diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.5555555555555556,
+  "epoch": 0.6666666666666666,
   "eval_steps": 9000,
-  "global_step": 25000,
+  "global_step": 30000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -175023,6 +175023,35014 @@
       "learning_rate": 8.890864636585908e-05,
       "loss": 1.0186,
       "step": 25000
+    },
+    {
+      "epoch": 0.5555777777777777,
+      "grad_norm": 1.6018104553222656,
+      "learning_rate": 8.890420093354079e-05,
+      "loss": 2.7778,
+      "step": 25001
+    },
+    {
+      "epoch": 0.5556,
+      "grad_norm": 1.4975850582122803,
+      "learning_rate": 8.88997555012225e-05,
+      "loss": 2.7011,
+      "step": 25002
+    },
+    {
+      "epoch": 0.5556222222222222,
+      "grad_norm": 1.3138421773910522,
+      "learning_rate": 8.889531006890421e-05,
+      "loss": 2.056,
+      "step": 25003
+    },
+    {
+      "epoch": 0.5556444444444445,
+      "grad_norm": 1.6650867462158203,
+      "learning_rate": 8.889086463658592e-05,
+      "loss": 2.3087,
+      "step": 25004
+    },
+    {
+      "epoch": 0.5556666666666666,
+      "grad_norm": 1.5537450313568115,
+      "learning_rate": 8.888641920426763e-05,
+      "loss": 2.1625,
+      "step": 25005
+    },
+    {
+      "epoch": 0.5556888888888889,
+      "grad_norm": 1.5213912725448608,
+      "learning_rate": 8.888197377194932e-05,
+      "loss": 2.3027,
+      "step": 25006
+    },
+    {
+      "epoch": 0.5557111111111112,
+      "grad_norm": 1.4587023258209229,
+      "learning_rate": 8.887752833963103e-05,
+      "loss": 1.7927,
+      "step": 25007
+    },
+    {
+      "epoch": 0.5557333333333333,
+      "grad_norm": 1.5978976488113403,
+      "learning_rate": 8.887308290731274e-05,
+      "loss": 1.9206,
+      "step": 25008
+    },
+    {
+      "epoch": 0.5557555555555556,
+      "grad_norm": 1.3036267757415771,
+      "learning_rate": 8.886863747499444e-05,
+      "loss": 1.5741,
+      "step": 25009
+    },
+    {
+      "epoch": 0.5557777777777778,
+      "grad_norm": 1.5057199001312256,
+      "learning_rate": 8.886419204267615e-05,
+      "loss": 1.9934,
+      "step": 25010
+    },
+    {
+      "epoch": 0.5558,
+      "grad_norm": 1.617224931716919,
+      "learning_rate": 8.885974661035786e-05,
+      "loss": 1.5776,
+      "step": 25011
+    },
+    {
+      "epoch": 0.5558222222222222,
+      "grad_norm": 1.4996505975723267,
+      "learning_rate": 8.885530117803957e-05,
+      "loss": 1.7387,
+      "step": 25012
+    },
+    {
+      "epoch": 0.5558444444444445,
+      "grad_norm": 1.9091581106185913,
+      "learning_rate": 8.885085574572128e-05,
+      "loss": 2.257,
+      "step": 25013
+    },
+    {
+      "epoch": 0.5558666666666666,
+      "grad_norm": 2.1841301918029785,
+      "learning_rate": 8.884641031340299e-05,
+      "loss": 1.6132,
+      "step": 25014
+    },
+    {
+      "epoch": 0.5558888888888889,
+      "grad_norm": 1.050742268562317,
+      "learning_rate": 8.88419648810847e-05,
+      "loss": 0.9285,
+      "step": 25015
+    },
+    {
+      "epoch": 0.5559111111111111,
+      "grad_norm": 1.4610708951950073,
+      "learning_rate": 8.88375194487664e-05,
+      "loss": 1.7434,
+      "step": 25016
+    },
+    {
+      "epoch": 0.5559333333333333,
+      "grad_norm": 1.5033701658248901,
+      "learning_rate": 8.88330740164481e-05,
+      "loss": 1.4852,
+      "step": 25017
+    },
+    {
+      "epoch": 0.5559555555555555,
+      "grad_norm": 1.5321849584579468,
+      "learning_rate": 8.882862858412981e-05,
+      "loss": 1.6994,
+      "step": 25018
+    },
+    {
+      "epoch": 0.5559777777777778,
+      "grad_norm": 1.53279709815979,
+      "learning_rate": 8.882418315181151e-05,
+      "loss": 1.9177,
+      "step": 25019
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 1.8359917402267456,
+      "learning_rate": 8.881973771949322e-05,
+      "loss": 2.1219,
+      "step": 25020
+    },
+    {
+      "epoch": 0.5560222222222222,
+      "grad_norm": 2.4665095806121826,
+      "learning_rate": 8.881529228717493e-05,
+      "loss": 2.1297,
+      "step": 25021
+    },
+    {
+      "epoch": 0.5560444444444445,
+      "grad_norm": 1.7346540689468384,
+      "learning_rate": 8.881084685485664e-05,
+      "loss": 1.7752,
+      "step": 25022
+    },
+    {
+      "epoch": 0.5560666666666667,
+      "grad_norm": 1.6376831531524658,
+      "learning_rate": 8.880640142253835e-05,
+      "loss": 2.2255,
+      "step": 25023
+    },
+    {
+      "epoch": 0.5560888888888889,
+      "grad_norm": 2.1054024696350098,
+      "learning_rate": 8.880195599022006e-05,
+      "loss": 2.4414,
+      "step": 25024
+    },
+    {
+      "epoch": 0.5561111111111111,
+      "grad_norm": 1.8348150253295898,
+      "learning_rate": 8.879751055790177e-05,
+      "loss": 1.9206,
+      "step": 25025
+    },
+    {
+      "epoch": 0.5561333333333334,
+      "grad_norm": 1.9463988542556763,
+      "learning_rate": 8.879306512558346e-05,
+      "loss": 2.6646,
+      "step": 25026
+    },
+    {
+      "epoch": 0.5561555555555555,
+      "grad_norm": 1.4555976390838623,
+      "learning_rate": 8.878861969326517e-05,
+      "loss": 1.7126,
+      "step": 25027
+    },
+    {
+      "epoch": 0.5561777777777778,
+      "grad_norm": 1.2904547452926636,
+      "learning_rate": 8.878417426094688e-05,
+      "loss": 1.0135,
+      "step": 25028
+    },
+    {
+      "epoch": 0.5562,
+      "grad_norm": 1.8080211877822876,
+      "learning_rate": 8.877972882862858e-05,
+      "loss": 2.229,
+      "step": 25029
+    },
+    {
+      "epoch": 0.5562222222222222,
+      "grad_norm": 1.6538316011428833,
+      "learning_rate": 8.877528339631029e-05,
+      "loss": 2.2293,
+      "step": 25030
+    },
+    {
+      "epoch": 0.5562444444444444,
+      "grad_norm": 1.7188304662704468,
+      "learning_rate": 8.877083796399201e-05,
+      "loss": 1.6404,
+      "step": 25031
+    },
+    {
+      "epoch": 0.5562666666666667,
+      "grad_norm": 1.7964560985565186,
+      "learning_rate": 8.876639253167372e-05,
+      "loss": 1.8025,
+      "step": 25032
+    },
+    {
+      "epoch": 0.5562888888888889,
+      "grad_norm": 1.527963399887085,
+      "learning_rate": 8.876194709935542e-05,
+      "loss": 1.8277,
+      "step": 25033
+    },
+    {
+      "epoch": 0.5563111111111111,
+      "grad_norm": 1.5908809900283813,
+      "learning_rate": 8.875750166703713e-05,
+      "loss": 1.4329,
+      "step": 25034
+    },
+    {
+      "epoch": 0.5563333333333333,
+      "grad_norm": 1.6306854486465454,
+      "learning_rate": 8.875305623471884e-05,
+      "loss": 2.056,
+      "step": 25035
+    },
+    {
+      "epoch": 0.5563555555555556,
+      "grad_norm": 1.3418653011322021,
+      "learning_rate": 8.874861080240053e-05,
+      "loss": 1.1482,
+      "step": 25036
+    },
+    {
+      "epoch": 0.5563777777777777,
+      "grad_norm": 1.7620158195495605,
+      "learning_rate": 8.874416537008224e-05,
+      "loss": 1.7257,
+      "step": 25037
+    },
+    {
+      "epoch": 0.5564,
+      "grad_norm": 1.6331011056900024,
+      "learning_rate": 8.873971993776395e-05,
+      "loss": 1.283,
+      "step": 25038
+    },
+    {
+      "epoch": 0.5564222222222223,
+      "grad_norm": 1.7324144840240479,
+      "learning_rate": 8.873527450544565e-05,
+      "loss": 1.9654,
+      "step": 25039
+    },
+    {
+      "epoch": 0.5564444444444444,
+      "grad_norm": 1.7590289115905762,
+      "learning_rate": 8.873082907312737e-05,
+      "loss": 1.881,
+      "step": 25040
+    },
+    {
+      "epoch": 0.5564666666666667,
+      "grad_norm": 1.5783190727233887,
+      "learning_rate": 8.872638364080908e-05,
+      "loss": 1.9342,
+      "step": 25041
+    },
+    {
+      "epoch": 0.5564888888888889,
+      "grad_norm": 1.7162874937057495,
+      "learning_rate": 8.872193820849079e-05,
+      "loss": 1.9172,
+      "step": 25042
+    },
+    {
+      "epoch": 0.5565111111111111,
+      "grad_norm": 1.4815313816070557,
+      "learning_rate": 8.871749277617249e-05,
+      "loss": 1.4027,
+      "step": 25043
+    },
+    {
+      "epoch": 0.5565333333333333,
+      "grad_norm": 1.5845388174057007,
+      "learning_rate": 8.87130473438542e-05,
+      "loss": 1.6193,
+      "step": 25044
+    },
+    {
+      "epoch": 0.5565555555555556,
+      "grad_norm": 1.7816119194030762,
+      "learning_rate": 8.87086019115359e-05,
+      "loss": 1.8064,
+      "step": 25045
+    },
+    {
+      "epoch": 0.5565777777777777,
+      "grad_norm": 1.853499174118042,
+      "learning_rate": 8.87041564792176e-05,
+      "loss": 1.5769,
+      "step": 25046
+    },
+    {
+      "epoch": 0.5566,
+      "grad_norm": 1.7572011947631836,
+      "learning_rate": 8.869971104689931e-05,
+      "loss": 1.5565,
+      "step": 25047
+    },
+    {
+      "epoch": 0.5566222222222222,
+      "grad_norm": 1.6061574220657349,
+      "learning_rate": 8.869526561458102e-05,
+      "loss": 1.5543,
+      "step": 25048
+    },
+    {
+      "epoch": 0.5566444444444445,
+      "grad_norm": 2.0534188747406006,
+      "learning_rate": 8.869082018226273e-05,
+      "loss": 2.089,
+      "step": 25049
+    },
+    {
+      "epoch": 0.5566666666666666,
+      "grad_norm": 1.7749592065811157,
+      "learning_rate": 8.868637474994444e-05,
+      "loss": 1.5679,
+      "step": 25050
+    },
+    {
+      "epoch": 0.5566888888888889,
+      "grad_norm": 1.4333471059799194,
+      "learning_rate": 8.868192931762615e-05,
+      "loss": 2.177,
+      "step": 25051
+    },
+    {
+      "epoch": 0.5567111111111112,
+      "grad_norm": 1.3617366552352905,
+      "learning_rate": 8.867748388530786e-05,
+      "loss": 2.0119,
+      "step": 25052
+    },
+    {
+      "epoch": 0.5567333333333333,
+      "grad_norm": 1.37482488155365,
+      "learning_rate": 8.867303845298955e-05,
+      "loss": 1.9341,
+      "step": 25053
+    },
+    {
+      "epoch": 0.5567555555555556,
+      "grad_norm": 1.5708028078079224,
+      "learning_rate": 8.866859302067126e-05,
+      "loss": 2.3296,
+      "step": 25054
+    },
+    {
+      "epoch": 0.5567777777777778,
+      "grad_norm": 1.42710280418396,
+      "learning_rate": 8.866414758835297e-05,
+      "loss": 2.0739,
+      "step": 25055
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 1.6513372659683228,
+      "learning_rate": 8.865970215603467e-05,
+      "loss": 2.1563,
+      "step": 25056
+    },
+    {
+      "epoch": 0.5568222222222222,
+      "grad_norm": 1.4783315658569336,
+      "learning_rate": 8.865525672371638e-05,
+      "loss": 1.7002,
+      "step": 25057
+    },
+    {
+      "epoch": 0.5568444444444445,
+      "grad_norm": 1.4798046350479126,
+      "learning_rate": 8.865081129139809e-05,
+      "loss": 2.2017,
+      "step": 25058
+    },
+    {
+      "epoch": 0.5568666666666666,
+      "grad_norm": 1.5378378629684448,
+      "learning_rate": 8.86463658590798e-05,
+      "loss": 2.1798,
+      "step": 25059
+    },
+    {
+      "epoch": 0.5568888888888889,
+      "grad_norm": 1.6290702819824219,
+      "learning_rate": 8.864192042676151e-05,
+      "loss": 1.8907,
+      "step": 25060
+    },
+    {
+      "epoch": 0.5569111111111111,
+      "grad_norm": 1.5109094381332397,
+      "learning_rate": 8.863747499444322e-05,
+      "loss": 2.0781,
+      "step": 25061
+    },
+    {
+      "epoch": 0.5569333333333333,
+      "grad_norm": 1.7988399267196655,
+      "learning_rate": 8.863302956212493e-05,
+      "loss": 2.292,
+      "step": 25062
+    },
+    {
+      "epoch": 0.5569555555555555,
+      "grad_norm": 1.8292760848999023,
+      "learning_rate": 8.862858412980662e-05,
+      "loss": 2.1439,
+      "step": 25063
+    },
+    {
+      "epoch": 0.5569777777777778,
+      "grad_norm": 1.753173828125,
+      "learning_rate": 8.862413869748833e-05,
+      "loss": 2.0589,
+      "step": 25064
+    },
+    {
+      "epoch": 0.557,
+      "grad_norm": 1.4849519729614258,
+      "learning_rate": 8.861969326517004e-05,
+      "loss": 1.8589,
+      "step": 25065
+    },
+    {
+      "epoch": 0.5570222222222222,
+      "grad_norm": 1.563999056816101,
+      "learning_rate": 8.861524783285174e-05,
+      "loss": 1.8839,
+      "step": 25066
+    },
+    {
+      "epoch": 0.5570444444444445,
+      "grad_norm": 2.108994245529175,
+      "learning_rate": 8.861080240053345e-05,
+      "loss": 1.4836,
+      "step": 25067
+    },
+    {
+      "epoch": 0.5570666666666667,
+      "grad_norm": 1.462251901626587,
+      "learning_rate": 8.860635696821517e-05,
+      "loss": 2.081,
+      "step": 25068
+    },
+    {
+      "epoch": 0.5570888888888889,
+      "grad_norm": 1.4455344676971436,
+      "learning_rate": 8.860191153589688e-05,
+      "loss": 1.5538,
+      "step": 25069
+    },
+    {
+      "epoch": 0.5571111111111111,
+      "grad_norm": 1.5053396224975586,
+      "learning_rate": 8.859746610357858e-05,
+      "loss": 2.1755,
+      "step": 25070
+    },
+    {
+      "epoch": 0.5571333333333334,
+      "grad_norm": 2.1108238697052,
+      "learning_rate": 8.859302067126029e-05,
+      "loss": 2.1607,
+      "step": 25071
+    },
+    {
+      "epoch": 0.5571555555555555,
+      "grad_norm": 1.5452802181243896,
+      "learning_rate": 8.8588575238942e-05,
+      "loss": 2.1198,
+      "step": 25072
+    },
+    {
+      "epoch": 0.5571777777777778,
+      "grad_norm": 1.305687427520752,
+      "learning_rate": 8.858412980662369e-05,
+      "loss": 1.3974,
+      "step": 25073
+    },
+    {
+      "epoch": 0.5572,
+      "grad_norm": 1.6314971446990967,
+      "learning_rate": 8.85796843743054e-05,
+      "loss": 1.9672,
+      "step": 25074
+    },
+    {
+      "epoch": 0.5572222222222222,
+      "grad_norm": 1.672951340675354,
+      "learning_rate": 8.857523894198711e-05,
+      "loss": 2.2149,
+      "step": 25075
+    },
+    {
+      "epoch": 0.5572444444444444,
+      "grad_norm": 1.5819165706634521,
+      "learning_rate": 8.857079350966881e-05,
+      "loss": 1.5749,
+      "step": 25076
+    },
+    {
+      "epoch": 0.5572666666666667,
+      "grad_norm": 1.9342191219329834,
+      "learning_rate": 8.856634807735053e-05,
+      "loss": 2.265,
+      "step": 25077
+    },
+    {
+      "epoch": 0.5572888888888888,
+      "grad_norm": 1.6798408031463623,
+      "learning_rate": 8.856190264503224e-05,
+      "loss": 2.0166,
+      "step": 25078
+    },
+    {
+      "epoch": 0.5573111111111111,
+      "grad_norm": 1.69929838180542,
+      "learning_rate": 8.855745721271395e-05,
+      "loss": 2.0431,
+      "step": 25079
+    },
+    {
+      "epoch": 0.5573333333333333,
+      "grad_norm": 1.7669845819473267,
+      "learning_rate": 8.855301178039565e-05,
+      "loss": 2.2433,
+      "step": 25080
+    },
+    {
+      "epoch": 0.5573555555555556,
+      "grad_norm": 1.5253289937973022,
+      "learning_rate": 8.854856634807736e-05,
+      "loss": 1.6538,
+      "step": 25081
+    },
+    {
+      "epoch": 0.5573777777777777,
+      "grad_norm": 1.7386434078216553,
+      "learning_rate": 8.854412091575907e-05,
+      "loss": 1.9486,
+      "step": 25082
+    },
+    {
+      "epoch": 0.5574,
+      "grad_norm": 1.0491255521774292,
+      "learning_rate": 8.853967548344076e-05,
+      "loss": 0.7682,
+      "step": 25083
+    },
+    {
+      "epoch": 0.5574222222222223,
+      "grad_norm": 1.6801350116729736,
+      "learning_rate": 8.853523005112247e-05,
+      "loss": 2.12,
+      "step": 25084
+    },
+    {
+      "epoch": 0.5574444444444444,
+      "grad_norm": 1.7507060766220093,
+      "learning_rate": 8.853078461880418e-05,
+      "loss": 2.1324,
+      "step": 25085
+    },
+    {
+      "epoch": 0.5574666666666667,
+      "grad_norm": 1.7973806858062744,
+      "learning_rate": 8.852633918648589e-05,
+      "loss": 2.0655,
+      "step": 25086
+    },
+    {
+      "epoch": 0.5574888888888889,
+      "grad_norm": 1.6366549730300903,
+      "learning_rate": 8.85218937541676e-05,
+      "loss": 2.0583,
+      "step": 25087
+    },
+    {
+      "epoch": 0.5575111111111111,
+      "grad_norm": 1.4576225280761719,
+      "learning_rate": 8.851744832184931e-05,
+      "loss": 1.3184,
+      "step": 25088
+    },
+    {
+      "epoch": 0.5575333333333333,
+      "grad_norm": 2.1194889545440674,
+      "learning_rate": 8.851300288953102e-05,
+      "loss": 1.7683,
+      "step": 25089
+    },
+    {
+      "epoch": 0.5575555555555556,
+      "grad_norm": 1.4802414178848267,
+      "learning_rate": 8.850855745721272e-05,
+      "loss": 1.4868,
+      "step": 25090
+    },
+    {
+      "epoch": 0.5575777777777777,
+      "grad_norm": 1.659393310546875,
+      "learning_rate": 8.850411202489443e-05,
+      "loss": 1.8494,
+      "step": 25091
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 1.835160493850708,
+      "learning_rate": 8.849966659257613e-05,
+      "loss": 1.7385,
+      "step": 25092
+    },
+    {
+      "epoch": 0.5576222222222222,
+      "grad_norm": 1.8487638235092163,
+      "learning_rate": 8.849522116025783e-05,
+      "loss": 1.9471,
+      "step": 25093
+    },
+    {
+      "epoch": 0.5576444444444445,
+      "grad_norm": 1.6787410974502563,
+      "learning_rate": 8.849077572793954e-05,
+      "loss": 1.6944,
+      "step": 25094
+    },
+    {
+      "epoch": 0.5576666666666666,
+      "grad_norm": 1.7631950378417969,
+      "learning_rate": 8.848633029562125e-05,
+      "loss": 1.5408,
+      "step": 25095
+    },
+    {
+      "epoch": 0.5576888888888889,
+      "grad_norm": 1.939127802848816,
+      "learning_rate": 8.848188486330296e-05,
+      "loss": 1.6063,
+      "step": 25096
+    },
+    {
+      "epoch": 0.5577111111111112,
+      "grad_norm": 1.4154397249221802,
+      "learning_rate": 8.847743943098467e-05,
+      "loss": 1.2231,
+      "step": 25097
+    },
+    {
+      "epoch": 0.5577333333333333,
+      "grad_norm": 1.8410276174545288,
+      "learning_rate": 8.847299399866638e-05,
+      "loss": 1.8084,
+      "step": 25098
+    },
+    {
+      "epoch": 0.5577555555555556,
+      "grad_norm": 1.5185855627059937,
+      "learning_rate": 8.846854856634809e-05,
+      "loss": 0.9481,
+      "step": 25099
+    },
+    {
+      "epoch": 0.5577777777777778,
+      "grad_norm": 1.788804292678833,
+      "learning_rate": 8.846410313402978e-05,
+      "loss": 1.6403,
+      "step": 25100
+    },
+    {
+      "epoch": 0.5578,
+      "grad_norm": 1.5832068920135498,
+      "learning_rate": 8.84596577017115e-05,
+      "loss": 1.812,
+      "step": 25101
+    },
+    {
+      "epoch": 0.5578222222222222,
+      "grad_norm": 1.5860427618026733,
+      "learning_rate": 8.84552122693932e-05,
+      "loss": 2.5241,
+      "step": 25102
+    },
+    {
+      "epoch": 0.5578444444444445,
+      "grad_norm": 1.4625405073165894,
+      "learning_rate": 8.84507668370749e-05,
+      "loss": 2.3494,
+      "step": 25103
+    },
+    {
+      "epoch": 0.5578666666666666,
+      "grad_norm": 1.3589277267456055,
+      "learning_rate": 8.844632140475661e-05,
+      "loss": 2.3455,
+      "step": 25104
+    },
+    {
+      "epoch": 0.5578888888888889,
+      "grad_norm": 1.4070804119110107,
+      "learning_rate": 8.844187597243833e-05,
+      "loss": 2.3014,
+      "step": 25105
+    },
+    {
+      "epoch": 0.5579111111111111,
+      "grad_norm": 1.4124137163162231,
+      "learning_rate": 8.843743054012003e-05,
+      "loss": 1.0,
+      "step": 25106
+    },
+    {
+      "epoch": 0.5579333333333333,
+      "grad_norm": 1.5017024278640747,
+      "learning_rate": 8.843298510780174e-05,
+      "loss": 1.8322,
+      "step": 25107
+    },
+    {
+      "epoch": 0.5579555555555555,
+      "grad_norm": 1.6320382356643677,
+      "learning_rate": 8.842853967548345e-05,
+      "loss": 2.1433,
+      "step": 25108
+    },
+    {
+      "epoch": 0.5579777777777778,
+      "grad_norm": 1.7392261028289795,
+      "learning_rate": 8.842409424316516e-05,
+      "loss": 2.0716,
+      "step": 25109
+    },
+    {
+      "epoch": 0.558,
+      "grad_norm": 1.5695531368255615,
+      "learning_rate": 8.841964881084685e-05,
+      "loss": 2.1222,
+      "step": 25110
+    },
+    {
+      "epoch": 0.5580222222222222,
+      "grad_norm": 1.5330922603607178,
+      "learning_rate": 8.841520337852856e-05,
+      "loss": 2.0254,
+      "step": 25111
+    },
+    {
+      "epoch": 0.5580444444444445,
+      "grad_norm": 1.40812087059021,
+      "learning_rate": 8.841075794621027e-05,
+      "loss": 1.6054,
+      "step": 25112
+    },
+    {
+      "epoch": 0.5580666666666667,
+      "grad_norm": 1.6731535196304321,
+      "learning_rate": 8.840631251389197e-05,
+      "loss": 1.9831,
+      "step": 25113
+    },
+    {
+      "epoch": 0.5580888888888889,
+      "grad_norm": 1.7269052267074585,
+      "learning_rate": 8.840186708157369e-05,
+      "loss": 2.197,
+      "step": 25114
+    },
+    {
+      "epoch": 0.5581111111111111,
+      "grad_norm": 1.8623415231704712,
+      "learning_rate": 8.83974216492554e-05,
+      "loss": 1.6876,
+      "step": 25115
+    },
+    {
+      "epoch": 0.5581333333333334,
+      "grad_norm": 1.8906834125518799,
+      "learning_rate": 8.839297621693711e-05,
+      "loss": 2.5045,
+      "step": 25116
+    },
+    {
+      "epoch": 0.5581555555555555,
+      "grad_norm": 1.070365071296692,
+      "learning_rate": 8.838853078461881e-05,
+      "loss": 1.1872,
+      "step": 25117
+    },
+    {
+      "epoch": 0.5581777777777778,
+      "grad_norm": 0.3382456600666046,
+      "learning_rate": 8.838408535230052e-05,
+      "loss": 0.0277,
+      "step": 25118
+    },
+    {
+      "epoch": 0.5582,
+      "grad_norm": 1.8913952112197876,
+      "learning_rate": 8.837963991998223e-05,
+      "loss": 2.2064,
+      "step": 25119
+    },
+    {
+      "epoch": 0.5582222222222222,
+      "grad_norm": 1.9757825136184692,
+      "learning_rate": 8.837519448766392e-05,
+      "loss": 2.3808,
+      "step": 25120
+    },
+    {
+      "epoch": 0.5582444444444444,
+      "grad_norm": 1.481068730354309,
+      "learning_rate": 8.837074905534563e-05,
+      "loss": 1.8185,
+      "step": 25121
+    },
+    {
+      "epoch": 0.5582666666666667,
+      "grad_norm": 1.7362627983093262,
+      "learning_rate": 8.836630362302734e-05,
+      "loss": 2.045,
+      "step": 25122
+    },
+    {
+      "epoch": 0.5582888888888888,
+      "grad_norm": 1.7237762212753296,
+      "learning_rate": 8.836185819070905e-05,
+      "loss": 1.9348,
+      "step": 25123
+    },
+    {
+      "epoch": 0.5583111111111111,
+      "grad_norm": 1.5723251104354858,
+      "learning_rate": 8.835741275839076e-05,
+      "loss": 1.824,
+      "step": 25124
+    },
+    {
+      "epoch": 0.5583333333333333,
+      "grad_norm": 2.0629472732543945,
+      "learning_rate": 8.835296732607247e-05,
+      "loss": 2.4707,
+      "step": 25125
+    },
+    {
+      "epoch": 0.5583555555555556,
+      "grad_norm": 1.853628396987915,
+      "learning_rate": 8.834852189375418e-05,
+      "loss": 1.6413,
+      "step": 25126
+    },
+    {
+      "epoch": 0.5583777777777778,
+      "grad_norm": 1.5292593240737915,
+      "learning_rate": 8.834407646143588e-05,
+      "loss": 1.7817,
+      "step": 25127
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 1.5123380422592163,
+      "learning_rate": 8.833963102911759e-05,
+      "loss": 1.7791,
+      "step": 25128
+    },
+    {
+      "epoch": 0.5584222222222223,
+      "grad_norm": 2.058326244354248,
+      "learning_rate": 8.83351855967993e-05,
+      "loss": 2.4383,
+      "step": 25129
+    },
+    {
+      "epoch": 0.5584444444444444,
+      "grad_norm": 1.7544281482696533,
+      "learning_rate": 8.833074016448099e-05,
+      "loss": 1.9952,
+      "step": 25130
+    },
+    {
+      "epoch": 0.5584666666666667,
+      "grad_norm": 1.7322680950164795,
+      "learning_rate": 8.83262947321627e-05,
+      "loss": 1.9922,
+      "step": 25131
+    },
+    {
+      "epoch": 0.5584888888888889,
+      "grad_norm": 1.798189640045166,
+      "learning_rate": 8.832184929984441e-05,
+      "loss": 2.1415,
+      "step": 25132
+    },
+    {
+      "epoch": 0.5585111111111111,
+      "grad_norm": 1.7430840730667114,
+      "learning_rate": 8.831740386752612e-05,
+      "loss": 1.8528,
+      "step": 25133
+    },
+    {
+      "epoch": 0.5585333333333333,
+      "grad_norm": 1.6595499515533447,
+      "learning_rate": 8.831295843520783e-05,
+      "loss": 1.6088,
+      "step": 25134
+    },
+    {
+      "epoch": 0.5585555555555556,
+      "grad_norm": 1.873802900314331,
+      "learning_rate": 8.830851300288954e-05,
+      "loss": 1.8963,
+      "step": 25135
+    },
+    {
+      "epoch": 0.5585777777777777,
+      "grad_norm": 1.6037418842315674,
+      "learning_rate": 8.830406757057125e-05,
+      "loss": 1.9936,
+      "step": 25136
+    },
+    {
+      "epoch": 0.5586,
+      "grad_norm": 1.7320588827133179,
+      "learning_rate": 8.829962213825295e-05,
+      "loss": 2.0704,
+      "step": 25137
+    },
+    {
+      "epoch": 0.5586222222222222,
+      "grad_norm": 1.5538798570632935,
+      "learning_rate": 8.829517670593466e-05,
+      "loss": 1.9575,
+      "step": 25138
+    },
+    {
+      "epoch": 0.5586444444444445,
+      "grad_norm": 1.7341487407684326,
+      "learning_rate": 8.829073127361636e-05,
+      "loss": 2.1159,
+      "step": 25139
+    },
+    {
+      "epoch": 0.5586666666666666,
+      "grad_norm": 1.3691928386688232,
+      "learning_rate": 8.828628584129806e-05,
+      "loss": 1.4764,
+      "step": 25140
+    },
+    {
+      "epoch": 0.5586888888888889,
+      "grad_norm": 1.517678141593933,
+      "learning_rate": 8.828184040897977e-05,
+      "loss": 1.6072,
+      "step": 25141
+    },
+    {
+      "epoch": 0.5587111111111112,
+      "grad_norm": 1.7343122959136963,
+      "learning_rate": 8.82773949766615e-05,
+      "loss": 2.3817,
+      "step": 25142
+    },
+    {
+      "epoch": 0.5587333333333333,
+      "grad_norm": 1.5004314184188843,
+      "learning_rate": 8.827294954434319e-05,
+      "loss": 1.9208,
+      "step": 25143
+    },
+    {
+      "epoch": 0.5587555555555556,
+      "grad_norm": 2.1103506088256836,
+      "learning_rate": 8.82685041120249e-05,
+      "loss": 2.2693,
+      "step": 25144
+    },
+    {
+      "epoch": 0.5587777777777778,
+      "grad_norm": 1.7467249631881714,
+      "learning_rate": 8.826405867970661e-05,
+      "loss": 2.183,
+      "step": 25145
+    },
+    {
+      "epoch": 0.5588,
+      "grad_norm": 1.4826949834823608,
+      "learning_rate": 8.825961324738832e-05,
+      "loss": 1.545,
+      "step": 25146
+    },
+    {
+      "epoch": 0.5588222222222222,
+      "grad_norm": 1.5557191371917725,
+      "learning_rate": 8.825516781507001e-05,
+      "loss": 1.5978,
+      "step": 25147
+    },
+    {
+      "epoch": 0.5588444444444445,
+      "grad_norm": 1.6098825931549072,
+      "learning_rate": 8.825072238275172e-05,
+      "loss": 1.5343,
+      "step": 25148
+    },
+    {
+      "epoch": 0.5588666666666666,
+      "grad_norm": 1.8505737781524658,
+      "learning_rate": 8.824627695043343e-05,
+      "loss": 1.8377,
+      "step": 25149
+    },
+    {
+      "epoch": 0.5588888888888889,
+      "grad_norm": 1.4390512704849243,
+      "learning_rate": 8.824183151811513e-05,
+      "loss": 1.3804,
+      "step": 25150
+    },
+    {
+      "epoch": 0.5589111111111111,
+      "grad_norm": 1.4105268716812134,
+      "learning_rate": 8.823738608579685e-05,
+      "loss": 2.5203,
+      "step": 25151
+    },
+    {
+      "epoch": 0.5589333333333333,
+      "grad_norm": 1.429821491241455,
+      "learning_rate": 8.823294065347856e-05,
+      "loss": 2.5985,
+      "step": 25152
+    },
+    {
+      "epoch": 0.5589555555555555,
+      "grad_norm": 1.441236138343811,
+      "learning_rate": 8.822849522116026e-05,
+      "loss": 0.6427,
+      "step": 25153
+    },
+    {
+      "epoch": 0.5589777777777778,
+      "grad_norm": 1.460187554359436,
+      "learning_rate": 8.822404978884197e-05,
+      "loss": 1.9902,
+      "step": 25154
+    },
+    {
+      "epoch": 0.559,
+      "grad_norm": 1.2783151865005493,
+      "learning_rate": 8.821960435652368e-05,
+      "loss": 1.0684,
+      "step": 25155
+    },
+    {
+      "epoch": 0.5590222222222222,
+      "grad_norm": 1.341304063796997,
+      "learning_rate": 8.821515892420539e-05,
+      "loss": 1.911,
+      "step": 25156
+    },
+    {
+      "epoch": 0.5590444444444445,
+      "grad_norm": 1.565393090248108,
+      "learning_rate": 8.821071349188708e-05,
+      "loss": 2.0663,
+      "step": 25157
+    },
+    {
+      "epoch": 0.5590666666666667,
+      "grad_norm": 2.0223257541656494,
+      "learning_rate": 8.82062680595688e-05,
+      "loss": 2.627,
+      "step": 25158
+    },
+    {
+      "epoch": 0.5590888888888889,
+      "grad_norm": 1.4590426683425903,
+      "learning_rate": 8.82018226272505e-05,
+      "loss": 1.6162,
+      "step": 25159
+    },
+    {
+      "epoch": 0.5591111111111111,
+      "grad_norm": 1.4461957216262817,
+      "learning_rate": 8.819737719493221e-05,
+      "loss": 1.7326,
+      "step": 25160
+    },
+    {
+      "epoch": 0.5591333333333334,
+      "grad_norm": 1.526092529296875,
+      "learning_rate": 8.819293176261392e-05,
+      "loss": 2.0117,
+      "step": 25161
+    },
+    {
+      "epoch": 0.5591555555555555,
+      "grad_norm": 1.4501690864562988,
+      "learning_rate": 8.818848633029563e-05,
+      "loss": 1.9252,
+      "step": 25162
+    },
+    {
+      "epoch": 0.5591777777777778,
+      "grad_norm": 1.68050217628479,
+      "learning_rate": 8.818404089797734e-05,
+      "loss": 2.2222,
+      "step": 25163
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 1.6856437921524048,
+      "learning_rate": 8.817959546565904e-05,
+      "loss": 1.499,
+      "step": 25164
+    },
+    {
+      "epoch": 0.5592222222222222,
+      "grad_norm": 1.6665334701538086,
+      "learning_rate": 8.817515003334075e-05,
+      "loss": 2.2639,
+      "step": 25165
+    },
+    {
+      "epoch": 0.5592444444444444,
+      "grad_norm": 1.4440217018127441,
+      "learning_rate": 8.817070460102246e-05,
+      "loss": 1.7205,
+      "step": 25166
+    },
+    {
+      "epoch": 0.5592666666666667,
+      "grad_norm": 1.6238607168197632,
+      "learning_rate": 8.816625916870415e-05,
+      "loss": 2.0203,
+      "step": 25167
+    },
+    {
+      "epoch": 0.5592888888888888,
+      "grad_norm": 1.5175899267196655,
+      "learning_rate": 8.816181373638586e-05,
+      "loss": 1.76,
+      "step": 25168
+    },
+    {
+      "epoch": 0.5593111111111111,
+      "grad_norm": 1.5204845666885376,
+      "learning_rate": 8.815736830406757e-05,
+      "loss": 2.231,
+      "step": 25169
+    },
+    {
+      "epoch": 0.5593333333333333,
+      "grad_norm": 1.8707369565963745,
+      "learning_rate": 8.815292287174928e-05,
+      "loss": 2.2178,
+      "step": 25170
+    },
+    {
+      "epoch": 0.5593555555555556,
+      "grad_norm": 1.6467576026916504,
+      "learning_rate": 8.814847743943099e-05,
+      "loss": 2.2364,
+      "step": 25171
+    },
+    {
