,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 0.11023186445236206,
+      "step": 35360
+    },
+    {
+      "epoch": 0.09846900283328554,
+      "grad_norm": 0.27402663230895996,
+      "learning_rate": 6.877772557204687e-06,
+      "loss": 0.07877730727195739,
+      "step": 35370
+    },
+    {
+      "epoch": 0.09876128244988774,
+      "grad_norm": 0.405137836933136,
+      "learning_rate": 6.8484448521053065e-06,
+      "loss": 0.08205050826072693,
+      "step": 35380
+    },
+    {
+      "epoch": 0.09905356206648996,
+      "grad_norm": 0.35444316267967224,
+      "learning_rate": 6.819177592038728e-06,
+      "loss": 0.10442720651626587,
+      "step": 35390
+    },
+    {
+      "epoch": 0.09934584168309217,
+      "grad_norm": 0.48142343759536743,
+      "learning_rate": 6.789970795996159e-06,
+      "loss": 0.0988464593887329,
+      "step": 35400
+    },
+    {
+      "epoch": 0.09963812129969439,
+      "grad_norm": 0.5209444165229797,
+      "learning_rate": 6.760824482929595e-06,
+      "loss": 0.09303090572357178,
+      "step": 35410
+    },
+    {
+      "epoch": 0.0999304009162966,
+      "grad_norm": 0.39319440722465515,
+      "learning_rate": 6.731738671751719e-06,
+      "loss": 0.097932630777359,
+      "step": 35420
+    },
+    {
+      "epoch": 0.10022268053289882,
+      "grad_norm": 0.3927178978919983,
+      "learning_rate": 6.702713381336023e-06,
+      "loss": 0.11360737085342407,
+      "step": 35430
+    },
+    {
+      "epoch": 0.10051496014950102,
+      "grad_norm": 0.3209347724914551,
+      "learning_rate": 6.673748630516719e-06,
+      "loss": 0.09337975978851318,
+      "step": 35440
+    },
+    {
+      "epoch": 0.10080723976610324,
+      "grad_norm": 0.4022107422351837,
+      "learning_rate": 6.644844438088682e-06,
+      "loss": 0.07966390252113342,
+      "step": 35450
+    },
+    {
+      "epoch": 0.10109951938270545,
+      "grad_norm": 0.3984311521053314,
+      "learning_rate": 6.616000822807544e-06,
+      "loss": 0.09620524644851684,
+      "step": 35460
+    },
+    {
+      "epoch": 0.10139179899930767,
+      "grad_norm": 0.4628983736038208,
+      "learning_rate": 6.587217803389634e-06,
+      "loss": 0.1000235915184021,
+      "step": 35470
+    },
+    {
+      "epoch": 0.10168407861590988,
+      "grad_norm": 0.442244291305542,
+      "learning_rate": 6.55849539851191e-06,
+      "loss": 0.06783375144004822,
+      "step": 35480
+    },
+    {
+      "epoch": 0.10197635823251208,
+      "grad_norm": 0.5233607292175293,
+      "learning_rate": 6.529833626812043e-06,
+      "loss": 0.12960963249206542,
+      "step": 35490
+    },
+    {
+      "epoch": 0.1022686378491143,
+      "grad_norm": 0.5155558586120605,
+      "learning_rate": 6.501232506888333e-06,
+      "loss": 0.09775227308273315,
+      "step": 35500
+    },
+    {
+      "epoch": 0.10256091746571651,
+      "grad_norm": 0.6757098436355591,
+      "learning_rate": 6.472692057299768e-06,
+      "loss": 0.11020770072937011,
+      "step": 35510
+    },
+    {
+      "epoch": 0.10285319708231873,
+      "grad_norm": 0.6915234923362732,
+      "learning_rate": 6.44421229656591e-06,
+      "loss": 0.09538257718086243,
+      "step": 35520
+    },
+    {
+      "epoch": 0.10314547669892093,
+      "grad_norm": 0.5162079930305481,
+      "learning_rate": 6.415793243166968e-06,
+      "loss": 0.07812063694000244,
+      "step": 35530
+    },
+    {
+      "epoch": 0.10343775631552315,
+      "grad_norm": 0.4993222653865814,
+      "learning_rate": 6.387434915543755e-06,
+      "loss": 0.1197195053100586,
+      "step": 35540
+    },
+    {
+      "epoch": 0.10373003593212536,
+      "grad_norm": 0.2711483836174011,
+      "learning_rate": 6.359137332097698e-06,
+      "loss": 0.09298015236854554,
+      "step": 35550
+    },
+    {
+      "epoch": 0.10402231554872758,
+      "grad_norm": 0.39711621403694153,
+      "learning_rate": 6.330900511190796e-06,
+      "loss": 0.09864088892936707,
+      "step": 35560
+    },
+    {
+      "epoch": 0.10431459516532979,
+      "grad_norm": 0.4574972987174988,
+      "learning_rate": 6.302724471145582e-06,
+      "loss": 0.07466109395027161,
+      "step": 35570
+    },
+    {
+      "epoch": 0.10460687478193201,
+      "grad_norm": 0.3749317526817322,
+      "learning_rate": 6.274609230245199e-06,
+      "loss": 0.0952911615371704,
+      "step": 35580
+    },
+    {
+      "epoch": 0.10489915439853421,
+      "grad_norm": 0.6553129553794861,
+      "learning_rate": 6.24655480673334e-06,
+      "loss": 0.08348474502563477,
+      "step": 35590
+    },
+    {
+      "epoch": 0.10519143401513643,
+      "grad_norm": 0.4482124149799347,
+      "learning_rate": 6.21856121881419e-06,
+      "loss": 0.09457226395606995,
+      "step": 35600
+    },
+    {
+      "epoch": 0.10548371363173864,
+      "grad_norm": 0.644478440284729,
+      "learning_rate": 6.190628484652494e-06,
+      "loss": 0.1126552939414978,
+      "step": 35610
+    },
+    {
+      "epoch": 0.10577599324834086,
+      "grad_norm": 0.4793469309806824,
+      "learning_rate": 6.1627566223735e-06,
+      "loss": 0.11904491186141967,
+      "step": 35620
+    },
+    {
+      "epoch": 0.10606827286494307,
+      "grad_norm": 0.4004933536052704,
+      "learning_rate": 6.134945650062962e-06,
+      "loss": 0.062014222145080566,
+      "step": 35630
+    },
+    {
+      "epoch": 0.10636055248154529,
+      "grad_norm": 0.6047390103340149,
+      "learning_rate": 6.107195585767111e-06,
+      "loss": 0.10721466541290284,
+      "step": 35640
+    },
+    {
+      "epoch": 0.1066528320981475,
+      "grad_norm": 0.35131752490997314,
+      "learning_rate": 6.079506447492667e-06,
+      "loss": 0.07750452160835267,
+      "step": 35650
+    },
+    {
+      "epoch": 0.10694511171474971,
+      "grad_norm": 0.38737979531288147,
+      "learning_rate": 6.051878253206811e-06,
+      "loss": 0.09712913632392883,
+      "step": 35660
+    },
+    {
+      "epoch": 0.10723739133135192,
+      "grad_norm": 0.27050936222076416,
+      "learning_rate": 6.024311020837159e-06,
+      "loss": 0.11708197593688965,
+      "step": 35670
+    },
+    {
+      "epoch": 0.10752967094795414,
+      "grad_norm": 0.4854159951210022,
+      "learning_rate": 5.996804768271835e-06,
+      "loss": 0.07446560859680176,
+      "step": 35680
+    },
+    {
+      "epoch": 0.10782195056455635,
+      "grad_norm": 0.4967937469482422,
+      "learning_rate": 5.969359513359307e-06,
+      "loss": 0.0958297610282898,
+      "step": 35690
+    },
+    {
+      "epoch": 0.10811423018115857,
+      "grad_norm": 0.5686248540878296,
+      "learning_rate": 5.941975273908496e-06,
+      "loss": 0.14568789005279542,
+      "step": 35700
+    },
+    {
+      "epoch": 0.10840650979776077,
+      "grad_norm": 0.4834304749965668,
+      "learning_rate": 5.914652067688786e-06,
+      "loss": 0.118193519115448,
+      "step": 35710
+    },
+    {
+      "epoch": 0.10869878941436299,
+      "grad_norm": 0.46335726976394653,
+      "learning_rate": 5.887389912429864e-06,
+      "loss": 0.08519103527069091,
+      "step": 35720
+    },
+    {
+      "epoch": 0.1089910690309652,
+      "grad_norm": 0.4636739492416382,
+      "learning_rate": 5.860188825821855e-06,
+      "loss": 0.07281619310379028,
+      "step": 35730
+    },
+    {
+      "epoch": 0.1092833486475674,
+      "grad_norm": 0.44078245759010315,
+      "learning_rate": 5.833048825515275e-06,
+      "loss": 0.11348780393600463,
+      "step": 35740
+    },
+    {
+      "epoch": 0.10957562826416962,
+      "grad_norm": 0.4571603238582611,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 0.11943268775939941,
+      "step": 35750
+    },
+    {
+      "epoch": 0.10986790788077183,
+      "grad_norm": 0.24379688501358032,
+      "learning_rate": 5.778952154210071e-06,
+      "loss": 0.11379902362823487,
+      "step": 35760
+    },
+    {
+      "epoch": 0.11016018749737405,
+      "grad_norm": 0.4096236228942871,
+      "learning_rate": 5.751995518314202e-06,
+      "loss": 0.10485565662384033,
+      "step": 35770
+    },
+    {
+      "epoch": 0.11045246711397626,
+      "grad_norm": 0.5383893847465515,
+      "learning_rate": 5.725100038925224e-06,
+      "loss": 0.15146160125732422,
+      "step": 35780
+    },
+    {
+      "epoch": 0.11074474673057848,
+      "grad_norm": 0.5122965574264526,
+      "learning_rate": 5.698265733495289e-06,
+      "loss": 0.08850286602973938,
+      "step": 35790
+    },
+    {
+      "epoch": 0.11103702634718068,
+      "grad_norm": 0.5089948177337646,
+      "learning_rate": 5.671492619436891e-06,
+      "loss": 0.11797703504562378,
+      "step": 35800
+    },
+    {
+      "epoch": 0.1113293059637829,
+      "grad_norm": 0.4202145040035248,
+      "learning_rate": 5.644780714122843e-06,
+      "loss": 0.11024415493011475,
+      "step": 35810
+    },
+    {
+      "epoch": 0.11162158558038511,
+      "grad_norm": 0.5743666887283325,
+      "learning_rate": 5.618130034886182e-06,
+      "loss": 0.07543653845787049,
+      "step": 35820
+    },
+    {
+      "epoch": 0.11191386519698733,
+      "grad_norm": 0.3306371867656708,
+      "learning_rate": 5.591540599020251e-06,
+      "loss": 0.11850699186325073,
+      "step": 35830
+    },
+    {
+      "epoch": 0.11220614481358954,
+      "grad_norm": 0.42529937624931335,
+      "learning_rate": 5.5650124237786375e-06,
+      "loss": 0.05562476515769958,
+      "step": 35840
+    },
+    {
+      "epoch": 0.11249842443019176,
+      "grad_norm": 0.5735292434692383,
+      "learning_rate": 5.538545526375216e-06,
+      "loss": 0.1364474892616272,
+      "step": 35850
+    },
+    {
+      "epoch": 0.11279070404679396,
+      "grad_norm": 0.42832493782043457,
+      "learning_rate": 5.512139923984039e-06,
+      "loss": 0.11328768730163574,
+      "step": 35860
+    },
+    {
+      "epoch": 0.11308298366339618,
+      "grad_norm": 0.458047479391098,
+      "learning_rate": 5.48579563373941e-06,
+      "loss": 0.12515199184417725,
+      "step": 35870
+    },
+    {
+      "epoch": 0.11337526327999839,
+      "grad_norm": 0.6304228901863098,
+      "learning_rate": 5.4595126727359e-06,
+      "loss": 0.12083259820938111,
+      "step": 35880
+    },
+    {
+      "epoch": 0.11366754289660061,
+      "grad_norm": 0.5135407447814941,
+      "learning_rate": 5.433291058028167e-06,
+      "loss": 0.10923534631729126,
+      "step": 35890
+    },
+    {
+      "epoch": 0.11395982251320282,
+      "grad_norm": 0.38851088285446167,
+      "learning_rate": 5.4071308066312e-06,
+      "loss": 0.13863390684127808,
+      "step": 35900
+    },
+    {
+      "epoch": 0.11425210212980504,
+      "grad_norm": 0.32888197898864746,
+      "learning_rate": 5.381031935520076e-06,
+      "loss": 0.1257638931274414,
+      "step": 35910
+    },
+    {
+      "epoch": 0.11454438174640724,
+      "grad_norm": 0.5303085446357727,
+      "learning_rate": 5.354994461630036e-06,
+      "loss": 0.1960148334503174,
+      "step": 35920
+    },
+    {
+      "epoch": 0.11483666136300946,
+      "grad_norm": 0.7478082180023193,
+      "learning_rate": 5.3290184018565785e-06,
+      "loss": 0.12432985305786133,
+      "step": 35930
+    },
+    {
+      "epoch": 0.11512894097961167,
+      "grad_norm": 0.3853866457939148,
+      "learning_rate": 5.303103773055229e-06,
+      "loss": 0.10508391857147217,
+      "step": 35940
+    },
+    {
+      "epoch": 0.11542122059621389,
+      "grad_norm": 0.5634444355964661,
+      "learning_rate": 5.277250592041727e-06,
+      "loss": 0.12553178071975707,
+      "step": 35950
+    },
+    {
+      "epoch": 0.1157135002128161,
+      "grad_norm": 0.4314773976802826,
+      "learning_rate": 5.251458875591953e-06,
+      "loss": 0.13623230457305907,
+      "step": 35960
+    },
+    {
+      "epoch": 0.11600577982941832,
+      "grad_norm": 0.47805464267730713,
+      "learning_rate": 5.225728640441818e-06,
+      "loss": 0.15281635522842407,
+      "step": 35970
+    },
+    {
+      "epoch": 0.11629805944602052,
+      "grad_norm": 0.5540910363197327,
+      "learning_rate": 5.200059903287424e-06,
+      "loss": 0.1353284239768982,
+      "step": 35980
+    },
+    {
+      "epoch": 0.11659033906262274,
+      "grad_norm": 0.2936714291572571,
+      "learning_rate": 5.174452680784936e-06,
+      "loss": 0.07588390707969665,
+      "step": 35990
+    },
+    {
+      "epoch": 0.11688261867922495,
+      "grad_norm": 0.4887588918209076,
+      "learning_rate": 5.148906989550606e-06,
+      "loss": 0.16103655099868774,
+      "step": 36000
+    },
+    {
+      "epoch": 0.11717489829582715,
+      "grad_norm": 0.31138041615486145,
+      "learning_rate": 5.1234228461607304e-06,
+      "loss": 0.10906132459640502,
+      "step": 36010
+    },
+    {
+      "epoch": 0.11746717791242937,
+      "grad_norm": 0.5944001078605652,
+      "learning_rate": 5.098000267151715e-06,