+      "epoch": 0.5593777777777778,
+      "grad_norm": 1.8275185823440552,
+      "learning_rate": 8.81440320071127e-05,
+      "loss": 2.1331,
+      "step": 25172
+    },
+    {
+      "epoch": 0.5594,
+      "grad_norm": 1.559929370880127,
+      "learning_rate": 8.813958657479441e-05,
+      "loss": 2.1326,
+      "step": 25173
+    },
+    {
+      "epoch": 0.5594222222222223,
+      "grad_norm": 2.1848270893096924,
+      "learning_rate": 8.81351411424761e-05,
+      "loss": 1.8675,
+      "step": 25174
+    },
+    {
+      "epoch": 0.5594444444444444,
+      "grad_norm": 1.3665002584457397,
+      "learning_rate": 8.813069571015782e-05,
+      "loss": 1.4474,
+      "step": 25175
+    },
+    {
+      "epoch": 0.5594666666666667,
+      "grad_norm": 1.7274236679077148,
+      "learning_rate": 8.812625027783953e-05,
+      "loss": 1.9424,
+      "step": 25176
+    },
+    {
+      "epoch": 0.5594888888888889,
+      "grad_norm": 1.6203581094741821,
+      "learning_rate": 8.812180484552122e-05,
+      "loss": 1.6601,
+      "step": 25177
+    },
+    {
+      "epoch": 0.5595111111111111,
+      "grad_norm": 1.8690255880355835,
+      "learning_rate": 8.811735941320293e-05,
+      "loss": 1.8835,
+      "step": 25178
+    },
+    {
+      "epoch": 0.5595333333333333,
+      "grad_norm": 1.4840466976165771,
+      "learning_rate": 8.811291398088465e-05,
+      "loss": 1.9987,
+      "step": 25179
+    },
+    {
+      "epoch": 0.5595555555555556,
+      "grad_norm": 1.9614216089248657,
+      "learning_rate": 8.810846854856635e-05,
+      "loss": 1.9445,
+      "step": 25180
+    },
+    {
+      "epoch": 0.5595777777777777,
+      "grad_norm": 1.560650110244751,
+      "learning_rate": 8.810402311624806e-05,
+      "loss": 1.6483,
+      "step": 25181
+    },
+    {
+      "epoch": 0.5596,
+      "grad_norm": 1.543728232383728,
+      "learning_rate": 8.809957768392977e-05,
+      "loss": 2.0516,
+      "step": 25182
+    },
+    {
+      "epoch": 0.5596222222222222,
+      "grad_norm": 2.026848316192627,
+      "learning_rate": 8.809513225161148e-05,
+      "loss": 2.276,
+      "step": 25183
+    },
+    {
+      "epoch": 0.5596444444444445,
+      "grad_norm": 2.0482680797576904,
+      "learning_rate": 8.809068681929318e-05,
+      "loss": 2.0603,
+      "step": 25184
+    },
+    {
+      "epoch": 0.5596666666666666,
+      "grad_norm": 1.650883674621582,
+      "learning_rate": 8.808624138697489e-05,
+      "loss": 1.6454,
+      "step": 25185
+    },
+    {
+      "epoch": 0.5596888888888889,
+      "grad_norm": 1.5135639905929565,
+      "learning_rate": 8.80817959546566e-05,
+      "loss": 1.7825,
+      "step": 25186
+    },
+    {
+      "epoch": 0.5597111111111112,
+      "grad_norm": 1.509371042251587,
+      "learning_rate": 8.807735052233829e-05,
+      "loss": 1.7063,
+      "step": 25187
+    },
+    {
+      "epoch": 0.5597333333333333,
+      "grad_norm": 1.7396836280822754,
+      "learning_rate": 8.807290509002001e-05,
+      "loss": 2.1483,
+      "step": 25188
+    },
+    {
+      "epoch": 0.5597555555555556,
+      "grad_norm": 1.742016315460205,
+      "learning_rate": 8.806845965770172e-05,
+      "loss": 2.0434,
+      "step": 25189
+    },
+    {
+      "epoch": 0.5597777777777778,
+      "grad_norm": 1.679200291633606,
+      "learning_rate": 8.806401422538342e-05,
+      "loss": 1.5045,
+      "step": 25190
+    },
+    {
+      "epoch": 0.5598,
+      "grad_norm": 1.8576488494873047,
+      "learning_rate": 8.805956879306513e-05,
+      "loss": 2.1748,
+      "step": 25191
+    },
+    {
+      "epoch": 0.5598222222222222,
+      "grad_norm": 1.8510833978652954,
+      "learning_rate": 8.805512336074684e-05,
+      "loss": 1.5863,
+      "step": 25192
+    },
+    {
+      "epoch": 0.5598444444444445,
+      "grad_norm": 2.1455936431884766,
+      "learning_rate": 8.805067792842855e-05,
+      "loss": 2.0357,
+      "step": 25193
+    },
+    {
+      "epoch": 0.5598666666666666,
+      "grad_norm": 1.841348648071289,
+      "learning_rate": 8.804623249611024e-05,
+      "loss": 1.9626,
+      "step": 25194
+    },
+    {
+      "epoch": 0.5598888888888889,
+      "grad_norm": 1.55734384059906,
+      "learning_rate": 8.804178706379195e-05,
+      "loss": 1.6384,
+      "step": 25195
+    },
+    {
+      "epoch": 0.5599111111111111,
+      "grad_norm": 2.034369945526123,
+      "learning_rate": 8.803734163147366e-05,
+      "loss": 1.8983,
+      "step": 25196
+    },
+    {
+      "epoch": 0.5599333333333333,
+      "grad_norm": 1.7569447755813599,
+      "learning_rate": 8.803289619915537e-05,
+      "loss": 1.879,
+      "step": 25197
+    },
+    {
+      "epoch": 0.5599555555555555,
+      "grad_norm": 2.063053607940674,
+      "learning_rate": 8.802845076683708e-05,
+      "loss": 1.8524,
+      "step": 25198
+    },
+    {
+      "epoch": 0.5599777777777778,
+      "grad_norm": 1.9506410360336304,
+      "learning_rate": 8.802400533451879e-05,
+      "loss": 2.1881,
+      "step": 25199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5537521839141846,
+      "learning_rate": 8.801955990220049e-05,
+      "loss": 1.5269,
+      "step": 25200
+    },
+    {
+      "epoch": 0.5600222222222222,
+      "grad_norm": 1.7377948760986328,
+      "learning_rate": 8.80151144698822e-05,
+      "loss": 2.5955,
+      "step": 25201
+    },
+    {
+      "epoch": 0.5600444444444445,
+      "grad_norm": 1.6233863830566406,
+      "learning_rate": 8.801066903756391e-05,
+      "loss": 2.6002,
+      "step": 25202
+    },
+    {
+      "epoch": 0.5600666666666667,
+      "grad_norm": 1.5851396322250366,
+      "learning_rate": 8.800622360524562e-05,
+      "loss": 2.5364,
+      "step": 25203
+    },
+    {
+      "epoch": 0.5600888888888889,
+      "grad_norm": 1.334736943244934,
+      "learning_rate": 8.800177817292731e-05,
+      "loss": 2.1957,
+      "step": 25204
+    },
+    {
+      "epoch": 0.5601111111111111,
+      "grad_norm": 1.3811289072036743,
+      "learning_rate": 8.799733274060902e-05,
+      "loss": 2.08,
+      "step": 25205
+    },
+    {
+      "epoch": 0.5601333333333334,
+      "grad_norm": 1.5644488334655762,
+      "learning_rate": 8.799288730829073e-05,
+      "loss": 2.6846,
+      "step": 25206
+    },
+    {
+      "epoch": 0.5601555555555555,
+      "grad_norm": 1.5639359951019287,
+      "learning_rate": 8.798844187597244e-05,
+      "loss": 2.0832,
+      "step": 25207
+    },
+    {
+      "epoch": 0.5601777777777778,
+      "grad_norm": 1.5768216848373413,
+      "learning_rate": 8.798399644365415e-05,
+      "loss": 2.1764,
+      "step": 25208
+    },
+    {
+      "epoch": 0.5602,
+      "grad_norm": 1.8076982498168945,
+      "learning_rate": 8.797955101133586e-05,
+      "loss": 2.2899,
+      "step": 25209
+    },
+    {
+      "epoch": 0.5602222222222222,
+      "grad_norm": 1.1235891580581665,
+      "learning_rate": 8.797510557901756e-05,
+      "loss": 1.1146,
+      "step": 25210
+    },
+    {
+      "epoch": 0.5602444444444444,
+      "grad_norm": 1.7415649890899658,
+      "learning_rate": 8.797066014669927e-05,
+      "loss": 2.2942,
+      "step": 25211
+    },
+    {
+      "epoch": 0.5602666666666667,
+      "grad_norm": 1.6006947755813599,
+      "learning_rate": 8.796621471438098e-05,
+      "loss": 2.2651,
+      "step": 25212
+    },
+    {
+      "epoch": 0.5602888888888888,
+      "grad_norm": 1.5128785371780396,
+      "learning_rate": 8.796176928206269e-05,
+      "loss": 1.7946,
+      "step": 25213
+    },
+    {
+      "epoch": 0.5603111111111111,
+      "grad_norm": 1.5808520317077637,
+      "learning_rate": 8.795732384974438e-05,
+      "loss": 2.2544,
+      "step": 25214
+    },
+    {
+      "epoch": 0.5603333333333333,
+      "grad_norm": 1.525814175605774,
+      "learning_rate": 8.795287841742609e-05,
+      "loss": 1.4875,
+      "step": 25215
+    },
+    {
+      "epoch": 0.5603555555555556,
+      "grad_norm": 1.5856590270996094,
+      "learning_rate": 8.794843298510782e-05,
+      "loss": 1.7909,
+      "step": 25216
+    },
+    {
+      "epoch": 0.5603777777777778,
+      "grad_norm": 1.6325963735580444,
+      "learning_rate": 8.794398755278951e-05,
+      "loss": 2.1416,
+      "step": 25217
+    },
+    {
+      "epoch": 0.5604,
+      "grad_norm": 2.059194564819336,
+      "learning_rate": 8.793954212047122e-05,
+      "loss": 1.8765,
+      "step": 25218
+    },
+    {
+      "epoch": 0.5604222222222223,
+      "grad_norm": 1.6079334020614624,
+      "learning_rate": 8.793509668815293e-05,
+      "loss": 2.1008,
+      "step": 25219
+    },
+    {
+      "epoch": 0.5604444444444444,
+      "grad_norm": 1.5019943714141846,
+      "learning_rate": 8.793065125583464e-05,
+      "loss": 1.961,
+      "step": 25220
+    },
+    {
+      "epoch": 0.5604666666666667,
+      "grad_norm": 1.7217684984207153,
+      "learning_rate": 8.792620582351634e-05,
+      "loss": 2.2064,
+      "step": 25221
+    },
+    {
+      "epoch": 0.5604888888888889,
+      "grad_norm": 1.7666687965393066,
+      "learning_rate": 8.792176039119805e-05,
+      "loss": 1.9191,
+      "step": 25222
+    },
+    {
+      "epoch": 0.5605111111111111,
+      "grad_norm": 1.4344292879104614,
+      "learning_rate": 8.791731495887976e-05,
+      "loss": 1.1761,
+      "step": 25223
+    },
+    {
+      "epoch": 0.5605333333333333,
+      "grad_norm": 2.339412212371826,
+      "learning_rate": 8.791286952656145e-05,
+      "loss": 1.822,
+      "step": 25224
+    },
+    {
+      "epoch": 0.5605555555555556,
+      "grad_norm": 1.6369621753692627,
+      "learning_rate": 8.790842409424318e-05,
+      "loss": 1.9765,
+      "step": 25225
+    },
+    {
+      "epoch": 0.5605777777777777,
+      "grad_norm": 1.9765450954437256,
+      "learning_rate": 8.790397866192488e-05,
+      "loss": 2.1667,
+      "step": 25226
+    },
+    {
+      "epoch": 0.5606,
+      "grad_norm": 1.5706331729888916,
+      "learning_rate": 8.789953322960658e-05,
+      "loss": 1.859,
+      "step": 25227
+    },
+    {
+      "epoch": 0.5606222222222222,
+      "grad_norm": 1.4491208791732788,
+      "learning_rate": 8.789508779728829e-05,
+      "loss": 1.7216,
+      "step": 25228
+    },
+    {
+      "epoch": 0.5606444444444444,
+      "grad_norm": 1.8563940525054932,
+      "learning_rate": 8.789064236497e-05,
+      "loss": 2.1766,
+      "step": 25229
+    },
+    {
+      "epoch": 0.5606666666666666,
+      "grad_norm": 1.293811559677124,
+      "learning_rate": 8.788619693265171e-05,
+      "loss": 0.8898,
+      "step": 25230
+    },
+    {
+      "epoch": 0.5606888888888889,
+      "grad_norm": 1.5508888959884644,
+      "learning_rate": 8.78817515003334e-05,
+      "loss": 1.8485,
+      "step": 25231
+    },
+    {
+      "epoch": 0.5607111111111112,
+      "grad_norm": 1.7579081058502197,
+      "learning_rate": 8.787730606801512e-05,
+      "loss": 2.1352,
+      "step": 25232
+    },
+    {
+      "epoch": 0.5607333333333333,
+      "grad_norm": 1.6988152265548706,
+      "learning_rate": 8.787286063569683e-05,
+      "loss": 1.8743,
+      "step": 25233
+    },
+    {
+      "epoch": 0.5607555555555556,
+      "grad_norm": 1.710771918296814,
+      "learning_rate": 8.786841520337853e-05,
+      "loss": 1.9996,
+      "step": 25234
+    },
+    {
+      "epoch": 0.5607777777777778,
+      "grad_norm": 1.6037424802780151,
+      "learning_rate": 8.786396977106024e-05,
+      "loss": 1.7988,
+      "step": 25235
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 1.7195894718170166,
+      "learning_rate": 8.785952433874195e-05,
+      "loss": 1.9299,
+      "step": 25236
+    },
+    {
+      "epoch": 0.5608222222222222,
+      "grad_norm": 1.5542490482330322,
+      "learning_rate": 8.785507890642365e-05,
+      "loss": 1.4587,
+      "step": 25237
+    },
+    {
+      "epoch": 0.5608444444444445,
+      "grad_norm": 1.9157278537750244,
+      "learning_rate": 8.785063347410536e-05,
+      "loss": 2.0123,
+      "step": 25238
+    },
+    {
+      "epoch": 0.5608666666666666,
+      "grad_norm": 1.5922707319259644,
+      "learning_rate": 8.784618804178707e-05,
+      "loss": 1.7134,
+      "step": 25239
+    },
+    {
+      "epoch": 0.5608888888888889,
+      "grad_norm": 1.407324194908142,
+      "learning_rate": 8.784174260946878e-05,
+      "loss": 1.3508,
+      "step": 25240
+    },
+    {
+      "epoch": 0.5609111111111111,
+      "grad_norm": 1.816657543182373,
+      "learning_rate": 8.783729717715047e-05,
+      "loss": 2.2265,
+      "step": 25241
+    },
+    {
+      "epoch": 0.5609333333333333,
+      "grad_norm": 1.7667323350906372,
+      "learning_rate": 8.783285174483218e-05,
+      "loss": 1.5843,
+      "step": 25242
+    },
+    {
+      "epoch": 0.5609555555555555,
+      "grad_norm": 1.7764583826065063,
+      "learning_rate": 8.78284063125139e-05,
+      "loss": 1.7084,
+      "step": 25243
+    },
+    {
+      "epoch": 0.5609777777777778,
+      "grad_norm": 1.8567525148391724,
+      "learning_rate": 8.78239608801956e-05,
+      "loss": 1.9548,
+      "step": 25244
+    },
+    {
+      "epoch": 0.561,
+      "grad_norm": 1.7268896102905273,
+      "learning_rate": 8.781951544787731e-05,
+      "loss": 2.1631,
+      "step": 25245
+    },
+    {
+      "epoch": 0.5610222222222222,
+      "grad_norm": 1.588640570640564,
+      "learning_rate": 8.781507001555902e-05,
+      "loss": 1.6146,
+      "step": 25246
+    },
+    {
+      "epoch": 0.5610444444444445,
+      "grad_norm": 2.132821798324585,
+      "learning_rate": 8.781062458324072e-05,
+      "loss": 1.9696,
+      "step": 25247
+    },
+    {
+      "epoch": 0.5610666666666667,
+      "grad_norm": 1.6377253532409668,
+      "learning_rate": 8.780617915092243e-05,
+      "loss": 1.4368,
+      "step": 25248
+    },
+    {
+      "epoch": 0.5610888888888889,
+      "grad_norm": 2.060591697692871,
+      "learning_rate": 8.780173371860414e-05,
+      "loss": 1.7435,
+      "step": 25249
+    },
+    {
+      "epoch": 0.5611111111111111,
+      "grad_norm": 1.711743712425232,
+      "learning_rate": 8.779728828628585e-05,
+      "loss": 1.2941,
+      "step": 25250
+    },
+    {
+      "epoch": 0.5611333333333334,
+      "grad_norm": 1.6078704595565796,
+      "learning_rate": 8.779284285396754e-05,
+      "loss": 2.4519,
+      "step": 25251
+    },
+    {
+      "epoch": 0.5611555555555555,
+      "grad_norm": 1.683768391609192,
+      "learning_rate": 8.778839742164925e-05,
+      "loss": 2.858,
+      "step": 25252
+    },
+    {
+      "epoch": 0.5611777777777778,
+      "grad_norm": 1.5357310771942139,
+      "learning_rate": 8.778395198933098e-05,
+      "loss": 2.8118,
+      "step": 25253
+    },
+    {
+      "epoch": 0.5612,
+      "grad_norm": 1.3293133974075317,
+      "learning_rate": 8.777950655701267e-05,
+      "loss": 2.5063,
+      "step": 25254
+    },
+    {
+      "epoch": 0.5612222222222222,
+      "grad_norm": 1.4129914045333862,
+      "learning_rate": 8.777506112469438e-05,
+      "loss": 2.2103,
+      "step": 25255
+    },
+    {
+      "epoch": 0.5612444444444444,
+      "grad_norm": 1.5258193016052246,
+      "learning_rate": 8.777061569237609e-05,
+      "loss": 1.8931,
+      "step": 25256
+    },
+    {
+      "epoch": 0.5612666666666667,
+      "grad_norm": 1.6168725490570068,
+      "learning_rate": 8.776617026005779e-05,
+      "loss": 1.9514,
+      "step": 25257
+    },
+    {
+      "epoch": 0.5612888888888888,
+      "grad_norm": 1.647147297859192,
+      "learning_rate": 8.77617248277395e-05,
+      "loss": 2.1524,
+      "step": 25258
+    },
+    {
+      "epoch": 0.5613111111111111,
+      "grad_norm": 1.4901888370513916,
+      "learning_rate": 8.775727939542121e-05,
+      "loss": 1.8784,
+      "step": 25259
+    },
+    {
+      "epoch": 0.5613333333333334,
+      "grad_norm": 1.7015953063964844,
+      "learning_rate": 8.775283396310292e-05,
+      "loss": 1.7413,
+      "step": 25260
+    },
+    {
+      "epoch": 0.5613555555555556,
+      "grad_norm": 1.8106613159179688,
+      "learning_rate": 8.774838853078463e-05,
+      "loss": 2.456,
+      "step": 25261
+    },
+    {
+      "epoch": 0.5613777777777778,
+      "grad_norm": 1.8653677701950073,
+      "learning_rate": 8.774394309846634e-05,
+      "loss": 1.8331,
+      "step": 25262
+    },
+    {
+      "epoch": 0.5614,
+      "grad_norm": 1.485082983970642,
+      "learning_rate": 8.773949766614805e-05,
+      "loss": 1.9771,
+      "step": 25263
+    },
+    {
+      "epoch": 0.5614222222222223,
+      "grad_norm": 1.6807743310928345,
+      "learning_rate": 8.773505223382974e-05,
+      "loss": 2.3118,
+      "step": 25264
+    },
+    {
+      "epoch": 0.5614444444444444,
+      "grad_norm": 1.4277927875518799,
+      "learning_rate": 8.773060680151145e-05,
+      "loss": 1.6863,
+      "step": 25265
+    },
+    {
+      "epoch": 0.5614666666666667,
+      "grad_norm": 1.5671919584274292,
+      "learning_rate": 8.772616136919316e-05,
+      "loss": 1.9608,
+      "step": 25266
+    },
+    {
+      "epoch": 0.5614888888888889,
+      "grad_norm": 1.6028257608413696,
+      "learning_rate": 8.772171593687486e-05,
+      "loss": 1.6542,
+      "step": 25267
+    },
+    {
+      "epoch": 0.5615111111111111,
+      "grad_norm": 1.4625035524368286,
+      "learning_rate": 8.771727050455657e-05,
+      "loss": 2.1563,
+      "step": 25268
+    },
+    {
+      "epoch": 0.5615333333333333,
+      "grad_norm": 1.9192852973937988,
+      "learning_rate": 8.771282507223828e-05,
+      "loss": 1.7663,
+      "step": 25269
+    },
+    {
+      "epoch": 0.5615555555555556,
+      "grad_norm": 1.596518635749817,
+      "learning_rate": 8.770837963991999e-05,
+      "loss": 1.8391,
+      "step": 25270
+    },
+    {
+      "epoch": 0.5615777777777777,
+      "grad_norm": 1.869960904121399,
+      "learning_rate": 8.77039342076017e-05,
+      "loss": 1.8417,
+      "step": 25271
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 1.4895312786102295,
+      "learning_rate": 8.76994887752834e-05,
+      "loss": 1.9274,
+      "step": 25272
+    },
+    {
+      "epoch": 0.5616222222222222,
+      "grad_norm": 1.7752124071121216,
+      "learning_rate": 8.769504334296511e-05,
+      "loss": 2.0309,
+      "step": 25273
+    },
+    {
+      "epoch": 0.5616444444444444,
+      "grad_norm": 1.4135031700134277,
+      "learning_rate": 8.769059791064681e-05,
+      "loss": 1.8481,
+      "step": 25274
+    },
+    {
+      "epoch": 0.5616666666666666,
+      "grad_norm": 1.489335298538208,
+      "learning_rate": 8.768615247832852e-05,
+      "loss": 1.9864,
+      "step": 25275
+    },
+    {
+      "epoch": 0.5616888888888889,
+      "grad_norm": 1.7148730754852295,
+      "learning_rate": 8.768170704601023e-05,
+      "loss": 2.2432,
+      "step": 25276
+    },
+    {
+      "epoch": 0.5617111111111112,
+      "grad_norm": 1.6796962022781372,
+      "learning_rate": 8.767726161369194e-05,
+      "loss": 2.3246,
+      "step": 25277
+    },
+    {
+      "epoch": 0.5617333333333333,
+      "grad_norm": 1.7491155862808228,
+      "learning_rate": 8.767281618137364e-05,
+      "loss": 1.9976,
+      "step": 25278
+    },
+    {
+      "epoch": 0.5617555555555556,
+      "grad_norm": 1.6207523345947266,
+      "learning_rate": 8.766837074905535e-05,
+      "loss": 1.8278,
+      "step": 25279
+    },
+    {
+      "epoch": 0.5617777777777778,
+      "grad_norm": 2.00871205329895,
+      "learning_rate": 8.766392531673706e-05,
+      "loss": 2.0722,
+      "step": 25280
+    },
+    {
+      "epoch": 0.5618,
+      "grad_norm": 1.0934069156646729,
+      "learning_rate": 8.765947988441876e-05,
+      "loss": 0.9379,
+      "step": 25281
+    },
+    {
+      "epoch": 0.5618222222222222,
+      "grad_norm": 1.9000760316848755,
+      "learning_rate": 8.765503445210047e-05,
+      "loss": 1.8411,
+      "step": 25282
+    },
+    {
+      "epoch": 0.5618444444444445,
+      "grad_norm": 1.752655029296875,
+      "learning_rate": 8.765058901978218e-05,
+      "loss": 1.9778,
+      "step": 25283
+    },
+    {
+      "epoch": 0.5618666666666666,
+      "grad_norm": 1.6022640466690063,
+      "learning_rate": 8.764614358746388e-05,
+      "loss": 1.6126,
+      "step": 25284
+    },
+    {
+      "epoch": 0.5618888888888889,
+      "grad_norm": 1.6593174934387207,
+      "learning_rate": 8.764169815514559e-05,
+      "loss": 2.3195,
+      "step": 25285
+    },
+    {
+      "epoch": 0.5619111111111111,
+      "grad_norm": 1.9099777936935425,
+      "learning_rate": 8.76372527228273e-05,
+      "loss": 2.2803,
+      "step": 25286
+    },
+    {
+      "epoch": 0.5619333333333333,
+      "grad_norm": 1.6091148853302002,
+      "learning_rate": 8.763280729050901e-05,
+      "loss": 1.8329,
+      "step": 25287
+    },
+    {
+      "epoch": 0.5619555555555555,
+      "grad_norm": 2.04919171333313,
+      "learning_rate": 8.76283618581907e-05,
+      "loss": 2.0668,
+      "step": 25288
+    },
+    {
+      "epoch": 0.5619777777777778,
+      "grad_norm": 1.6717642545700073,
+      "learning_rate": 8.762391642587241e-05,
+      "loss": 1.7885,
+      "step": 25289
+    },
+    {
+      "epoch": 0.562,
+      "grad_norm": 2.0663769245147705,
+      "learning_rate": 8.761947099355414e-05,
+      "loss": 2.1753,
+      "step": 25290
+    },
+    {
+      "epoch": 0.5620222222222222,
+      "grad_norm": 1.6406569480895996,
+      "learning_rate": 8.761502556123583e-05,
+      "loss": 2.0932,
+      "step": 25291
+    },
+    {
+      "epoch": 0.5620444444444445,
+      "grad_norm": 1.6592798233032227,
+      "learning_rate": 8.761058012891754e-05,
+      "loss": 1.8012,
+      "step": 25292
+    },
+    {
+      "epoch": 0.5620666666666667,
+      "grad_norm": 1.3389942646026611,
+      "learning_rate": 8.760613469659925e-05,
+      "loss": 1.4157,
+      "step": 25293
+    },
+    {
+      "epoch": 0.5620888888888889,
+      "grad_norm": 1.5179102420806885,
+      "learning_rate": 8.760168926428095e-05,
+      "loss": 1.7008,
+      "step": 25294
+    },
+    {
+      "epoch": 0.5621111111111111,
+      "grad_norm": 2.3237197399139404,
+      "learning_rate": 8.759724383196266e-05,
+      "loss": 2.5221,
+      "step": 25295
+    },
+    {
+      "epoch": 0.5621333333333334,
+      "grad_norm": 1.9417719841003418,
+      "learning_rate": 8.759279839964437e-05,
+      "loss": 1.0572,
+      "step": 25296
+    },
+    {
+      "epoch": 0.5621555555555555,
+      "grad_norm": 1.9588453769683838,
+      "learning_rate": 8.758835296732608e-05,
+      "loss": 1.8536,
+      "step": 25297
+    },
+    {
+      "epoch": 0.5621777777777778,
+      "grad_norm": 1.9292761087417603,
+      "learning_rate": 8.758390753500779e-05,
+      "loss": 1.6143,
+      "step": 25298
+    },
+    {
+      "epoch": 0.5622,
+      "grad_norm": 1.8616777658462524,
+      "learning_rate": 8.75794621026895e-05,
+      "loss": 1.2399,
+      "step": 25299
+    },
+    {
+      "epoch": 0.5622222222222222,
+      "grad_norm": 1.3126881122589111,
+      "learning_rate": 8.757501667037121e-05,
+      "loss": 0.8374,
+      "step": 25300
+    },
+    {
+      "epoch": 0.5622444444444444,
+      "grad_norm": 1.904229998588562,
+      "learning_rate": 8.75705712380529e-05,
+      "loss": 2.7687,
+      "step": 25301
+    },
+    {
+      "epoch": 0.5622666666666667,
+      "grad_norm": 1.7405216693878174,
+      "learning_rate": 8.756612580573461e-05,
+      "loss": 2.6446,
+      "step": 25302
+    },
+    {
+      "epoch": 0.5622888888888888,
+      "grad_norm": 1.3169325590133667,
+      "learning_rate": 8.756168037341632e-05,
+      "loss": 1.3684,
+      "step": 25303
+    },
+    {
+      "epoch": 0.5623111111111111,
+      "grad_norm": 1.3349109888076782,
+      "learning_rate": 8.755723494109802e-05,
+      "loss": 0.9687,
+      "step": 25304
+    },
+    {
+      "epoch": 0.5623333333333334,
+      "grad_norm": 1.5008281469345093,
+      "learning_rate": 8.755278950877973e-05,
+      "loss": 2.0184,
+      "step": 25305
+    },
+    {
+      "epoch": 0.5623555555555556,
+      "grad_norm": 1.422032356262207,
+      "learning_rate": 8.754834407646144e-05,
+      "loss": 2.0879,
+      "step": 25306
+    },
+    {
+      "epoch": 0.5623777777777778,
+      "grad_norm": 1.5428098440170288,
+      "learning_rate": 8.754389864414315e-05,
+      "loss": 2.2109,
+      "step": 25307
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 1.5082614421844482,
+      "learning_rate": 8.753945321182486e-05,
+      "loss": 1.8282,
+      "step": 25308
+    },
+    {
+      "epoch": 0.5624222222222223,
+      "grad_norm": 1.5564109086990356,
+      "learning_rate": 8.753500777950657e-05,
+      "loss": 2.2583,
+      "step": 25309
+    },
+    {
+      "epoch": 0.5624444444444444,
+      "grad_norm": 1.4597630500793457,
+      "learning_rate": 8.753056234718828e-05,
+      "loss": 1.7068,
+      "step": 25310
+    },
+    {
+      "epoch": 0.5624666666666667,
+      "grad_norm": 1.603948950767517,
+      "learning_rate": 8.752611691486997e-05,
+      "loss": 2.1259,
+      "step": 25311
+    },
+    {
+      "epoch": 0.5624888888888889,
+      "grad_norm": 1.4440677165985107,
+      "learning_rate": 8.752167148255168e-05,
+      "loss": 1.6405,
+      "step": 25312
+    },
+    {
+      "epoch": 0.5625111111111111,
+      "grad_norm": 0.9187580943107605,
+      "learning_rate": 8.751722605023339e-05,
+      "loss": 0.5134,
+      "step": 25313
+    },
+    {
+      "epoch": 0.5625333333333333,
+      "grad_norm": 1.7306079864501953,
+      "learning_rate": 8.751278061791509e-05,
+      "loss": 1.9311,
+      "step": 25314
+    },
+    {
+      "epoch": 0.5625555555555556,
+      "grad_norm": 1.932061791419983,
+      "learning_rate": 8.75083351855968e-05,
+      "loss": 2.4453,
+      "step": 25315
+    },
+    {
+      "epoch": 0.5625777777777777,
+      "grad_norm": 1.70633065700531,
+      "learning_rate": 8.75038897532785e-05,
+      "loss": 2.3742,
+      "step": 25316
+    },
+    {
+      "epoch": 0.5626,
+      "grad_norm": 1.711786150932312,
+      "learning_rate": 8.749944432096022e-05,
+      "loss": 1.9739,
+      "step": 25317
+    },
+    {
+      "epoch": 0.5626222222222222,
+      "grad_norm": 1.4621399641036987,
+      "learning_rate": 8.749499888864193e-05,
+      "loss": 1.5956,
+      "step": 25318
+    },
+    {
+      "epoch": 0.5626444444444444,
+      "grad_norm": 1.4943276643753052,
+      "learning_rate": 8.749055345632364e-05,
+      "loss": 1.9649,
+      "step": 25319
+    },
+    {
+      "epoch": 0.5626666666666666,
+      "grad_norm": 2.1974916458129883,
+      "learning_rate": 8.748610802400535e-05,
+      "loss": 2.1227,
+      "step": 25320
+    },
+    {
+      "epoch": 0.5626888888888889,
+      "grad_norm": 1.8154371976852417,
+      "learning_rate": 8.748166259168704e-05,
+      "loss": 2.022,
+      "step": 25321
+    },
+    {
+      "epoch": 0.5627111111111112,
+      "grad_norm": 2.9448294639587402,
+      "learning_rate": 8.747721715936875e-05,
+      "loss": 2.1413,
+      "step": 25322
+    },
+    {
+      "epoch": 0.5627333333333333,
+      "grad_norm": 1.6197272539138794,
+      "learning_rate": 8.747277172705046e-05,
+      "loss": 1.4278,
+      "step": 25323
+    },
+    {
+      "epoch": 0.5627555555555556,
+      "grad_norm": 1.4971791505813599,
+      "learning_rate": 8.746832629473217e-05,
+      "loss": 2.0016,
+      "step": 25324
+    },
+    {
+      "epoch": 0.5627777777777778,
+      "grad_norm": 1.629438042640686,
+      "learning_rate": 8.746388086241387e-05,
+      "loss": 1.9129,
+      "step": 25325
+    },
+    {
+      "epoch": 0.5628,
+      "grad_norm": 1.8357394933700562,
+      "learning_rate": 8.745943543009558e-05,
+      "loss": 2.146,
+      "step": 25326
+    },
+    {
+      "epoch": 0.5628222222222222,
+      "grad_norm": 2.109930992126465,
+      "learning_rate": 8.74549899977773e-05,
+      "loss": 2.1673,
+      "step": 25327
+    },
+    {
+      "epoch": 0.5628444444444445,
+      "grad_norm": 1.9761508703231812,
+      "learning_rate": 8.7450544565459e-05,
+      "loss": 1.6631,
+      "step": 25328
+    },
+    {
+      "epoch": 0.5628666666666666,
+      "grad_norm": 2.0955917835235596,
+      "learning_rate": 8.74460991331407e-05,
+      "loss": 1.9176,
+      "step": 25329
+    },
+    {
+      "epoch": 0.5628888888888889,
+      "grad_norm": 1.5698069334030151,
+      "learning_rate": 8.744165370082241e-05,
+      "loss": 1.761,
+      "step": 25330
+    },
+    {
+      "epoch": 0.5629111111111111,
+      "grad_norm": 1.6463110446929932,
+      "learning_rate": 8.743720826850411e-05,
+      "loss": 1.8677,
+      "step": 25331
+    },
+    {
+      "epoch": 0.5629333333333333,
+      "grad_norm": 1.896437406539917,
+      "learning_rate": 8.743276283618582e-05,
+      "loss": 2.139,
+      "step": 25332
+    },
+    {
+      "epoch": 0.5629555555555555,
+      "grad_norm": 1.5840256214141846,
+      "learning_rate": 8.742831740386753e-05,
+      "loss": 1.9418,
+      "step": 25333
+    },
+    {
+      "epoch": 0.5629777777777778,
+      "grad_norm": 1.5916732549667358,
+      "learning_rate": 8.742387197154924e-05,
+      "loss": 1.9866,
+      "step": 25334
+    },
+    {
+      "epoch": 0.563,
+      "grad_norm": 1.9358857870101929,
+      "learning_rate": 8.741942653923095e-05,
+      "loss": 1.913,
+      "step": 25335
+    },
+    {
+      "epoch": 0.5630222222222222,
+      "grad_norm": 1.5716407299041748,
+      "learning_rate": 8.741498110691266e-05,
+      "loss": 1.6528,
+      "step": 25336
+    },
+    {
+      "epoch": 0.5630444444444445,
+      "grad_norm": 1.7800015211105347,
+      "learning_rate": 8.741053567459437e-05,
+      "loss": 2.1893,
+      "step": 25337
+    },
+    {
+      "epoch": 0.5630666666666667,
+      "grad_norm": 1.4720046520233154,
+      "learning_rate": 8.740609024227606e-05,
+      "loss": 1.8069,
+      "step": 25338
+    },
+    {
+      "epoch": 0.5630888888888889,
+      "grad_norm": 2.0705015659332275,
+      "learning_rate": 8.740164480995777e-05,
+      "loss": 2.1206,
+      "step": 25339
+    },
+    {
+      "epoch": 0.5631111111111111,
+      "grad_norm": 1.7749475240707397,
+      "learning_rate": 8.739719937763948e-05,
+      "loss": 1.988,
+      "step": 25340
+    },
+    {
+      "epoch": 0.5631333333333334,
+      "grad_norm": 1.589619517326355,
+      "learning_rate": 8.739275394532118e-05,
+      "loss": 1.2761,
+      "step": 25341
+    },
+    {
+      "epoch": 0.5631555555555555,
+      "grad_norm": 2.2986977100372314,
+      "learning_rate": 8.738830851300289e-05,
+      "loss": 2.5454,
+      "step": 25342
+    },
+    {
+      "epoch": 0.5631777777777778,
+      "grad_norm": 1.552807331085205,
+      "learning_rate": 8.73838630806846e-05,
+      "loss": 1.3278,
+      "step": 25343
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 1.748751163482666,
+      "learning_rate": 8.737941764836631e-05,
+      "loss": 2.0745,
+      "step": 25344
+    },
+    {
+      "epoch": 0.5632222222222222,