+      "loss": 0.16141773462295533,
+      "step": 36020
+    },
+    {
+      "epoch": 0.11775945752903158,
+      "grad_norm": 0.49215975403785706,
+      "learning_rate": 5.072639269020007e-06,
+      "loss": 0.10081020593643189,
+      "step": 36030
+    },
+    {
+      "epoch": 0.1180517371456338,
+      "grad_norm": 0.4139394760131836,
+      "learning_rate": 5.047339868222046e-06,
+      "loss": 0.11932525634765626,
+      "step": 36040
+    },
+    {
+      "epoch": 0.118344016762236,
+      "grad_norm": 0.3580107092857361,
+      "learning_rate": 5.0221020811743516e-06,
+      "loss": 0.1097977638244629,
+      "step": 36050
+    },
+    {
+      "epoch": 0.11863629637883823,
+      "grad_norm": 0.5650413036346436,
+      "learning_rate": 4.99692592425347e-06,
+      "loss": 0.13751375675201416,
+      "step": 36060
+    },
+    {
+      "epoch": 0.11892857599544043,
+      "grad_norm": 0.42039206624031067,
+      "learning_rate": 4.971811413795924e-06,
+      "loss": 0.13672657012939454,
+      "step": 36070
+    },
+    {
+      "epoch": 0.11922085561204265,
+      "grad_norm": 0.3474499583244324,
+      "learning_rate": 4.946758566098242e-06,
+      "loss": 0.12606663703918458,
+      "step": 36080
+    },
+    {
+      "epoch": 0.11951313522864486,
+      "grad_norm": 0.47652584314346313,
+      "learning_rate": 4.921767397416943e-06,
+      "loss": 0.15218812227249146,
+      "step": 36090
+    },
+    {
+      "epoch": 0.11980541484524708,
+      "grad_norm": 0.5100054144859314,
+      "learning_rate": 4.896837923968545e-06,
+      "loss": 0.09665951728820801,
+      "step": 36100
+    },
+    {
+      "epoch": 0.12009769446184929,
+      "grad_norm": 0.5017781257629395,
+      "learning_rate": 4.871970161929474e-06,
+      "loss": 0.10842703580856324,
+      "step": 36110
+    },
+    {
+      "epoch": 0.1203899740784515,
+      "grad_norm": 0.38195595145225525,
+      "learning_rate": 4.847164127436199e-06,
+      "loss": 0.12413654327392579,
+      "step": 36120
+    },
+    {
+      "epoch": 0.12068225369505371,
+      "grad_norm": 0.6844592094421387,
+      "learning_rate": 4.822419836585057e-06,
+      "loss": 0.11847741603851318,
+      "step": 36130
+    },
+    {
+      "epoch": 0.12097453331165593,
+      "grad_norm": 0.4024202823638916,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 0.1469767689704895,
+      "step": 36140
+    },
+    {
+      "epoch": 0.12126681292825814,
+      "grad_norm": 0.47260984778404236,
+      "learning_rate": 4.773116549994306e-06,
+      "loss": 0.11846110820770264,
+      "step": 36150
+    },
+    {
+      "epoch": 0.12155909254486036,
+      "grad_norm": 0.3946249485015869,
+      "learning_rate": 4.74855758624706e-06,
+      "loss": 0.10292274951934814,
+      "step": 36160
+    },
+    {
+      "epoch": 0.12185137216146257,
+      "grad_norm": 0.22267845273017883,
+      "learning_rate": 4.7240604301266755e-06,
+      "loss": 0.10845946073532105,
+      "step": 36170
+    },
+    {
+      "epoch": 0.12214365177806479,
+      "grad_norm": 0.5327144861221313,
+      "learning_rate": 4.699625097529103e-06,
+      "loss": 0.1529550790786743,
+      "step": 36180
+    },
+    {
+      "epoch": 0.12243593139466699,
+      "grad_norm": 0.4343515634536743,
+      "learning_rate": 4.675251604310116e-06,
+      "loss": 0.1345563769340515,
+      "step": 36190
+    },
+    {
+      "epoch": 0.12272821101126921,
+      "grad_norm": 0.48489269614219666,
+      "learning_rate": 4.650939966285439e-06,
+      "loss": 0.08664263486862182,
+      "step": 36200
+    },
+    {
+      "epoch": 0.12302049062787142,
+      "grad_norm": 0.41283851861953735,
+      "learning_rate": 4.6266901992306325e-06,
+      "loss": 0.11822358369827271,
+      "step": 36210
+    },
+    {
+      "epoch": 0.12331277024447364,
+      "grad_norm": 0.37946948409080505,
+      "learning_rate": 4.602502318881119e-06,
+      "loss": 0.10721222162246705,
+      "step": 36220
+    },
+    {
+      "epoch": 0.12360504986107584,
+      "grad_norm": 0.5040225982666016,
+      "learning_rate": 4.578376340932123e-06,
+      "loss": 0.21010353565216064,
+      "step": 36230
+    },
+    {
+      "epoch": 0.12389732947767806,
+      "grad_norm": 0.34812042117118835,
+      "learning_rate": 4.55431228103873e-06,
+      "loss": 0.1397791624069214,
+      "step": 36240
+    },
+    {
+      "epoch": 0.12418960909428027,
+      "grad_norm": 0.4657280445098877,
+      "learning_rate": 4.5303101548158755e-06,
+      "loss": 0.17170748710632325,
+      "step": 36250
+    },
+    {
+      "epoch": 0.12448188871088248,
+      "grad_norm": 0.931492269039154,
+      "learning_rate": 4.506369977838254e-06,
+      "loss": 0.1901220917701721,
+      "step": 36260
+    },
+    {
+      "epoch": 0.1247741683274847,
+      "grad_norm": 0.646689236164093,
+      "learning_rate": 4.482491765640395e-06,
+      "loss": 0.09852958917617798,
+      "step": 36270
+    },
+    {
+      "epoch": 0.12506644794408692,
+      "grad_norm": 0.528256893157959,
+      "learning_rate": 4.458675533716605e-06,
+      "loss": 0.15075598955154418,
+      "step": 36280
+    },
+    {
+      "epoch": 0.1253587275606891,
+      "grad_norm": 0.6852626204490662,
+      "learning_rate": 4.434921297521e-06,
+      "loss": 0.14908047914505004,
+      "step": 36290
+    },
+    {
+      "epoch": 0.12565100717729133,
+      "grad_norm": 0.5609096884727478,
+      "learning_rate": 4.4112290724674154e-06,
+      "loss": 0.1450022578239441,
+      "step": 36300
+    },
+    {
+      "epoch": 0.12594328679389355,
+      "grad_norm": 0.6882448196411133,
+      "learning_rate": 4.387598873929499e-06,
+      "loss": 0.16208202838897706,
+      "step": 36310
+    },
+    {
+      "epoch": 0.12623556641049577,
+      "grad_norm": 0.6265836358070374,
+      "learning_rate": 4.364030717240631e-06,
+      "loss": 0.1573137640953064,
+      "step": 36320
+    },
+    {
+      "epoch": 0.12652784602709796,
+      "grad_norm": 0.48220837116241455,
+      "learning_rate": 4.340524617693897e-06,
+      "loss": 0.12417106628417969,
+      "step": 36330
+    },
+    {
+      "epoch": 0.12682012564370018,
+      "grad_norm": 0.5380669832229614,
+      "learning_rate": 4.317080590542211e-06,
+      "loss": 0.1818043112754822,
+      "step": 36340
+    },
+    {
+      "epoch": 0.1271124052603024,
+      "grad_norm": 0.3300946354866028,
+      "learning_rate": 4.293698650998112e-06,
+      "loss": 0.09733134508132935,
+      "step": 36350
+    },
+    {
+      "epoch": 0.12740468487690462,
+      "grad_norm": 0.2969580590724945,
+      "learning_rate": 4.270378814233866e-06,
+      "loss": 0.09416690468788147,
+      "step": 36360
+    },
+    {
+      "epoch": 0.12769696449350681,
+      "grad_norm": 0.5740458369255066,
+      "learning_rate": 4.2471210953815125e-06,
+      "loss": 0.13978064060211182,
+      "step": 36370
+    },
+    {
+      "epoch": 0.12798924411010904,
+      "grad_norm": 0.4273037910461426,
+      "learning_rate": 4.223925509532689e-06,
+      "loss": 0.14492490291595458,
+      "step": 36380
+    },
+    {
+      "epoch": 0.12828152372671126,
+      "grad_norm": 0.5599185228347778,
+      "learning_rate": 4.20079207173878e-06,
+      "loss": 0.10717883110046386,
+      "step": 36390
+    },
+    {
+      "epoch": 0.12857380334331348,
+      "grad_norm": 0.3812043368816376,
+      "learning_rate": 4.177720797010831e-06,
+      "loss": 0.09398279190063477,
+      "step": 36400
+    },
+    {
+      "epoch": 0.12886608295991567,
+      "grad_norm": 0.5027279257774353,
+      "learning_rate": 4.1547117003195244e-06,
+      "loss": 0.14286571741104126,
+      "step": 36410
+    },
+    {
+      "epoch": 0.1291583625765179,
+      "grad_norm": 0.5669628977775574,
+      "learning_rate": 4.1317647965952056e-06,
+      "loss": 0.07535434365272523,
+      "step": 36420
+    },
+    {
+      "epoch": 0.1294506421931201,
+      "grad_norm": 0.5710353851318359,
+      "learning_rate": 4.108880100727885e-06,
+      "loss": 0.167682945728302,
+      "step": 36430
+    },
+    {
+      "epoch": 0.12974292180972233,
+      "grad_norm": 0.4249417185783386,
+      "learning_rate": 4.086057627567208e-06,
+      "loss": 0.14893131256103515,
+      "step": 36440
+    },
+    {
+      "epoch": 0.13003520142632452,
+      "grad_norm": 0.32661864161491394,
+      "learning_rate": 4.0632973919224046e-06,
+      "loss": 0.11757429838180541,
+      "step": 36450
+    },
+    {
+      "epoch": 0.13032748104292674,
+      "grad_norm": 0.3804573118686676,
+      "learning_rate": 4.040599408562351e-06,
+      "loss": 0.11865862607955932,
+      "step": 36460
+    },
+    {
+      "epoch": 0.13061976065952896,
+      "grad_norm": 0.45721957087516785,
+      "learning_rate": 4.017963692215532e-06,
+      "loss": 0.11639668941497802,
+      "step": 36470
+    },
+    {
+      "epoch": 0.13091204027613118,
+      "grad_norm": 0.32086679339408875,
+      "learning_rate": 3.995390257570009e-06,
+      "loss": 0.13833825588226317,
+      "step": 36480
+    },
+    {
+      "epoch": 0.13120431989273337,
+      "grad_norm": 0.35988515615463257,
+      "learning_rate": 3.972879119273431e-06,
+      "loss": 0.11501911878585816,
+      "step": 36490
+    },
+    {
+      "epoch": 0.1314965995093356,
+      "grad_norm": 0.5318719148635864,
+      "learning_rate": 3.95043029193306e-06,
+      "loss": 0.1376611351966858,
+      "step": 36500
+    },
+    {
+      "epoch": 0.1317888791259378,
+      "grad_norm": 0.34844911098480225,
+      "learning_rate": 3.928043790115654e-06,
+      "loss": 0.1480000138282776,
+      "step": 36510
+    },
+    {
+      "epoch": 0.13208115874254,
+      "grad_norm": 0.5595666766166687,
+      "learning_rate": 3.905719628347604e-06,
+      "loss": 0.12417954206466675,
+      "step": 36520
+    },
+    {
+      "epoch": 0.13237343835914223,
+      "grad_norm": 0.559913158416748,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 0.14163187742233277,
+      "step": 36530
+    },
+    {
+      "epoch": 0.13266571797574445,
+      "grad_norm": 0.3621932566165924,
+      "learning_rate": 3.861258382862721e-06,
+      "loss": 0.15181828737258912,
+      "step": 36540
+    },
+    {
+      "epoch": 0.13295799759234667,
+      "grad_norm": 0.5108553767204285,
+      "learning_rate": 3.839121327996275e-06,
+      "loss": 0.1439156413078308,
+      "step": 36550
+    },
+    {
+      "epoch": 0.13325027720894886,
+      "grad_norm": 0.4530680775642395,
+      "learning_rate": 3.817046670880031e-06,
+      "loss": 0.10560730695724488,
+      "step": 36560
+    },
+    {
+      "epoch": 0.13354255682555108,
+      "grad_norm": 0.5120121836662292,
+      "learning_rate": 3.795034425837951e-06,
+      "loss": 0.13006887435913086,
+      "step": 36570
+    },
+    {
+      "epoch": 0.1338348364421533,
+      "grad_norm": 0.3312967121601105,
+      "learning_rate": 3.773084607153521e-06,
+      "loss": 0.15030533075332642,
+      "step": 36580
+    },
+    {
+      "epoch": 0.13412711605875552,
+      "grad_norm": 0.4514562785625458,
+      "learning_rate": 3.751197229069803e-06,
+      "loss": 0.12111517190933227,
+      "step": 36590
+    },
+    {
+      "epoch": 0.1344193956753577,
+      "grad_norm": 0.4732087552547455,
+      "learning_rate": 3.729372305789225e-06,
+      "loss": 0.1283136010169983,
+      "step": 36600
+    },
+    {
+      "epoch": 0.13471167529195993,
+      "grad_norm": 0.5528901815414429,
+      "learning_rate": 3.7076098514737702e-06,
+      "loss": 0.12460596561431884,
+      "step": 36610
+    },
+    {
+      "epoch": 0.13500395490856215,
+      "grad_norm": 0.4937191605567932,
+      "learning_rate": 3.6859098802448645e-06,
+      "loss": 0.11625337600708008,
+      "step": 36620
+    },
+    {
+      "epoch": 0.13529623452516437,
+      "grad_norm": 0.5909830927848816,
+      "learning_rate": 3.6642724061833444e-06,
+      "loss": 0.13173744678497315,
+      "step": 36630
+    },
+    {
+      "epoch": 0.13558851414176656,
+      "grad_norm": 0.45072680711746216,
+      "learning_rate": 3.642697443329568e-06,
+      "loss": 0.17165329456329345,
+      "step": 36640
+    },
+    {
+      "epoch": 0.13588079375836878,
+      "grad_norm": 0.38801681995391846,
+      "learning_rate": 3.6211850056832806e-06,
+      "loss": 0.13696666955947875,
+      "step": 36650
+    },
+    {
+      "epoch": 0.136173073374971,
+      "grad_norm": 0.5639668703079224,
+      "learning_rate": 3.5997351072036945e-06,
+      "loss": 0.11051816940307617,
+      "step": 36660
+    },
+    {
+      "epoch": 0.13646535299157322,
+      "grad_norm": 0.49362123012542725,
+      "learning_rate": 3.578347761809375e-06,
+      "loss": 0.10341551303863525,
+      "step": 36670
+    },
+    {
+      "epoch": 0.13675763260817542,
+      "grad_norm": 0.7838643789291382,
+      "learning_rate": 3.5570229833783553e-06,
+      "loss": 0.09792516827583313,