+      "grad_norm": 1.6754921674728394,
+      "learning_rate": 8.737497221604802e-05,
+      "loss": 1.5605,
+      "step": 25345
+    },
+    {
+      "epoch": 0.5632444444444444,
+      "grad_norm": 1.5707887411117554,
+      "learning_rate": 8.737052678372973e-05,
+      "loss": 1.4676,
+      "step": 25346
+    },
+    {
+      "epoch": 0.5632666666666667,
+      "grad_norm": 1.522444486618042,
+      "learning_rate": 8.736608135141144e-05,
+      "loss": 1.5384,
+      "step": 25347
+    },
+    {
+      "epoch": 0.5632888888888888,
+      "grad_norm": 1.6658782958984375,
+      "learning_rate": 8.736163591909313e-05,
+      "loss": 1.5798,
+      "step": 25348
+    },
+    {
+      "epoch": 0.5633111111111111,
+      "grad_norm": 1.4224662780761719,
+      "learning_rate": 8.735719048677484e-05,
+      "loss": 1.1295,
+      "step": 25349
+    },
+    {
+      "epoch": 0.5633333333333334,
+      "grad_norm": 1.9147154092788696,
+      "learning_rate": 8.735274505445655e-05,
+      "loss": 1.8996,
+      "step": 25350
+    },
+    {
+      "epoch": 0.5633555555555556,
+      "grad_norm": 1.0534144639968872,
+      "learning_rate": 8.734829962213825e-05,
+      "loss": 1.1374,
+      "step": 25351
+    },
+    {
+      "epoch": 0.5633777777777778,
+      "grad_norm": 1.0356683731079102,
+      "learning_rate": 8.734385418981996e-05,
+      "loss": 0.9505,
+      "step": 25352
+    },
+    {
+      "epoch": 0.5634,
+      "grad_norm": 1.3017328977584839,
+      "learning_rate": 8.733940875750167e-05,
+      "loss": 2.0527,
+      "step": 25353
+    },
+    {
+      "epoch": 0.5634222222222223,
+      "grad_norm": 1.7526426315307617,
+      "learning_rate": 8.733496332518338e-05,
+      "loss": 2.1973,
+      "step": 25354
+    },
+    {
+      "epoch": 0.5634444444444444,
+      "grad_norm": 1.5464191436767578,
+      "learning_rate": 8.733051789286509e-05,
+      "loss": 2.1379,
+      "step": 25355
+    },
+    {
+      "epoch": 0.5634666666666667,
+      "grad_norm": 1.9003633260726929,
+      "learning_rate": 8.73260724605468e-05,
+      "loss": 2.214,
+      "step": 25356
+    },
+    {
+      "epoch": 0.5634888888888889,
+      "grad_norm": 1.7008228302001953,
+      "learning_rate": 8.73216270282285e-05,
+      "loss": 2.2851,
+      "step": 25357
+    },
+    {
+      "epoch": 0.5635111111111111,
+      "grad_norm": 1.871749997138977,
+      "learning_rate": 8.73171815959102e-05,
+      "loss": 2.1141,
+      "step": 25358
+    },
+    {
+      "epoch": 0.5635333333333333,
+      "grad_norm": 1.8267793655395508,
+      "learning_rate": 8.731273616359191e-05,
+      "loss": 2.3976,
+      "step": 25359
+    },
+    {
+      "epoch": 0.5635555555555556,
+      "grad_norm": 1.6229313611984253,
+      "learning_rate": 8.730829073127362e-05,
+      "loss": 2.0573,
+      "step": 25360
+    },
+    {
+      "epoch": 0.5635777777777777,
+      "grad_norm": 1.5267568826675415,
+      "learning_rate": 8.730384529895532e-05,
+      "loss": 1.8809,
+      "step": 25361
+    },
+    {
+      "epoch": 0.5636,
+      "grad_norm": 2.0085060596466064,
+      "learning_rate": 8.729939986663703e-05,
+      "loss": 2.2086,
+      "step": 25362
+    },
+    {
+      "epoch": 0.5636222222222222,
+      "grad_norm": 1.1693706512451172,
+      "learning_rate": 8.729495443431874e-05,
+      "loss": 1.1021,
+      "step": 25363
+    },
+    {
+      "epoch": 0.5636444444444444,
+      "grad_norm": 1.5982754230499268,
+      "learning_rate": 8.729050900200046e-05,
+      "loss": 1.9899,
+      "step": 25364
+    },
+    {
+      "epoch": 0.5636666666666666,
+      "grad_norm": 1.4585381746292114,
+      "learning_rate": 8.728606356968216e-05,
+      "loss": 1.9564,
+      "step": 25365
+    },
+    {
+      "epoch": 0.5636888888888889,
+      "grad_norm": 1.5413857698440552,
+      "learning_rate": 8.728161813736387e-05,
+      "loss": 2.1478,
+      "step": 25366
+    },
+    {
+      "epoch": 0.5637111111111112,
+      "grad_norm": 1.5745688676834106,
+      "learning_rate": 8.727717270504558e-05,
+      "loss": 1.9242,
+      "step": 25367
+    },
+    {
+      "epoch": 0.5637333333333333,
+      "grad_norm": 1.760370135307312,
+      "learning_rate": 8.727272727272727e-05,
+      "loss": 1.9975,
+      "step": 25368
+    },
+    {
+      "epoch": 0.5637555555555556,
+      "grad_norm": 1.6587774753570557,
+      "learning_rate": 8.726828184040898e-05,
+      "loss": 2.0057,
+      "step": 25369
+    },
+    {
+      "epoch": 0.5637777777777778,
+      "grad_norm": 1.70685613155365,
+      "learning_rate": 8.726383640809069e-05,
+      "loss": 1.92,
+      "step": 25370
+    },
+    {
+      "epoch": 0.5638,
+      "grad_norm": 1.9009977579116821,
+      "learning_rate": 8.725939097577239e-05,
+      "loss": 1.9608,
+      "step": 25371
+    },
+    {
+      "epoch": 0.5638222222222222,
+      "grad_norm": 1.8968544006347656,
+      "learning_rate": 8.725494554345411e-05,
+      "loss": 1.8163,
+      "step": 25372
+    },
+    {
+      "epoch": 0.5638444444444445,
+      "grad_norm": 1.626792311668396,
+      "learning_rate": 8.725050011113582e-05,
+      "loss": 2.1957,
+      "step": 25373
+    },
+    {
+      "epoch": 0.5638666666666666,
+      "grad_norm": 1.4381468296051025,
+      "learning_rate": 8.724605467881753e-05,
+      "loss": 1.5413,
+      "step": 25374
+    },
+    {
+      "epoch": 0.5638888888888889,
+      "grad_norm": 2.1948907375335693,
+      "learning_rate": 8.724160924649923e-05,
+      "loss": 2.0981,
+      "step": 25375
+    },
+    {
+      "epoch": 0.5639111111111111,
+      "grad_norm": 1.748815894126892,
+      "learning_rate": 8.723716381418093e-05,
+      "loss": 2.2593,
+      "step": 25376
+    },
+    {
+      "epoch": 0.5639333333333333,
+      "grad_norm": 1.7379071712493896,
+      "learning_rate": 8.723271838186264e-05,
+      "loss": 2.4922,
+      "step": 25377
+    },
+    {
+      "epoch": 0.5639555555555555,
+      "grad_norm": 1.5309514999389648,
+      "learning_rate": 8.722827294954434e-05,
+      "loss": 1.9083,
+      "step": 25378
+    },
+    {
+      "epoch": 0.5639777777777778,
+      "grad_norm": 1.3660348653793335,
+      "learning_rate": 8.722382751722605e-05,
+      "loss": 0.8973,
+      "step": 25379
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 1.711879849433899,
+      "learning_rate": 8.721938208490776e-05,
+      "loss": 1.8977,
+      "step": 25380
+    },
+    {
+      "epoch": 0.5640222222222222,
+      "grad_norm": 1.8451247215270996,
+      "learning_rate": 8.721493665258947e-05,
+      "loss": 2.0085,
+      "step": 25381
+    },
+    {
+      "epoch": 0.5640444444444445,
+      "grad_norm": 1.7080289125442505,
+      "learning_rate": 8.721049122027118e-05,
+      "loss": 2.0674,
+      "step": 25382
+    },
+    {
+      "epoch": 0.5640666666666667,
+      "grad_norm": 1.6591728925704956,
+      "learning_rate": 8.720604578795289e-05,
+      "loss": 1.9249,
+      "step": 25383
+    },
+    {
+      "epoch": 0.5640888888888889,
+      "grad_norm": 1.966254711151123,
+      "learning_rate": 8.72016003556346e-05,
+      "loss": 1.6539,
+      "step": 25384
+    },
+    {
+      "epoch": 0.5641111111111111,
+      "grad_norm": 1.779379963874817,
+      "learning_rate": 8.71971549233163e-05,
+      "loss": 2.3801,
+      "step": 25385
+    },
+    {
+      "epoch": 0.5641333333333334,
+      "grad_norm": 1.5594086647033691,
+      "learning_rate": 8.7192709490998e-05,
+      "loss": 1.9107,
+      "step": 25386
+    },
+    {
+      "epoch": 0.5641555555555555,
+      "grad_norm": 2.0875155925750732,
+      "learning_rate": 8.718826405867971e-05,
+      "loss": 2.1634,
+      "step": 25387
+    },
+    {
+      "epoch": 0.5641777777777778,
+      "grad_norm": 1.5301467180252075,
+      "learning_rate": 8.718381862636141e-05,
+      "loss": 1.6208,
+      "step": 25388
+    },
+    {
+      "epoch": 0.5642,
+      "grad_norm": 1.5342662334442139,
+      "learning_rate": 8.717937319404312e-05,
+      "loss": 1.7199,
+      "step": 25389
+    },
+    {
+      "epoch": 0.5642222222222222,
+      "grad_norm": 1.5771845579147339,
+      "learning_rate": 8.717492776172483e-05,
+      "loss": 1.8793,
+      "step": 25390
+    },
+    {
+      "epoch": 0.5642444444444444,
+      "grad_norm": 1.7096185684204102,
+      "learning_rate": 8.717048232940654e-05,
+      "loss": 1.5729,
+      "step": 25391
+    },
+    {
+      "epoch": 0.5642666666666667,
+      "grad_norm": 1.581945538520813,
+      "learning_rate": 8.716603689708825e-05,
+      "loss": 1.6892,
+      "step": 25392
+    },
+    {
+      "epoch": 0.5642888888888888,
+      "grad_norm": 1.7134122848510742,
+      "learning_rate": 8.716159146476996e-05,
+      "loss": 1.5713,
+      "step": 25393
+    },
+    {
+      "epoch": 0.5643111111111111,
+      "grad_norm": 1.9367319345474243,
+      "learning_rate": 8.715714603245167e-05,
+      "loss": 1.9274,
+      "step": 25394
+    },
+    {
+      "epoch": 0.5643333333333334,
+      "grad_norm": 1.7920246124267578,
+      "learning_rate": 8.715270060013336e-05,
+      "loss": 1.7583,
+      "step": 25395
+    },
+    {
+      "epoch": 0.5643555555555556,
+      "grad_norm": 1.793280839920044,
+      "learning_rate": 8.714825516781507e-05,
+      "loss": 1.9587,
+      "step": 25396
+    },
+    {
+      "epoch": 0.5643777777777778,
+      "grad_norm": 1.8033382892608643,
+      "learning_rate": 8.714380973549678e-05,
+      "loss": 1.6978,
+      "step": 25397
+    },
+    {
+      "epoch": 0.5644,
+      "grad_norm": 1.66734778881073,
+      "learning_rate": 8.713936430317848e-05,
+      "loss": 1.7054,
+      "step": 25398
+    },
+    {
+      "epoch": 0.5644222222222223,
+      "grad_norm": 1.5701909065246582,
+      "learning_rate": 8.713491887086019e-05,
+      "loss": 1.5448,
+      "step": 25399
+    },
+    {
+      "epoch": 0.5644444444444444,
+      "grad_norm": 1.4017804861068726,
+      "learning_rate": 8.71304734385419e-05,
+      "loss": 0.9378,
+      "step": 25400
+    },
+    {
+      "epoch": 0.5644666666666667,
+      "grad_norm": 1.4009686708450317,
+      "learning_rate": 8.712602800622362e-05,
+      "loss": 2.3716,
+      "step": 25401
+    },
+    {
+      "epoch": 0.5644888888888889,
+      "grad_norm": 1.7983325719833374,
+      "learning_rate": 8.712158257390532e-05,
+      "loss": 2.7815,
+      "step": 25402
+    },
+    {
+      "epoch": 0.5645111111111111,
+      "grad_norm": 1.7624199390411377,
+      "learning_rate": 8.711713714158703e-05,
+      "loss": 1.9048,
+      "step": 25403
+    },
+    {
+      "epoch": 0.5645333333333333,
+      "grad_norm": 1.4177815914154053,
+      "learning_rate": 8.711269170926874e-05,
+      "loss": 2.0578,
+      "step": 25404
+    },
+    {
+      "epoch": 0.5645555555555556,
+      "grad_norm": 1.2785425186157227,
+      "learning_rate": 8.710824627695043e-05,
+      "loss": 1.5198,
+      "step": 25405
+    },
+    {
+      "epoch": 0.5645777777777777,
+      "grad_norm": 1.5148561000823975,
+      "learning_rate": 8.710380084463214e-05,
+      "loss": 2.316,
+      "step": 25406
+    },
+    {
+      "epoch": 0.5646,
+      "grad_norm": 1.6512395143508911,
+      "learning_rate": 8.709935541231385e-05,
+      "loss": 2.118,
+      "step": 25407
+    },
+    {
+      "epoch": 0.5646222222222222,
+      "grad_norm": 1.554484248161316,
+      "learning_rate": 8.709490997999555e-05,
+      "loss": 2.4742,
+      "step": 25408
+    },
+    {
+      "epoch": 0.5646444444444444,
+      "grad_norm": 1.5429123640060425,
+      "learning_rate": 8.709046454767727e-05,
+      "loss": 2.1788,
+      "step": 25409
+    },
+    {
+      "epoch": 0.5646666666666667,
+      "grad_norm": 1.8350896835327148,
+      "learning_rate": 8.708601911535898e-05,
+      "loss": 2.4694,
+      "step": 25410
+    },
+    {
+      "epoch": 0.5646888888888889,
+      "grad_norm": 1.7665791511535645,
+      "learning_rate": 8.708157368304069e-05,
+      "loss": 2.2558,
+      "step": 25411
+    },
+    {
+      "epoch": 0.5647111111111112,
+      "grad_norm": 1.6597291231155396,
+      "learning_rate": 8.707712825072239e-05,
+      "loss": 2.2376,
+      "step": 25412
+    },
+    {
+      "epoch": 0.5647333333333333,
+      "grad_norm": 1.2979716062545776,
+      "learning_rate": 8.70726828184041e-05,
+      "loss": 1.1761,
+      "step": 25413
+    },
+    {
+      "epoch": 0.5647555555555556,
+      "grad_norm": 1.4011353254318237,
+      "learning_rate": 8.70682373860858e-05,
+      "loss": 1.821,
+      "step": 25414
+    },
+    {
+      "epoch": 0.5647777777777778,
+      "grad_norm": 1.4270352125167847,
+      "learning_rate": 8.70637919537675e-05,
+      "loss": 1.8507,
+      "step": 25415
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 1.5780997276306152,
+      "learning_rate": 8.705934652144921e-05,
+      "loss": 1.836,
+      "step": 25416
+    },
+    {
+      "epoch": 0.5648222222222222,
+      "grad_norm": 1.5082387924194336,
+      "learning_rate": 8.705490108913092e-05,
+      "loss": 2.2373,
+      "step": 25417
+    },
+    {
+      "epoch": 0.5648444444444445,
+      "grad_norm": 1.6113547086715698,
+      "learning_rate": 8.705045565681263e-05,
+      "loss": 1.6091,
+      "step": 25418
+    },
+    {
+      "epoch": 0.5648666666666666,
+      "grad_norm": 1.0369369983673096,
+      "learning_rate": 8.704601022449434e-05,
+      "loss": 0.5911,
+      "step": 25419
+    },
+    {
+      "epoch": 0.5648888888888889,
+      "grad_norm": 1.4776184558868408,
+      "learning_rate": 8.704156479217605e-05,
+      "loss": 1.1912,
+      "step": 25420
+    },
+    {
+      "epoch": 0.5649111111111111,
+      "grad_norm": 1.6801823377609253,
+      "learning_rate": 8.703711935985776e-05,
+      "loss": 2.1828,
+      "step": 25421
+    },
+    {
+      "epoch": 0.5649333333333333,
+      "grad_norm": 1.4651553630828857,
+      "learning_rate": 8.703267392753946e-05,
+      "loss": 1.4728,
+      "step": 25422
+    },
+    {
+      "epoch": 0.5649555555555555,
+      "grad_norm": 1.9302335977554321,
+      "learning_rate": 8.702822849522116e-05,
+      "loss": 2.0986,
+      "step": 25423
+    },
+    {
+      "epoch": 0.5649777777777778,
+      "grad_norm": 1.7884409427642822,
+      "learning_rate": 8.702378306290287e-05,
+      "loss": 2.0985,
+      "step": 25424
+    },
+    {
+      "epoch": 0.565,
+      "grad_norm": 1.600262999534607,
+      "learning_rate": 8.701933763058457e-05,
+      "loss": 2.1894,
+      "step": 25425
+    },
+    {
+      "epoch": 0.5650222222222222,
+      "grad_norm": 2.0447311401367188,
+      "learning_rate": 8.701489219826628e-05,
+      "loss": 2.156,
+      "step": 25426
+    },
+    {
+      "epoch": 0.5650444444444445,
+      "grad_norm": 1.5858628749847412,
+      "learning_rate": 8.701044676594799e-05,
+      "loss": 1.6714,
+      "step": 25427
+    },
+    {
+      "epoch": 0.5650666666666667,
+      "grad_norm": 1.7606112957000732,
+      "learning_rate": 8.70060013336297e-05,
+      "loss": 1.4105,
+      "step": 25428
+    },
+    {
+      "epoch": 0.5650888888888889,
+      "grad_norm": 1.3921977281570435,
+      "learning_rate": 8.700155590131141e-05,
+      "loss": 1.7857,
+      "step": 25429
+    },
+    {
+      "epoch": 0.5651111111111111,
+      "grad_norm": 1.7770886421203613,
+      "learning_rate": 8.699711046899312e-05,
+      "loss": 1.7963,
+      "step": 25430
+    },
+    {
+      "epoch": 0.5651333333333334,
+      "grad_norm": 1.8601276874542236,
+      "learning_rate": 8.699266503667483e-05,
+      "loss": 2.1159,
+      "step": 25431
+    },
+    {
+      "epoch": 0.5651555555555555,
+      "grad_norm": 1.8692203760147095,
+      "learning_rate": 8.698821960435652e-05,
+      "loss": 1.8525,
+      "step": 25432
+    },
+    {
+      "epoch": 0.5651777777777778,
+      "grad_norm": 1.6445821523666382,
+      "learning_rate": 8.698377417203823e-05,
+      "loss": 0.6555,
+      "step": 25433
+    },
+    {
+      "epoch": 0.5652,
+      "grad_norm": 1.7919769287109375,
+      "learning_rate": 8.697932873971994e-05,
+      "loss": 2.061,
+      "step": 25434
+    },
+    {
+      "epoch": 0.5652222222222222,
+      "grad_norm": 2.170060396194458,
+      "learning_rate": 8.697488330740164e-05,
+      "loss": 2.4497,
+      "step": 25435
+    },
+    {
+      "epoch": 0.5652444444444444,
+      "grad_norm": 1.3361024856567383,
+      "learning_rate": 8.697043787508335e-05,
+      "loss": 1.0493,
+      "step": 25436
+    },
+    {
+      "epoch": 0.5652666666666667,
+      "grad_norm": 1.9073244333267212,
+      "learning_rate": 8.696599244276506e-05,
+      "loss": 2.2122,
+      "step": 25437
+    },
+    {
+      "epoch": 0.5652888888888888,
+      "grad_norm": 1.7902787923812866,
+      "learning_rate": 8.696154701044678e-05,
+      "loss": 1.9999,
+      "step": 25438
+    },
+    {
+      "epoch": 0.5653111111111111,
+      "grad_norm": 2.0787904262542725,
+      "learning_rate": 8.695710157812848e-05,
+      "loss": 2.064,
+      "step": 25439
+    },
+    {
+      "epoch": 0.5653333333333334,
+      "grad_norm": 1.44843327999115,
+      "learning_rate": 8.695265614581019e-05,
+      "loss": 1.5858,
+      "step": 25440
+    },
+    {
+      "epoch": 0.5653555555555555,
+      "grad_norm": 1.5655252933502197,
+      "learning_rate": 8.69482107134919e-05,
+      "loss": 1.8586,
+      "step": 25441
+    },
+    {
+      "epoch": 0.5653777777777778,
+      "grad_norm": 1.646780014038086,
+      "learning_rate": 8.69437652811736e-05,
+      "loss": 1.7439,
+      "step": 25442
+    },
+    {
+      "epoch": 0.5654,
+      "grad_norm": 1.7771084308624268,
+      "learning_rate": 8.69393198488553e-05,
+      "loss": 1.6262,
+      "step": 25443
+    },
+    {
+      "epoch": 0.5654222222222223,
+      "grad_norm": 1.6789112091064453,
+      "learning_rate": 8.693487441653701e-05,
+      "loss": 1.8269,
+      "step": 25444
+    },
+    {
+      "epoch": 0.5654444444444444,
+      "grad_norm": 2.0198254585266113,
+      "learning_rate": 8.693042898421871e-05,
+      "loss": 1.6695,
+      "step": 25445
+    },
+    {
+      "epoch": 0.5654666666666667,
+      "grad_norm": 1.8060046434402466,
+      "learning_rate": 8.692598355190043e-05,
+      "loss": 1.9623,
+      "step": 25446
+    },
+    {
+      "epoch": 0.5654888888888889,
+      "grad_norm": 1.9680043458938599,
+      "learning_rate": 8.692153811958214e-05,
+      "loss": 1.9075,
+      "step": 25447
+    },
+    {
+      "epoch": 0.5655111111111111,
+      "grad_norm": 1.6656614542007446,
+      "learning_rate": 8.691709268726385e-05,
+      "loss": 1.648,
+      "step": 25448
+    },
+    {
+      "epoch": 0.5655333333333333,
+      "grad_norm": 1.683111548423767,
+      "learning_rate": 8.691264725494555e-05,
+      "loss": 1.0773,
+      "step": 25449
+    },
+    {
+      "epoch": 0.5655555555555556,
+      "grad_norm": 1.7862014770507812,
+      "learning_rate": 8.690820182262726e-05,
+      "loss": 1.8669,
+      "step": 25450
+    },
+    {
+      "epoch": 0.5655777777777777,
+      "grad_norm": 1.6154983043670654,
+      "learning_rate": 8.690375639030897e-05,
+      "loss": 1.9956,
+      "step": 25451
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 1.482670783996582,
+      "learning_rate": 8.689931095799066e-05,
+      "loss": 2.263,
+      "step": 25452
+    },
+    {
+      "epoch": 0.5656222222222222,
+      "grad_norm": 1.6822906732559204,
+      "learning_rate": 8.689486552567237e-05,
+      "loss": 0.0431,
+      "step": 25453
+    },
+    {
+      "epoch": 0.5656444444444444,
+      "grad_norm": 1.6634759902954102,
+      "learning_rate": 8.689042009335408e-05,
+      "loss": 2.292,
+      "step": 25454
+    },
+    {
+      "epoch": 0.5656666666666667,
+      "grad_norm": 1.638785481452942,
+      "learning_rate": 8.688597466103579e-05,
+      "loss": 2.1165,
+      "step": 25455
+    },
+    {
+      "epoch": 0.5656888888888889,
+      "grad_norm": 1.6883729696273804,
+      "learning_rate": 8.68815292287175e-05,
+      "loss": 2.2838,
+      "step": 25456
+    },
+    {
+      "epoch": 0.5657111111111112,
+      "grad_norm": 1.673374056816101,
+      "learning_rate": 8.687708379639921e-05,
+      "loss": 2.4126,
+      "step": 25457
+    },
+    {
+      "epoch": 0.5657333333333333,
+      "grad_norm": 1.664180040359497,
+      "learning_rate": 8.687263836408092e-05,
+      "loss": 2.2606,
+      "step": 25458
+    },
+    {
+      "epoch": 0.5657555555555556,
+      "grad_norm": 1.4526246786117554,
+      "learning_rate": 8.686819293176262e-05,
+      "loss": 2.0279,
+      "step": 25459
+    },
+    {
+      "epoch": 0.5657777777777778,
+      "grad_norm": 1.7229489088058472,
+      "learning_rate": 8.686374749944433e-05,
+      "loss": 2.5118,
+      "step": 25460
+    },
+    {
+      "epoch": 0.5658,
+      "grad_norm": 1.9437999725341797,
+      "learning_rate": 8.685930206712604e-05,
+      "loss": 2.7655,
+      "step": 25461
+    },
+    {
+      "epoch": 0.5658222222222222,
+      "grad_norm": 1.7487212419509888,
+      "learning_rate": 8.685485663480773e-05,
+      "loss": 2.2664,
+      "step": 25462
+    },
+    {
+      "epoch": 0.5658444444444445,
+      "grad_norm": 1.682439923286438,
+      "learning_rate": 8.685041120248944e-05,
+      "loss": 2.0246,
+      "step": 25463
+    },
+    {
+      "epoch": 0.5658666666666666,
+      "grad_norm": 1.7988824844360352,
+      "learning_rate": 8.684596577017115e-05,
+      "loss": 2.1121,
+      "step": 25464
+    },
+    {
+      "epoch": 0.5658888888888889,
+      "grad_norm": 1.8382154703140259,
+      "learning_rate": 8.684152033785286e-05,
+      "loss": 2.0848,
+      "step": 25465
+    },
+    {
+      "epoch": 0.5659111111111111,
+      "grad_norm": 1.9211585521697998,
+      "learning_rate": 8.683707490553457e-05,
+      "loss": 2.1939,
+      "step": 25466
+    },
+    {
+      "epoch": 0.5659333333333333,
+      "grad_norm": 1.4006104469299316,
+      "learning_rate": 8.683262947321628e-05,
+      "loss": 1.895,
+      "step": 25467
+    },
+    {
+      "epoch": 0.5659555555555555,
+      "grad_norm": 1.5469164848327637,
+      "learning_rate": 8.682818404089799e-05,
+      "loss": 2.3618,
+      "step": 25468
+    },
+    {
+      "epoch": 0.5659777777777778,
+      "grad_norm": 1.4230570793151855,
+      "learning_rate": 8.682373860857969e-05,
+      "loss": 1.7998,
+      "step": 25469
+    },
+    {
+      "epoch": 0.566,
+      "grad_norm": 1.1125915050506592,
+      "learning_rate": 8.68192931762614e-05,
+      "loss": 0.9492,
+      "step": 25470
+    },
+    {
+      "epoch": 0.5660222222222222,
+      "grad_norm": 1.9148786067962646,
+      "learning_rate": 8.68148477439431e-05,
+      "loss": 1.5855,
+      "step": 25471
+    },
+    {
+      "epoch": 0.5660444444444445,
+      "grad_norm": 1.9745311737060547,
+      "learning_rate": 8.68104023116248e-05,
+      "loss": 2.2945,
+      "step": 25472
+    },
+    {
+      "epoch": 0.5660666666666667,
+      "grad_norm": 1.7374578714370728,
+      "learning_rate": 8.680595687930651e-05,
+      "loss": 2.0763,
+      "step": 25473
+    },
+    {
+      "epoch": 0.5660888888888889,
+      "grad_norm": 1.493938684463501,
+      "learning_rate": 8.680151144698822e-05,
+      "loss": 2.2095,
+      "step": 25474
+    },
+    {
+      "epoch": 0.5661111111111111,
+      "grad_norm": 1.388858675956726,
+      "learning_rate": 8.679706601466993e-05,
+      "loss": 1.9734,
+      "step": 25475
+    },
+    {
+      "epoch": 0.5661333333333334,
+      "grad_norm": 1.6583508253097534,
+      "learning_rate": 8.679262058235164e-05,
+      "loss": 2.0741,
+      "step": 25476
+    },
+    {
+      "epoch": 0.5661555555555555,
+      "grad_norm": 1.6548030376434326,
+      "learning_rate": 8.678817515003335e-05,
+      "loss": 2.3124,
+      "step": 25477
+    },
+    {
+      "epoch": 0.5661777777777778,
+      "grad_norm": 1.4862264394760132,
+      "learning_rate": 8.678372971771506e-05,
+      "loss": 1.6101,
+      "step": 25478
+    },
+    {
+      "epoch": 0.5662,
+      "grad_norm": 1.6298364400863647,
+      "learning_rate": 8.677928428539675e-05,
+      "loss": 1.9753,
+      "step": 25479
+    },
+    {
+      "epoch": 0.5662222222222222,
+      "grad_norm": 1.9583399295806885,
+      "learning_rate": 8.677483885307846e-05,
+      "loss": 1.9028,
+      "step": 25480
+    },
+    {
+      "epoch": 0.5662444444444444,
+      "grad_norm": 1.9671558141708374,
+      "learning_rate": 8.677039342076017e-05,
+      "loss": 2.0369,
+      "step": 25481
+    },
+    {
+      "epoch": 0.5662666666666667,
+      "grad_norm": 1.336354374885559,
+      "learning_rate": 8.676594798844187e-05,
+      "loss": 1.4759,
+      "step": 25482
+    },
+    {
+      "epoch": 0.5662888888888888,
+      "grad_norm": 1.658531665802002,
+      "learning_rate": 8.676150255612359e-05,
+      "loss": 2.0151,
+      "step": 25483
+    },
+    {
+      "epoch": 0.5663111111111111,
+      "grad_norm": 1.5404964685440063,
+      "learning_rate": 8.67570571238053e-05,
+      "loss": 2.1178,
+      "step": 25484
+    },
+    {
+      "epoch": 0.5663333333333334,
+      "grad_norm": 1.5054473876953125,
+      "learning_rate": 8.6752611691487e-05,
+      "loss": 1.8238,
+      "step": 25485
+    },
+    {
+      "epoch": 0.5663555555555555,
+      "grad_norm": 1.724499225616455,
+      "learning_rate": 8.674816625916871e-05,
+      "loss": 1.7396,
+      "step": 25486
+    },
+    {
+      "epoch": 0.5663777777777778,
+      "grad_norm": 1.8233646154403687,
+      "learning_rate": 8.674372082685042e-05,
+      "loss": 1.731,
+      "step": 25487
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 1.9924031496047974,
+      "learning_rate": 8.673927539453213e-05,
+      "loss": 1.9948,
+      "step": 25488
+    },
+    {
+      "epoch": 0.5664222222222223,
+      "grad_norm": 1.707899808883667,
+      "learning_rate": 8.673482996221382e-05,
+      "loss": 2.2374,
+      "step": 25489
+    },
+    {
+      "epoch": 0.5664444444444444,
+      "grad_norm": 2.0084919929504395,
+      "learning_rate": 8.673038452989553e-05,
+      "loss": 2.199,
+      "step": 25490
+    },
+    {
+      "epoch": 0.5664666666666667,
+      "grad_norm": 1.6072585582733154,
+      "learning_rate": 8.672593909757724e-05,
+      "loss": 1.6715,
+      "step": 25491
+    },
+    {
+      "epoch": 0.5664888888888889,
+      "grad_norm": 1.8686779737472534,
+      "learning_rate": 8.672149366525895e-05,
+      "loss": 1.6971,
+      "step": 25492
+    },
+    {
+      "epoch": 0.5665111111111111,
+      "grad_norm": 1.6931346654891968,
+      "learning_rate": 8.671704823294066e-05,
+      "loss": 1.9617,
+      "step": 25493
+    },
+    {
+      "epoch": 0.5665333333333333,
+      "grad_norm": 1.8643832206726074,
+      "learning_rate": 8.671260280062237e-05,
+      "loss": 2.1704,
+      "step": 25494
+    },
+    {
+      "epoch": 0.5665555555555556,
+      "grad_norm": 2.1833972930908203,
+      "learning_rate": 8.670815736830408e-05,
+      "loss": 2.103,
+      "step": 25495
+    },
+    {
+      "epoch": 0.5665777777777777,
+      "grad_norm": 2.137716770172119,
+      "learning_rate": 8.670371193598578e-05,
+      "loss": 1.9764,
+      "step": 25496
+    },
+    {
+      "epoch": 0.5666,
+      "grad_norm": 1.7726776599884033,
+      "learning_rate": 8.669926650366749e-05,
+      "loss": 1.8766,
+      "step": 25497
+    },
+    {
+      "epoch": 0.5666222222222222,
+      "grad_norm": 2.3358371257781982,
+      "learning_rate": 8.66948210713492e-05,
+      "loss": 2.0126,
+      "step": 25498
+    },
+    {
+      "epoch": 0.5666444444444444,
+      "grad_norm": 1.7990282773971558,
+      "learning_rate": 8.669037563903089e-05,
+      "loss": 1.8925,
+      "step": 25499
+    },
+    {
+      "epoch": 0.5666666666666667,
+      "grad_norm": 2.0586588382720947,
+      "learning_rate": 8.66859302067126e-05,
+      "loss": 1.8382,
+      "step": 25500
+    },
+    {
+      "epoch": 0.5666888888888889,
+      "grad_norm": 1.2641650438308716,
+      "learning_rate": 8.668148477439431e-05,
+      "loss": 2.1141,
+      "step": 25501
+    },
+    {
+      "epoch": 0.5667111111111112,
+      "grad_norm": 1.080187201499939,
+      "learning_rate": 8.667703934207602e-05,
+      "loss": 1.0952,
+      "step": 25502
+    },
+    {
+      "epoch": 0.5667333333333333,
+      "grad_norm": 1.647226095199585,
+      "learning_rate": 8.667259390975773e-05,
+      "loss": 2.2651,
+      "step": 25503
+    },
+    {
+      "epoch": 0.5667555555555556,
+      "grad_norm": 1.6892226934432983,
+      "learning_rate": 8.666814847743944e-05,
+      "loss": 2.348,
+      "step": 25504
+    },
+    {
+      "epoch": 0.5667777777777778,
+      "grad_norm": 1.515511155128479,
+      "learning_rate": 8.666370304512115e-05,
+      "loss": 2.0383,
+      "step": 25505
+    },
+    {
+      "epoch": 0.5668,
+      "grad_norm": 1.4195060729980469,
+      "learning_rate": 8.665925761280285e-05,
+      "loss": 1.9549,
+      "step": 25506
+    },
+    {
+      "epoch": 0.5668222222222222,
+      "grad_norm": 0.9295563697814941,
+      "learning_rate": 8.665481218048456e-05,
+      "loss": 0.0285,
+      "step": 25507
+    },
+    {
+      "epoch": 0.5668444444444445,
+      "grad_norm": 1.2143354415893555,
+      "learning_rate": 8.665036674816627e-05,
+      "loss": 1.6016,
+      "step": 25508
+    },
+    {
+      "epoch": 0.5668666666666666,
+      "grad_norm": 1.9575049877166748,
+      "learning_rate": 8.664592131584796e-05,
+      "loss": 2.2171,
+      "step": 25509
+    },
+    {
+      "epoch": 0.5668888888888889,
+      "grad_norm": 1.8051297664642334,
+      "learning_rate": 8.664147588352967e-05,
+      "loss": 2.6714,
+      "step": 25510
+    },
+    {
+      "epoch": 0.5669111111111111,
+      "grad_norm": 1.5467302799224854,
+      "learning_rate": 8.663703045121138e-05,
+      "loss": 1.867,
+      "step": 25511
+    },
+    {
+      "epoch": 0.5669333333333333,
+      "grad_norm": 1.3129912614822388,
+      "learning_rate": 8.663258501889309e-05,
+      "loss": 1.8731,
+      "step": 25512
+    },
+    {
+      "epoch": 0.5669555555555555,
+      "grad_norm": 1.4830926656723022,
+      "learning_rate": 8.66281395865748e-05,
+      "loss": 1.9213,
+      "step": 25513
+    },
+    {
+      "epoch": 0.5669777777777778,
+      "grad_norm": 1.6727464199066162,
+      "learning_rate": 8.662369415425651e-05,
+      "loss": 2.009,
+      "step": 25514
+    },
+    {
+      "epoch": 0.567,
+      "grad_norm": 1.5978381633758545,
+      "learning_rate": 8.661924872193822e-05,
+      "loss": 2.2661,
+      "step": 25515
+    },
+    {
+      "epoch": 0.5670222222222222,
+      "grad_norm": 1.3395562171936035,
+      "learning_rate": 8.661480328961992e-05,
+      "loss": 1.7839,
+      "step": 25516
+    },
+    {
+      "epoch": 0.5670444444444445,
+      "grad_norm": 1.7934595346450806,
+      "learning_rate": 8.661035785730162e-05,
+      "loss": 2.2136,
+      "step": 25517
+    },
+    {
+      "epoch": 0.5670666666666667,