+      "step": 36680
+    },
+    {
+      "epoch": 0.13704991222477764,
+      "grad_norm": 0.49488916993141174,
+      "learning_rate": 3.5357607857480767e-06,
+      "loss": 0.17155238389968872,
+      "step": 36690
+    },
+    {
+      "epoch": 0.13734219184137986,
+      "grad_norm": 0.5723786354064941,
+      "learning_rate": 3.5145611827153034e-06,
+      "loss": 0.13445619344711304,
+      "step": 36700
+    },
+    {
+      "epoch": 0.13763447145798208,
+      "grad_norm": 0.38268840312957764,
+      "learning_rate": 3.4934241880362652e-06,
+      "loss": 0.12913683652877808,
+      "step": 36710
+    },
+    {
+      "epoch": 0.13792675107458427,
+      "grad_norm": 0.36367568373680115,
+      "learning_rate": 3.472349815426512e-06,
+      "loss": 0.13956739902496337,
+      "step": 36720
+    },
+    {
+      "epoch": 0.1382190306911865,
+      "grad_norm": 0.5901442766189575,
+      "learning_rate": 3.4513380785609617e-06,
+      "loss": 0.1865132689476013,
+      "step": 36730
+    },
+    {
+      "epoch": 0.1385113103077887,
+      "grad_norm": 0.4285356104373932,
+      "learning_rate": 3.4303889910739407e-06,
+      "loss": 0.1700521230697632,
+      "step": 36740
+    },
+    {
+      "epoch": 0.13880358992439093,
+      "grad_norm": 0.5145702362060547,
+      "learning_rate": 3.409502566559053e-06,
+      "loss": 0.07043490409851075,
+      "step": 36750
+    },
+    {
+      "epoch": 0.13909586954099312,
+      "grad_norm": 0.40982016921043396,
+      "learning_rate": 3.3886788185692797e-06,
+      "loss": 0.11828362941741943,
+      "step": 36760
+    },
+    {
+      "epoch": 0.13938814915759534,
+      "grad_norm": 0.6130247116088867,
+      "learning_rate": 3.3679177606169233e-06,
+      "loss": 0.18113901615142822,
+      "step": 36770
+    },
+    {
+      "epoch": 0.13968042877419756,
+      "grad_norm": 0.5512343049049377,
+      "learning_rate": 3.3472194061736297e-06,
+      "loss": 0.11042602062225342,
+      "step": 36780
+    },
+    {
+      "epoch": 0.13997270839079975,
+      "grad_norm": 0.6605775356292725,
+      "learning_rate": 3.3265837686703106e-06,
+      "loss": 0.21308441162109376,
+      "step": 36790
+    },
+    {
+      "epoch": 0.14026498800740198,
+      "grad_norm": 0.45100513100624084,
+      "learning_rate": 3.3060108614972106e-06,
+      "loss": 0.1455661177635193,
+      "step": 36800
+    },
+    {
+      "epoch": 0.1405572676240042,
+      "grad_norm": 0.5949532985687256,
+      "learning_rate": 3.285500698003874e-06,
+      "loss": 0.13783363103866578,
+      "step": 36810
+    },
+    {
+      "epoch": 0.14084954724060642,
+      "grad_norm": 0.715850830078125,
+      "learning_rate": 3.2650532914991205e-06,
+      "loss": 0.15755876302719116,
+      "step": 36820
+    },
+    {
+      "epoch": 0.1411418268572086,
+      "grad_norm": 0.6192346215248108,
+      "learning_rate": 3.2446686552510486e-06,
+      "loss": 0.13309574127197266,
+      "step": 36830
+    },
+    {
+      "epoch": 0.14143410647381083,
+      "grad_norm": 0.3527209460735321,
+      "learning_rate": 3.224346802487044e-06,
+      "loss": 0.13381574153900147,
+      "step": 36840
+    },
+    {
+      "epoch": 0.14172638609041305,
+      "grad_norm": 0.4954020380973816,
+      "learning_rate": 3.2040877463937024e-06,
+      "loss": 0.13602094650268554,
+      "step": 36850
+    },
+    {
+      "epoch": 0.14201866570701527,
+      "grad_norm": 0.46751344203948975,
+      "learning_rate": 3.18389150011692e-06,
+      "loss": 0.16339893341064454,
+      "step": 36860
+    },
+    {
+      "epoch": 0.14231094532361746,
+      "grad_norm": 0.5027575492858887,
+      "learning_rate": 3.163758076761836e-06,
+      "loss": 0.12312878370285034,
+      "step": 36870
+    },
+    {
+      "epoch": 0.14260322494021968,
+      "grad_norm": 0.8673272728919983,
+      "learning_rate": 3.143687489392799e-06,
+      "loss": 0.1692042350769043,
+      "step": 36880
+    },
+    {
+      "epoch": 0.1428955045568219,
+      "grad_norm": 0.4589051306247711,
+      "learning_rate": 3.1236797510334038e-06,
+      "loss": 0.15570673942565919,
+      "step": 36890
+    },
+    {
+      "epoch": 0.14318778417342412,
+      "grad_norm": 0.44759368896484375,
+      "learning_rate": 3.103734874666431e-06,
+      "loss": 0.14395030736923217,
+      "step": 36900
+    },
+    {
+      "epoch": 0.1434800637900263,
+      "grad_norm": 0.5873273015022278,
+      "learning_rate": 3.083852873233928e-06,
+      "loss": 0.15299232006073,
+      "step": 36910
+    },
+    {
+      "epoch": 0.14377234340662853,
+      "grad_norm": 0.35499775409698486,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 0.20767335891723632,
+      "step": 36920
+    },
+    {
+      "epoch": 0.14406462302323075,
+      "grad_norm": 0.34108811616897583,
+      "learning_rate": 3.044277546736307e-06,
+      "loss": 0.13296178579330445,
+      "step": 36930
+    },
+    {
+      "epoch": 0.14435690263983297,
+      "grad_norm": 0.49674472212791443,
+      "learning_rate": 3.024584247351203e-06,
+      "loss": 0.15179860591888428,
+      "step": 36940
+    },
+    {
+      "epoch": 0.14464918225643517,
+      "grad_norm": 0.48936575651168823,
+      "learning_rate": 3.0049538742605298e-06,
+      "loss": 0.15211007595062256,
+      "step": 36950
+    },
+    {
+      "epoch": 0.14494146187303739,
+      "grad_norm": 0.49640047550201416,
+      "learning_rate": 2.9853864402022314e-06,
+      "loss": 0.15404913425445557,
+      "step": 36960
+    },
+    {
+      "epoch": 0.1452337414896396,
+      "grad_norm": 0.43722882866859436,
+      "learning_rate": 2.9658819578733955e-06,
+      "loss": 0.1817418098449707,
+      "step": 36970
+    },
+    {
+      "epoch": 0.14552602110624183,
+      "grad_norm": 0.5036656260490417,
+      "learning_rate": 2.9464404399302757e-06,
+      "loss": 0.1368376135826111,
+      "step": 36980
+    },
+    {
+      "epoch": 0.14581830072284402,
+      "grad_norm": 0.6423154473304749,
+      "learning_rate": 2.92706189898827e-06,
+      "loss": 0.16258559226989747,
+      "step": 36990
+    },
+    {
+      "epoch": 0.14611058033944624,
+      "grad_norm": 0.44569382071495056,
+      "learning_rate": 2.9077463476218757e-06,
+      "loss": 0.136662757396698,
+      "step": 37000
+    },
+    {
+      "epoch": 1.000263051654942,
+      "grad_norm": 1.1045737266540527,
+      "learning_rate": 2.8884937983647664e-06,
+      "loss": 0.497054311964247,
+      "step": 37010