+      "grad_norm": 1.6741819381713867,
+      "learning_rate": 8.660591242498333e-05,
+      "loss": 1.6916,
+      "step": 25518
+    },
+    {
+      "epoch": 0.5670888888888889,
+      "grad_norm": 1.6226348876953125,
+      "learning_rate": 8.660146699266503e-05,
+      "loss": 1.9363,
+      "step": 25519
+    },
+    {
+      "epoch": 0.5671111111111111,
+      "grad_norm": 1.6335357427597046,
+      "learning_rate": 8.659702156034675e-05,
+      "loss": 1.9537,
+      "step": 25520
+    },
+    {
+      "epoch": 0.5671333333333334,
+      "grad_norm": 1.6441919803619385,
+      "learning_rate": 8.659257612802846e-05,
+      "loss": 1.8217,
+      "step": 25521
+    },
+    {
+      "epoch": 0.5671555555555555,
+      "grad_norm": 1.528879165649414,
+      "learning_rate": 8.658813069571016e-05,
+      "loss": 1.9105,
+      "step": 25522
+    },
+    {
+      "epoch": 0.5671777777777778,
+      "grad_norm": 1.4856935739517212,
+      "learning_rate": 8.658368526339187e-05,
+      "loss": 1.7963,
+      "step": 25523
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 1.4754873514175415,
+      "learning_rate": 8.657923983107358e-05,
+      "loss": 1.2109,
+      "step": 25524
+    },
+    {
+      "epoch": 0.5672222222222222,
+      "grad_norm": 1.4310427904129028,
+      "learning_rate": 8.657479439875529e-05,
+      "loss": 1.24,
+      "step": 25525
+    },
+    {
+      "epoch": 0.5672444444444444,
+      "grad_norm": 1.9599642753601074,
+      "learning_rate": 8.657034896643698e-05,
+      "loss": 2.1241,
+      "step": 25526
+    },
+    {
+      "epoch": 0.5672666666666667,
+      "grad_norm": 1.650848150253296,
+      "learning_rate": 8.65659035341187e-05,
+      "loss": 1.9539,
+      "step": 25527
+    },
+    {
+      "epoch": 0.5672888888888888,
+      "grad_norm": 1.6401448249816895,
+      "learning_rate": 8.65614581018004e-05,
+      "loss": 2.0195,
+      "step": 25528
+    },
+    {
+      "epoch": 0.5673111111111111,
+      "grad_norm": 1.3462976217269897,
+      "learning_rate": 8.655701266948211e-05,
+      "loss": 1.5376,
+      "step": 25529
+    },
+    {
+      "epoch": 0.5673333333333334,
+      "grad_norm": 1.7441065311431885,
+      "learning_rate": 8.655256723716382e-05,
+      "loss": 1.9041,
+      "step": 25530
+    },
+    {
+      "epoch": 0.5673555555555555,
+      "grad_norm": 1.5822638273239136,
+      "learning_rate": 8.654812180484553e-05,
+      "loss": 1.932,
+      "step": 25531
+    },
+    {
+      "epoch": 0.5673777777777778,
+      "grad_norm": 1.9012054204940796,
+      "learning_rate": 8.654367637252723e-05,
+      "loss": 1.9656,
+      "step": 25532
+    },
+    {
+      "epoch": 0.5674,
+      "grad_norm": 1.0392942428588867,
+      "learning_rate": 8.653923094020894e-05,
+      "loss": 0.8862,
+      "step": 25533
+    },
+    {
+      "epoch": 0.5674222222222223,
+      "grad_norm": 1.622781753540039,
+      "learning_rate": 8.653478550789065e-05,
+      "loss": 1.5679,
+      "step": 25534
+    },
+    {
+      "epoch": 0.5674444444444444,
+      "grad_norm": 1.4929344654083252,
+      "learning_rate": 8.653034007557236e-05,
+      "loss": 1.7067,
+      "step": 25535
+    },
+    {
+      "epoch": 0.5674666666666667,
+      "grad_norm": 1.837020993232727,
+      "learning_rate": 8.652589464325405e-05,
+      "loss": 1.6447,
+      "step": 25536
+    },
+    {
+      "epoch": 0.5674888888888889,
+      "grad_norm": 1.5858163833618164,
+      "learning_rate": 8.652144921093576e-05,
+      "loss": 1.742,
+      "step": 25537
+    },
+    {
+      "epoch": 0.5675111111111111,
+      "grad_norm": 1.4231958389282227,
+      "learning_rate": 8.651700377861747e-05,
+      "loss": 1.7786,
+      "step": 25538
+    },
+    {
+      "epoch": 0.5675333333333333,
+      "grad_norm": 1.6966981887817383,
+      "learning_rate": 8.651255834629918e-05,
+      "loss": 1.8127,
+      "step": 25539
+    },
+    {
+      "epoch": 0.5675555555555556,
+      "grad_norm": 1.7579383850097656,
+      "learning_rate": 8.650811291398089e-05,
+      "loss": 1.7069,
+      "step": 25540
+    },
+    {
+      "epoch": 0.5675777777777777,
+      "grad_norm": 1.9080697298049927,
+      "learning_rate": 8.65036674816626e-05,
+      "loss": 1.595,
+      "step": 25541
+    },
+    {
+      "epoch": 0.5676,
+      "grad_norm": 1.6535087823867798,
+      "learning_rate": 8.649922204934431e-05,
+      "loss": 1.7829,
+      "step": 25542
+    },
+    {
+      "epoch": 0.5676222222222223,
+      "grad_norm": 1.5289617776870728,
+      "learning_rate": 8.649477661702601e-05,
+      "loss": 1.5355,
+      "step": 25543
+    },
+    {
+      "epoch": 0.5676444444444444,
+      "grad_norm": 1.6726182699203491,
+      "learning_rate": 8.649033118470772e-05,
+      "loss": 1.6021,
+      "step": 25544
+    },
+    {
+      "epoch": 0.5676666666666667,
+      "grad_norm": 1.7910929918289185,
+      "learning_rate": 8.648588575238943e-05,
+      "loss": 1.7157,
+      "step": 25545
+    },
+    {
+      "epoch": 0.5676888888888889,
+      "grad_norm": 1.7049421072006226,
+      "learning_rate": 8.648144032007112e-05,
+      "loss": 2.0987,
+      "step": 25546
+    },
+    {
+      "epoch": 0.5677111111111112,
+      "grad_norm": 1.732585072517395,
+      "learning_rate": 8.647699488775283e-05,
+      "loss": 1.8902,
+      "step": 25547
+    },
+    {
+      "epoch": 0.5677333333333333,
+      "grad_norm": 1.5966678857803345,
+      "learning_rate": 8.647254945543454e-05,
+      "loss": 1.7052,
+      "step": 25548
+    },
+    {
+      "epoch": 0.5677555555555556,
+      "grad_norm": 2.260141372680664,
+      "learning_rate": 8.646810402311625e-05,
+      "loss": 1.7403,
+      "step": 25549
+    },
+    {
+      "epoch": 0.5677777777777778,
+      "grad_norm": 0.8711251020431519,
+      "learning_rate": 8.646365859079796e-05,
+      "loss": 0.4839,
+      "step": 25550
+    },
+    {
+      "epoch": 0.5678,
+      "grad_norm": 1.83405601978302,
+      "learning_rate": 8.645921315847967e-05,
+      "loss": 2.4266,
+      "step": 25551
+    },
+    {
+      "epoch": 0.5678222222222222,
+      "grad_norm": 1.4117273092269897,
+      "learning_rate": 8.645476772616138e-05,
+      "loss": 2.6935,
+      "step": 25552
+    },
+    {
+      "epoch": 0.5678444444444445,
+      "grad_norm": 1.2493290901184082,
+      "learning_rate": 8.645032229384308e-05,
+      "loss": 2.2871,
+      "step": 25553
+    },
+    {
+      "epoch": 0.5678666666666666,
+      "grad_norm": 1.019303798675537,
+      "learning_rate": 8.644587686152479e-05,
+      "loss": 1.1308,
+      "step": 25554
+    },
+    {
+      "epoch": 0.5678888888888889,
+      "grad_norm": 1.352089524269104,
+      "learning_rate": 8.64414314292065e-05,
+      "loss": 2.0598,
+      "step": 25555
+    },
+    {
+      "epoch": 0.5679111111111111,
+      "grad_norm": 1.6639633178710938,
+      "learning_rate": 8.643698599688819e-05,
+      "loss": 1.6875,
+      "step": 25556
+    },
+    {
+      "epoch": 0.5679333333333333,
+      "grad_norm": 1.5251015424728394,
+      "learning_rate": 8.643254056456991e-05,
+      "loss": 2.0902,
+      "step": 25557
+    },
+    {
+      "epoch": 0.5679555555555555,
+      "grad_norm": 1.4009424448013306,
+      "learning_rate": 8.642809513225162e-05,
+      "loss": 2.28,
+      "step": 25558
+    },
+    {
+      "epoch": 0.5679777777777778,
+      "grad_norm": 1.5966070890426636,
+      "learning_rate": 8.642364969993332e-05,
+      "loss": 2.2866,
+      "step": 25559
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 1.0532625913619995,
+      "learning_rate": 8.641920426761503e-05,
+      "loss": 1.2138,
+      "step": 25560
+    },
+    {
+      "epoch": 0.5680222222222222,
+      "grad_norm": 1.5685704946517944,
+      "learning_rate": 8.641475883529674e-05,
+      "loss": 2.1516,
+      "step": 25561
+    },
+    {
+      "epoch": 0.5680444444444445,
+      "grad_norm": 1.348705768585205,
+      "learning_rate": 8.641031340297845e-05,
+      "loss": 1.7455,
+      "step": 25562
+    },
+    {
+      "epoch": 0.5680666666666667,
+      "grad_norm": 1.7225929498672485,
+      "learning_rate": 8.640586797066015e-05,
+      "loss": 2.1124,
+      "step": 25563
+    },
+    {
+      "epoch": 0.5680888888888889,
+      "grad_norm": 1.5577679872512817,
+      "learning_rate": 8.640142253834186e-05,
+      "loss": 1.857,
+      "step": 25564
+    },
+    {
+      "epoch": 0.5681111111111111,
+      "grad_norm": 2.3156497478485107,
+      "learning_rate": 8.639697710602356e-05,
+      "loss": 1.6415,
+      "step": 25565
+    },
+    {
+      "epoch": 0.5681333333333334,
+      "grad_norm": 1.426537036895752,
+      "learning_rate": 8.639253167370527e-05,
+      "loss": 1.5825,
+      "step": 25566
+    },
+    {
+      "epoch": 0.5681555555555555,
+      "grad_norm": 1.494095802307129,
+      "learning_rate": 8.638808624138698e-05,
+      "loss": 1.8617,
+      "step": 25567
+    },
+    {
+      "epoch": 0.5681777777777778,
+      "grad_norm": 1.7394038438796997,
+      "learning_rate": 8.63836408090687e-05,
+      "loss": 2.2626,
+      "step": 25568
+    },
+    {
+      "epoch": 0.5682,
+      "grad_norm": 1.7255122661590576,
+      "learning_rate": 8.637919537675039e-05,
+      "loss": 2.1076,
+      "step": 25569
+    },
+    {
+      "epoch": 0.5682222222222222,
+      "grad_norm": 1.3343063592910767,
+      "learning_rate": 8.63747499444321e-05,
+      "loss": 1.4298,
+      "step": 25570
+    },
+    {
+      "epoch": 0.5682444444444444,
+      "grad_norm": 1.4318054914474487,
+      "learning_rate": 8.637030451211381e-05,
+      "loss": 1.8676,
+      "step": 25571
+    },
+    {
+      "epoch": 0.5682666666666667,
+      "grad_norm": 1.3568811416625977,
+      "learning_rate": 8.636585907979552e-05,
+      "loss": 2.1097,
+      "step": 25572
+    },
+    {
+      "epoch": 0.5682888888888888,
+      "grad_norm": 1.5402088165283203,
+      "learning_rate": 8.636141364747721e-05,
+      "loss": 1.8708,
+      "step": 25573
+    },
+    {
+      "epoch": 0.5683111111111111,
+      "grad_norm": 1.8865468502044678,
+      "learning_rate": 8.635696821515892e-05,
+      "loss": 2.1948,
+      "step": 25574
+    },
+    {
+      "epoch": 0.5683333333333334,
+      "grad_norm": 1.824686884880066,
+      "learning_rate": 8.635252278284063e-05,
+      "loss": 2.2085,
+      "step": 25575
+    },
+    {
+      "epoch": 0.5683555555555555,
+      "grad_norm": 1.5992943048477173,
+      "learning_rate": 8.634807735052234e-05,
+      "loss": 1.893,
+      "step": 25576
+    },
+    {
+      "epoch": 0.5683777777777778,
+      "grad_norm": 1.4310848712921143,
+      "learning_rate": 8.634363191820405e-05,
+      "loss": 1.7461,
+      "step": 25577
+    },
+    {
+      "epoch": 0.5684,
+      "grad_norm": 1.4332364797592163,
+      "learning_rate": 8.633918648588576e-05,
+      "loss": 1.5877,
+      "step": 25578
+    },
+    {
+      "epoch": 0.5684222222222223,
+      "grad_norm": 1.5231653451919556,
+      "learning_rate": 8.633474105356746e-05,
+      "loss": 1.9179,
+      "step": 25579
+    },
+    {
+      "epoch": 0.5684444444444444,
+      "grad_norm": 1.6611709594726562,
+      "learning_rate": 8.633029562124917e-05,
+      "loss": 1.3753,
+      "step": 25580
+    },
+    {
+      "epoch": 0.5684666666666667,
+      "grad_norm": 1.5533298254013062,
+      "learning_rate": 8.632585018893088e-05,
+      "loss": 1.7141,
+      "step": 25581
+    },
+    {
+      "epoch": 0.5684888888888889,
+      "grad_norm": 1.7611892223358154,
+      "learning_rate": 8.632140475661259e-05,
+      "loss": 1.9242,
+      "step": 25582
+    },
+    {
+      "epoch": 0.5685111111111111,
+      "grad_norm": 1.7563116550445557,
+      "learning_rate": 8.631695932429428e-05,
+      "loss": 2.3091,
+      "step": 25583
+    },
+    {
+      "epoch": 0.5685333333333333,
+      "grad_norm": 1.78745436668396,
+      "learning_rate": 8.6312513891976e-05,
+      "loss": 1.7655,
+      "step": 25584
+    },
+    {
+      "epoch": 0.5685555555555556,
+      "grad_norm": 1.9148340225219727,
+      "learning_rate": 8.63080684596577e-05,
+      "loss": 1.9374,
+      "step": 25585
+    },
+    {
+      "epoch": 0.5685777777777777,
+      "grad_norm": 1.1031118631362915,
+      "learning_rate": 8.630362302733941e-05,
+      "loss": 0.9773,
+      "step": 25586
+    },
+    {
+      "epoch": 0.5686,
+      "grad_norm": 1.5591325759887695,
+      "learning_rate": 8.629917759502112e-05,
+      "loss": 1.8767,
+      "step": 25587
+    },
+    {
+      "epoch": 0.5686222222222223,
+      "grad_norm": 1.592635989189148,
+      "learning_rate": 8.629473216270283e-05,
+      "loss": 1.7193,
+      "step": 25588
+    },
+    {
+      "epoch": 0.5686444444444444,
+      "grad_norm": 1.505963683128357,
+      "learning_rate": 8.629028673038453e-05,
+      "loss": 1.8382,
+      "step": 25589
+    },
+    {
+      "epoch": 0.5686666666666667,
+      "grad_norm": 1.7800480127334595,
+      "learning_rate": 8.628584129806624e-05,
+      "loss": 1.7844,
+      "step": 25590
+    },
+    {
+      "epoch": 0.5686888888888889,
+      "grad_norm": 1.703181505203247,
+      "learning_rate": 8.628139586574795e-05,
+      "loss": 1.8816,
+      "step": 25591
+    },
+    {
+      "epoch": 0.5687111111111111,
+      "grad_norm": 1.9718616008758545,
+      "learning_rate": 8.627695043342966e-05,
+      "loss": 2.0569,
+      "step": 25592
+    },
+    {
+      "epoch": 0.5687333333333333,
+      "grad_norm": 2.016561508178711,
+      "learning_rate": 8.627250500111135e-05,
+      "loss": 2.1293,
+      "step": 25593
+    },
+    {
+      "epoch": 0.5687555555555556,
+      "grad_norm": 1.6165771484375,
+      "learning_rate": 8.626805956879308e-05,
+      "loss": 1.7433,
+      "step": 25594
+    },
+    {
+      "epoch": 0.5687777777777778,
+      "grad_norm": 1.7712388038635254,
+      "learning_rate": 8.626361413647479e-05,
+      "loss": 1.6944,
+      "step": 25595
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 1.5624693632125854,
+      "learning_rate": 8.625916870415648e-05,
+      "loss": 0.7768,
+      "step": 25596
+    },
+    {
+      "epoch": 0.5688222222222222,
+      "grad_norm": 1.7532920837402344,
+      "learning_rate": 8.625472327183819e-05,
+      "loss": 1.9357,
+      "step": 25597
+    },
+    {
+      "epoch": 0.5688444444444445,
+      "grad_norm": 1.4923877716064453,
+      "learning_rate": 8.62502778395199e-05,
+      "loss": 1.3285,
+      "step": 25598
+    },
+    {
+      "epoch": 0.5688666666666666,
+      "grad_norm": 1.7593766450881958,
+      "learning_rate": 8.624583240720161e-05,
+      "loss": 1.4869,
+      "step": 25599
+    },
+    {
+      "epoch": 0.5688888888888889,
+      "grad_norm": 2.146467447280884,
+      "learning_rate": 8.62413869748833e-05,
+      "loss": 1.6363,
+      "step": 25600
+    },
+    {
+      "epoch": 0.5689111111111111,
+      "grad_norm": 1.5496854782104492,
+      "learning_rate": 8.623694154256502e-05,
+      "loss": 2.6966,
+      "step": 25601
+    },
+    {
+      "epoch": 0.5689333333333333,
+      "grad_norm": 0.8772847652435303,
+      "learning_rate": 8.623249611024673e-05,
+      "loss": 1.1829,
+      "step": 25602
+    },
+    {
+      "epoch": 0.5689555555555555,
+      "grad_norm": 1.0426076650619507,
+      "learning_rate": 8.622805067792844e-05,
+      "loss": 1.0857,
+      "step": 25603
+    },
+    {
+      "epoch": 0.5689777777777778,
+      "grad_norm": 1.534347414970398,
+      "learning_rate": 8.622360524561015e-05,
+      "loss": 2.4706,
+      "step": 25604
+    },
+    {
+      "epoch": 0.569,
+      "grad_norm": 1.4586025476455688,
+      "learning_rate": 8.621915981329185e-05,
+      "loss": 2.4975,
+      "step": 25605
+    },
+    {
+      "epoch": 0.5690222222222222,
+      "grad_norm": 1.736374020576477,
+      "learning_rate": 8.621471438097355e-05,
+      "loss": 2.3636,
+      "step": 25606
+    },
+    {
+      "epoch": 0.5690444444444445,
+      "grad_norm": 1.303686499595642,
+      "learning_rate": 8.621026894865526e-05,
+      "loss": 1.356,
+      "step": 25607
+    },
+    {
+      "epoch": 0.5690666666666667,
+      "grad_norm": 1.3341532945632935,
+      "learning_rate": 8.620582351633697e-05,
+      "loss": 1.9148,
+      "step": 25608
+    },
+    {
+      "epoch": 0.5690888888888889,
+      "grad_norm": 1.4841862916946411,
+      "learning_rate": 8.620137808401868e-05,
+      "loss": 1.8911,
+      "step": 25609
+    },
+    {
+      "epoch": 0.5691111111111111,
+      "grad_norm": 1.6789429187774658,
+      "learning_rate": 8.619693265170038e-05,
+      "loss": 2.2117,
+      "step": 25610
+    },
+    {
+      "epoch": 0.5691333333333334,
+      "grad_norm": 0.1538681536912918,
+      "learning_rate": 8.619248721938209e-05,
+      "loss": 0.0183,
+      "step": 25611
+    },
+    {
+      "epoch": 0.5691555555555555,
+      "grad_norm": 1.7826683521270752,
+      "learning_rate": 8.61880417870638e-05,
+      "loss": 2.0759,
+      "step": 25612
+    },
+    {
+      "epoch": 0.5691777777777778,
+      "grad_norm": 1.508593201637268,
+      "learning_rate": 8.61835963547455e-05,
+      "loss": 2.2247,
+      "step": 25613
+    },
+    {
+      "epoch": 0.5692,
+      "grad_norm": 1.4140818119049072,
+      "learning_rate": 8.617915092242721e-05,
+      "loss": 1.613,
+      "step": 25614
+    },
+    {
+      "epoch": 0.5692222222222222,
+      "grad_norm": 1.6381505727767944,
+      "learning_rate": 8.617470549010892e-05,
+      "loss": 2.0254,
+      "step": 25615
+    },
+    {
+      "epoch": 0.5692444444444444,
+      "grad_norm": 1.8345551490783691,
+      "learning_rate": 8.617026005779062e-05,
+      "loss": 2.5182,
+      "step": 25616
+    },
+    {
+      "epoch": 0.5692666666666667,
+      "grad_norm": 1.2683861255645752,
+      "learning_rate": 8.616581462547233e-05,
+      "loss": 0.8326,
+      "step": 25617
+    },
+    {
+      "epoch": 0.5692888888888888,
+      "grad_norm": 1.8822568655014038,
+      "learning_rate": 8.616136919315404e-05,
+      "loss": 1.6483,
+      "step": 25618
+    },
+    {
+      "epoch": 0.5693111111111111,
+      "grad_norm": 1.725543737411499,
+      "learning_rate": 8.615692376083575e-05,
+      "loss": 2.064,
+      "step": 25619
+    },
+    {
+      "epoch": 0.5693333333333334,
+      "grad_norm": 2.3417868614196777,
+      "learning_rate": 8.615247832851744e-05,
+      "loss": 2.7043,
+      "step": 25620
+    },
+    {
+      "epoch": 0.5693555555555555,
+      "grad_norm": 1.4956707954406738,
+      "learning_rate": 8.614803289619915e-05,
+      "loss": 1.6743,
+      "step": 25621
+    },
+    {
+      "epoch": 0.5693777777777778,
+      "grad_norm": 2.0377418994903564,
+      "learning_rate": 8.614358746388086e-05,
+      "loss": 1.9123,
+      "step": 25622
+    },
+    {
+      "epoch": 0.5694,
+      "grad_norm": 1.8327478170394897,
+      "learning_rate": 8.613914203156257e-05,
+      "loss": 1.9568,
+      "step": 25623
+    },
+    {
+      "epoch": 0.5694222222222223,
+      "grad_norm": 1.4546701908111572,
+      "learning_rate": 8.613469659924428e-05,
+      "loss": 1.9495,
+      "step": 25624
+    },
+    {
+      "epoch": 0.5694444444444444,
+      "grad_norm": 1.475493311882019,
+      "learning_rate": 8.613025116692599e-05,
+      "loss": 1.5293,
+      "step": 25625
+    },
+    {
+      "epoch": 0.5694666666666667,
+      "grad_norm": 1.9478663206100464,
+      "learning_rate": 8.612580573460769e-05,
+      "loss": 2.33,
+      "step": 25626
+    },
+    {
+      "epoch": 0.5694888888888889,
+      "grad_norm": 1.5463484525680542,
+      "learning_rate": 8.61213603022894e-05,
+      "loss": 1.9062,
+      "step": 25627
+    },
+    {
+      "epoch": 0.5695111111111111,
+      "grad_norm": 1.953561544418335,
+      "learning_rate": 8.611691486997111e-05,
+      "loss": 1.9784,
+      "step": 25628
+    },
+    {
+      "epoch": 0.5695333333333333,
+      "grad_norm": 1.514666199684143,
+      "learning_rate": 8.611246943765282e-05,
+      "loss": 1.4286,
+      "step": 25629
+    },
+    {
+      "epoch": 0.5695555555555556,
+      "grad_norm": 1.5759919881820679,
+      "learning_rate": 8.610802400533451e-05,
+      "loss": 1.9392,
+      "step": 25630
+    },
+    {
+      "epoch": 0.5695777777777777,
+      "grad_norm": 0.9849022030830383,
+      "learning_rate": 8.610357857301624e-05,
+      "loss": 0.8068,
+      "step": 25631
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 1.7830392122268677,
+      "learning_rate": 8.609913314069795e-05,
+      "loss": 1.6613,
+      "step": 25632
+    },
+    {
+      "epoch": 0.5696222222222223,
+      "grad_norm": 1.3821425437927246,
+      "learning_rate": 8.609468770837964e-05,
+      "loss": 0.7226,
+      "step": 25633
+    },
+    {
+      "epoch": 0.5696444444444444,
+      "grad_norm": 0.15262767672538757,
+      "learning_rate": 8.609024227606135e-05,
+      "loss": 0.0273,
+      "step": 25634
+    },
+    {
+      "epoch": 0.5696666666666667,
+      "grad_norm": 1.5773595571517944,
+      "learning_rate": 8.608579684374306e-05,
+      "loss": 1.8314,
+      "step": 25635
+    },
+    {
+      "epoch": 0.5696888888888889,
+      "grad_norm": 1.7459397315979004,
+      "learning_rate": 8.608135141142476e-05,
+      "loss": 1.9914,
+      "step": 25636
+    },
+    {
+      "epoch": 0.5697111111111111,
+      "grad_norm": 1.83330237865448,
+      "learning_rate": 8.607690597910647e-05,
+      "loss": 1.8253,
+      "step": 25637
+    },
+    {
+      "epoch": 0.5697333333333333,
+      "grad_norm": 1.7723551988601685,
+      "learning_rate": 8.607246054678818e-05,
+      "loss": 2.0525,
+      "step": 25638
+    },
+    {
+      "epoch": 0.5697555555555556,
+      "grad_norm": 2.0471293926239014,
+      "learning_rate": 8.606801511446989e-05,
+      "loss": 1.9065,
+      "step": 25639
+    },
+    {
+      "epoch": 0.5697777777777778,
+      "grad_norm": 1.8356201648712158,
+      "learning_rate": 8.60635696821516e-05,
+      "loss": 2.0246,
+      "step": 25640
+    },
+    {
+      "epoch": 0.5698,
+      "grad_norm": 1.7239627838134766,
+      "learning_rate": 8.60591242498333e-05,
+      "loss": 2.2547,
+      "step": 25641
+    },
+    {
+      "epoch": 0.5698222222222222,
+      "grad_norm": 1.7103173732757568,
+      "learning_rate": 8.605467881751502e-05,
+      "loss": 1.8929,
+      "step": 25642
+    },
+    {
+      "epoch": 0.5698444444444445,
+      "grad_norm": 1.6156359910964966,
+      "learning_rate": 8.605023338519671e-05,
+      "loss": 1.8456,
+      "step": 25643
+    },
+    {
+      "epoch": 0.5698666666666666,
+      "grad_norm": 1.5640565156936646,
+      "learning_rate": 8.604578795287842e-05,
+      "loss": 1.6381,
+      "step": 25644
+    },
+    {
+      "epoch": 0.5698888888888889,
+      "grad_norm": 1.4334174394607544,
+      "learning_rate": 8.604134252056013e-05,
+      "loss": 1.3684,
+      "step": 25645
+    },
+    {
+      "epoch": 0.5699111111111111,
+      "grad_norm": 1.7092499732971191,
+      "learning_rate": 8.603689708824183e-05,
+      "loss": 1.8991,
+      "step": 25646
+    },
+    {
+      "epoch": 0.5699333333333333,
+      "grad_norm": 2.41092586517334,
+      "learning_rate": 8.603245165592354e-05,
+      "loss": 1.7,
+      "step": 25647
+    },
+    {
+      "epoch": 0.5699555555555555,
+      "grad_norm": 1.7601580619812012,
+      "learning_rate": 8.602800622360525e-05,
+      "loss": 1.7763,
+      "step": 25648
+    },
+    {
+      "epoch": 0.5699777777777778,
+      "grad_norm": 1.8733407258987427,
+      "learning_rate": 8.602356079128696e-05,
+      "loss": 1.5541,
+      "step": 25649
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.374967575073242,
+      "learning_rate": 8.601911535896867e-05,
+      "loss": 1.8792,
+      "step": 25650
+    },
+    {
+      "epoch": 0.5700222222222222,
+      "grad_norm": 1.2829123735427856,
+      "learning_rate": 8.601466992665038e-05,
+      "loss": 2.0867,
+      "step": 25651
+    },
+    {
+      "epoch": 0.5700444444444445,
+      "grad_norm": 1.1918680667877197,
+      "learning_rate": 8.601022449433208e-05,
+      "loss": 1.3664,
+      "step": 25652
+    },
+    {
+      "epoch": 0.5700666666666667,
+      "grad_norm": 1.6968246698379517,
+      "learning_rate": 8.600577906201378e-05,
+      "loss": 2.6106,
+      "step": 25653
+    },
+    {
+      "epoch": 0.5700888888888889,
+      "grad_norm": 1.7334518432617188,
+      "learning_rate": 8.600133362969549e-05,
+      "loss": 2.3642,
+      "step": 25654
+    },
+    {
+      "epoch": 0.5701111111111111,
+      "grad_norm": 1.4948846101760864,
+      "learning_rate": 8.59968881973772e-05,
+      "loss": 1.7472,
+      "step": 25655
+    },
+    {
+      "epoch": 0.5701333333333334,
+      "grad_norm": 1.5138331651687622,
+      "learning_rate": 8.599244276505891e-05,
+      "loss": 1.9211,
+      "step": 25656
+    },
+    {
+      "epoch": 0.5701555555555555,
+      "grad_norm": 1.4922399520874023,
+      "learning_rate": 8.59879973327406e-05,
+      "loss": 1.9827,
+      "step": 25657
+    },
+    {
+      "epoch": 0.5701777777777778,
+      "grad_norm": 1.9236634969711304,
+      "learning_rate": 8.598355190042232e-05,
+      "loss": 1.9192,
+      "step": 25658
+    },
+    {
+      "epoch": 0.5702,
+      "grad_norm": 1.7816553115844727,
+      "learning_rate": 8.597910646810402e-05,
+      "loss": 2.12,
+      "step": 25659
+    },
+    {
+      "epoch": 0.5702222222222222,
+      "grad_norm": 1.5962737798690796,
+      "learning_rate": 8.597466103578573e-05,
+      "loss": 1.9601,
+      "step": 25660
+    },
+    {
+      "epoch": 0.5702444444444444,
+      "grad_norm": 1.6761876344680786,
+      "learning_rate": 8.597021560346744e-05,
+      "loss": 2.4774,
+      "step": 25661
+    },
+    {
+      "epoch": 0.5702666666666667,
+      "grad_norm": 1.5479689836502075,
+      "learning_rate": 8.596577017114915e-05,
+      "loss": 1.9717,
+      "step": 25662
+    },
+    {
+      "epoch": 0.5702888888888888,
+      "grad_norm": 1.867241621017456,
+      "learning_rate": 8.596132473883085e-05,
+      "loss": 1.8237,
+      "step": 25663
+    },
+    {
+      "epoch": 0.5703111111111111,
+      "grad_norm": 1.7697581052780151,
+      "learning_rate": 8.595687930651256e-05,
+      "loss": 2.1297,
+      "step": 25664
+    },
+    {
+      "epoch": 0.5703333333333334,
+      "grad_norm": 1.597586989402771,
+      "learning_rate": 8.595243387419427e-05,
+      "loss": 1.8777,
+      "step": 25665
+    },
+    {
+      "epoch": 0.5703555555555555,
+      "grad_norm": 1.3696242570877075,
+      "learning_rate": 8.594798844187598e-05,
+      "loss": 1.1399,
+      "step": 25666
+    },
+    {
+      "epoch": 0.5703777777777778,
+      "grad_norm": 1.4491081237792969,
+      "learning_rate": 8.594354300955767e-05,
+      "loss": 1.5301,
+      "step": 25667
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 1.6953177452087402,
+      "learning_rate": 8.59390975772394e-05,
+      "loss": 2.0117,
+      "step": 25668
+    },
+    {
+      "epoch": 0.5704222222222223,
+      "grad_norm": 1.524505376815796,
+      "learning_rate": 8.593465214492111e-05,
+      "loss": 1.8162,
+      "step": 25669
+    },
+    {
+      "epoch": 0.5704444444444444,
+      "grad_norm": 1.861527919769287,
+      "learning_rate": 8.59302067126028e-05,
+      "loss": 1.4832,
+      "step": 25670
+    },
+    {
+      "epoch": 0.5704666666666667,
+      "grad_norm": 1.8651635646820068,
+      "learning_rate": 8.592576128028451e-05,
+      "loss": 1.6583,
+      "step": 25671
+    },
+    {
+      "epoch": 0.5704888888888889,
+      "grad_norm": 2.0610780715942383,
+      "learning_rate": 8.592131584796622e-05,
+      "loss": 2.4729,
+      "step": 25672
+    },
+    {
+      "epoch": 0.5705111111111111,
+      "grad_norm": 1.5690243244171143,
+      "learning_rate": 8.591687041564792e-05,
+      "loss": 1.846,
+      "step": 25673
+    },
+    {
+      "epoch": 0.5705333333333333,
+      "grad_norm": 1.7951501607894897,
+      "learning_rate": 8.591242498332963e-05,
+      "loss": 2.2787,
+      "step": 25674
+    },
+    {
+      "epoch": 0.5705555555555556,
+      "grad_norm": 1.767663598060608,
+      "learning_rate": 8.590797955101134e-05,
+      "loss": 2.1076,
+      "step": 25675
+    },
+    {
+      "epoch": 0.5705777777777777,
+      "grad_norm": 1.3374791145324707,
+      "learning_rate": 8.590353411869305e-05,
+      "loss": 1.065,
+      "step": 25676
+    },
+    {
+      "epoch": 0.5706,
+      "grad_norm": 1.9195759296417236,
+      "learning_rate": 8.589908868637476e-05,
+      "loss": 2.0019,
+      "step": 25677
+    },
+    {
+      "epoch": 0.5706222222222223,
+      "grad_norm": 1.859966516494751,
+      "learning_rate": 8.589464325405647e-05,
+      "loss": 1.7332,
+      "step": 25678
+    },
+    {
+      "epoch": 0.5706444444444444,
+      "grad_norm": 1.7546013593673706,
+      "learning_rate": 8.589019782173818e-05,
+      "loss": 2.2428,
+      "step": 25679
+    },
+    {
+      "epoch": 0.5706666666666667,
+      "grad_norm": 1.5671427249908447,
+      "learning_rate": 8.588575238941987e-05,
+      "loss": 1.851,
+      "step": 25680
+    },
+    {
+      "epoch": 0.5706888888888889,
+      "grad_norm": 2.0177090167999268,
+      "learning_rate": 8.588130695710158e-05,
+      "loss": 2.1053,
+      "step": 25681
+    },
+    {
+      "epoch": 0.5707111111111111,
+      "grad_norm": 1.8132013082504272,
+      "learning_rate": 8.587686152478329e-05,
+      "loss": 2.0969,
+      "step": 25682
+    },
+    {
+      "epoch": 0.5707333333333333,
+      "grad_norm": 1.8457845449447632,
+      "learning_rate": 8.587241609246499e-05,
+      "loss": 1.8723,
+      "step": 25683
+    },
+    {
+      "epoch": 0.5707555555555556,
+      "grad_norm": 1.5866272449493408,
+      "learning_rate": 8.58679706601467e-05,
+      "loss": 1.4801,
+      "step": 25684
+    },
+    {
+      "epoch": 0.5707777777777778,
+      "grad_norm": 1.7102094888687134,
+      "learning_rate": 8.586352522782841e-05,
+      "loss": 1.9712,
+      "step": 25685
+    },
+    {
+      "epoch": 0.5708,
+      "grad_norm": 1.9478623867034912,
+      "learning_rate": 8.585907979551012e-05,
+      "loss": 2.2846,
+      "step": 25686
+    },
+    {
+      "epoch": 0.5708222222222222,
+      "grad_norm": 1.900810718536377,
+      "learning_rate": 8.585463436319183e-05,
+      "loss": 2.1442,
+      "step": 25687
+    },
+    {
+      "epoch": 0.5708444444444445,
+      "grad_norm": 1.5752862691879272,
+      "learning_rate": 8.585018893087354e-05,
+      "loss": 1.7731,
+      "step": 25688
+    },
+    {
+      "epoch": 0.5708666666666666,
+      "grad_norm": 1.7529854774475098,
+      "learning_rate": 8.584574349855525e-05,
+      "loss": 1.9615,
+      "step": 25689
+    },
+    {
+      "epoch": 0.5708888888888889,
+      "grad_norm": 1.776617407798767,
+      "learning_rate": 8.584129806623694e-05,
+      "loss": 1.8497,
+      "step": 25690
+    },
+    {
+      "epoch": 0.5709111111111111,
+      "grad_norm": 1.5052199363708496,
+      "learning_rate": 8.583685263391865e-05,
+      "loss": 1.7508,
+      "step": 25691
+    },
+    {
+      "epoch": 0.5709333333333333,
+      "grad_norm": 1.5465022325515747,
+      "learning_rate": 8.583240720160036e-05,
+      "loss": 1.6559,
+      "step": 25692
+    },
+    {
+      "epoch": 0.5709555555555555,
+      "grad_norm": 2.8852856159210205,
+      "learning_rate": 8.582796176928206e-05,
+      "loss": 1.7696,
+      "step": 25693
+    },
+    {
+      "epoch": 0.5709777777777778,
+      "grad_norm": 1.6870107650756836,
+      "learning_rate": 8.582351633696377e-05,
+      "loss": 1.3945,
+      "step": 25694
+    },
+    {
+      "epoch": 0.571,
+      "grad_norm": 1.3791446685791016,
+      "learning_rate": 8.581907090464548e-05,
+      "loss": 1.3369,
+      "step": 25695
+    },
+    {
+      "epoch": 0.5710222222222222,
+      "grad_norm": 1.6492143869400024,
+      "learning_rate": 8.581462547232719e-05,
+      "loss": 1.6589,
+      "step": 25696
+    },
+    {
+      "epoch": 0.5710444444444445,
+      "grad_norm": 1.7930784225463867,
+      "learning_rate": 8.58101800400089e-05,
+      "loss": 1.7328,
+      "step": 25697
+    },
+    {
+      "epoch": 0.5710666666666666,
+      "grad_norm": 1.76011061668396,
+      "learning_rate": 8.58057346076906e-05,
+      "loss": 1.3635,
+      "step": 25698
+    },
+    {
+      "epoch": 0.5710888888888889,
+      "grad_norm": 1.642624020576477,
+      "learning_rate": 8.580128917537231e-05,
+      "loss": 1.6686,
+      "step": 25699
+    },
+    {
+      "epoch": 0.5711111111111111,
+      "grad_norm": 1.4365448951721191,
+      "learning_rate": 8.579684374305401e-05,
+      "loss": 0.064,
+      "step": 25700
+    },
+    {
+      "epoch": 0.5711333333333334,
+      "grad_norm": 1.3209048509597778,
+      "learning_rate": 8.579239831073572e-05,
+      "loss": 2.0327,
+      "step": 25701
+    },
+    {
+      "epoch": 0.5711555555555555,
+      "grad_norm": 1.3216556310653687,
+      "learning_rate": 8.578795287841743e-05,
+      "loss": 2.2464,
+      "step": 25702
+    },
+    {
+      "epoch": 0.5711777777777778,
+      "grad_norm": 1.5542192459106445,
+      "learning_rate": 8.578350744609913e-05,
+      "loss": 2.3578,
+      "step": 25703
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 1.7194514274597168,
+      "learning_rate": 8.577906201378084e-05,
+      "loss": 2.7349,
+      "step": 25704
+    },
+    {
+      "epoch": 0.5712222222222222,
+      "grad_norm": 1.194679856300354,
+      "learning_rate": 8.577461658146256e-05,
+      "loss": 1.3133,
+      "step": 25705
+    },
+    {
+      "epoch": 0.5712444444444444,
+      "grad_norm": 1.6330630779266357,
+      "learning_rate": 8.577017114914427e-05,
+      "loss": 2.2848,
+      "step": 25706
+    },
+    {
+      "epoch": 0.5712666666666667,
+      "grad_norm": 2.1747875213623047,
+      "learning_rate": 8.576572571682596e-05,
+      "loss": 2.2944,
+      "step": 25707
+    },
+    {
+      "epoch": 0.5712888888888888,
+      "grad_norm": 2.29811429977417,
+      "learning_rate": 8.576128028450767e-05,
+      "loss": 2.1093,
+      "step": 25708
+    },
+    {
+      "epoch": 0.5713111111111111,
+      "grad_norm": 1.3702226877212524,
+      "learning_rate": 8.575683485218938e-05,
+      "loss": 1.4981,
+      "step": 25709
+    },
+    {
+      "epoch": 0.5713333333333334,
+      "grad_norm": 1.570976734161377,
+      "learning_rate": 8.575238941987108e-05,
+      "loss": 1.6642,
+      "step": 25710
+    },
+    {
+      "epoch": 0.5713555555555555,
+      "grad_norm": 1.8886852264404297,
+      "learning_rate": 8.574794398755279e-05,
+      "loss": 1.6982,
+      "step": 25711
+    },
+    {
+      "epoch": 0.5713777777777778,
+      "grad_norm": 1.566110610961914,
+      "learning_rate": 8.57434985552345e-05,
+      "loss": 1.504,
+      "step": 25712
+    },
+    {
+      "epoch": 0.5714,
+      "grad_norm": 1.0063490867614746,
+      "learning_rate": 8.573905312291621e-05,
+      "loss": 0.7414,
+      "step": 25713
+    },
+    {
+      "epoch": 0.5714222222222223,
+      "grad_norm": 2.0926096439361572,
+      "learning_rate": 8.573460769059792e-05,
+      "loss": 1.762,
+      "step": 25714
+    },
+    {
+      "epoch": 0.5714444444444444,
+      "grad_norm": 1.5999740362167358,
+      "learning_rate": 8.573016225827963e-05,
+      "loss": 2.0492,
+      "step": 25715
+    },
+    {
+      "epoch": 0.5714666666666667,
+      "grad_norm": 2.4232420921325684,
+      "learning_rate": 8.572571682596134e-05,
+      "loss": 1.1813,
+      "step": 25716
+    },
+    {
+      "epoch": 0.5714888888888889,
+      "grad_norm": 1.5547337532043457,
+      "learning_rate": 8.572127139364303e-05,
+      "loss": 1.7063,
+      "step": 25717
+    },
+    {
+      "epoch": 0.5715111111111111,
+      "grad_norm": 1.738980770111084,
+      "learning_rate": 8.571682596132474e-05,
+      "loss": 2.3619,
+      "step": 25718
+    },
+    {
+      "epoch": 0.5715333333333333,
+      "grad_norm": 1.4263365268707275,
+      "learning_rate": 8.571238052900645e-05,
+      "loss": 1.4464,
+      "step": 25719
+    },
+    {
+      "epoch": 0.5715555555555556,
+      "grad_norm": 1.5924991369247437,
+      "learning_rate": 8.570793509668815e-05,
+      "loss": 1.8442,
+      "step": 25720
+    },
+    {
+      "epoch": 0.5715777777777777,
+      "grad_norm": 1.668595314025879,
+      "learning_rate": 8.570348966436986e-05,
+      "loss": 1.5966,
+      "step": 25721
+    },
+    {
+      "epoch": 0.5716,
+      "grad_norm": 1.7621865272521973,
+      "learning_rate": 8.569904423205157e-05,
+      "loss": 1.8259,
+      "step": 25722
+    },
+    {
+      "epoch": 0.5716222222222223,
+      "grad_norm": 1.6255216598510742,
+      "learning_rate": 8.569459879973328e-05,
+      "loss": 2.0912,
+      "step": 25723
+    },
+    {
+      "epoch": 0.5716444444444444,
+      "grad_norm": 1.5257017612457275,
+      "learning_rate": 8.569015336741499e-05,
+      "loss": 1.9858,
+      "step": 25724
+    },
+    {
+      "epoch": 0.5716666666666667,
+      "grad_norm": 1.5033650398254395,
+      "learning_rate": 8.56857079350967e-05,
+      "loss": 1.5099,
+      "step": 25725
+    },
+    {
+      "epoch": 0.5716888888888889,
+      "grad_norm": 1.9311522245407104,
+      "learning_rate": 8.568126250277841e-05,
+      "loss": 2.0615,
+      "step": 25726
+    },
+    {
+      "epoch": 0.5717111111111111,
+      "grad_norm": 1.7489691972732544,
+      "learning_rate": 8.56768170704601e-05,
+      "loss": 2.4009,
+      "step": 25727
+    },
+    {
+      "epoch": 0.5717333333333333,
+      "grad_norm": 1.7931649684906006,
+      "learning_rate": 8.567237163814181e-05,
+      "loss": 2.1929,
+      "step": 25728
+    },
+    {
+      "epoch": 0.5717555555555556,
+      "grad_norm": 1.84356689453125,
+      "learning_rate": 8.566792620582352e-05,
+      "loss": 1.9432,
+      "step": 25729
+    },
+    {
+      "epoch": 0.5717777777777778,
+      "grad_norm": 1.5121773481369019,
+      "learning_rate": 8.566348077350522e-05,
+      "loss": 1.9769,
+      "step": 25730
+    },
+    {
+      "epoch": 0.5718,
+      "grad_norm": 1.5745084285736084,
+      "learning_rate": 8.565903534118693e-05,
+      "loss": 1.8561,
+      "step": 25731
+    },
+    {
+      "epoch": 0.5718222222222222,
+      "grad_norm": 1.262855887413025,
+      "learning_rate": 8.565458990886864e-05,
+      "loss": 1.272,
+      "step": 25732
+    },
+    {
+      "epoch": 0.5718444444444445,
+      "grad_norm": 1.2928316593170166,
+      "learning_rate": 8.565014447655035e-05,
+      "loss": 1.0977,
+      "step": 25733
+    },
+    {
+      "epoch": 0.5718666666666666,
+      "grad_norm": 0.956841766834259,
+      "learning_rate": 8.564569904423206e-05,
+      "loss": 0.7857,
+      "step": 25734
+    },
+    {
+      "epoch": 0.5718888888888889,
+      "grad_norm": 0.14648161828517914,
+      "learning_rate": 8.564125361191377e-05,
+      "loss": 0.0258,
+      "step": 25735
+    },
+    {
+      "epoch": 0.5719111111111111,
+      "grad_norm": 0.2709256112575531,
+      "learning_rate": 8.563680817959548e-05,
+      "loss": 0.0263,
+      "step": 25736
+    },
+    {
+      "epoch": 0.5719333333333333,
+      "grad_norm": 1.778686761856079,
+      "learning_rate": 8.563236274727717e-05,
+      "loss": 2.0404,
+      "step": 25737
+    },
+    {
+      "epoch": 0.5719555555555556,
+      "grad_norm": 1.7901958227157593,
+      "learning_rate": 8.562791731495888e-05,
+      "loss": 1.5201,
+      "step": 25738
+    },
+    {
+      "epoch": 0.5719777777777778,
+      "grad_norm": 1.8837966918945312,
+      "learning_rate": 8.562347188264059e-05,
+      "loss": 2.0632,
+      "step": 25739
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 1.2224178314208984,
+      "learning_rate": 8.561902645032229e-05,
+      "loss": 0.9937,
+      "step": 25740
+    },
+    {
+      "epoch": 0.5720222222222222,
+      "grad_norm": 2.0332577228546143,
+      "learning_rate": 8.5614581018004e-05,
+      "loss": 2.1465,
+      "step": 25741
+    },
+    {
+      "epoch": 0.5720444444444445,
+      "grad_norm": 1.7775815725326538,
+      "learning_rate": 8.561013558568572e-05,
+      "loss": 1.8859,
+      "step": 25742
+    },
+    {
+      "epoch": 0.5720666666666666,
+      "grad_norm": 1.7891740798950195,
+      "learning_rate": 8.560569015336743e-05,
+      "loss": 1.9876,
+      "step": 25743
+    },
+    {
+      "epoch": 0.5720888888888889,
+      "grad_norm": 1.9831715822219849,
+      "learning_rate": 8.560124472104913e-05,
+      "loss": 1.9548,
+      "step": 25744
+    },
+    {
+      "epoch": 0.5721111111111111,
+      "grad_norm": 2.164729356765747,
+      "learning_rate": 8.559679928873084e-05,
+      "loss": 2.2877,
+      "step": 25745
+    },
+    {
+      "epoch": 0.5721333333333334,
+      "grad_norm": 1.6351230144500732,
+      "learning_rate": 8.559235385641254e-05,
+      "loss": 1.478,
+      "step": 25746
+    },
+    {
+      "epoch": 0.5721555555555555,
+      "grad_norm": 2.017198085784912,
+      "learning_rate": 8.558790842409424e-05,
+      "loss": 1.6519,
+      "step": 25747
+    },
+    {
+      "epoch": 0.5721777777777778,
+      "grad_norm": 1.556087851524353,
+      "learning_rate": 8.558346299177595e-05,
+      "loss": 1.5572,
+      "step": 25748
+    },
+    {
+      "epoch": 0.5722,
+      "grad_norm": 1.82135009765625,
+      "learning_rate": 8.557901755945766e-05,
+      "loss": 1.771,
+      "step": 25749
+    },
+    {
+      "epoch": 0.5722222222222222,
+      "grad_norm": 2.233006238937378,
+      "learning_rate": 8.557457212713936e-05,
+      "loss": 1.7638,
+      "step": 25750
+    },
+    {
+      "epoch": 0.5722444444444444,
+      "grad_norm": 1.3495924472808838,
+      "learning_rate": 8.557012669482108e-05,
+      "loss": 2.3135,
+      "step": 25751
+    },
+    {
+      "epoch": 0.5722666666666667,
+      "grad_norm": 1.718217134475708,
+      "learning_rate": 8.556568126250279e-05,
+      "loss": 2.3093,
+      "step": 25752
+    },
+    {
+      "epoch": 0.5722888888888888,
+      "grad_norm": 1.4839550256729126,
+      "learning_rate": 8.55612358301845e-05,
+      "loss": 2.6725,
+      "step": 25753
+    },
+    {
+      "epoch": 0.5723111111111111,
+      "grad_norm": 1.6164158582687378,
+      "learning_rate": 8.55567903978662e-05,
+      "loss": 2.3814,
+      "step": 25754
+    },
+    {
+      "epoch": 0.5723333333333334,
+      "grad_norm": 1.6461539268493652,
+      "learning_rate": 8.55523449655479e-05,
+      "loss": 2.1208,
+      "step": 25755
+    },
+    {
+      "epoch": 0.5723555555555555,
+      "grad_norm": 1.5425986051559448,
+      "learning_rate": 8.554789953322961e-05,
+      "loss": 2.1488,
+      "step": 25756
+    },
+    {
+      "epoch": 0.5723777777777778,
+      "grad_norm": 1.6818119287490845,
+      "learning_rate": 8.554345410091131e-05,
+      "loss": 2.2954,
+      "step": 25757
+    },
+    {
+      "epoch": 0.5724,
+      "grad_norm": 1.653141975402832,
+      "learning_rate": 8.553900866859302e-05,
+      "loss": 1.5584,
+      "step": 25758
+    },
+    {
+      "epoch": 0.5724222222222223,
+      "grad_norm": 1.3412193059921265,
+      "learning_rate": 8.553456323627473e-05,
+      "loss": 2.0158,
+      "step": 25759
+    },
+    {
+      "epoch": 0.5724444444444444,
+      "grad_norm": 1.7241566181182861,
+      "learning_rate": 8.553011780395644e-05,
+      "loss": 2.6476,
+      "step": 25760
+    },
+    {
+      "epoch": 0.5724666666666667,
+      "grad_norm": 1.53416907787323,
+      "learning_rate": 8.552567237163815e-05,
+      "loss": 2.087,
+      "step": 25761
+    },
+    {
+      "epoch": 0.5724888888888889,
+      "grad_norm": 1.8664326667785645,
+      "learning_rate": 8.552122693931986e-05,
+      "loss": 1.9608,
+      "step": 25762
+    },
+    {
+      "epoch": 0.5725111111111111,
+      "grad_norm": 1.7230284214019775,
+      "learning_rate": 8.551678150700157e-05,
+      "loss": 2.4159,
+      "step": 25763
+    },
+    {
+      "epoch": 0.5725333333333333,
+      "grad_norm": 1.5559327602386475,
+      "learning_rate": 8.551233607468326e-05,
+      "loss": 2.2237,
+      "step": 25764
+    },
+    {
+      "epoch": 0.5725555555555556,
+      "grad_norm": 1.481768012046814,
+      "learning_rate": 8.550789064236497e-05,
+      "loss": 2.1765,
+      "step": 25765
+    },
+    {
+      "epoch": 0.5725777777777777,
+      "grad_norm": 1.5719056129455566,
+      "learning_rate": 8.550344521004668e-05,
+      "loss": 1.9083,
+      "step": 25766
+    },
+    {
+      "epoch": 0.5726,
+      "grad_norm": 1.3396432399749756,
+      "learning_rate": 8.549899977772838e-05,
+      "loss": 1.7808,
+      "step": 25767
+    },
+    {
+      "epoch": 0.5726222222222223,
+      "grad_norm": 1.4139114618301392,
+      "learning_rate": 8.549455434541009e-05,
+      "loss": 1.8634,
+      "step": 25768
+    },
+    {
+      "epoch": 0.5726444444444444,
+      "grad_norm": 1.6188405752182007,
+      "learning_rate": 8.54901089130918e-05,
+      "loss": 2.0596,
+      "step": 25769
+    },
+    {
+      "epoch": 0.5726666666666667,
+      "grad_norm": 1.9198650121688843,
+      "learning_rate": 8.548566348077351e-05,
+      "loss": 1.7277,
+      "step": 25770
+    },
+    {
+      "epoch": 0.5726888888888889,
+      "grad_norm": 1.4434568881988525,
+      "learning_rate": 8.548121804845522e-05,
+      "loss": 1.742,
+      "step": 25771
+    },
+    {
+      "epoch": 0.5727111111111111,
+      "grad_norm": 1.572055697441101,
+      "learning_rate": 8.547677261613693e-05,
+      "loss": 1.8451,
+      "step": 25772
+    },
+    {
+      "epoch": 0.5727333333333333,
+      "grad_norm": 1.3459651470184326,
+      "learning_rate": 8.547232718381864e-05,
+      "loss": 1.3596,
+      "step": 25773
+    },
+    {
+      "epoch": 0.5727555555555556,
+      "grad_norm": 1.3569782972335815,
+      "learning_rate": 8.546788175150033e-05,
+      "loss": 1.7197,
+      "step": 25774
+    },
+    {
+      "epoch": 0.5727777777777778,
+      "grad_norm": 1.5443956851959229,
+      "learning_rate": 8.546343631918204e-05,
+      "loss": 1.6795,
+      "step": 25775
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 1.878814935684204,
+      "learning_rate": 8.545899088686375e-05,
+      "loss": 1.721,
+      "step": 25776
+    },
+    {
+      "epoch": 0.5728222222222222,
+      "grad_norm": 1.5908725261688232,
+      "learning_rate": 8.545454545454545e-05,
+      "loss": 2.0399,
+      "step": 25777
+    },
+    {
+      "epoch": 0.5728444444444445,
+      "grad_norm": 1.5259735584259033,
+      "learning_rate": 8.545010002222716e-05,
+      "loss": 1.6809,
+      "step": 25778
+    },
+    {
+      "epoch": 0.5728666666666666,
+      "grad_norm": 1.7970151901245117,
+      "learning_rate": 8.544565458990888e-05,
+      "loss": 2.0895,
+      "step": 25779
+    },
+    {
+      "epoch": 0.5728888888888889,
+      "grad_norm": 1.7510429620742798,
+      "learning_rate": 8.544120915759059e-05,
+      "loss": 2.0142,
+      "step": 25780
+    },
+    {
+      "epoch": 0.5729111111111111,
+      "grad_norm": 1.514723539352417,
+      "learning_rate": 8.543676372527229e-05,
+      "loss": 1.5025,
+      "step": 25781
+    },
+    {
+      "epoch": 0.5729333333333333,
+      "grad_norm": 1.241104245185852,
+      "learning_rate": 8.5432318292954e-05,
+      "loss": 1.2434,
+      "step": 25782
+    },
+    {
+      "epoch": 0.5729555555555556,
+      "grad_norm": 1.9197041988372803,
+      "learning_rate": 8.54278728606357e-05,
+      "loss": 1.7614,
+      "step": 25783
+    },
+    {
+      "epoch": 0.5729777777777778,
+      "grad_norm": 1.567199468612671,
+      "learning_rate": 8.54234274283174e-05,
+      "loss": 1.6021,
+      "step": 25784
+    },
+    {
+      "epoch": 0.573,
+      "grad_norm": 1.7817846536636353,
+      "learning_rate": 8.541898199599911e-05,
+      "loss": 1.8707,
+      "step": 25785
+    },
+    {
+      "epoch": 0.5730222222222222,
+      "grad_norm": 1.7348031997680664,
+      "learning_rate": 8.541453656368082e-05,
+      "loss": 1.771,
+      "step": 25786
+    },
+    {
+      "epoch": 0.5730444444444445,
+      "grad_norm": 2.0711631774902344,
+      "learning_rate": 8.541009113136252e-05,
+      "loss": 2.2206,
+      "step": 25787
+    },
+    {
+      "epoch": 0.5730666666666666,
+      "grad_norm": 1.8702232837677002,
+      "learning_rate": 8.540564569904424e-05,
+      "loss": 2.1554,
+      "step": 25788
+    },
+    {
+      "epoch": 0.5730888888888889,
+      "grad_norm": 2.070375680923462,
+      "learning_rate": 8.540120026672595e-05,
+      "loss": 1.9594,
+      "step": 25789
+    },
+    {
+      "epoch": 0.5731111111111111,
+      "grad_norm": 1.8700990676879883,
+      "learning_rate": 8.539675483440766e-05,
+      "loss": 1.8745,
+      "step": 25790
+    },
+    {
+      "epoch": 0.5731333333333334,
+      "grad_norm": 1.6505186557769775,
+      "learning_rate": 8.539230940208936e-05,
+      "loss": 1.6517,
+      "step": 25791
+    },
+    {
+      "epoch": 0.5731555555555555,
+      "grad_norm": 1.9102988243103027,
+      "learning_rate": 8.538786396977107e-05,
+      "loss": 2.0135,
+      "step": 25792
+    },
+    {
+      "epoch": 0.5731777777777778,
+      "grad_norm": 1.8339016437530518,
+      "learning_rate": 8.538341853745278e-05,
+      "loss": 1.7311,
+      "step": 25793
+    },
+    {
+      "epoch": 0.5732,
+      "grad_norm": 1.8846791982650757,
+      "learning_rate": 8.537897310513447e-05,
+      "loss": 2.0282,
+      "step": 25794
+    },
+    {
+      "epoch": 0.5732222222222222,
+      "grad_norm": 1.6500928401947021,
+      "learning_rate": 8.537452767281618e-05,
+      "loss": 1.734,
+      "step": 25795
+    },
+    {
+      "epoch": 0.5732444444444444,
+      "grad_norm": 1.7177919149398804,
+      "learning_rate": 8.537008224049789e-05,
+      "loss": 1.4664,
+      "step": 25796
+    },
+    {
+      "epoch": 0.5732666666666667,
+      "grad_norm": 1.9090465307235718,
+      "learning_rate": 8.53656368081796e-05,
+      "loss": 1.9601,
+      "step": 25797
+    },
+    {
+      "epoch": 0.5732888888888888,
+      "grad_norm": 1.3065292835235596,
+      "learning_rate": 8.536119137586131e-05,
+      "loss": 0.8583,
+      "step": 25798
+    },
+    {
+      "epoch": 0.5733111111111111,
+      "grad_norm": 1.5860507488250732,
+      "learning_rate": 8.535674594354302e-05,
+      "loss": 1.3665,
+      "step": 25799
+    },
+    {
+      "epoch": 0.5733333333333334,
+      "grad_norm": 1.4950746297836304,
+      "learning_rate": 8.535230051122473e-05,
+      "loss": 0.974,
+      "step": 25800
+    },
+    {
+      "epoch": 0.5733555555555555,
+      "grad_norm": 1.5080500841140747,
+      "learning_rate": 8.534785507890642e-05,
+      "loss": 2.2007,
+      "step": 25801
+    },
+    {
+      "epoch": 0.5733777777777778,
+      "grad_norm": 1.8065223693847656,
+      "learning_rate": 8.534340964658813e-05,
+      "loss": 2.7459,
+      "step": 25802
+    },
+    {
+      "epoch": 0.5734,
+      "grad_norm": 1.1576310396194458,
+      "learning_rate": 8.533896421426984e-05,
+      "loss": 1.1024,
+      "step": 25803
+    },
+    {
+      "epoch": 0.5734222222222223,
+      "grad_norm": 1.37685227394104,
+      "learning_rate": 8.533451878195154e-05,
+      "loss": 2.098,
+      "step": 25804
+    },
+    {
+      "epoch": 0.5734444444444444,
+      "grad_norm": 1.4093037843704224,
+      "learning_rate": 8.533007334963325e-05,
+      "loss": 2.2798,
+      "step": 25805
+    },
+    {
+      "epoch": 0.5734666666666667,
+      "grad_norm": 1.5947177410125732,
+      "learning_rate": 8.532562791731496e-05,
+      "loss": 2.0701,
+      "step": 25806
+    },
+    {
+      "epoch": 0.5734888888888889,
+      "grad_norm": 2.1120340824127197,
+      "learning_rate": 8.532118248499667e-05,
+      "loss": 1.6964,
+      "step": 25807
+    },
+    {
+      "epoch": 0.5735111111111111,
+      "grad_norm": 1.8550314903259277,
+      "learning_rate": 8.531673705267838e-05,
+      "loss": 2.5111,
+      "step": 25808
+    },
+    {
+      "epoch": 0.5735333333333333,
+      "grad_norm": 2.037273645401001,
+      "learning_rate": 8.531229162036009e-05,
+      "loss": 1.858,
+      "step": 25809
+    },
+    {
+      "epoch": 0.5735555555555556,
+      "grad_norm": 1.458878517150879,
+      "learning_rate": 8.53078461880418e-05,
+      "loss": 1.8525,
+      "step": 25810
+    },
+    {
+      "epoch": 0.5735777777777777,
+      "grad_norm": 0.2872360348701477,
+      "learning_rate": 8.53034007557235e-05,
+      "loss": 0.0209,
+      "step": 25811
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 1.529571294784546,
+      "learning_rate": 8.52989553234052e-05,
+      "loss": 1.8411,
+      "step": 25812
+    },
+    {
+      "epoch": 0.5736222222222223,
+      "grad_norm": 1.5923453569412231,
+      "learning_rate": 8.529450989108691e-05,
+      "loss": 1.8027,
+      "step": 25813
+    },
+    {
+      "epoch": 0.5736444444444444,
+      "grad_norm": 1.7380670309066772,
+      "learning_rate": 8.529006445876861e-05,
+      "loss": 2.4903,
+      "step": 25814
+    },
+    {
+      "epoch": 0.5736666666666667,
+      "grad_norm": 2.0399043560028076,
+      "learning_rate": 8.528561902645032e-05,
+      "loss": 2.3516,
+      "step": 25815
+    },
+    {
+      "epoch": 0.5736888888888889,
+      "grad_norm": 1.5405837297439575,
+      "learning_rate": 8.528117359413204e-05,
+      "loss": 1.7951,
+      "step": 25816
+    },
+    {
+      "epoch": 0.5737111111111111,
+      "grad_norm": 1.6100730895996094,
+      "learning_rate": 8.527672816181375e-05,
+      "loss": 1.955,
+      "step": 25817
+    },
+    {
+      "epoch": 0.5737333333333333,
+      "grad_norm": 1.479048252105713,
+      "learning_rate": 8.527228272949545e-05,
+      "loss": 1.6049,
+      "step": 25818
+    },
+    {
+      "epoch": 0.5737555555555556,
+      "grad_norm": 1.6415687799453735,
+      "learning_rate": 8.526783729717716e-05,
+      "loss": 2.1413,
+      "step": 25819
+    },
+    {
+      "epoch": 0.5737777777777778,
+      "grad_norm": 1.633242130279541,
+      "learning_rate": 8.526339186485887e-05,
+      "loss": 1.9074,
+      "step": 25820
+    },
+    {
+      "epoch": 0.5738,
+      "grad_norm": 1.722865343093872,
+      "learning_rate": 8.525894643254056e-05,
+      "loss": 2.2105,
+      "step": 25821
+    },
+    {
+      "epoch": 0.5738222222222222,
+      "grad_norm": 1.6807760000228882,
+      "learning_rate": 8.525450100022227e-05,
+      "loss": 2.1285,
+      "step": 25822
+    },
+    {
+      "epoch": 0.5738444444444445,
+      "grad_norm": 1.3529731035232544,
+      "learning_rate": 8.525005556790398e-05,
+      "loss": 1.7808,
+      "step": 25823
+    },
+    {
+      "epoch": 0.5738666666666666,
+      "grad_norm": 1.7612981796264648,
+      "learning_rate": 8.524561013558568e-05,
+      "loss": 1.8677,
+      "step": 25824
+    },
+    {
+      "epoch": 0.5738888888888889,
+      "grad_norm": 1.543408751487732,
+      "learning_rate": 8.52411647032674e-05,
+      "loss": 1.9944,
+      "step": 25825
+    },
+    {
+      "epoch": 0.5739111111111111,
+      "grad_norm": 1.546540379524231,
+      "learning_rate": 8.523671927094911e-05,
+      "loss": 1.942,
+      "step": 25826
+    },
+    {
+      "epoch": 0.5739333333333333,
+      "grad_norm": 1.7032814025878906,
+      "learning_rate": 8.523227383863082e-05,
+      "loss": 1.7724,
+      "step": 25827
+    },
+    {
+      "epoch": 0.5739555555555556,
+      "grad_norm": 1.6998419761657715,
+      "learning_rate": 8.522782840631252e-05,
+      "loss": 2.0003,
+      "step": 25828
+    },
+    {
+      "epoch": 0.5739777777777778,
+      "grad_norm": 1.5656496286392212,
+      "learning_rate": 8.522338297399423e-05,
+      "loss": 1.8617,
+      "step": 25829
+    },
+    {
+      "epoch": 0.574,
+      "grad_norm": 1.3194990158081055,
+      "learning_rate": 8.521893754167594e-05,
+      "loss": 1.3988,
+      "step": 25830
+    },
+    {
+      "epoch": 0.5740222222222222,
+      "grad_norm": 1.5632909536361694,
+      "learning_rate": 8.521449210935763e-05,
+      "loss": 1.8376,
+      "step": 25831
+    },
+    {
+      "epoch": 0.5740444444444445,
+      "grad_norm": 1.4915846586227417,
+      "learning_rate": 8.521004667703934e-05,
+      "loss": 1.9154,
+      "step": 25832
+    },
+    {
+      "epoch": 0.5740666666666666,
+      "grad_norm": 1.458807349205017,
+      "learning_rate": 8.520560124472105e-05,
+      "loss": 1.8826,
+      "step": 25833
+    },
+    {
+      "epoch": 0.5740888888888889,
+      "grad_norm": 1.8433349132537842,
+      "learning_rate": 8.520115581240276e-05,
+      "loss": 2.4107,
+      "step": 25834
+    },
+    {
+      "epoch": 0.5741111111111111,
+      "grad_norm": 1.5098316669464111,
+      "learning_rate": 8.519671038008447e-05,
+      "loss": 1.8326,
+      "step": 25835
+    },
+    {
+      "epoch": 0.5741333333333334,
+      "grad_norm": 1.8357150554656982,
+      "learning_rate": 8.519226494776618e-05,
+      "loss": 1.5214,
+      "step": 25836
+    },
+    {
+      "epoch": 0.5741555555555555,
+      "grad_norm": 1.6379412412643433,
+      "learning_rate": 8.518781951544789e-05,
+      "loss": 1.651,
+      "step": 25837
+    },
+    {
+      "epoch": 0.5741777777777778,
+      "grad_norm": 1.544201374053955,
+      "learning_rate": 8.518337408312959e-05,
+      "loss": 1.8596,
+      "step": 25838
+    },
+    {
+      "epoch": 0.5742,
+      "grad_norm": 1.7644819021224976,
+      "learning_rate": 8.51789286508113e-05,
+      "loss": 2.0177,
+      "step": 25839
+    },
+    {
+      "epoch": 0.5742222222222222,
+      "grad_norm": 1.0269098281860352,
+      "learning_rate": 8.5174483218493e-05,
+      "loss": 0.6541,
+      "step": 25840
+    },
+    {
+      "epoch": 0.5742444444444444,
+      "grad_norm": 2.171380043029785,
+      "learning_rate": 8.51700377861747e-05,
+      "loss": 2.2374,
+      "step": 25841
+    },
+    {
+      "epoch": 0.5742666666666667,
+      "grad_norm": 1.542111873626709,
+      "learning_rate": 8.516559235385641e-05,
+      "loss": 1.855,
+      "step": 25842
+    },
+    {
+      "epoch": 0.5742888888888888,
+      "grad_norm": 2.249403238296509,
+      "learning_rate": 8.516114692153812e-05,
+      "loss": 1.9984,
+      "step": 25843
+    },
+    {
+      "epoch": 0.5743111111111111,
+      "grad_norm": 1.7134921550750732,
+      "learning_rate": 8.515670148921983e-05,
+      "loss": 1.8807,
+      "step": 25844
+    },
+    {
+      "epoch": 0.5743333333333334,
+      "grad_norm": 1.8306804895401,
+      "learning_rate": 8.515225605690154e-05,
+      "loss": 2.2559,
+      "step": 25845
+    },
+    {
+      "epoch": 0.5743555555555555,
+      "grad_norm": 1.6717336177825928,
+      "learning_rate": 8.514781062458325e-05,
+      "loss": 1.7139,
+      "step": 25846
+    },
+    {
+      "epoch": 0.5743777777777778,
+      "grad_norm": 1.6805630922317505,
+      "learning_rate": 8.514336519226496e-05,
+      "loss": 1.9044,
+      "step": 25847
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 1.1804941892623901,
+      "learning_rate": 8.513891975994666e-05,
+      "loss": 0.8815,
+      "step": 25848
+    },
+    {
+      "epoch": 0.5744222222222222,
+      "grad_norm": 1.9663548469543457,
+      "learning_rate": 8.513447432762836e-05,
+      "loss": 1.888,
+      "step": 25849
+    },
+    {
+      "epoch": 0.5744444444444444,
+      "grad_norm": 1.8257575035095215,
+      "learning_rate": 8.513002889531007e-05,
+      "loss": 1.6505,
+      "step": 25850
+    },
+    {
+      "epoch": 0.5744666666666667,
+      "grad_norm": 1.527230143547058,
+      "learning_rate": 8.512558346299177e-05,
+      "loss": 2.4513,
+      "step": 25851
+    },
+    {
+      "epoch": 0.5744888888888889,
+      "grad_norm": 1.4793553352355957,
+      "learning_rate": 8.512113803067348e-05,
+      "loss": 2.4797,
+      "step": 25852
+    },
+    {
+      "epoch": 0.5745111111111111,
+      "grad_norm": 1.4620087146759033,
+      "learning_rate": 8.51166925983552e-05,
+      "loss": 2.487,
+      "step": 25853
+    },
+    {
+      "epoch": 0.5745333333333333,
+      "grad_norm": 1.2276864051818848,
+      "learning_rate": 8.51122471660369e-05,
+      "loss": 1.3152,
+      "step": 25854
+    },
+    {
+      "epoch": 0.5745555555555556,
+      "grad_norm": 1.73105788230896,
+      "learning_rate": 8.510780173371861e-05,
+      "loss": 1.9796,
+      "step": 25855
+    },
+    {
+      "epoch": 0.5745777777777777,
+      "grad_norm": 1.2348138093948364,
+      "learning_rate": 8.510335630140032e-05,
+      "loss": 1.4845,
+      "step": 25856
+    },
+    {
+      "epoch": 0.5746,
+      "grad_norm": 1.5831315517425537,
+      "learning_rate": 8.509891086908203e-05,
+      "loss": 2.0515,
+      "step": 25857
+    },
+    {
+      "epoch": 0.5746222222222223,
+      "grad_norm": 1.4946595430374146,
+      "learning_rate": 8.509446543676372e-05,
+      "loss": 2.227,
+      "step": 25858
+    },
+    {
+      "epoch": 0.5746444444444444,
+      "grad_norm": 1.8229634761810303,
+      "learning_rate": 8.509002000444543e-05,
+      "loss": 2.4852,
+      "step": 25859
+    },
+    {
+      "epoch": 0.5746666666666667,
+      "grad_norm": 1.4324060678482056,
+      "learning_rate": 8.508557457212714e-05,
+      "loss": 2.1294,
+      "step": 25860
+    },
+    {
+      "epoch": 0.5746888888888889,
+      "grad_norm": 1.3680497407913208,
+      "learning_rate": 8.508112913980884e-05,
+      "loss": 1.8083,
+      "step": 25861
+    },
+    {
+      "epoch": 0.5747111111111111,
+      "grad_norm": 1.4841502904891968,
+      "learning_rate": 8.507668370749056e-05,
+      "loss": 2.3155,
+      "step": 25862
+    },
+    {
+      "epoch": 0.5747333333333333,
+      "grad_norm": 1.6526727676391602,
+      "learning_rate": 8.507223827517227e-05,
+      "loss": 2.0219,
+      "step": 25863
+    },
+    {
+      "epoch": 0.5747555555555556,