+    },
+    {
+      "epoch": 1.0005553312715443,
+      "grad_norm": 1.6590017080307007,
+      "learning_rate": 2.8693042637097045e-06,
+      "loss": 0.4164711952209473,
+      "step": 37020
+    },
+    {
+      "epoch": 1.0008476108881463,
+      "grad_norm": 0.8531391620635986,
+      "learning_rate": 2.850177756108585e-06,
+      "loss": 0.3985168695449829,
+      "step": 37030
+    },
+    {
+      "epoch": 1.0011398905047486,
+      "grad_norm": 0.9973299503326416,
+      "learning_rate": 2.8311142879723694e-06,
+      "loss": 0.365405797958374,
+      "step": 37040
+    },
+    {
+      "epoch": 1.0014321701213509,
+      "grad_norm": 0.8639014363288879,
+      "learning_rate": 2.8121138716711404e-06,
+      "loss": 0.36868724822998045,
+      "step": 37050
+    },
+    {
+      "epoch": 1.0017244497379532,
+      "grad_norm": 0.6027888059616089,
+      "learning_rate": 2.79317651953408e-06,
+      "loss": 0.3076002597808838,
+      "step": 37060
+    },
+    {
+      "epoch": 1.0020167293545552,
+      "grad_norm": 0.9892123937606812,
+      "learning_rate": 2.7743022438494136e-06,
+      "loss": 0.38199853897094727,
+      "step": 37070
+    },
+    {
+      "epoch": 1.0023090089711575,
+      "grad_norm": 0.7430927157402039,
+      "learning_rate": 2.7554910568644453e-06,
+      "loss": 0.3823023080825806,
+      "step": 37080
+    },
+    {
+      "epoch": 1.0026012885877598,
+      "grad_norm": 0.691180944442749,
+      "learning_rate": 2.736742970785566e-06,
+      "loss": 0.3490960121154785,
+      "step": 37090
+    },
+    {
+      "epoch": 1.0028935682043618,
+      "grad_norm": 0.6556819677352905,
+      "learning_rate": 2.718057997778212e-06,
+      "loss": 0.3388651371002197,
+      "step": 37100
+    },
+    {
+      "epoch": 1.003185847820964,
+      "grad_norm": 0.6700540781021118,
+      "learning_rate": 2.6994361499668387e-06,
+      "loss": 0.32860729694366453,
+      "step": 37110
+    },
+    {
+      "epoch": 1.0034781274375664,
+      "grad_norm": 0.5523332953453064,
+      "learning_rate": 2.68087743943497e-06,
+      "loss": 0.28271989822387694,
+      "step": 37120
+    },
+    {
+      "epoch": 1.0037704070541686,
+      "grad_norm": 0.6754739284515381,
+      "learning_rate": 2.662381878225162e-06,
+      "loss": 0.4046600341796875,
+      "step": 37130
+    },
+    {
+      "epoch": 1.0040626866707707,
+      "grad_norm": 0.8212656378746033,
+      "learning_rate": 2.6439494783389473e-06,
+      "loss": 0.34784755706787107,
+      "step": 37140
+    },
+    {
+      "epoch": 1.004354966287373,
+      "grad_norm": 0.8965246081352234,
+      "learning_rate": 2.62558025173697e-06,
+      "loss": 0.3726522922515869,
+      "step": 37150
+    },
+    {
+      "epoch": 1.0046472459039753,
+      "grad_norm": 0.5125054121017456,
+      "learning_rate": 2.607274210338784e-06,
+      "loss": 0.30754828453063965,
+      "step": 37160
+    },
+    {
+      "epoch": 1.0049395255205773,
+      "grad_norm": 0.7094303369522095,
+      "learning_rate": 2.5890313660229535e-06,
+      "loss": 0.2727017402648926,
+      "step": 37170
+    },
+    {
+      "epoch": 1.0052318051371796,
+      "grad_norm": 0.8346501588821411,
+      "learning_rate": 2.570851730627122e-06,
+      "loss": 0.3407344579696655,
+      "step": 37180
+    },
+    {
+      "epoch": 1.0055240847537819,
+      "grad_norm": 0.7564747333526611,
+      "learning_rate": 2.5527353159478295e-06,
+      "loss": 0.3224808216094971,
+      "step": 37190
+    },
+    {
+      "epoch": 1.0058163643703841,
+      "grad_norm": 0.5166079998016357,
+      "learning_rate": 2.534682133740607e-06,
+      "loss": 0.3936962604522705,
+      "step": 37200
+    },
+    {
+      "epoch": 1.0061086439869862,
+      "grad_norm": 0.8116148114204407,
+      "learning_rate": 2.5166921957199942e-06,
+      "loss": 0.36410577297210694,
+      "step": 37210
+    },
+    {
+      "epoch": 1.0064009236035885,
+      "grad_norm": 0.5255950689315796,
+      "learning_rate": 2.4987655135594532e-06,
+      "loss": 0.37209329605102537,
+      "step": 37220
+    },
+    {
+      "epoch": 1.0066932032201907,
+      "grad_norm": 0.8289549946784973,
+      "learning_rate": 2.4809020988914023e-06,
+      "loss": 0.3080328941345215,
+      "step": 37230
+    },
+    {
+      "epoch": 1.0069854828367928,
+      "grad_norm": 0.5827141404151917,
+      "learning_rate": 2.463101963307235e-06,
+      "loss": 0.3151430606842041,
+      "step": 37240
+    },
+    {
+      "epoch": 1.007277762453395,
+      "grad_norm": 0.7312423586845398,
+      "learning_rate": 2.445365118357279e-06,
+      "loss": 0.3921444654464722,
+      "step": 37250
+    },
+    {
+      "epoch": 1.0075700420699973,
+      "grad_norm": 0.7592278718948364,
+      "learning_rate": 2.427691575550739e-06,
+      "loss": 0.3230127811431885,
+      "step": 37260
+    },
+    {
+      "epoch": 1.0078623216865996,
+      "grad_norm": 0.6767473816871643,
+      "learning_rate": 2.410081346355819e-06,
+      "loss": 0.35103812217712405,
+      "step": 37270
+    },
+    {
+      "epoch": 1.0081546013032017,
+      "grad_norm": 0.7218213081359863,
+      "learning_rate": 2.392534442199601e-06,
+      "loss": 0.36617794036865237,
+      "step": 37280
+    },
+    {
+      "epoch": 1.008446880919804,
+      "grad_norm": 0.5789059996604919,
+      "learning_rate": 2.3750508744680545e-06,
+      "loss": 0.3292683124542236,
+      "step": 37290
+    },
+    {
+      "epoch": 1.0087391605364062,
+      "grad_norm": 0.49887555837631226,
+      "learning_rate": 2.3576306545061044e-06,
+      "loss": 0.357650899887085,
+      "step": 37300
+    },
+    {
+      "epoch": 1.0090314401530083,
+      "grad_norm": 0.5848112106323242,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 0.32222397327423097,
+      "step": 37310
+    },
+    {
+      "epoch": 1.0093237197696106,
+      "grad_norm": 0.6707907915115356,
+      "learning_rate": 2.3229803030650367e-06,
+      "loss": 0.33132352828979494,
+      "step": 37320
+    },
+    {
+      "epoch": 1.0096159993862128,
+      "grad_norm": 0.47345343232154846,
+      "learning_rate": 2.3057501940701445e-06,
+      "loss": 0.3688469409942627,
+      "step": 37330
+    },
+    {
+      "epoch": 1.009908279002815,
+      "grad_norm": 0.7968910932540894,
+      "learning_rate": 2.2885834778132995e-06,
+      "loss": 0.32636706829071044,
+      "step": 37340
+    },
+    {
+      "epoch": 1.0102005586194172,
+      "grad_norm": 0.5237778425216675,
+      "learning_rate": 2.271480165433815e-06,
+      "loss": 0.33920166492462156,
+      "step": 37350
+    },
+    {
+      "epoch": 1.0104928382360194,
+      "grad_norm": 0.4079629182815552,
+      "learning_rate": 2.2544402680298007e-06,
+      "loss": 0.3020551443099976,
+      "step": 37360
+    },
+    {
+      "epoch": 1.0107851178526217,
+      "grad_norm": 0.43953534960746765,
+      "learning_rate": 2.2374637966583124e-06,
+      "loss": 0.4204400539398193,
+      "step": 37370
+    },
+    {
+      "epoch": 1.0110773974692238,
+      "grad_norm": 0.39976966381073,
+      "learning_rate": 2.2205507623351606e-06,
+      "loss": 0.29968283176422117,
+      "step": 37380
+    },
+    {
+      "epoch": 1.011369677085826,
+      "grad_norm": 0.7851330041885376,
+      "learning_rate": 2.2037011760350335e-06,
+      "loss": 0.3519948244094849,
+      "step": 37390
+    },
+    {
+      "epoch": 1.0116619567024283,
+      "grad_norm": 0.6299710273742676,
+      "learning_rate": 2.1869150486914734e-06,
+      "loss": 0.3716180086135864,
+      "step": 37400
+    },
+    {
+      "epoch": 1.0119542363190306,
+      "grad_norm": 0.5160528421401978,
+      "learning_rate": 2.1701923911967903e-06,
+      "loss": 0.2992725133895874,
+      "step": 37410
+    },
+    {
+      "epoch": 1.0122465159356326,
+      "grad_norm": 0.8031061887741089,
+      "learning_rate": 2.153533214402137e-06,
+      "loss": 0.33388044834136965,
+      "step": 37420
+    },
+    {
+      "epoch": 1.012538795552235,
+      "grad_norm": 0.6910454034805298,
+      "learning_rate": 2.136937529117489e-06,
+      "loss": 0.4381894588470459,
+      "step": 37430
+    },
+    {
+      "epoch": 1.0128310751688372,
+      "grad_norm": 0.4338637590408325,
+      "learning_rate": 2.120405346111576e-06,
+      "loss": 0.29959099292755126,
+      "step": 37440
+    },
+    {
+      "epoch": 1.0131233547854395,
+      "grad_norm": 0.42582347989082336,
+      "learning_rate": 2.103936676111984e-06,
+      "loss": 0.38194172382354735,
+      "step": 37450
+    },
+    {
+      "epoch": 1.0134156344020415,
+      "grad_norm": 0.6750156283378601,
+      "learning_rate": 2.0875315298050314e-06,
+      "loss": 0.3878228425979614,
+      "step": 37460
+    },
+    {
+      "epoch": 1.0137079140186438,
+      "grad_norm": 0.40968164801597595,
+      "learning_rate": 2.0711899178358584e-06,
+      "loss": 0.32732059955596926,
+      "step": 37470
+    },
+    {
+      "epoch": 1.014000193635246,
+      "grad_norm": 0.5351089239120483,
+      "learning_rate": 2.05491185080835e-06,
+      "loss": 0.2601869344711304,
+      "step": 37480
+    },
+    {
+      "epoch": 1.0142924732518481,
+      "grad_norm": 0.7036119103431702,
+      "learning_rate": 2.038697339285156e-06,
+      "loss": 0.344208288192749,
+      "step": 37490
+    },
+    {
+      "epoch": 1.0145847528684504,
+      "grad_norm": 0.37079668045043945,
+      "learning_rate": 2.0225463937877167e-06,
+      "loss": 0.3513414144515991,
+      "step": 37500
+    },
+    {
+      "epoch": 1.0148770324850527,
+      "grad_norm": 0.5795589089393616,
+      "learning_rate": 2.006459024796192e-06,
+      "loss": 0.32459030151367185,
+      "step": 37510
+    },
+    {
+      "epoch": 1.015169312101655,
+      "grad_norm": 0.5426260232925415,
+      "learning_rate": 1.990435242749511e-06,
+      "loss": 0.38744094371795657,
+      "step": 37520
+    },
+    {
+      "epoch": 1.015461591718257,
+      "grad_norm": 0.3786596655845642,
+      "learning_rate": 1.974475058045322e-06,
+      "loss": 0.3555562734603882,
+      "step": 37530
+    },
+    {
+      "epoch": 1.0157538713348593,
+      "grad_norm": 0.45258283615112305,
+      "learning_rate": 1.958578481040019e-06,
+      "loss": 0.39402992725372316,
+      "step": 37540
+    },
+    {
+      "epoch": 1.0160461509514616,
+      "grad_norm": 1.139046549797058,
+      "learning_rate": 1.942745522048728e-06,
+      "loss": 0.36787264347076415,
+      "step": 37550
+    },
+    {
+      "epoch": 1.0163384305680636,
+      "grad_norm": 0.49008938670158386,
+      "learning_rate": 1.9269761913452642e-06,
+      "loss": 0.35456874370574953,
+      "step": 37560
+    },
+    {
+      "epoch": 1.0166307101846659,
+      "grad_norm": 0.5575117468833923,
+      "learning_rate": 1.911270499162199e-06,
+      "loss": 0.3592578172683716,
+      "step": 37570
+    },
+    {
+      "epoch": 1.0169229898012682,
+      "grad_norm": 0.6462110280990601,
+      "learning_rate": 1.8956284556907455e-06,
+      "loss": 0.36372694969177244,
+      "step": 37580
+    },
+    {
+      "epoch": 1.0172152694178704,
+      "grad_norm": 0.439432293176651,
+      "learning_rate": 1.8800500710809188e-06,
+      "loss": 0.2770462274551392,
+      "step": 37590
+    },
+    {
+      "epoch": 1.0175075490344725,
+      "grad_norm": 0.3852550983428955,
+      "learning_rate": 1.8645353554413102e-06,
+      "loss": 0.3770076036453247,
+      "step": 37600
+    },
+    {
+      "epoch": 1.0177998286510748,
+      "grad_norm": 0.44788751006126404,
+      "learning_rate": 1.849084318839256e-06,
+      "loss": 0.3554239749908447,
+      "step": 37610
+    },
+    {
+      "epoch": 1.018092108267677,
+      "grad_norm": 0.5140615105628967,
+      "learning_rate": 1.8336969713008024e-06,
+      "loss": 0.3046652555465698,
+      "step": 37620
+    },
+    {
+      "epoch": 1.018384387884279,
+      "grad_norm": 0.4333508312702179,
+      "learning_rate": 1.818373322810596e-06,
+      "loss": 0.35647175312042234,
+      "step": 37630
+    },
+    {
+      "epoch": 1.0186766675008814,
+      "grad_norm": 0.5301503539085388,
+      "learning_rate": 1.8031133833119939e-06,
+      "loss": 0.36941120624542234,
+      "step": 37640
+    },
+    {
+      "epoch": 1.0189689471174836,
+      "grad_norm": 0.6098165512084961,
+      "learning_rate": 1.7879171627070202e-06,
+      "loss": 0.33992278575897217,
+      "step": 37650
+    },
+    {
+      "epoch": 1.019261226734086,
+      "grad_norm": 0.5786981582641602,
+      "learning_rate": 1.77278467085632e-06,
+      "loss": 0.33217971324920653,
+      "step": 37660
+    },
+    {
+      "epoch": 1.019553506350688,
+      "grad_norm": 0.4249230623245239,
+      "learning_rate": 1.7577159175792057e-06,
+      "loss": 0.30268595218658445,
+      "step": 37670
+    },
+    {
+      "epoch": 1.0198457859672903,
+      "grad_norm": 0.3292296528816223,