+      "grad_norm": 1.7587121725082397,
+      "learning_rate": 8.506779284285397e-05,
+      "loss": 1.9555,
+      "step": 25864
+    },
+    {
+      "epoch": 0.5747777777777778,
+      "grad_norm": 1.8311957120895386,
+      "learning_rate": 8.506334741053568e-05,
+      "loss": 2.3431,
+      "step": 25865
+    },
+    {
+      "epoch": 0.5748,
+      "grad_norm": 1.590606451034546,
+      "learning_rate": 8.505890197821739e-05,
+      "loss": 1.7907,
+      "step": 25866
+    },
+    {
+      "epoch": 0.5748222222222222,
+      "grad_norm": 1.7074474096298218,
+      "learning_rate": 8.50544565458991e-05,
+      "loss": 2.0113,
+      "step": 25867
+    },
+    {
+      "epoch": 0.5748444444444445,
+      "grad_norm": 1.7187414169311523,
+      "learning_rate": 8.505001111358079e-05,
+      "loss": 1.7066,
+      "step": 25868
+    },
+    {
+      "epoch": 0.5748666666666666,
+      "grad_norm": 1.5686930418014526,
+      "learning_rate": 8.50455656812625e-05,
+      "loss": 1.9829,
+      "step": 25869
+    },
+    {
+      "epoch": 0.5748888888888889,
+      "grad_norm": 1.6503736972808838,
+      "learning_rate": 8.504112024894421e-05,
+      "loss": 2.0525,
+      "step": 25870
+    },
+    {
+      "epoch": 0.5749111111111112,
+      "grad_norm": 1.5381848812103271,
+      "learning_rate": 8.503667481662592e-05,
+      "loss": 2.1522,
+      "step": 25871
+    },
+    {
+      "epoch": 0.5749333333333333,
+      "grad_norm": 1.5714423656463623,
+      "learning_rate": 8.503222938430763e-05,
+      "loss": 1.8523,
+      "step": 25872
+    },
+    {
+      "epoch": 0.5749555555555556,
+      "grad_norm": 1.486332893371582,
+      "learning_rate": 8.502778395198934e-05,
+      "loss": 1.8179,
+      "step": 25873
+    },
+    {
+      "epoch": 0.5749777777777778,
+      "grad_norm": 1.8666901588439941,
+      "learning_rate": 8.502333851967105e-05,
+      "loss": 1.9923,
+      "step": 25874
+    },
+    {
+      "epoch": 0.575,
+      "grad_norm": 1.815557837486267,
+      "learning_rate": 8.501889308735275e-05,
+      "loss": 2.1112,
+      "step": 25875
+    },
+    {
+      "epoch": 0.5750222222222222,
+      "grad_norm": 1.8054728507995605,
+      "learning_rate": 8.501444765503446e-05,
+      "loss": 1.5906,
+      "step": 25876
+    },
+    {
+      "epoch": 0.5750444444444445,
+      "grad_norm": 1.6779707670211792,
+      "learning_rate": 8.501000222271617e-05,
+      "loss": 1.8326,
+      "step": 25877
+    },
+    {
+      "epoch": 0.5750666666666666,
+      "grad_norm": 1.5057170391082764,
+      "learning_rate": 8.500555679039786e-05,
+      "loss": 1.8389,
+      "step": 25878
+    },
+    {
+      "epoch": 0.5750888888888889,
+      "grad_norm": 2.190169334411621,
+      "learning_rate": 8.500111135807957e-05,
+      "loss": 2.0456,
+      "step": 25879
+    },
+    {
+      "epoch": 0.5751111111111111,
+      "grad_norm": 1.7067064046859741,
+      "learning_rate": 8.499666592576128e-05,
+      "loss": 1.9427,
+      "step": 25880
+    },
+    {
+      "epoch": 0.5751333333333334,
+      "grad_norm": 1.3328043222427368,
+      "learning_rate": 8.499222049344299e-05,
+      "loss": 1.436,
+      "step": 25881
+    },
+    {
+      "epoch": 0.5751555555555555,
+      "grad_norm": 1.5364593267440796,
+      "learning_rate": 8.49877750611247e-05,
+      "loss": 1.7917,
+      "step": 25882
+    },
+    {
+      "epoch": 0.5751777777777778,
+      "grad_norm": 1.9151912927627563,
+      "learning_rate": 8.498332962880641e-05,
+      "loss": 2.1354,
+      "step": 25883
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 1.7001210451126099,
+      "learning_rate": 8.497888419648812e-05,
+      "loss": 2.0169,
+      "step": 25884
+    },
+    {
+      "epoch": 0.5752222222222222,
+      "grad_norm": 1.3206290006637573,
+      "learning_rate": 8.497443876416982e-05,
+      "loss": 1.386,
+      "step": 25885
+    },
+    {
+      "epoch": 0.5752444444444444,
+      "grad_norm": 1.776057243347168,
+      "learning_rate": 8.496999333185153e-05,
+      "loss": 2.0317,
+      "step": 25886
+    },
+    {
+      "epoch": 0.5752666666666667,
+      "grad_norm": 1.4158116579055786,
+      "learning_rate": 8.496554789953324e-05,
+      "loss": 1.4969,
+      "step": 25887
+    },
+    {
+      "epoch": 0.5752888888888888,
+      "grad_norm": 1.621314287185669,
+      "learning_rate": 8.496110246721493e-05,
+      "loss": 1.7737,
+      "step": 25888
+    },
+    {
+      "epoch": 0.5753111111111111,
+      "grad_norm": 1.9011842012405396,
+      "learning_rate": 8.495665703489664e-05,
+      "loss": 2.5415,
+      "step": 25889
+    },
+    {
+      "epoch": 0.5753333333333334,
+      "grad_norm": 2.2260749340057373,
+      "learning_rate": 8.495221160257836e-05,
+      "loss": 1.8284,
+      "step": 25890
+    },
+    {
+      "epoch": 0.5753555555555555,
+      "grad_norm": 1.7880183458328247,
+      "learning_rate": 8.494776617026006e-05,
+      "loss": 1.5225,
+      "step": 25891
+    },
+    {
+      "epoch": 0.5753777777777778,
+      "grad_norm": 1.56460702419281,
+      "learning_rate": 8.494332073794177e-05,
+      "loss": 1.9047,
+      "step": 25892
+    },
+    {
+      "epoch": 0.5754,
+      "grad_norm": 1.4668025970458984,
+      "learning_rate": 8.493887530562348e-05,
+      "loss": 1.4527,
+      "step": 25893
+    },
+    {
+      "epoch": 0.5754222222222222,
+      "grad_norm": 1.743921160697937,
+      "learning_rate": 8.493442987330519e-05,
+      "loss": 1.5777,
+      "step": 25894
+    },
+    {
+      "epoch": 0.5754444444444444,
+      "grad_norm": 1.5977262258529663,
+      "learning_rate": 8.492998444098689e-05,
+      "loss": 1.4955,
+      "step": 25895
+    },
+    {
+      "epoch": 0.5754666666666667,
+      "grad_norm": 2.3820276260375977,
+      "learning_rate": 8.49255390086686e-05,
+      "loss": 2.1317,
+      "step": 25896
+    },
+    {
+      "epoch": 0.5754888888888889,
+      "grad_norm": 1.623916745185852,
+      "learning_rate": 8.49210935763503e-05,
+      "loss": 1.6631,
+      "step": 25897
+    },
+    {
+      "epoch": 0.5755111111111111,
+      "grad_norm": 2.083130359649658,
+      "learning_rate": 8.4916648144032e-05,
+      "loss": 2.0471,
+      "step": 25898
+    },
+    {
+      "epoch": 0.5755333333333333,
+      "grad_norm": 1.6949437856674194,
+      "learning_rate": 8.491220271171372e-05,
+      "loss": 1.7325,
+      "step": 25899
+    },
+    {
+      "epoch": 0.5755555555555556,
+      "grad_norm": 1.465520977973938,
+      "learning_rate": 8.490775727939543e-05,
+      "loss": 1.5087,
+      "step": 25900
+    },
+    {
+      "epoch": 0.5755777777777777,
+      "grad_norm": 2.0363380908966064,
+      "learning_rate": 8.490331184707713e-05,
+      "loss": 2.9875,
+      "step": 25901
+    },
+    {
+      "epoch": 0.5756,
+      "grad_norm": 1.3037781715393066,
+      "learning_rate": 8.489886641475884e-05,
+      "loss": 2.2702,
+      "step": 25902
+    },
+    {
+      "epoch": 0.5756222222222223,
+      "grad_norm": 1.4280723333358765,
+      "learning_rate": 8.489442098244055e-05,
+      "loss": 2.4732,
+      "step": 25903
+    },
+    {
+      "epoch": 0.5756444444444444,
+      "grad_norm": 0.8468064665794373,
+      "learning_rate": 8.488997555012226e-05,
+      "loss": 1.0019,
+      "step": 25904
+    },
+    {
+      "epoch": 0.5756666666666667,
+      "grad_norm": 1.372684121131897,
+      "learning_rate": 8.488553011780395e-05,
+      "loss": 2.4093,
+      "step": 25905
+    },
+    {
+      "epoch": 0.5756888888888889,
+      "grad_norm": 1.5047813653945923,
+      "learning_rate": 8.488108468548566e-05,
+      "loss": 2.5721,
+      "step": 25906
+    },
+    {
+      "epoch": 0.5757111111111111,
+      "grad_norm": 1.5781632661819458,
+      "learning_rate": 8.487663925316737e-05,
+      "loss": 2.0417,
+      "step": 25907
+    },
+    {
+      "epoch": 0.5757333333333333,
+      "grad_norm": 1.6170930862426758,
+      "learning_rate": 8.487219382084908e-05,
+      "loss": 2.135,
+      "step": 25908
+    },
+    {
+      "epoch": 0.5757555555555556,
+      "grad_norm": 1.679280400276184,
+      "learning_rate": 8.486774838853079e-05,
+      "loss": 2.2844,
+      "step": 25909
+    },
+    {
+      "epoch": 0.5757777777777778,
+      "grad_norm": 1.9144145250320435,
+      "learning_rate": 8.48633029562125e-05,
+      "loss": 2.1399,
+      "step": 25910
+    },
+    {
+      "epoch": 0.5758,
+      "grad_norm": 1.6415542364120483,
+      "learning_rate": 8.48588575238942e-05,
+      "loss": 1.2934,
+      "step": 25911
+    },
+    {
+      "epoch": 0.5758222222222222,
+      "grad_norm": 1.952968955039978,
+      "learning_rate": 8.485441209157591e-05,
+      "loss": 2.2187,
+      "step": 25912
+    },
+    {
+      "epoch": 0.5758444444444445,
+      "grad_norm": 1.1734278202056885,
+      "learning_rate": 8.484996665925762e-05,
+      "loss": 1.1814,
+      "step": 25913
+    },
+    {
+      "epoch": 0.5758666666666666,
+      "grad_norm": 1.1765111684799194,
+      "learning_rate": 8.484552122693933e-05,
+      "loss": 1.1921,
+      "step": 25914
+    },
+    {
+      "epoch": 0.5758888888888889,
+      "grad_norm": 1.738513708114624,
+      "learning_rate": 8.484107579462102e-05,
+      "loss": 1.9746,
+      "step": 25915
+    },
+    {
+      "epoch": 0.5759111111111112,
+      "grad_norm": 1.661946177482605,
+      "learning_rate": 8.483663036230273e-05,
+      "loss": 1.9189,
+      "step": 25916
+    },
+    {
+      "epoch": 0.5759333333333333,
+      "grad_norm": 1.8266701698303223,
+      "learning_rate": 8.483218492998444e-05,
+      "loss": 2.4225,
+      "step": 25917
+    },
+    {
+      "epoch": 0.5759555555555556,
+      "grad_norm": 1.9770067930221558,
+      "learning_rate": 8.482773949766615e-05,
+      "loss": 2.1569,
+      "step": 25918
+    },
+    {
+      "epoch": 0.5759777777777778,
+      "grad_norm": 1.2214926481246948,
+      "learning_rate": 8.482329406534786e-05,
+      "loss": 1.1254,
+      "step": 25919
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.5936335325241089,
+      "learning_rate": 8.481884863302957e-05,
+      "loss": 1.7393,
+      "step": 25920
+    },
+    {
+      "epoch": 0.5760222222222222,
+      "grad_norm": 1.5994678735733032,
+      "learning_rate": 8.481440320071128e-05,
+      "loss": 1.8985,
+      "step": 25921
+    },
+    {
+      "epoch": 0.5760444444444445,
+      "grad_norm": 1.9072351455688477,
+      "learning_rate": 8.480995776839298e-05,
+      "loss": 2.0587,
+      "step": 25922
+    },
+    {
+      "epoch": 0.5760666666666666,
+      "grad_norm": 1.9863942861557007,
+      "learning_rate": 8.480551233607469e-05,
+      "loss": 1.991,
+      "step": 25923
+    },
+    {
+      "epoch": 0.5760888888888889,
+      "grad_norm": 1.942941427230835,
+      "learning_rate": 8.48010669037564e-05,
+      "loss": 2.2744,
+      "step": 25924
+    },
+    {
+      "epoch": 0.5761111111111111,
+      "grad_norm": 1.8738967180252075,
+      "learning_rate": 8.479662147143809e-05,
+      "loss": 1.9959,
+      "step": 25925
+    },
+    {
+      "epoch": 0.5761333333333334,
+      "grad_norm": 1.520711064338684,
+      "learning_rate": 8.47921760391198e-05,
+      "loss": 1.8296,
+      "step": 25926
+    },
+    {
+      "epoch": 0.5761555555555555,
+      "grad_norm": 1.6332484483718872,
+      "learning_rate": 8.478773060680153e-05,
+      "loss": 1.8679,
+      "step": 25927
+    },
+    {
+      "epoch": 0.5761777777777778,
+      "grad_norm": 1.8324124813079834,
+      "learning_rate": 8.478328517448322e-05,
+      "loss": 1.9311,
+      "step": 25928
+    },
+    {
+      "epoch": 0.5762,
+      "grad_norm": 1.7656601667404175,
+      "learning_rate": 8.477883974216493e-05,
+      "loss": 1.8967,
+      "step": 25929
+    },
+    {
+      "epoch": 0.5762222222222222,
+      "grad_norm": 1.652225375175476,
+      "learning_rate": 8.477439430984664e-05,
+      "loss": 1.8419,
+      "step": 25930
+    },
+    {
+      "epoch": 0.5762444444444444,
+      "grad_norm": 2.149226665496826,
+      "learning_rate": 8.476994887752835e-05,
+      "loss": 2.0951,
+      "step": 25931
+    },
+    {
+      "epoch": 0.5762666666666667,
+      "grad_norm": 1.4774209260940552,
+      "learning_rate": 8.476550344521005e-05,
+      "loss": 1.8889,
+      "step": 25932
+    },
+    {
+      "epoch": 0.5762888888888889,
+      "grad_norm": 1.4891711473464966,
+      "learning_rate": 8.476105801289176e-05,
+      "loss": 1.5859,
+      "step": 25933
+    },
+    {
+      "epoch": 0.5763111111111111,
+      "grad_norm": 2.0572304725646973,
+      "learning_rate": 8.475661258057347e-05,
+      "loss": 2.1509,
+      "step": 25934
+    },
+    {
+      "epoch": 0.5763333333333334,
+      "grad_norm": 2.207815647125244,
+      "learning_rate": 8.475216714825516e-05,
+      "loss": 2.0996,
+      "step": 25935
+    },
+    {
+      "epoch": 0.5763555555555555,
+      "grad_norm": 2.040614128112793,
+      "learning_rate": 8.474772171593688e-05,
+      "loss": 2.0951,
+      "step": 25936
+    },
+    {
+      "epoch": 0.5763777777777778,
+      "grad_norm": 1.738759160041809,
+      "learning_rate": 8.47432762836186e-05,
+      "loss": 2.4787,
+      "step": 25937
+    },
+    {
+      "epoch": 0.5764,
+      "grad_norm": 2.2333452701568604,
+      "learning_rate": 8.473883085130029e-05,
+      "loss": 2.0454,
+      "step": 25938
+    },
+    {
+      "epoch": 0.5764222222222222,
+      "grad_norm": 1.803543210029602,
+      "learning_rate": 8.4734385418982e-05,
+      "loss": 1.9241,
+      "step": 25939
+    },
+    {
+      "epoch": 0.5764444444444444,
+      "grad_norm": 1.8549774885177612,
+      "learning_rate": 8.472993998666371e-05,
+      "loss": 2.135,
+      "step": 25940
+    },
+    {
+      "epoch": 0.5764666666666667,
+      "grad_norm": 2.093702554702759,
+      "learning_rate": 8.472549455434542e-05,
+      "loss": 1.8778,
+      "step": 25941
+    },
+    {
+      "epoch": 0.5764888888888889,
+      "grad_norm": 1.755881428718567,
+      "learning_rate": 8.472104912202712e-05,
+      "loss": 2.0415,
+      "step": 25942
+    },
+    {
+      "epoch": 0.5765111111111111,
+      "grad_norm": 2.4870405197143555,
+      "learning_rate": 8.471660368970882e-05,
+      "loss": 1.9493,
+      "step": 25943
+    },
+    {
+      "epoch": 0.5765333333333333,
+      "grad_norm": 1.4329767227172852,
+      "learning_rate": 8.471215825739053e-05,
+      "loss": 1.6417,
+      "step": 25944
+    },
+    {
+      "epoch": 0.5765555555555556,
+      "grad_norm": 1.9608339071273804,
+      "learning_rate": 8.470771282507224e-05,
+      "loss": 2.0607,
+      "step": 25945
+    },
+    {
+      "epoch": 0.5765777777777777,
+      "grad_norm": 1.8790415525436401,
+      "learning_rate": 8.470326739275395e-05,
+      "loss": 1.8563,
+      "step": 25946
+    },
+    {
+      "epoch": 0.5766,
+      "grad_norm": 2.0572495460510254,
+      "learning_rate": 8.469882196043566e-05,
+      "loss": 2.0796,
+      "step": 25947
+    },
+    {
+      "epoch": 0.5766222222222223,
+      "grad_norm": 2.031140089035034,
+      "learning_rate": 8.469437652811736e-05,
+      "loss": 1.7421,
+      "step": 25948
+    },
+    {
+      "epoch": 0.5766444444444444,
+      "grad_norm": 1.3088346719741821,
+      "learning_rate": 8.468993109579907e-05,
+      "loss": 1.0353,
+      "step": 25949
+    },
+    {
+      "epoch": 0.5766666666666667,
+      "grad_norm": 1.5769364833831787,
+      "learning_rate": 8.468548566348078e-05,
+      "loss": 0.9108,
+      "step": 25950
+    },
+    {
+      "epoch": 0.5766888888888889,
+      "grad_norm": 1.0300639867782593,
+      "learning_rate": 8.468104023116249e-05,
+      "loss": 1.2338,
+      "step": 25951
+    },
+    {
+      "epoch": 0.5767111111111111,
+      "grad_norm": 1.5113728046417236,
+      "learning_rate": 8.467659479884418e-05,
+      "loss": 2.6188,
+      "step": 25952
+    },
+    {
+      "epoch": 0.5767333333333333,
+      "grad_norm": 1.1564249992370605,
+      "learning_rate": 8.46721493665259e-05,
+      "loss": 1.4414,
+      "step": 25953
+    },
+    {
+      "epoch": 0.5767555555555556,
+      "grad_norm": 3.3172104358673096,
+      "learning_rate": 8.46677039342076e-05,
+      "loss": 2.6221,
+      "step": 25954
+    },
+    {
+      "epoch": 0.5767777777777777,
+      "grad_norm": 1.5539910793304443,
+      "learning_rate": 8.466325850188931e-05,
+      "loss": 2.5753,
+      "step": 25955
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 1.4837034940719604,
+      "learning_rate": 8.465881306957102e-05,
+      "loss": 2.2811,
+      "step": 25956
+    },
+    {
+      "epoch": 0.5768222222222222,
+      "grad_norm": 1.4640401601791382,
+      "learning_rate": 8.465436763725273e-05,
+      "loss": 1.1172,
+      "step": 25957
+    },
+    {
+      "epoch": 0.5768444444444445,
+      "grad_norm": 1.5787912607192993,
+      "learning_rate": 8.464992220493443e-05,
+      "loss": 2.1486,
+      "step": 25958
+    },
+    {
+      "epoch": 0.5768666666666666,
+      "grad_norm": 1.4304194450378418,
+      "learning_rate": 8.464547677261614e-05,
+      "loss": 2.031,
+      "step": 25959
+    },
+    {
+      "epoch": 0.5768888888888889,
+      "grad_norm": 1.6707733869552612,
+      "learning_rate": 8.464103134029785e-05,
+      "loss": 2.2478,
+      "step": 25960
+    },
+    {
+      "epoch": 0.5769111111111112,
+      "grad_norm": 1.6346321105957031,
+      "learning_rate": 8.463658590797956e-05,
+      "loss": 2.3574,
+      "step": 25961
+    },
+    {
+      "epoch": 0.5769333333333333,
+      "grad_norm": 1.6079306602478027,
+      "learning_rate": 8.463214047566125e-05,
+      "loss": 2.1815,
+      "step": 25962
+    },
+    {
+      "epoch": 0.5769555555555556,
+      "grad_norm": 1.5131382942199707,
+      "learning_rate": 8.462769504334296e-05,
+      "loss": 2.0589,
+      "step": 25963
+    },
+    {
+      "epoch": 0.5769777777777778,
+      "grad_norm": 1.4120280742645264,
+      "learning_rate": 8.462324961102469e-05,
+      "loss": 1.2998,
+      "step": 25964
+    },
+    {
+      "epoch": 0.577,
+      "grad_norm": 1.5342144966125488,
+      "learning_rate": 8.461880417870638e-05,
+      "loss": 1.4498,
+      "step": 25965
+    },
+    {
+      "epoch": 0.5770222222222222,
+      "grad_norm": 1.4350162744522095,
+      "learning_rate": 8.461435874638809e-05,
+      "loss": 1.8269,
+      "step": 25966
+    },
+    {
+      "epoch": 0.5770444444444445,
+      "grad_norm": 1.6343975067138672,
+      "learning_rate": 8.46099133140698e-05,
+      "loss": 2.0455,
+      "step": 25967
+    },
+    {
+      "epoch": 0.5770666666666666,
+      "grad_norm": 1.7373617887496948,
+      "learning_rate": 8.46054678817515e-05,
+      "loss": 2.2496,
+      "step": 25968
+    },
+    {
+      "epoch": 0.5770888888888889,
+      "grad_norm": 1.7621031999588013,
+      "learning_rate": 8.460102244943321e-05,
+      "loss": 2.0521,
+      "step": 25969
+    },
+    {
+      "epoch": 0.5771111111111111,
+      "grad_norm": 2.0356478691101074,
+      "learning_rate": 8.459657701711492e-05,
+      "loss": 1.8139,
+      "step": 25970
+    },
+    {
+      "epoch": 0.5771333333333334,
+      "grad_norm": 1.6037461757659912,
+      "learning_rate": 8.459213158479663e-05,
+      "loss": 1.9601,
+      "step": 25971
+    },
+    {
+      "epoch": 0.5771555555555555,
+      "grad_norm": 1.9582102298736572,
+      "learning_rate": 8.458768615247832e-05,
+      "loss": 2.2421,
+      "step": 25972
+    },
+    {
+      "epoch": 0.5771777777777778,
+      "grad_norm": 1.8982878923416138,
+      "learning_rate": 8.458324072016005e-05,
+      "loss": 2.4973,
+      "step": 25973
+    },
+    {
+      "epoch": 0.5772,
+      "grad_norm": 1.451770544052124,
+      "learning_rate": 8.457879528784176e-05,
+      "loss": 1.8959,
+      "step": 25974
+    },
+    {
+      "epoch": 0.5772222222222222,
+      "grad_norm": 1.4310729503631592,
+      "learning_rate": 8.457434985552345e-05,
+      "loss": 1.3622,
+      "step": 25975
+    },
+    {
+      "epoch": 0.5772444444444444,
+      "grad_norm": 1.6004141569137573,
+      "learning_rate": 8.456990442320516e-05,
+      "loss": 1.6239,
+      "step": 25976
+    },
+    {
+      "epoch": 0.5772666666666667,
+      "grad_norm": 1.7305545806884766,
+      "learning_rate": 8.456545899088687e-05,
+      "loss": 2.0589,
+      "step": 25977
+    },
+    {
+      "epoch": 0.5772888888888889,
+      "grad_norm": 1.3793563842773438,
+      "learning_rate": 8.456101355856858e-05,
+      "loss": 1.6397,
+      "step": 25978
+    },
+    {
+      "epoch": 0.5773111111111111,
+      "grad_norm": 1.5642708539962769,
+      "learning_rate": 8.455656812625028e-05,
+      "loss": 1.9601,
+      "step": 25979
+    },
+    {
+      "epoch": 0.5773333333333334,
+      "grad_norm": 1.8084590435028076,
+      "learning_rate": 8.455212269393199e-05,
+      "loss": 1.5212,
+      "step": 25980
+    },
+    {
+      "epoch": 0.5773555555555555,
+      "grad_norm": 1.7321137189865112,
+      "learning_rate": 8.45476772616137e-05,
+      "loss": 0.9447,
+      "step": 25981
+    },
+    {
+      "epoch": 0.5773777777777778,
+      "grad_norm": 1.569234013557434,
+      "learning_rate": 8.45432318292954e-05,
+      "loss": 1.8806,
+      "step": 25982
+    },
+    {
+      "epoch": 0.5774,
+      "grad_norm": 1.749790072441101,
+      "learning_rate": 8.453878639697711e-05,
+      "loss": 1.9686,
+      "step": 25983
+    },
+    {
+      "epoch": 0.5774222222222222,
+      "grad_norm": 1.4470043182373047,
+      "learning_rate": 8.453434096465882e-05,
+      "loss": 1.5933,
+      "step": 25984
+    },
+    {
+      "epoch": 0.5774444444444444,
+      "grad_norm": 1.4855726957321167,
+      "learning_rate": 8.452989553234052e-05,
+      "loss": 1.76,
+      "step": 25985
+    },
+    {
+      "epoch": 0.5774666666666667,
+      "grad_norm": 1.7785967588424683,
+      "learning_rate": 8.452545010002223e-05,
+      "loss": 1.9297,
+      "step": 25986
+    },
+    {
+      "epoch": 0.5774888888888889,
+      "grad_norm": 1.0004416704177856,
+      "learning_rate": 8.452100466770394e-05,
+      "loss": 0.8414,
+      "step": 25987
+    },
+    {
+      "epoch": 0.5775111111111111,
+      "grad_norm": 17.7835693359375,
+      "learning_rate": 8.451655923538565e-05,
+      "loss": 2.2178,
+      "step": 25988
+    },
+    {
+      "epoch": 0.5775333333333333,
+      "grad_norm": 1.5151816606521606,
+      "learning_rate": 8.451211380306735e-05,
+      "loss": 1.4501,
+      "step": 25989
+    },
+    {
+      "epoch": 0.5775555555555556,
+      "grad_norm": 1.9943801164627075,
+      "learning_rate": 8.450766837074906e-05,
+      "loss": 2.2157,
+      "step": 25990
+    },
+    {
+      "epoch": 0.5775777777777777,
+      "grad_norm": 1.8530950546264648,
+      "learning_rate": 8.450322293843076e-05,
+      "loss": 1.7795,
+      "step": 25991
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 1.7538650035858154,
+      "learning_rate": 8.449877750611247e-05,
+      "loss": 1.6754,
+      "step": 25992
+    },
+    {
+      "epoch": 0.5776222222222223,
+      "grad_norm": 1.8652743101119995,
+      "learning_rate": 8.449433207379418e-05,
+      "loss": 2.1591,
+      "step": 25993
+    },
+    {
+      "epoch": 0.5776444444444444,
+      "grad_norm": 1.8382536172866821,
+      "learning_rate": 8.44898866414759e-05,
+      "loss": 1.8605,
+      "step": 25994
+    },
+    {
+      "epoch": 0.5776666666666667,
+      "grad_norm": 1.5919816493988037,
+      "learning_rate": 8.448544120915759e-05,
+      "loss": 1.7085,
+      "step": 25995
+    },
+    {
+      "epoch": 0.5776888888888889,
+      "grad_norm": 1.91946542263031,
+      "learning_rate": 8.44809957768393e-05,
+      "loss": 1.7697,
+      "step": 25996
+    },
+    {
+      "epoch": 0.5777111111111111,
+      "grad_norm": 1.5796613693237305,
+      "learning_rate": 8.447655034452101e-05,
+      "loss": 1.3837,
+      "step": 25997
+    },
+    {
+      "epoch": 0.5777333333333333,
+      "grad_norm": 1.1964937448501587,
+      "learning_rate": 8.447210491220272e-05,
+      "loss": 0.9521,
+      "step": 25998
+    },
+    {
+      "epoch": 0.5777555555555556,
+      "grad_norm": 1.6924588680267334,
+      "learning_rate": 8.446765947988441e-05,
+      "loss": 1.7327,
+      "step": 25999
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 1.842859148979187,
+      "learning_rate": 8.446321404756612e-05,
+      "loss": 1.6703,
+      "step": 26000
+    },
+    {
+      "epoch": 0.5778,
+      "grad_norm": 0.8927009701728821,
+      "learning_rate": 8.445876861524785e-05,
+      "loss": 0.8949,
+      "step": 26001
+    },
+    {
+      "epoch": 0.5778222222222222,
+      "grad_norm": 1.79087233543396,
+      "learning_rate": 8.445432318292954e-05,
+      "loss": 2.1933,
+      "step": 26002
+    },
+    {
+      "epoch": 0.5778444444444445,
+      "grad_norm": 1.7325570583343506,
+      "learning_rate": 8.444987775061125e-05,
+      "loss": 2.4312,
+      "step": 26003
+    },
+    {
+      "epoch": 0.5778666666666666,
+      "grad_norm": 1.6615060567855835,
+      "learning_rate": 8.444543231829296e-05,
+      "loss": 1.5557,
+      "step": 26004
+    },
+    {
+      "epoch": 0.5778888888888889,
+      "grad_norm": 1.3299299478530884,
+      "learning_rate": 8.444098688597466e-05,
+      "loss": 1.6282,
+      "step": 26005
+    },
+    {
+      "epoch": 0.5779111111111112,
+      "grad_norm": 1.5845272541046143,
+      "learning_rate": 8.443654145365637e-05,
+      "loss": 2.1,
+      "step": 26006
+    },
+    {
+      "epoch": 0.5779333333333333,
+      "grad_norm": 1.890878438949585,
+      "learning_rate": 8.443209602133808e-05,
+      "loss": 1.3777,
+      "step": 26007
+    },
+    {
+      "epoch": 0.5779555555555556,
+      "grad_norm": 1.5573099851608276,
+      "learning_rate": 8.442765058901979e-05,
+      "loss": 2.2133,
+      "step": 26008
+    },
+    {
+      "epoch": 0.5779777777777778,
+      "grad_norm": 1.5288937091827393,
+      "learning_rate": 8.442320515670148e-05,
+      "loss": 1.7278,
+      "step": 26009
+    },
+    {
+      "epoch": 0.578,
+      "grad_norm": 1.4313246011734009,
+      "learning_rate": 8.44187597243832e-05,
+      "loss": 2.0185,
+      "step": 26010
+    },
+    {
+      "epoch": 0.5780222222222222,
+      "grad_norm": 1.7366185188293457,
+      "learning_rate": 8.441431429206492e-05,
+      "loss": 2.2675,
+      "step": 26011
+    },
+    {
+      "epoch": 0.5780444444444445,
+      "grad_norm": 1.7425129413604736,
+      "learning_rate": 8.440986885974661e-05,
+      "loss": 2.218,
+      "step": 26012
+    },
+    {
+      "epoch": 0.5780666666666666,
+      "grad_norm": 1.5490167140960693,
+      "learning_rate": 8.440542342742832e-05,
+      "loss": 1.7977,
+      "step": 26013
+    },
+    {
+      "epoch": 0.5780888888888889,
+      "grad_norm": 1.307113528251648,
+      "learning_rate": 8.440097799511003e-05,
+      "loss": 1.146,
+      "step": 26014
+    },
+    {
+      "epoch": 0.5781111111111111,
+      "grad_norm": 1.814853549003601,
+      "learning_rate": 8.439653256279173e-05,
+      "loss": 1.9841,
+      "step": 26015
+    },
+    {
+      "epoch": 0.5781333333333334,
+      "grad_norm": 1.8090673685073853,
+      "learning_rate": 8.439208713047344e-05,
+      "loss": 1.7952,
+      "step": 26016
+    },
+    {
+      "epoch": 0.5781555555555555,
+      "grad_norm": 1.5758663415908813,
+      "learning_rate": 8.438764169815515e-05,
+      "loss": 2.2281,
+      "step": 26017
+    },
+    {
+      "epoch": 0.5781777777777778,
+      "grad_norm": 1.8426498174667358,
+      "learning_rate": 8.438319626583686e-05,
+      "loss": 2.4655,
+      "step": 26018
+    },
+    {
+      "epoch": 0.5782,
+      "grad_norm": 1.5860536098480225,
+      "learning_rate": 8.437875083351857e-05,
+      "loss": 1.9361,
+      "step": 26019
+    },
+    {
+      "epoch": 0.5782222222222222,
+      "grad_norm": 1.7160062789916992,
+      "learning_rate": 8.437430540120028e-05,
+      "loss": 1.8582,
+      "step": 26020
+    },
+    {
+      "epoch": 0.5782444444444444,
+      "grad_norm": 1.8908746242523193,
+      "learning_rate": 8.436985996888199e-05,
+      "loss": 2.1268,
+      "step": 26021
+    },
+    {
+      "epoch": 0.5782666666666667,
+      "grad_norm": 2.027647018432617,
+      "learning_rate": 8.436541453656368e-05,
+      "loss": 2.4584,
+      "step": 26022
+    },
+    {
+      "epoch": 0.5782888888888889,
+      "grad_norm": 1.639888882637024,
+      "learning_rate": 8.436096910424539e-05,
+      "loss": 1.3977,
+      "step": 26023
+    },
+    {
+      "epoch": 0.5783111111111111,
+      "grad_norm": 1.7778737545013428,
+      "learning_rate": 8.43565236719271e-05,
+      "loss": 2.2876,
+      "step": 26024
+    },
+    {
+      "epoch": 0.5783333333333334,
+      "grad_norm": 2.1147873401641846,
+      "learning_rate": 8.43520782396088e-05,
+      "loss": 1.7371,
+      "step": 26025
+    },
+    {
+      "epoch": 0.5783555555555555,
+      "grad_norm": 2.118032217025757,
+      "learning_rate": 8.43476328072905e-05,
+      "loss": 2.0833,
+      "step": 26026
+    },
+    {
+      "epoch": 0.5783777777777778,
+      "grad_norm": 1.5896843671798706,
+      "learning_rate": 8.434318737497222e-05,
+      "loss": 2.176,
+      "step": 26027
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 1.406416893005371,
+      "learning_rate": 8.433874194265393e-05,
+      "loss": 1.3327,
+      "step": 26028
+    },
+    {
+      "epoch": 0.5784222222222222,
+      "grad_norm": 1.8351752758026123,
+      "learning_rate": 8.433429651033564e-05,
+      "loss": 1.8609,
+      "step": 26029
+    },
+    {
+      "epoch": 0.5784444444444444,
+      "grad_norm": 2.0752804279327393,
+      "learning_rate": 8.432985107801734e-05,
+      "loss": 2.5484,
+      "step": 26030
+    },
+    {
+      "epoch": 0.5784666666666667,
+      "grad_norm": 1.6778042316436768,
+      "learning_rate": 8.432540564569905e-05,
+      "loss": 1.3947,
+      "step": 26031
+    },
+    {
+      "epoch": 0.5784888888888889,
+      "grad_norm": 2.0801656246185303,
+      "learning_rate": 8.432096021338075e-05,
+      "loss": 2.1063,
+      "step": 26032
+    },
+    {
+      "epoch": 0.5785111111111111,
+      "grad_norm": 1.9608205556869507,
+      "learning_rate": 8.431651478106246e-05,
+      "loss": 2.6487,
+      "step": 26033
+    },
+    {
+      "epoch": 0.5785333333333333,
+      "grad_norm": 1.5958486795425415,
+      "learning_rate": 8.431206934874417e-05,
+      "loss": 1.9128,
+      "step": 26034
+    },
+    {
+      "epoch": 0.5785555555555556,
+      "grad_norm": 1.6694389581680298,
+      "learning_rate": 8.430762391642588e-05,
+      "loss": 1.1725,
+      "step": 26035
+    },
+    {
+      "epoch": 0.5785777777777777,
+      "grad_norm": 1.5564417839050293,
+      "learning_rate": 8.430317848410758e-05,
+      "loss": 0.9956,
+      "step": 26036
+    },
+    {
+      "epoch": 0.5786,
+      "grad_norm": 1.6584241390228271,
+      "learning_rate": 8.429873305178929e-05,
+      "loss": 1.8983,
+      "step": 26037
+    },
+    {
+      "epoch": 0.5786222222222223,
+      "grad_norm": 1.9169220924377441,
+      "learning_rate": 8.429428761947101e-05,
+      "loss": 1.8601,
+      "step": 26038
+    },
+    {
+      "epoch": 0.5786444444444444,
+      "grad_norm": 1.9622453451156616,
+      "learning_rate": 8.42898421871527e-05,
+      "loss": 2.2768,
+      "step": 26039
+    },
+    {
+      "epoch": 0.5786666666666667,
+      "grad_norm": 1.8803428411483765,
+      "learning_rate": 8.428539675483441e-05,
+      "loss": 2.0619,
+      "step": 26040
+    },
+    {
+      "epoch": 0.5786888888888889,
+      "grad_norm": 1.8330425024032593,
+      "learning_rate": 8.428095132251612e-05,
+      "loss": 1.8456,
+      "step": 26041
+    },
+    {
+      "epoch": 0.5787111111111111,
+      "grad_norm": 1.7811545133590698,
+      "learning_rate": 8.427650589019782e-05,
+      "loss": 1.6961,
+      "step": 26042
+    },
+    {
+      "epoch": 0.5787333333333333,
+      "grad_norm": 1.984512448310852,
+      "learning_rate": 8.427206045787953e-05,
+      "loss": 1.7993,
+      "step": 26043
+    },
+    {
+      "epoch": 0.5787555555555556,
+      "grad_norm": 2.2069578170776367,
+      "learning_rate": 8.426761502556124e-05,
+      "loss": 1.8305,
+      "step": 26044
+    },
+    {
+      "epoch": 0.5787777777777777,
+      "grad_norm": 1.84796941280365,
+      "learning_rate": 8.426316959324295e-05,
+      "loss": 1.3626,
+      "step": 26045
+    },
+    {
+      "epoch": 0.5788,
+      "grad_norm": 2.032742977142334,
+      "learning_rate": 8.425872416092464e-05,
+      "loss": 1.8709,
+      "step": 26046
+    },
+    {
+      "epoch": 0.5788222222222222,
+      "grad_norm": 1.9718986749649048,
+      "learning_rate": 8.425427872860637e-05,
+      "loss": 1.7734,
+      "step": 26047
+    },
+    {
+      "epoch": 0.5788444444444445,
+      "grad_norm": 1.4757591485977173,
+      "learning_rate": 8.424983329628808e-05,
+      "loss": 1.3125,
+      "step": 26048
+    },
+    {
+      "epoch": 0.5788666666666666,
+      "grad_norm": 1.3887970447540283,
+      "learning_rate": 8.424538786396977e-05,
+      "loss": 0.9535,
+      "step": 26049
+    },
+    {
+      "epoch": 0.5788888888888889,
+      "grad_norm": 2.0439767837524414,
+      "learning_rate": 8.424094243165148e-05,
+      "loss": 0.9072,
+      "step": 26050
+    },
+    {
+      "epoch": 0.5789111111111112,
+      "grad_norm": 1.327715277671814,
+      "learning_rate": 8.423649699933319e-05,
+      "loss": 2.4519,
+      "step": 26051
+    },
+    {
+      "epoch": 0.5789333333333333,
+      "grad_norm": 1.6525444984436035,
+      "learning_rate": 8.423205156701489e-05,
+      "loss": 2.2627,
+      "step": 26052
+    },
+    {
+      "epoch": 0.5789555555555556,
+      "grad_norm": 1.309954047203064,
+      "learning_rate": 8.42276061346966e-05,
+      "loss": 2.5536,
+      "step": 26053
+    },
+    {
+      "epoch": 0.5789777777777778,
+      "grad_norm": 0.5188814997673035,
+      "learning_rate": 8.422316070237831e-05,
+      "loss": 0.0196,
+      "step": 26054
+    },
+    {
+      "epoch": 0.579,
+      "grad_norm": 1.337967872619629,
+      "learning_rate": 8.421871527006002e-05,
+      "loss": 2.1331,
+      "step": 26055
+    },
+    {
+      "epoch": 0.5790222222222222,
+      "grad_norm": 1.9190223217010498,
+      "learning_rate": 8.421426983774173e-05,
+      "loss": 2.3599,
+      "step": 26056
+    },
+    {
+      "epoch": 0.5790444444444445,
+      "grad_norm": 1.5431402921676636,
+      "learning_rate": 8.420982440542344e-05,
+      "loss": 2.2485,
+      "step": 26057
+    },
+    {
+      "epoch": 0.5790666666666666,
+      "grad_norm": 1.8881802558898926,
+      "learning_rate": 8.420537897310515e-05,
+      "loss": 2.4807,
+      "step": 26058
+    },
+    {
+      "epoch": 0.5790888888888889,
+      "grad_norm": 1.7573673725128174,
+      "learning_rate": 8.420093354078684e-05,
+      "loss": 1.9549,
+      "step": 26059
+    },
+    {
+      "epoch": 0.5791111111111111,
+      "grad_norm": 1.707239031791687,
+      "learning_rate": 8.419648810846855e-05,
+      "loss": 2.4954,
+      "step": 26060
+    },
+    {
+      "epoch": 0.5791333333333334,
+      "grad_norm": 1.5896327495574951,
+      "learning_rate": 8.419204267615026e-05,
+      "loss": 1.9372,
+      "step": 26061
+    },
+    {
+      "epoch": 0.5791555555555555,
+      "grad_norm": 1.5502749681472778,
+      "learning_rate": 8.418759724383196e-05,
+      "loss": 2.0857,
+      "step": 26062
+    },
+    {
+      "epoch": 0.5791777777777778,
+      "grad_norm": 1.7027544975280762,
+      "learning_rate": 8.418315181151367e-05,
+      "loss": 1.9132,
+      "step": 26063
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 1.5180604457855225,
+      "learning_rate": 8.417870637919538e-05,
+      "loss": 1.9358,
+      "step": 26064
+    },
+    {
+      "epoch": 0.5792222222222222,
+      "grad_norm": 1.7091317176818848,
+      "learning_rate": 8.417426094687709e-05,
+      "loss": 2.1182,
+      "step": 26065
+    },
+    {
+      "epoch": 0.5792444444444445,
+      "grad_norm": 1.6138535737991333,
+      "learning_rate": 8.41698155145588e-05,
+      "loss": 2.0723,
+      "step": 26066
+    },
+    {
+      "epoch": 0.5792666666666667,
+      "grad_norm": 1.7695465087890625,
+      "learning_rate": 8.41653700822405e-05,
+      "loss": 2.4244,
+      "step": 26067
+    },
+    {
+      "epoch": 0.5792888888888889,
+      "grad_norm": 1.6330814361572266,
+      "learning_rate": 8.416092464992222e-05,
+      "loss": 1.7658,
+      "step": 26068
+    },
+    {
+      "epoch": 0.5793111111111111,
+      "grad_norm": 1.3131344318389893,
+      "learning_rate": 8.415647921760391e-05,
+      "loss": 1.6413,
+      "step": 26069
+    },
+    {
+      "epoch": 0.5793333333333334,
+      "grad_norm": 1.4656513929367065,
+      "learning_rate": 8.415203378528562e-05,
+      "loss": 1.9481,
+      "step": 26070
+    },
+    {
+      "epoch": 0.5793555555555555,
+      "grad_norm": 1.9465343952178955,
+      "learning_rate": 8.414758835296733e-05,
+      "loss": 1.8912,
+      "step": 26071
+    },
+    {
+      "epoch": 0.5793777777777778,
+      "grad_norm": 1.7784497737884521,
+      "learning_rate": 8.414314292064903e-05,
+      "loss": 1.6295,
+      "step": 26072
+    },
+    {
+      "epoch": 0.5794,
+      "grad_norm": 1.859359860420227,
+      "learning_rate": 8.413869748833074e-05,
+      "loss": 2.1273,
+      "step": 26073
+    },
+    {
+      "epoch": 0.5794222222222222,
+      "grad_norm": 1.8585387468338013,
+      "learning_rate": 8.413425205601245e-05,
+      "loss": 2.2089,
+      "step": 26074
+    },
+    {
+      "epoch": 0.5794444444444444,
+      "grad_norm": 1.4101988077163696,
+      "learning_rate": 8.412980662369417e-05,
+      "loss": 1.3788,
+      "step": 26075
+    },
+    {
+      "epoch": 0.5794666666666667,
+      "grad_norm": 1.579291582107544,
+      "learning_rate": 8.412536119137587e-05,
+      "loss": 1.9922,
+      "step": 26076
+    },
+    {
+      "epoch": 0.5794888888888889,
+      "grad_norm": 1.8342472314834595,
+      "learning_rate": 8.412091575905758e-05,
+      "loss": 1.5453,
+      "step": 26077
+    },
+    {
+      "epoch": 0.5795111111111111,
+      "grad_norm": 1.6948658227920532,
+      "learning_rate": 8.411647032673928e-05,
+      "loss": 2.0403,
+      "step": 26078
+    },
+    {
+      "epoch": 0.5795333333333333,
+      "grad_norm": 1.831552505493164,
+      "learning_rate": 8.411202489442098e-05,
+      "loss": 1.327,
+      "step": 26079
+    },
+    {
+      "epoch": 0.5795555555555556,
+      "grad_norm": 1.650087594985962,
+      "learning_rate": 8.410757946210269e-05,
+      "loss": 1.6081,
+      "step": 26080
+    },
+    {
+      "epoch": 0.5795777777777777,
+      "grad_norm": 1.7019950151443481,
+      "learning_rate": 8.41031340297844e-05,
+      "loss": 2.2084,
+      "step": 26081
+    },
+    {
+      "epoch": 0.5796,
+      "grad_norm": 1.4378479719161987,
+      "learning_rate": 8.40986885974661e-05,
+      "loss": 1.6135,
+      "step": 26082
+    },
+    {
+      "epoch": 0.5796222222222223,
+      "grad_norm": 2.8144371509552,
+      "learning_rate": 8.409424316514782e-05,
+      "loss": 1.8706,
+      "step": 26083
+    },
+    {
+      "epoch": 0.5796444444444444,
+      "grad_norm": 1.7685115337371826,
+      "learning_rate": 8.408979773282953e-05,
+      "loss": 1.8798,
+      "step": 26084
+    },
+    {
+      "epoch": 0.5796666666666667,
+      "grad_norm": 1.2805466651916504,
+      "learning_rate": 8.408535230051124e-05,
+      "loss": 0.905,
+      "step": 26085
+    },
+    {
+      "epoch": 0.5796888888888889,
+      "grad_norm": 1.7151274681091309,
+      "learning_rate": 8.408090686819293e-05,
+      "loss": 2.0463,
+      "step": 26086
+    },
+    {
+      "epoch": 0.5797111111111111,
+      "grad_norm": 1.7159134149551392,
+      "learning_rate": 8.407646143587464e-05,
+      "loss": 1.8117,
+      "step": 26087
+    },
+    {
+      "epoch": 0.5797333333333333,
+      "grad_norm": 1.7387936115264893,
+      "learning_rate": 8.407201600355635e-05,
+      "loss": 1.6479,
+      "step": 26088
+    },
+    {
+      "epoch": 0.5797555555555556,
+      "grad_norm": 1.5587831735610962,
+      "learning_rate": 8.406757057123805e-05,
+      "loss": 1.6855,
+      "step": 26089
+    },
+    {
+      "epoch": 0.5797777777777777,
+      "grad_norm": 1.9584736824035645,
+      "learning_rate": 8.406312513891976e-05,
+      "loss": 1.9174,
+      "step": 26090
+    },
+    {
+      "epoch": 0.5798,
+      "grad_norm": 2.032670497894287,
+      "learning_rate": 8.405867970660147e-05,
+      "loss": 2.5423,
+      "step": 26091
+    },
+    {
+      "epoch": 0.5798222222222222,
+      "grad_norm": 1.73224937915802,
+      "learning_rate": 8.405423427428318e-05,
+      "loss": 1.8867,
+      "step": 26092
+    },
+    {
+      "epoch": 0.5798444444444445,
+      "grad_norm": 1.6868783235549927,
+      "learning_rate": 8.404978884196489e-05,
+      "loss": 2.0453,
+      "step": 26093
+    },
+    {
+      "epoch": 0.5798666666666666,
+      "grad_norm": 1.8552583456039429,
+      "learning_rate": 8.40453434096466e-05,
+      "loss": 2.0552,
+      "step": 26094
+    },
+    {
+      "epoch": 0.5798888888888889,
+      "grad_norm": 2.3586230278015137,
+      "learning_rate": 8.404089797732831e-05,
+      "loss": 2.2544,
+      "step": 26095
+    },
+    {
+      "epoch": 0.5799111111111112,
+      "grad_norm": 1.809834599494934,
+      "learning_rate": 8.403645254501e-05,
+      "loss": 1.547,
+      "step": 26096
+    },
+    {
+      "epoch": 0.5799333333333333,
+      "grad_norm": 1.7929770946502686,
+      "learning_rate": 8.403200711269171e-05,
+      "loss": 1.7148,
+      "step": 26097
+    },
+    {
+      "epoch": 0.5799555555555556,
+      "grad_norm": 1.786232352256775,
+      "learning_rate": 8.402756168037342e-05,
+      "loss": 1.4739,
+      "step": 26098
+    },
+    {
+      "epoch": 0.5799777777777778,
+      "grad_norm": 1.915028691291809,
+      "learning_rate": 8.402311624805512e-05,
+      "loss": 1.9315,
+      "step": 26099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8769526481628418,
+      "learning_rate": 8.401867081573683e-05,
+      "loss": 1.0471,
+      "step": 26100
+    },
+    {
+      "epoch": 0.5800222222222222,
+      "grad_norm": 1.4817572832107544,
+      "learning_rate": 8.401422538341854e-05,
+      "loss": 2.7088,
+      "step": 26101
+    },
+    {
+      "epoch": 0.5800444444444445,
+      "grad_norm": 1.4244190454483032,
+      "learning_rate": 8.400977995110025e-05,
+      "loss": 1.9826,
+      "step": 26102
+    },
+    {
+      "epoch": 0.5800666666666666,
+      "grad_norm": 1.5300400257110596,
+      "learning_rate": 8.400533451878196e-05,
+      "loss": 2.388,
+      "step": 26103
+    },
+    {
+      "epoch": 0.5800888888888889,
+      "grad_norm": 1.3785556554794312,
+      "learning_rate": 8.400088908646367e-05,
+      "loss": 2.2987,
+      "step": 26104
+    },
+    {
+      "epoch": 0.5801111111111111,
+      "grad_norm": 1.6409354209899902,
+      "learning_rate": 8.399644365414538e-05,
+      "loss": 2.3494,
+      "step": 26105
+    },
+    {
+      "epoch": 0.5801333333333333,
+      "grad_norm": 1.6343281269073486,
+      "learning_rate": 8.399199822182707e-05,
+      "loss": 2.0914,
+      "step": 26106
+    },
+    {
+      "epoch": 0.5801555555555555,
+      "grad_norm": 1.7643954753875732,
+      "learning_rate": 8.398755278950878e-05,
+      "loss": 2.4257,
+      "step": 26107
+    },
+    {
+      "epoch": 0.5801777777777778,
+      "grad_norm": 1.5621308088302612,
+      "learning_rate": 8.398310735719049e-05,
+      "loss": 1.8084,
+      "step": 26108
+    },
+    {
+      "epoch": 0.5802,
+      "grad_norm": 1.443436622619629,
+      "learning_rate": 8.397866192487219e-05,
+      "loss": 1.6621,
+      "step": 26109
+    },
+    {
+      "epoch": 0.5802222222222222,
+      "grad_norm": 1.5355186462402344,
+      "learning_rate": 8.39742164925539e-05,
+      "loss": 2.015,
+      "step": 26110
+    },
+    {
+      "epoch": 0.5802444444444445,
+      "grad_norm": 1.5714714527130127,
+      "learning_rate": 8.396977106023561e-05,
+      "loss": 1.6738,
+      "step": 26111
+    },
+    {
+      "epoch": 0.5802666666666667,
+      "grad_norm": 1.5971449613571167,
+      "learning_rate": 8.396532562791733e-05,
+      "loss": 2.032,
+      "step": 26112
+    },
+    {
+      "epoch": 0.5802888888888889,
+      "grad_norm": 1.4106720685958862,
+      "learning_rate": 8.396088019559903e-05,
+      "loss": 1.8086,
+      "step": 26113
+    },
+    {
+      "epoch": 0.5803111111111111,
+      "grad_norm": 1.46250319480896,
+      "learning_rate": 8.395643476328074e-05,
+      "loss": 2.0213,
+      "step": 26114
+    },
+    {
+      "epoch": 0.5803333333333334,
+      "grad_norm": 1.6457831859588623,
+      "learning_rate": 8.395198933096245e-05,
+      "loss": 2.228,
+      "step": 26115
+    },
+    {
+      "epoch": 0.5803555555555555,
+      "grad_norm": 1.6616058349609375,
+      "learning_rate": 8.394754389864414e-05,
+      "loss": 2.2746,
+      "step": 26116
+    },
+    {
+      "epoch": 0.5803777777777778,
+      "grad_norm": 1.7403942346572876,
+      "learning_rate": 8.394309846632585e-05,
+      "loss": 2.221,
+      "step": 26117
+    },
+    {
+      "epoch": 0.5804,
+      "grad_norm": 1.6714938879013062,
+      "learning_rate": 8.393865303400756e-05,
+      "loss": 1.8509,
+      "step": 26118
+    },
+    {
+      "epoch": 0.5804222222222222,
+      "grad_norm": 1.7840502262115479,
+      "learning_rate": 8.393420760168926e-05,
+      "loss": 2.1026,
+      "step": 26119
+    },
+    {
+      "epoch": 0.5804444444444444,
+      "grad_norm": 1.6288597583770752,
+      "learning_rate": 8.392976216937098e-05,
+      "loss": 1.978,
+      "step": 26120
+    },
+    {
+      "epoch": 0.5804666666666667,
+      "grad_norm": 1.6369847059249878,
+      "learning_rate": 8.392531673705269e-05,
+      "loss": 2.2848,
+      "step": 26121
+    },
+    {
+      "epoch": 0.5804888888888889,
+      "grad_norm": 1.6995961666107178,
+      "learning_rate": 8.39208713047344e-05,
+      "loss": 2.2823,
+      "step": 26122
+    },
+    {
+      "epoch": 0.5805111111111111,
+      "grad_norm": 1.77353835105896,
+      "learning_rate": 8.39164258724161e-05,
+      "loss": 2.4403,
+      "step": 26123
+    },
+    {
+      "epoch": 0.5805333333333333,
+      "grad_norm": 1.6163816452026367,
+      "learning_rate": 8.39119804400978e-05,
+      "loss": 2.1503,
+      "step": 26124
+    },
+    {
+      "epoch": 0.5805555555555556,
+      "grad_norm": 1.5171163082122803,
+      "learning_rate": 8.390753500777951e-05,
+      "loss": 1.5474,
+      "step": 26125
+    },
+    {
+      "epoch": 0.5805777777777777,
+      "grad_norm": 1.759602665901184,
+      "learning_rate": 8.390308957546121e-05,
+      "loss": 1.611,
+      "step": 26126
+    },
+    {
+      "epoch": 0.5806,
+      "grad_norm": 1.63189697265625,
+      "learning_rate": 8.389864414314292e-05,
+      "loss": 2.0526,
+      "step": 26127
+    },
+    {
+      "epoch": 0.5806222222222223,
+      "grad_norm": 2.072521686553955,
+      "learning_rate": 8.389419871082463e-05,
+      "loss": 2.1334,
+      "step": 26128
+    },
+    {
+      "epoch": 0.5806444444444444,
+      "grad_norm": 1.7085297107696533,
+      "learning_rate": 8.388975327850634e-05,
+      "loss": 1.8668,
+      "step": 26129
+    },
+    {
+      "epoch": 0.5806666666666667,
+      "grad_norm": 0.3806315064430237,
+      "learning_rate": 8.388530784618805e-05,
+      "loss": 0.0278,
+      "step": 26130
+    },
+    {
+      "epoch": 0.5806888888888889,
+      "grad_norm": 1.724603533744812,
+      "learning_rate": 8.388086241386976e-05,
+      "loss": 1.9284,
+      "step": 26131
+    },
+    {
+      "epoch": 0.5807111111111111,
+      "grad_norm": 1.47846519947052,
+      "learning_rate": 8.387641698155147e-05,
+      "loss": 1.4473,
+      "step": 26132
+    },
+    {
+      "epoch": 0.5807333333333333,
+      "grad_norm": 1.6238040924072266,
+      "learning_rate": 8.387197154923316e-05,
+      "loss": 2.0136,
+      "step": 26133
+    },
+    {
+      "epoch": 0.5807555555555556,
+      "grad_norm": 1.7846760749816895,
+      "learning_rate": 8.386752611691487e-05,
+      "loss": 1.83,
+      "step": 26134
+    },
+    {
+      "epoch": 0.5807777777777777,
+      "grad_norm": 1.350911021232605,
+      "learning_rate": 8.386308068459658e-05,
+      "loss": 1.5452,
+      "step": 26135
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 1.2389070987701416,
+      "learning_rate": 8.385863525227828e-05,
+      "loss": 0.9817,
+      "step": 26136
+    },
+    {
+      "epoch": 0.5808222222222222,
+      "grad_norm": 1.82937753200531,
+      "learning_rate": 8.385418981995999e-05,
+      "loss": 2.1897,
+      "step": 26137
+    },
+    {
+      "epoch": 0.5808444444444445,
+      "grad_norm": 1.6447229385375977,
+      "learning_rate": 8.38497443876417e-05,
+      "loss": 1.5815,
+      "step": 26138
+    },
+    {
+      "epoch": 0.5808666666666666,
+      "grad_norm": 1.963240385055542,
+      "learning_rate": 8.384529895532341e-05,
+      "loss": 1.8421,
+      "step": 26139
+    },
+    {
+      "epoch": 0.5808888888888889,
+      "grad_norm": 1.9662257432937622,
+      "learning_rate": 8.384085352300512e-05,
+      "loss": 2.0641,
+      "step": 26140
+    },
+    {
+      "epoch": 0.5809111111111112,
+      "grad_norm": 1.741540789604187,
+      "learning_rate": 8.383640809068683e-05,
+      "loss": 2.1514,
+      "step": 26141
+    },
+    {
+      "epoch": 0.5809333333333333,
+      "grad_norm": 1.692657470703125,
+      "learning_rate": 8.383196265836854e-05,
+      "loss": 1.6779,
+      "step": 26142
+    },
+    {
+      "epoch": 0.5809555555555556,
+      "grad_norm": 1.5241749286651611,
+      "learning_rate": 8.382751722605023e-05,
+      "loss": 1.7482,
+      "step": 26143
+    },
+    {
+      "epoch": 0.5809777777777778,
+      "grad_norm": 1.9232747554779053,
+      "learning_rate": 8.382307179373194e-05,
+      "loss": 1.9883,
+      "step": 26144
+    },
+    {
+      "epoch": 0.581,
+      "grad_norm": 2.1884825229644775,
+      "learning_rate": 8.381862636141365e-05,
+      "loss": 1.8161,
+      "step": 26145
+    },
+    {
+      "epoch": 0.5810222222222222,
+      "grad_norm": 1.8066474199295044,
+      "learning_rate": 8.381418092909535e-05,
+      "loss": 1.4866,
+      "step": 26146
+    },
+    {
+      "epoch": 0.5810444444444445,
+      "grad_norm": 1.9330378770828247,
+      "learning_rate": 8.380973549677706e-05,
+      "loss": 1.9459,
+      "step": 26147
+    },
+    {
+      "epoch": 0.5810666666666666,
+      "grad_norm": 1.8063390254974365,
+      "learning_rate": 8.380529006445877e-05,
+      "loss": 1.8505,
+      "step": 26148
+    },
+    {
+      "epoch": 0.5810888888888889,
+      "grad_norm": 1.8492445945739746,
+      "learning_rate": 8.380084463214049e-05,
+      "loss": 1.6854,
+      "step": 26149
+    },
+    {
+      "epoch": 0.5811111111111111,
+      "grad_norm": 0.2438049167394638,
+      "learning_rate": 8.379639919982219e-05,
+      "loss": 0.0497,
+      "step": 26150
+    },
+    {
+      "epoch": 0.5811333333333333,
+      "grad_norm": 1.6242479085922241,
+      "learning_rate": 8.37919537675039e-05,
+      "loss": 2.1615,
+      "step": 26151
+    },
+    {
+      "epoch": 0.5811555555555555,
+      "grad_norm": 1.438272476196289,
+      "learning_rate": 8.37875083351856e-05,
+      "loss": 2.4119,
+      "step": 26152
+    },
+    {
+      "epoch": 0.5811777777777778,
+      "grad_norm": 0.1465410739183426,
+      "learning_rate": 8.37830629028673e-05,
+      "loss": 0.017,
+      "step": 26153
+    },
+    {
+      "epoch": 0.5812,
+      "grad_norm": 1.5817630290985107,
+      "learning_rate": 8.377861747054901e-05,
+      "loss": 2.1741,
+      "step": 26154
+    },
+    {
+      "epoch": 0.5812222222222222,
+      "grad_norm": 1.5494940280914307,
+      "learning_rate": 8.377417203823072e-05,
+      "loss": 2.602,
+      "step": 26155
+    },
+    {
+      "epoch": 0.5812444444444445,
+      "grad_norm": 1.5453170537948608,
+      "learning_rate": 8.376972660591242e-05,
+      "loss": 1.8616,
+      "step": 26156
+    },
+    {
+      "epoch": 0.5812666666666667,
+      "grad_norm": 1.8508343696594238,
+      "learning_rate": 8.376528117359414e-05,
+      "loss": 2.2297,
+      "step": 26157
+    },
+    {
+      "epoch": 0.5812888888888889,
+      "grad_norm": 1.8826130628585815,
+      "learning_rate": 8.376083574127585e-05,
+      "loss": 2.3587,
+      "step": 26158
+    },
+    {
+      "epoch": 0.5813111111111111,
+      "grad_norm": 1.9748693704605103,
+      "learning_rate": 8.375639030895756e-05,
+      "loss": 2.1999,
+      "step": 26159
+    },
+    {
+      "epoch": 0.5813333333333334,
+      "grad_norm": 1.8187892436981201,
+      "learning_rate": 8.375194487663926e-05,
+      "loss": 1.4834,
+      "step": 26160
+    },
+    {
+      "epoch": 0.5813555555555555,
+      "grad_norm": 1.8985687494277954,
+      "learning_rate": 8.374749944432097e-05,
+      "loss": 2.2664,
+      "step": 26161
+    },
+    {
+      "epoch": 0.5813777777777778,
+      "grad_norm": 1.7321456670761108,
+      "learning_rate": 8.374305401200268e-05,
+      "loss": 2.1491,
+      "step": 26162
+    },
+    {
+      "epoch": 0.5814,
+      "grad_norm": 1.2643800973892212,
+      "learning_rate": 8.373860857968437e-05,
+      "loss": 1.0812,
+      "step": 26163
+    },
+    {
+      "epoch": 0.5814222222222222,
+      "grad_norm": 1.6095198392868042,
+      "learning_rate": 8.373416314736608e-05,
+      "loss": 2.0306,
+      "step": 26164
+    },
+    {
+      "epoch": 0.5814444444444444,
+      "grad_norm": 1.8974038362503052,
+      "learning_rate": 8.372971771504779e-05,
+      "loss": 1.6857,
+      "step": 26165
+    },
+    {
+      "epoch": 0.5814666666666667,
+      "grad_norm": 1.8022187948226929,
+      "learning_rate": 8.37252722827295e-05,
+      "loss": 1.6285,
+      "step": 26166
+    },
+    {
+      "epoch": 0.5814888888888889,
+      "grad_norm": 1.73734450340271,
+      "learning_rate": 8.372082685041121e-05,
+      "loss": 1.926,
+      "step": 26167
+    },
+    {
+      "epoch": 0.5815111111111111,
+      "grad_norm": 1.4879589080810547,
+      "learning_rate": 8.371638141809292e-05,
+      "loss": 1.7018,
+      "step": 26168
+    },
+    {
+      "epoch": 0.5815333333333333,
+      "grad_norm": 1.672507882118225,
+      "learning_rate": 8.371193598577463e-05,
+      "loss": 2.0912,
+      "step": 26169
+    },
+    {
+      "epoch": 0.5815555555555556,
+      "grad_norm": 1.5402486324310303,
+      "learning_rate": 8.370749055345633e-05,
+      "loss": 2.1801,
+      "step": 26170
+    },
+    {
+      "epoch": 0.5815777777777777,
+      "grad_norm": 1.5981429815292358,
+      "learning_rate": 8.370304512113804e-05,
+      "loss": 2.0361,
+      "step": 26171
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 1.615914225578308,
+      "learning_rate": 8.369859968881974e-05,
+      "loss": 1.7654,
+      "step": 26172
+    },
+    {
+      "epoch": 0.5816222222222223,
+      "grad_norm": 1.6028060913085938,
+      "learning_rate": 8.369415425650144e-05,
+      "loss": 2.3957,
+      "step": 26173
+    },
+    {
+      "epoch": 0.5816444444444444,
+      "grad_norm": 1.8624187707901,
+      "learning_rate": 8.368970882418315e-05,
+      "loss": 2.0461,
+      "step": 26174
+    },
+    {
+      "epoch": 0.5816666666666667,
+      "grad_norm": 1.5869486331939697,
+      "learning_rate": 8.368526339186486e-05,
+      "loss": 1.9551,
+      "step": 26175
+    },
+    {
+      "epoch": 0.5816888888888889,
+      "grad_norm": 1.9387506246566772,
+      "learning_rate": 8.368081795954657e-05,
+      "loss": 2.1581,
+      "step": 26176
+    },
+    {
+      "epoch": 0.5817111111111111,
+      "grad_norm": 1.8396356105804443,
+      "learning_rate": 8.367637252722828e-05,
+      "loss": 1.9605,
+      "step": 26177
+    },
+    {
+      "epoch": 0.5817333333333333,
+      "grad_norm": 1.579355001449585,
+      "learning_rate": 8.367192709490999e-05,
+      "loss": 1.8826,
+      "step": 26178
+    },
+    {
+      "epoch": 0.5817555555555556,
+      "grad_norm": 2.3210432529449463,
+      "learning_rate": 8.36674816625917e-05,
+      "loss": 2.2686,
+      "step": 26179
+    },
+    {
+      "epoch": 0.5817777777777777,
+      "grad_norm": 1.7856831550598145,
+      "learning_rate": 8.36630362302734e-05,
+      "loss": 2.0718,
+      "step": 26180
+    },
+    {
+      "epoch": 0.5818,
+      "grad_norm": 1.4405708312988281,
+      "learning_rate": 8.36585907979551e-05,
+      "loss": 1.4981,
+      "step": 26181
+    },
+    {
+      "epoch": 0.5818222222222222,
+      "grad_norm": 2.206476926803589,
+      "learning_rate": 8.365414536563681e-05,
+      "loss": 1.9997,
+      "step": 26182
+    },
+    {
+      "epoch": 0.5818444444444445,
+      "grad_norm": 1.7495813369750977,
+      "learning_rate": 8.364969993331851e-05,
+      "loss": 1.6429,
+      "step": 26183
+    },
+    {
+      "epoch": 0.5818666666666666,
+      "grad_norm": 1.9488918781280518,
+      "learning_rate": 8.364525450100022e-05,
+      "loss": 2.0269,
+      "step": 26184
+    },
+    {
+      "epoch": 0.5818888888888889,
+      "grad_norm": 1.624327540397644,
+      "learning_rate": 8.364080906868193e-05,
+      "loss": 1.9028,
+      "step": 26185
+    },
+    {
+      "epoch": 0.5819111111111112,
+      "grad_norm": 1.8741378784179688,
+      "learning_rate": 8.363636363636364e-05,
+      "loss": 2.2326,
+      "step": 26186
+    },
+    {
+      "epoch": 0.5819333333333333,
+      "grad_norm": 1.7637832164764404,
+      "learning_rate": 8.363191820404535e-05,
+      "loss": 1.9232,
+      "step": 26187
+    },
+    {
+      "epoch": 0.5819555555555556,
+      "grad_norm": 1.5867350101470947,
+      "learning_rate": 8.362747277172706e-05,
+      "loss": 1.7418,
+      "step": 26188
+    },
+    {
+      "epoch": 0.5819777777777778,
+      "grad_norm": 1.8549751043319702,
+      "learning_rate": 8.362302733940877e-05,
+      "loss": 2.3625,
+      "step": 26189
+    },
+    {
+      "epoch": 0.582,
+      "grad_norm": 1.6489076614379883,
+      "learning_rate": 8.361858190709046e-05,
+      "loss": 1.7887,
+      "step": 26190
+    },
+    {
+      "epoch": 0.5820222222222222,
+      "grad_norm": 1.6363847255706787,
+      "learning_rate": 8.361413647477217e-05,
+      "loss": 1.5879,
+      "step": 26191
+    },
+    {
+      "epoch": 0.5820444444444445,
+      "grad_norm": 1.8551013469696045,
+      "learning_rate": 8.360969104245388e-05,
+      "loss": 2.251,
+      "step": 26192
+    },
+    {
+      "epoch": 0.5820666666666666,
+      "grad_norm": 1.7150312662124634,
+      "learning_rate": 8.360524561013558e-05,
+      "loss": 1.7961,
+      "step": 26193
+    },
+    {
+      "epoch": 0.5820888888888889,
+      "grad_norm": 1.5141977071762085,
+      "learning_rate": 8.36008001778173e-05,
+      "loss": 1.3297,
+      "step": 26194
+    },
+    {
+      "epoch": 0.5821111111111111,
+      "grad_norm": 1.8479887247085571,
+      "learning_rate": 8.359635474549901e-05,
+      "loss": 1.8324,
+      "step": 26195
+    },
+    {
+      "epoch": 0.5821333333333333,
+      "grad_norm": 1.1966216564178467,
+      "learning_rate": 8.359190931318072e-05,
+      "loss": 0.838,
+      "step": 26196
+    },
+    {
+      "epoch": 0.5821555555555555,
+      "grad_norm": 1.891556978225708,
+      "learning_rate": 8.358746388086242e-05,
+      "loss": 2.0078,
+      "step": 26197
+    },
+    {
+      "epoch": 0.5821777777777778,
+      "grad_norm": 1.81918203830719,
+      "learning_rate": 8.358301844854413e-05,
+      "loss": 1.7798,
+      "step": 26198
+    },
+    {
+      "epoch": 0.5822,
+      "grad_norm": 1.9150341749191284,
+      "learning_rate": 8.357857301622584e-05,
+      "loss": 1.8632,
+      "step": 26199
+    },
+    {
+      "epoch": 0.5822222222222222,
+      "grad_norm": 1.4459576606750488,
+      "learning_rate": 8.357412758390753e-05,
+      "loss": 0.9049,
+      "step": 26200
+    },
+    {
+      "epoch": 0.5822444444444445,
+      "grad_norm": 1.3595621585845947,
+      "learning_rate": 8.356968215158924e-05,
+      "loss": 2.1102,
+      "step": 26201
+    },
+    {
+      "epoch": 0.5822666666666667,
+      "grad_norm": 1.6290918588638306,
+      "learning_rate": 8.356523671927095e-05,
+      "loss": 1.3813,
+      "step": 26202
+    },
+    {
+      "epoch": 0.5822888888888889,
+      "grad_norm": 1.4016199111938477,
+      "learning_rate": 8.356079128695266e-05,
+      "loss": 2.1448,
+      "step": 26203
+    },
+    {
+      "epoch": 0.5823111111111111,
+      "grad_norm": 1.6045475006103516,
+      "learning_rate": 8.355634585463437e-05,
+      "loss": 2.1319,
+      "step": 26204
+    },
+    {
+      "epoch": 0.5823333333333334,
+      "grad_norm": 1.6283369064331055,
+      "learning_rate": 8.355190042231608e-05,
+      "loss": 2.0471,
+      "step": 26205
+    },
+    {
+      "epoch": 0.5823555555555555,
+      "grad_norm": 1.6710686683654785,
+      "learning_rate": 8.354745498999779e-05,
+      "loss": 1.8345,
+      "step": 26206
+    },
+    {
+      "epoch": 0.5823777777777778,
+      "grad_norm": 1.1778653860092163,
+      "learning_rate": 8.354300955767949e-05,
+      "loss": 1.231,
+      "step": 26207
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 1.6818339824676514,
+      "learning_rate": 8.35385641253612e-05,
+      "loss": 2.5075,
+      "step": 26208
+    },
+    {
+      "epoch": 0.5824222222222222,
+      "grad_norm": 1.7084901332855225,
+      "learning_rate": 8.35341186930429e-05,
+      "loss": 2.2311,
+      "step": 26209
+    },
+    {
+      "epoch": 0.5824444444444444,
+      "grad_norm": 1.6797622442245483,
+      "learning_rate": 8.35296732607246e-05,