+      "learning_rate": 1.7427109126536333e-06,
+      "loss": 0.31471521854400636,
+      "step": 37680
+    },
+    {
+      "epoch": 1.0201380655838925,
+      "grad_norm": 0.42141005396842957,
+      "learning_rate": 1.7277696658161924e-06,
+      "loss": 0.3457106351852417,
+      "step": 37690
+    },
+    {
+      "epoch": 1.0204303452004946,
+      "grad_norm": 0.5575936436653137,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 0.33511412143707275,
+      "step": 37700
+    },
+    {
+      "epoch": 1.0207226248170969,
+      "grad_norm": 0.5327735543251038,
+      "learning_rate": 1.6980784851451381e-06,
+      "loss": 0.3659277200698853,
+      "step": 37710
+    },
+    {
+      "epoch": 1.0210149044336991,
+      "grad_norm": 0.3761720359325409,
+      "learning_rate": 1.6833285705778356e-06,
+      "loss": 0.28815226554870604,
+      "step": 37720
+    },
+    {
+      "epoch": 1.0213071840503014,
+      "grad_norm": 0.5779837369918823,
+      "learning_rate": 1.6686424526311971e-06,
+      "loss": 0.3303104639053345,
+      "step": 37730
+    },
+    {
+      "epoch": 1.0215994636669035,
+      "grad_norm": 0.5631416440010071,
+      "learning_rate": 1.654020140834911e-06,
+      "loss": 0.3239152908325195,
+      "step": 37740
+    },
+    {
+      "epoch": 1.0218917432835057,
+      "grad_norm": 0.4757640063762665,
+      "learning_rate": 1.6394616446772203e-06,
+      "loss": 0.40555434226989745,
+      "step": 37750
+    },
+    {
+      "epoch": 1.022184022900108,
+      "grad_norm": 0.32517361640930176,
+      "learning_rate": 1.6249669736050022e-06,
+      "loss": 0.3327082633972168,
+      "step": 37760
+    },
+    {
+      "epoch": 1.02247630251671,
+      "grad_norm": 0.5137649774551392,
+      "learning_rate": 1.6105361370236549e-06,
+      "loss": 0.26106321811676025,
+      "step": 37770
+    },
+    {
+      "epoch": 1.0227685821333123,
+      "grad_norm": 0.5665978789329529,
+      "learning_rate": 1.5961691442972105e-06,
+      "loss": 0.3650979518890381,
+      "step": 37780
+    },
+    {
+      "epoch": 1.0230608617499146,
+      "grad_norm": 0.5606212615966797,
+      "learning_rate": 1.581866004748267e-06,
+      "loss": 0.28970022201538087,
+      "step": 37790
+    },
+    {
+      "epoch": 1.023353141366517,
+      "grad_norm": 0.6590144038200378,
+      "learning_rate": 1.5676267276579448e-06,
+      "loss": 0.33405213356018065,
+      "step": 37800
+    },
+    {
+      "epoch": 1.023645420983119,
+      "grad_norm": 0.5126269459724426,
+      "learning_rate": 1.5534513222659864e-06,
+      "loss": 0.3366661787033081,
+      "step": 37810
+    },
+    {
+      "epoch": 1.0239377005997212,
+      "grad_norm": 0.5179246664047241,
+      "learning_rate": 1.5393397977706448e-06,
+      "loss": 0.29231624603271483,
+      "step": 37820
+    },
+    {
+      "epoch": 1.0242299802163235,
+      "grad_norm": 0.4854632616043091,
+      "learning_rate": 1.5252921633287177e-06,
+      "loss": 0.27622959613800047,
+      "step": 37830
+    },
+    {
+      "epoch": 1.0245222598329256,
+      "grad_norm": 0.6228880882263184,
+      "learning_rate": 1.5113084280556023e-06,
+      "loss": 0.3524168968200684,
+      "step": 37840
+    },
+    {
+      "epoch": 1.0248145394495278,
+      "grad_norm": 0.4980785548686981,
+      "learning_rate": 1.4973886010251737e-06,
+      "loss": 0.3284755706787109,
+      "step": 37850
+    },
+    {
+      "epoch": 1.02510681906613,
+      "grad_norm": 0.4193580448627472,
+      "learning_rate": 1.4835326912698399e-06,
+      "loss": 0.2962606430053711,
+      "step": 37860
+    },
+    {
+      "epoch": 1.0253990986827324,
+      "grad_norm": 0.4399389922618866,
+      "learning_rate": 1.4697407077805759e-06,
+      "loss": 0.306509804725647,
+      "step": 37870
+    },
+    {
+      "epoch": 1.0256913782993344,
+      "grad_norm": 0.39663103222846985,
+      "learning_rate": 1.4560126595068336e-06,
+      "loss": 0.3600484609603882,
+      "step": 37880
+    },
+    {
+      "epoch": 1.0259836579159367,
+      "grad_norm": 0.4279068410396576,
+      "learning_rate": 1.4423485553565985e-06,
+      "loss": 0.35353043079376223,
+      "step": 37890
+    },
+    {
+      "epoch": 1.026275937532539,
+      "grad_norm": 0.37096720933914185,
+      "learning_rate": 1.4287484041963672e-06,
+      "loss": 0.31028294563293457,
+      "step": 37900
+    },
+    {
+      "epoch": 1.0265682171491413,
+      "grad_norm": 0.5079541206359863,
+      "learning_rate": 1.4152122148511248e-06,
+      "loss": 0.32904372215270994,
+      "step": 37910
+    },
+    {
+      "epoch": 1.0268604967657433,
+      "grad_norm": 0.4463079273700714,
+      "learning_rate": 1.4017399961043454e-06,
+      "loss": 0.35142791271209717,
+      "step": 37920
+    },
+    {
+      "epoch": 1.0271527763823456,
+      "grad_norm": 0.4648186266422272,
+      "learning_rate": 1.3883317566980247e-06,
+      "loss": 0.31124053001403806,
+      "step": 37930
+    },
+    {
+      "epoch": 1.0274450559989479,
+      "grad_norm": 0.41876286268234253,
+      "learning_rate": 1.3749875053326144e-06,
+      "loss": 0.3389892101287842,
+      "step": 37940
+    },
+    {
+      "epoch": 1.02773733561555,
+      "grad_norm": 0.4941273629665375,
+      "learning_rate": 1.3617072506670548e-06,
+      "loss": 0.3293140888214111,
+      "step": 37950
+    },
+    {
+      "epoch": 1.0280296152321522,
+      "grad_norm": 0.4562717080116272,
+      "learning_rate": 1.3484910013187524e-06,
+      "loss": 0.33832240104675293,
+      "step": 37960
+    },
+    {
+      "epoch": 1.0283218948487545,
+      "grad_norm": 0.4882437586784363,
+      "learning_rate": 1.335338765863614e-06,
+      "loss": 0.364630126953125,
+      "step": 37970
+    },
+    {
+      "epoch": 1.0286141744653567,
+      "grad_norm": 0.6000685095787048,
+      "learning_rate": 1.3222505528359351e-06,
+      "loss": 0.37566494941711426,
+      "step": 37980
+    },
+    {
+      "epoch": 1.0289064540819588,
+      "grad_norm": 0.45487022399902344,
+      "learning_rate": 1.3092263707285668e-06,
+      "loss": 0.324053955078125,
+      "step": 37990
+    },
+    {
+      "epoch": 1.029198733698561,
+      "grad_norm": 0.43845951557159424,
+      "learning_rate": 1.2962662279927262e-06,
+      "loss": 0.3586984395980835,
+      "step": 38000
+    }
+  ],
+  "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": 4.052489394878153e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}