+      "loss": 2.0646,
+      "step": 26210
+    },
+    {
+      "epoch": 0.5824666666666667,
+      "grad_norm": 1.6726146936416626,
+      "learning_rate": 8.352522782840631e-05,
+      "loss": 1.8802,
+      "step": 26211
+    },
+    {
+      "epoch": 0.5824888888888888,
+      "grad_norm": 1.711694359779358,
+      "learning_rate": 8.352078239608802e-05,
+      "loss": 2.2033,
+      "step": 26212
+    },
+    {
+      "epoch": 0.5825111111111111,
+      "grad_norm": 1.5540680885314941,
+      "learning_rate": 8.351633696376973e-05,
+      "loss": 2.0245,
+      "step": 26213
+    },
+    {
+      "epoch": 0.5825333333333333,
+      "grad_norm": 2.045577049255371,
+      "learning_rate": 8.351189153145144e-05,
+      "loss": 2.1828,
+      "step": 26214
+    },
+    {
+      "epoch": 0.5825555555555556,
+      "grad_norm": 1.543917179107666,
+      "learning_rate": 8.350744609913315e-05,
+      "loss": 1.8969,
+      "step": 26215
+    },
+    {
+      "epoch": 0.5825777777777777,
+      "grad_norm": 1.5529899597167969,
+      "learning_rate": 8.350300066681486e-05,
+      "loss": 2.1522,
+      "step": 26216
+    },
+    {
+      "epoch": 0.5826,
+      "grad_norm": 1.8591609001159668,
+      "learning_rate": 8.349855523449656e-05,
+      "loss": 2.0281,
+      "step": 26217
+    },
+    {
+      "epoch": 0.5826222222222223,
+      "grad_norm": 1.2038625478744507,
+      "learning_rate": 8.349410980217827e-05,
+      "loss": 0.4999,
+      "step": 26218
+    },
+    {
+      "epoch": 0.5826444444444444,
+      "grad_norm": 1.7365057468414307,
+      "learning_rate": 8.348966436985998e-05,
+      "loss": 2.2447,
+      "step": 26219
+    },
+    {
+      "epoch": 0.5826666666666667,
+      "grad_norm": 1.4882011413574219,
+      "learning_rate": 8.348521893754167e-05,
+      "loss": 1.5969,
+      "step": 26220
+    },
+    {
+      "epoch": 0.5826888888888889,
+      "grad_norm": 1.4881402254104614,
+      "learning_rate": 8.348077350522338e-05,
+      "loss": 1.8631,
+      "step": 26221
+    },
+    {
+      "epoch": 0.5827111111111111,
+      "grad_norm": 1.7825554609298706,
+      "learning_rate": 8.347632807290509e-05,
+      "loss": 1.9985,
+      "step": 26222
+    },
+    {
+      "epoch": 0.5827333333333333,
+      "grad_norm": 1.7529854774475098,
+      "learning_rate": 8.34718826405868e-05,
+      "loss": 1.7588,
+      "step": 26223
+    },
+    {
+      "epoch": 0.5827555555555556,
+      "grad_norm": 1.7051653861999512,
+      "learning_rate": 8.346743720826851e-05,
+      "loss": 1.8225,
+      "step": 26224
+    },
+    {
+      "epoch": 0.5827777777777777,
+      "grad_norm": 2.4538986682891846,
+      "learning_rate": 8.346299177595022e-05,
+      "loss": 1.5975,
+      "step": 26225
+    },
+    {
+      "epoch": 0.5828,
+      "grad_norm": 1.4693490266799927,
+      "learning_rate": 8.345854634363193e-05,
+      "loss": 1.5349,
+      "step": 26226
+    },
+    {
+      "epoch": 0.5828222222222222,
+      "grad_norm": 2.286280632019043,
+      "learning_rate": 8.345410091131362e-05,
+      "loss": 2.2337,
+      "step": 26227
+    },
+    {
+      "epoch": 0.5828444444444445,
+      "grad_norm": 2.219813346862793,
+      "learning_rate": 8.344965547899533e-05,
+      "loss": 1.8498,
+      "step": 26228
+    },
+    {
+      "epoch": 0.5828666666666666,
+      "grad_norm": 1.8761414289474487,
+      "learning_rate": 8.344521004667704e-05,
+      "loss": 1.7299,
+      "step": 26229
+    },
+    {
+      "epoch": 0.5828888888888889,
+      "grad_norm": 2.0295374393463135,
+      "learning_rate": 8.344076461435874e-05,
+      "loss": 2.1673,
+      "step": 26230
+    },
+    {
+      "epoch": 0.5829111111111112,
+      "grad_norm": 1.9511631727218628,
+      "learning_rate": 8.343631918204046e-05,
+      "loss": 1.7042,
+      "step": 26231
+    },
+    {
+      "epoch": 0.5829333333333333,
+      "grad_norm": 1.9228847026824951,
+      "learning_rate": 8.343187374972217e-05,
+      "loss": 2.0077,
+      "step": 26232
+    },
+    {
+      "epoch": 0.5829555555555556,
+      "grad_norm": 1.6258912086486816,
+      "learning_rate": 8.342742831740387e-05,
+      "loss": 1.3767,
+      "step": 26233
+    },
+    {
+      "epoch": 0.5829777777777778,
+      "grad_norm": 2.137866497039795,
+      "learning_rate": 8.342298288508558e-05,
+      "loss": 1.6094,
+      "step": 26234
+    },
+    {
+      "epoch": 0.583,
+      "grad_norm": 2.144888162612915,
+      "learning_rate": 8.341853745276729e-05,
+      "loss": 1.6773,
+      "step": 26235
+    },
+    {
+      "epoch": 0.5830222222222222,
+      "grad_norm": 1.9768238067626953,
+      "learning_rate": 8.3414092020449e-05,
+      "loss": 1.8659,
+      "step": 26236
+    },
+    {
+      "epoch": 0.5830444444444445,
+      "grad_norm": 1.9847415685653687,
+      "learning_rate": 8.34096465881307e-05,
+      "loss": 2.4647,
+      "step": 26237
+    },
+    {
+      "epoch": 0.5830666666666666,
+      "grad_norm": 2.0302507877349854,
+      "learning_rate": 8.34052011558124e-05,
+      "loss": 2.0827,
+      "step": 26238
+    },
+    {
+      "epoch": 0.5830888888888889,
+      "grad_norm": 2.288281202316284,
+      "learning_rate": 8.340075572349411e-05,
+      "loss": 2.3895,
+      "step": 26239
+    },
+    {
+      "epoch": 0.5831111111111111,
+      "grad_norm": 2.0385663509368896,
+      "learning_rate": 8.339631029117582e-05,
+      "loss": 1.967,
+      "step": 26240
+    },
+    {
+      "epoch": 0.5831333333333333,
+      "grad_norm": 1.8273963928222656,
+      "learning_rate": 8.339186485885753e-05,
+      "loss": 1.7718,
+      "step": 26241
+    },
+    {
+      "epoch": 0.5831555555555555,
+      "grad_norm": 1.6473088264465332,
+      "learning_rate": 8.338741942653924e-05,
+      "loss": 1.8773,
+      "step": 26242
+    },
+    {
+      "epoch": 0.5831777777777778,
+      "grad_norm": 1.4469784498214722,
+      "learning_rate": 8.338297399422094e-05,
+      "loss": 1.0654,
+      "step": 26243
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 1.5022889375686646,
+      "learning_rate": 8.337852856190265e-05,
+      "loss": 1.5355,
+      "step": 26244
+    },
+    {
+      "epoch": 0.5832222222222222,
+      "grad_norm": 1.4862998723983765,
+      "learning_rate": 8.337408312958436e-05,
+      "loss": 1.4346,
+      "step": 26245
+    },
+    {
+      "epoch": 0.5832444444444445,
+      "grad_norm": 1.8965848684310913,
+      "learning_rate": 8.336963769726607e-05,
+      "loss": 1.6884,
+      "step": 26246
+    },
+    {
+      "epoch": 0.5832666666666667,
+      "grad_norm": 1.659008502960205,
+      "learning_rate": 8.336519226494776e-05,
+      "loss": 1.9439,
+      "step": 26247
+    },
+    {
+      "epoch": 0.5832888888888889,
+      "grad_norm": 1.8410717248916626,
+      "learning_rate": 8.336074683262947e-05,
+      "loss": 1.9104,
+      "step": 26248
+    },
+    {
+      "epoch": 0.5833111111111111,
+      "grad_norm": 1.6476742029190063,
+      "learning_rate": 8.335630140031118e-05,
+      "loss": 1.7227,
+      "step": 26249
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 2.0273349285125732,
+      "learning_rate": 8.335185596799289e-05,
+      "loss": 1.697,
+      "step": 26250
+    },
+    {
+      "epoch": 0.5833555555555555,
+      "grad_norm": 1.3179007768630981,
+      "learning_rate": 8.33474105356746e-05,
+      "loss": 1.6495,
+      "step": 26251
+    },
+    {
+      "epoch": 0.5833777777777778,
+      "grad_norm": 1.3309334516525269,
+      "learning_rate": 8.334296510335631e-05,
+      "loss": 2.062,
+      "step": 26252
+    },
+    {
+      "epoch": 0.5834,
+      "grad_norm": 0.9220927953720093,
+      "learning_rate": 8.333851967103802e-05,
+      "loss": 1.149,
+      "step": 26253
+    },
+    {
+      "epoch": 0.5834222222222222,
+      "grad_norm": 1.560506820678711,
+      "learning_rate": 8.333407423871972e-05,
+      "loss": 2.3106,
+      "step": 26254
+    },
+    {
+      "epoch": 0.5834444444444444,
+      "grad_norm": 1.5376288890838623,
+      "learning_rate": 8.332962880640143e-05,
+      "loss": 1.7624,
+      "step": 26255
+    },
+    {
+      "epoch": 0.5834666666666667,
+      "grad_norm": 1.712773084640503,
+      "learning_rate": 8.332518337408314e-05,
+      "loss": 1.9538,
+      "step": 26256
+    },
+    {
+      "epoch": 0.5834888888888888,
+      "grad_norm": 1.3960744142532349,
+      "learning_rate": 8.332073794176483e-05,
+      "loss": 1.8866,
+      "step": 26257
+    },
+    {
+      "epoch": 0.5835111111111111,
+      "grad_norm": 0.9395681023597717,
+      "learning_rate": 8.331629250944654e-05,
+      "loss": 0.9285,
+      "step": 26258
+    },
+    {
+      "epoch": 0.5835333333333333,
+      "grad_norm": 1.9319789409637451,
+      "learning_rate": 8.331184707712825e-05,
+      "loss": 2.5569,
+      "step": 26259
+    },
+    {
+      "epoch": 0.5835555555555556,
+      "grad_norm": 1.4398587942123413,
+      "learning_rate": 8.330740164480996e-05,
+      "loss": 2.1114,
+      "step": 26260
+    },
+    {
+      "epoch": 0.5835777777777778,
+      "grad_norm": 1.695851445198059,
+      "learning_rate": 8.330295621249167e-05,
+      "loss": 2.5833,
+      "step": 26261
+    },
+    {
+      "epoch": 0.5836,
+      "grad_norm": 1.396723747253418,
+      "learning_rate": 8.329851078017338e-05,
+      "loss": 1.66,
+      "step": 26262
+    },
+    {
+      "epoch": 0.5836222222222223,
+      "grad_norm": 1.5797001123428345,
+      "learning_rate": 8.329406534785509e-05,
+      "loss": 2.074,
+      "step": 26263
+    },
+    {
+      "epoch": 0.5836444444444444,
+      "grad_norm": 1.6313436031341553,
+      "learning_rate": 8.328961991553679e-05,
+      "loss": 2.5088,
+      "step": 26264
+    },
+    {
+      "epoch": 0.5836666666666667,
+      "grad_norm": 1.7688790559768677,
+      "learning_rate": 8.32851744832185e-05,
+      "loss": 1.7651,
+      "step": 26265
+    },
+    {
+      "epoch": 0.5836888888888889,
+      "grad_norm": 1.5532402992248535,
+      "learning_rate": 8.32807290509002e-05,
+      "loss": 2.171,
+      "step": 26266
+    },
+    {
+      "epoch": 0.5837111111111111,
+      "grad_norm": 1.6320881843566895,
+      "learning_rate": 8.32762836185819e-05,
+      "loss": 2.4216,
+      "step": 26267
+    },
+    {
+      "epoch": 0.5837333333333333,
+      "grad_norm": 1.4460666179656982,
+      "learning_rate": 8.327183818626362e-05,
+      "loss": 1.8043,
+      "step": 26268
+    },
+    {
+      "epoch": 0.5837555555555556,
+      "grad_norm": 1.8040688037872314,
+      "learning_rate": 8.326739275394533e-05,
+      "loss": 1.9625,
+      "step": 26269
+    },
+    {
+      "epoch": 0.5837777777777777,
+      "grad_norm": 1.0893901586532593,
+      "learning_rate": 8.326294732162703e-05,
+      "loss": 0.9989,
+      "step": 26270
+    },
+    {
+      "epoch": 0.5838,
+      "grad_norm": 1.9131560325622559,
+      "learning_rate": 8.325850188930874e-05,
+      "loss": 2.5563,
+      "step": 26271
+    },
+    {
+      "epoch": 0.5838222222222222,
+      "grad_norm": 1.5659085512161255,
+      "learning_rate": 8.325405645699045e-05,
+      "loss": 2.1676,
+      "step": 26272
+    },
+    {
+      "epoch": 0.5838444444444445,
+      "grad_norm": 1.5888844728469849,
+      "learning_rate": 8.324961102467216e-05,
+      "loss": 2.1525,
+      "step": 26273
+    },
+    {
+      "epoch": 0.5838666666666666,
+      "grad_norm": 1.691839575767517,
+      "learning_rate": 8.324516559235385e-05,
+      "loss": 1.9559,
+      "step": 26274
+    },
+    {
+      "epoch": 0.5838888888888889,
+      "grad_norm": 1.625472903251648,
+      "learning_rate": 8.324072016003556e-05,
+      "loss": 2.118,
+      "step": 26275
+    },
+    {
+      "epoch": 0.5839111111111112,
+      "grad_norm": 1.6146228313446045,
+      "learning_rate": 8.323627472771727e-05,
+      "loss": 1.9359,
+      "step": 26276
+    },
+    {
+      "epoch": 0.5839333333333333,
+      "grad_norm": 1.5743327140808105,
+      "learning_rate": 8.323182929539898e-05,
+      "loss": 1.9866,
+      "step": 26277
+    },
+    {
+      "epoch": 0.5839555555555556,
+      "grad_norm": 1.5054017305374146,
+      "learning_rate": 8.32273838630807e-05,
+      "loss": 1.2454,
+      "step": 26278
+    },
+    {
+      "epoch": 0.5839777777777778,
+      "grad_norm": 1.793137788772583,
+      "learning_rate": 8.32229384307624e-05,
+      "loss": 1.9548,
+      "step": 26279
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 1.9058033227920532,
+      "learning_rate": 8.32184929984441e-05,
+      "loss": 2.566,
+      "step": 26280
+    },
+    {
+      "epoch": 0.5840222222222222,
+      "grad_norm": 1.8583147525787354,
+      "learning_rate": 8.321404756612581e-05,
+      "loss": 2.0726,
+      "step": 26281
+    },
+    {
+      "epoch": 0.5840444444444445,
+      "grad_norm": 1.6046953201293945,
+      "learning_rate": 8.320960213380752e-05,
+      "loss": 2.0069,
+      "step": 26282
+    },
+    {
+      "epoch": 0.5840666666666666,
+      "grad_norm": 1.7175462245941162,
+      "learning_rate": 8.320515670148923e-05,
+      "loss": 1.9083,
+      "step": 26283
+    },
+    {
+      "epoch": 0.5840888888888889,
+      "grad_norm": 1.4772974252700806,
+      "learning_rate": 8.320071126917092e-05,
+      "loss": 1.239,
+      "step": 26284
+    },
+    {
+      "epoch": 0.5841111111111111,
+      "grad_norm": 1.5468647480010986,
+      "learning_rate": 8.319626583685263e-05,
+      "loss": 1.6306,
+      "step": 26285
+    },
+    {
+      "epoch": 0.5841333333333333,
+      "grad_norm": 1.430153489112854,
+      "learning_rate": 8.319182040453434e-05,
+      "loss": 1.7573,
+      "step": 26286
+    },
+    {
+      "epoch": 0.5841555555555555,
+      "grad_norm": 2.067629098892212,
+      "learning_rate": 8.318737497221605e-05,
+      "loss": 2.143,
+      "step": 26287
+    },
+    {
+      "epoch": 0.5841777777777778,
+      "grad_norm": 1.7656430006027222,
+      "learning_rate": 8.318292953989776e-05,
+      "loss": 2.0896,
+      "step": 26288
+    },
+    {
+      "epoch": 0.5842,
+      "grad_norm": 1.560054063796997,
+      "learning_rate": 8.317848410757947e-05,
+      "loss": 1.7963,
+      "step": 26289
+    },
+    {
+      "epoch": 0.5842222222222222,
+      "grad_norm": 1.8045637607574463,
+      "learning_rate": 8.317403867526117e-05,
+      "loss": 1.7165,
+      "step": 26290
+    },
+    {
+      "epoch": 0.5842444444444445,
+      "grad_norm": 1.740210771560669,
+      "learning_rate": 8.316959324294288e-05,
+      "loss": 2.3107,
+      "step": 26291
+    },
+    {
+      "epoch": 0.5842666666666667,
+      "grad_norm": 1.606910228729248,
+      "learning_rate": 8.316514781062459e-05,
+      "loss": 1.9242,
+      "step": 26292
+    },
+    {
+      "epoch": 0.5842888888888889,
+      "grad_norm": 1.793172836303711,
+      "learning_rate": 8.31607023783063e-05,
+      "loss": 1.7542,
+      "step": 26293
+    },
+    {
+      "epoch": 0.5843111111111111,
+      "grad_norm": 1.8749566078186035,
+      "learning_rate": 8.315625694598799e-05,
+      "loss": 2.1248,
+      "step": 26294
+    },
+    {
+      "epoch": 0.5843333333333334,
+      "grad_norm": 2.0712335109710693,
+      "learning_rate": 8.31518115136697e-05,
+      "loss": 1.9189,
+      "step": 26295
+    },
+    {
+      "epoch": 0.5843555555555555,
+      "grad_norm": 1.778473973274231,
+      "learning_rate": 8.314736608135141e-05,
+      "loss": 1.6249,
+      "step": 26296
+    },
+    {
+      "epoch": 0.5843777777777778,
+      "grad_norm": 1.758664846420288,
+      "learning_rate": 8.314292064903312e-05,
+      "loss": 1.7072,
+      "step": 26297
+    },
+    {
+      "epoch": 0.5844,
+      "grad_norm": 1.531131386756897,
+      "learning_rate": 8.313847521671483e-05,
+      "loss": 1.216,
+      "step": 26298
+    },
+    {
+      "epoch": 0.5844222222222222,
+      "grad_norm": 2.2430763244628906,
+      "learning_rate": 8.313402978439654e-05,
+      "loss": 2.2344,
+      "step": 26299
+    },
+    {
+      "epoch": 0.5844444444444444,
+      "grad_norm": 1.7089288234710693,
+      "learning_rate": 8.312958435207825e-05,
+      "loss": 1.3861,
+      "step": 26300
+    },
+    {
+      "epoch": 0.5844666666666667,
+      "grad_norm": 1.5989187955856323,
+      "learning_rate": 8.312513891975995e-05,
+      "loss": 2.6309,
+      "step": 26301
+    },
+    {
+      "epoch": 0.5844888888888888,
+      "grad_norm": 1.3423744440078735,
+      "learning_rate": 8.312069348744166e-05,
+      "loss": 1.9436,
+      "step": 26302
+    },
+    {
+      "epoch": 0.5845111111111111,
+      "grad_norm": 1.9386184215545654,
+      "learning_rate": 8.311624805512337e-05,
+      "loss": 1.9313,
+      "step": 26303
+    },
+    {
+      "epoch": 0.5845333333333333,
+      "grad_norm": 1.4838286638259888,
+      "learning_rate": 8.311180262280506e-05,
+      "loss": 1.6181,
+      "step": 26304
+    },
+    {
+      "epoch": 0.5845555555555556,
+      "grad_norm": 1.5339857339859009,
+      "learning_rate": 8.310735719048679e-05,
+      "loss": 1.8669,
+      "step": 26305
+    },
+    {
+      "epoch": 0.5845777777777778,
+      "grad_norm": 1.5780014991760254,
+      "learning_rate": 8.31029117581685e-05,
+      "loss": 1.8785,
+      "step": 26306
+    },
+    {
+      "epoch": 0.5846,
+      "grad_norm": 2.025686502456665,
+      "learning_rate": 8.309846632585019e-05,
+      "loss": 1.478,
+      "step": 26307
+    },
+    {
+      "epoch": 0.5846222222222223,
+      "grad_norm": 1.8832579851150513,
+      "learning_rate": 8.30940208935319e-05,
+      "loss": 2.0248,
+      "step": 26308
+    },
+    {
+      "epoch": 0.5846444444444444,
+      "grad_norm": 1.3615474700927734,
+      "learning_rate": 8.308957546121361e-05,
+      "loss": 1.3794,
+      "step": 26309
+    },
+    {
+      "epoch": 0.5846666666666667,
+      "grad_norm": 1.5770617723464966,
+      "learning_rate": 8.308513002889532e-05,
+      "loss": 1.5432,
+      "step": 26310
+    },
+    {
+      "epoch": 0.5846888888888889,
+      "grad_norm": 1.994457721710205,
+      "learning_rate": 8.308068459657702e-05,
+      "loss": 2.5888,
+      "step": 26311
+    },
+    {
+      "epoch": 0.5847111111111111,
+      "grad_norm": 1.5343011617660522,
+      "learning_rate": 8.307623916425873e-05,
+      "loss": 1.5649,
+      "step": 26312
+    },
+    {
+      "epoch": 0.5847333333333333,
+      "grad_norm": 1.5088410377502441,
+      "learning_rate": 8.307179373194044e-05,
+      "loss": 1.9576,
+      "step": 26313
+    },
+    {
+      "epoch": 0.5847555555555556,
+      "grad_norm": 1.646855354309082,
+      "learning_rate": 8.306734829962214e-05,
+      "loss": 1.722,
+      "step": 26314
+    },
+    {
+      "epoch": 0.5847777777777777,
+      "grad_norm": 1.5722808837890625,
+      "learning_rate": 8.306290286730385e-05,
+      "loss": 1.6116,
+      "step": 26315
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 1.5221266746520996,
+      "learning_rate": 8.305845743498556e-05,
+      "loss": 1.7647,
+      "step": 26316
+    },
+    {
+      "epoch": 0.5848222222222222,
+      "grad_norm": 1.6071947813034058,
+      "learning_rate": 8.305401200266726e-05,
+      "loss": 2.1635,
+      "step": 26317
+    },
+    {
+      "epoch": 0.5848444444444444,
+      "grad_norm": 1.9324707984924316,
+      "learning_rate": 8.304956657034897e-05,
+      "loss": 2.3949,
+      "step": 26318
+    },
+    {
+      "epoch": 0.5848666666666666,
+      "grad_norm": 1.6427240371704102,
+      "learning_rate": 8.304512113803068e-05,
+      "loss": 1.7098,
+      "step": 26319
+    },
+    {
+      "epoch": 0.5848888888888889,
+      "grad_norm": 1.8649818897247314,
+      "learning_rate": 8.304067570571239e-05,
+      "loss": 2.0969,
+      "step": 26320
+    },
+    {
+      "epoch": 0.5849111111111112,
+      "grad_norm": 1.9790477752685547,
+      "learning_rate": 8.303623027339409e-05,
+      "loss": 1.9604,
+      "step": 26321
+    },
+    {
+      "epoch": 0.5849333333333333,
+      "grad_norm": 1.46359384059906,
+      "learning_rate": 8.30317848410758e-05,
+      "loss": 1.4153,
+      "step": 26322
+    },
+    {
+      "epoch": 0.5849555555555556,
+      "grad_norm": 1.8081319332122803,
+      "learning_rate": 8.30273394087575e-05,
+      "loss": 1.932,
+      "step": 26323
+    },
+    {
+      "epoch": 0.5849777777777778,
+      "grad_norm": 1.7537386417388916,
+      "learning_rate": 8.302289397643921e-05,
+      "loss": 2.3716,
+      "step": 26324
+    },
+    {
+      "epoch": 0.585,
+      "grad_norm": 1.5806208848953247,
+      "learning_rate": 8.301844854412092e-05,
+      "loss": 1.8761,
+      "step": 26325
+    },
+    {
+      "epoch": 0.5850222222222222,
+      "grad_norm": 1.8859691619873047,
+      "learning_rate": 8.301400311180263e-05,
+      "loss": 1.997,
+      "step": 26326
+    },
+    {
+      "epoch": 0.5850444444444445,
+      "grad_norm": 1.6441972255706787,
+      "learning_rate": 8.300955767948433e-05,
+      "loss": 1.5677,
+      "step": 26327
+    },
+    {
+      "epoch": 0.5850666666666666,
+      "grad_norm": 1.6208170652389526,
+      "learning_rate": 8.300511224716604e-05,
+      "loss": 1.8993,
+      "step": 26328
+    },
+    {
+      "epoch": 0.5850888888888889,
+      "grad_norm": 1.696751356124878,
+      "learning_rate": 8.300066681484775e-05,
+      "loss": 2.3392,
+      "step": 26329
+    },
+    {
+      "epoch": 0.5851111111111111,
+      "grad_norm": 1.7575993537902832,
+      "learning_rate": 8.299622138252946e-05,
+      "loss": 2.1465,
+      "step": 26330
+    },
+    {
+      "epoch": 0.5851333333333333,
+      "grad_norm": 1.0533134937286377,
+      "learning_rate": 8.299177595021115e-05,
+      "loss": 0.8847,
+      "step": 26331
+    },
+    {
+      "epoch": 0.5851555555555555,
+      "grad_norm": 1.5377509593963623,
+      "learning_rate": 8.298733051789286e-05,
+      "loss": 1.452,
+      "step": 26332
+    },
+    {
+      "epoch": 0.5851777777777778,
+      "grad_norm": 1.8691941499710083,
+      "learning_rate": 8.298288508557457e-05,
+      "loss": 1.5431,
+      "step": 26333
+    },
+    {
+      "epoch": 0.5852,
+      "grad_norm": 1.8643362522125244,
+      "learning_rate": 8.297843965325628e-05,
+      "loss": 1.906,
+      "step": 26334
+    },
+    {
+      "epoch": 0.5852222222222222,
+      "grad_norm": 1.832690715789795,
+      "learning_rate": 8.297399422093799e-05,
+      "loss": 1.5813,
+      "step": 26335
+    },
+    {
+      "epoch": 0.5852444444444445,
+      "grad_norm": 1.643566608428955,
+      "learning_rate": 8.29695487886197e-05,
+      "loss": 1.9578,
+      "step": 26336
+    },
+    {
+      "epoch": 0.5852666666666667,
+      "grad_norm": 1.567914605140686,
+      "learning_rate": 8.29651033563014e-05,
+      "loss": 1.4821,
+      "step": 26337
+    },
+    {
+      "epoch": 0.5852888888888889,
+      "grad_norm": 1.438464879989624,
+      "learning_rate": 8.296065792398311e-05,
+      "loss": 1.47,
+      "step": 26338
+    },
+    {
+      "epoch": 0.5853111111111111,
+      "grad_norm": 1.6269550323486328,
+      "learning_rate": 8.295621249166482e-05,
+      "loss": 2.0255,
+      "step": 26339
+    },
+    {
+      "epoch": 0.5853333333333334,
+      "grad_norm": 1.3445321321487427,
+      "learning_rate": 8.295176705934653e-05,
+      "loss": 1.2979,
+      "step": 26340
+    },
+    {
+      "epoch": 0.5853555555555555,
+      "grad_norm": 1.9354634284973145,
+      "learning_rate": 8.294732162702822e-05,
+      "loss": 2.2849,
+      "step": 26341
+    },
+    {
+      "epoch": 0.5853777777777778,
+      "grad_norm": 1.5181694030761719,
+      "learning_rate": 8.294287619470995e-05,
+      "loss": 1.4888,
+      "step": 26342
+    },
+    {
+      "epoch": 0.5854,
+      "grad_norm": 1.9285415410995483,
+      "learning_rate": 8.293843076239166e-05,
+      "loss": 1.7749,
+      "step": 26343
+    },
+    {
+      "epoch": 0.5854222222222222,
+      "grad_norm": 2.097766160964966,
+      "learning_rate": 8.293398533007335e-05,
+      "loss": 2.3797,
+      "step": 26344
+    },
+    {
+      "epoch": 0.5854444444444444,
+      "grad_norm": 1.821632742881775,
+      "learning_rate": 8.292953989775506e-05,
+      "loss": 1.6612,
+      "step": 26345
+    },
+    {
+      "epoch": 0.5854666666666667,
+      "grad_norm": 1.9774261713027954,
+      "learning_rate": 8.292509446543677e-05,
+      "loss": 1.6549,
+      "step": 26346
+    },
+    {
+      "epoch": 0.5854888888888888,
+      "grad_norm": 1.726208209991455,
+      "learning_rate": 8.292064903311847e-05,
+      "loss": 2.0537,
+      "step": 26347
+    },
+    {
+      "epoch": 0.5855111111111111,
+      "grad_norm": 1.1972006559371948,
+      "learning_rate": 8.291620360080018e-05,
+      "loss": 0.6462,
+      "step": 26348
+    },
+    {
+      "epoch": 0.5855333333333334,
+      "grad_norm": 0.35262811183929443,
+      "learning_rate": 8.291175816848189e-05,
+      "loss": 0.0556,
+      "step": 26349
+    },
+    {
+      "epoch": 0.5855555555555556,
+      "grad_norm": 1.5765961408615112,
+      "learning_rate": 8.29073127361636e-05,
+      "loss": 0.817,
+      "step": 26350
+    },
+    {
+      "epoch": 0.5855777777777778,
+      "grad_norm": 1.6506803035736084,
+      "learning_rate": 8.29028673038453e-05,
+      "loss": 2.6258,
+      "step": 26351
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 1.4478061199188232,
+      "learning_rate": 8.289842187152702e-05,
+      "loss": 1.9368,
+      "step": 26352
+    },
+    {
+      "epoch": 0.5856222222222223,
+      "grad_norm": 1.4829243421554565,
+      "learning_rate": 8.289397643920873e-05,
+      "loss": 1.2137,
+      "step": 26353
+    },
+    {
+      "epoch": 0.5856444444444444,
+      "grad_norm": 1.551202654838562,
+      "learning_rate": 8.288953100689042e-05,
+      "loss": 2.54,
+      "step": 26354
+    },
+    {
+      "epoch": 0.5856666666666667,
+      "grad_norm": 1.680566430091858,
+      "learning_rate": 8.288508557457213e-05,
+      "loss": 2.3305,
+      "step": 26355
+    },
+    {
+      "epoch": 0.5856888888888889,
+      "grad_norm": 1.6335614919662476,
+      "learning_rate": 8.288064014225384e-05,
+      "loss": 1.6486,
+      "step": 26356
+    },
+    {
+      "epoch": 0.5857111111111111,
+      "grad_norm": 1.60475492477417,
+      "learning_rate": 8.287619470993555e-05,
+      "loss": 2.0379,
+      "step": 26357
+    },
+    {
+      "epoch": 0.5857333333333333,
+      "grad_norm": 1.900649905204773,
+      "learning_rate": 8.287174927761725e-05,
+      "loss": 2.5529,
+      "step": 26358
+    },
+    {
+      "epoch": 0.5857555555555556,
+      "grad_norm": 1.625902533531189,
+      "learning_rate": 8.286730384529896e-05,
+      "loss": 2.1112,
+      "step": 26359
+    },
+    {
+      "epoch": 0.5857777777777777,
+      "grad_norm": 1.5074518918991089,
+      "learning_rate": 8.286285841298067e-05,
+      "loss": 2.1067,
+      "step": 26360
+    },
+    {
+      "epoch": 0.5858,
+      "grad_norm": 1.585428237915039,
+      "learning_rate": 8.285841298066237e-05,
+      "loss": 1.8526,
+      "step": 26361
+    },
+    {
+      "epoch": 0.5858222222222222,
+      "grad_norm": 1.7944355010986328,
+      "learning_rate": 8.285396754834408e-05,
+      "loss": 2.4323,
+      "step": 26362
+    },
+    {
+      "epoch": 0.5858444444444444,
+      "grad_norm": 1.4390625953674316,
+      "learning_rate": 8.28495221160258e-05,
+      "loss": 1.7495,
+      "step": 26363
+    },
+    {
+      "epoch": 0.5858666666666666,
+      "grad_norm": 1.7868386507034302,
+      "learning_rate": 8.284507668370749e-05,
+      "loss": 1.6837,
+      "step": 26364
+    },
+    {
+      "epoch": 0.5858888888888889,
+      "grad_norm": 1.818138837814331,
+      "learning_rate": 8.28406312513892e-05,
+      "loss": 1.7053,
+      "step": 26365
+    },
+    {
+      "epoch": 0.5859111111111112,
+      "grad_norm": 2.056191921234131,
+      "learning_rate": 8.283618581907091e-05,
+      "loss": 2.0468,
+      "step": 26366
+    },
+    {
+      "epoch": 0.5859333333333333,
+      "grad_norm": 1.6591705083847046,
+      "learning_rate": 8.283174038675262e-05,
+      "loss": 2.3547,
+      "step": 26367
+    },
+    {
+      "epoch": 0.5859555555555556,
+      "grad_norm": 1.5027168989181519,
+      "learning_rate": 8.282729495443432e-05,
+      "loss": 1.9793,
+      "step": 26368
+    },
+    {
+      "epoch": 0.5859777777777778,
+      "grad_norm": 1.5128077268600464,
+      "learning_rate": 8.282284952211602e-05,
+      "loss": 1.5506,
+      "step": 26369
+    },
+    {
+      "epoch": 0.586,
+      "grad_norm": 1.9915423393249512,
+      "learning_rate": 8.281840408979773e-05,
+      "loss": 2.1444,
+      "step": 26370
+    },
+    {
+      "epoch": 0.5860222222222222,
+      "grad_norm": 1.9194926023483276,
+      "learning_rate": 8.281395865747944e-05,
+      "loss": 2.1724,
+      "step": 26371
+    },
+    {
+      "epoch": 0.5860444444444445,
+      "grad_norm": 2.282923936843872,
+      "learning_rate": 8.280951322516115e-05,
+      "loss": 2.2116,
+      "step": 26372
+    },
+    {
+      "epoch": 0.5860666666666666,
+      "grad_norm": 1.53107488155365,
+      "learning_rate": 8.280506779284286e-05,
+      "loss": 2.062,
+      "step": 26373
+    },
+    {
+      "epoch": 0.5860888888888889,
+      "grad_norm": 1.673590898513794,
+      "learning_rate": 8.280062236052456e-05,
+      "loss": 1.5491,
+      "step": 26374
+    },
+    {
+      "epoch": 0.5861111111111111,
+      "grad_norm": 2.9730846881866455,
+      "learning_rate": 8.279617692820627e-05,
+      "loss": 1.2786,
+      "step": 26375
+    },
+    {
+      "epoch": 0.5861333333333333,
+      "grad_norm": 1.4333029985427856,
+      "learning_rate": 8.279173149588798e-05,
+      "loss": 1.6072,
+      "step": 26376
+    },
+    {
+      "epoch": 0.5861555555555555,
+      "grad_norm": 1.7236535549163818,
+      "learning_rate": 8.278728606356969e-05,
+      "loss": 1.5908,
+      "step": 26377
+    },
+    {
+      "epoch": 0.5861777777777778,
+      "grad_norm": 1.8024083375930786,
+      "learning_rate": 8.278284063125138e-05,
+      "loss": 2.0472,
+      "step": 26378
+    },
+    {
+      "epoch": 0.5862,
+      "grad_norm": 1.3262346982955933,
+      "learning_rate": 8.277839519893311e-05,
+      "loss": 1.7635,
+      "step": 26379
+    },
+    {
+      "epoch": 0.5862222222222222,
+      "grad_norm": 1.9368542432785034,
+      "learning_rate": 8.277394976661482e-05,
+      "loss": 2.1713,
+      "step": 26380
+    },
+    {
+      "epoch": 0.5862444444444445,
+      "grad_norm": 1.776813268661499,
+      "learning_rate": 8.276950433429651e-05,
+      "loss": 2.2592,
+      "step": 26381
+    },
+    {
+      "epoch": 0.5862666666666667,
+      "grad_norm": 2.261319160461426,
+      "learning_rate": 8.276505890197822e-05,
+      "loss": 2.4149,
+      "step": 26382
+    },
+    {
+      "epoch": 0.5862888888888889,
+      "grad_norm": 1.633979320526123,
+      "learning_rate": 8.276061346965993e-05,
+      "loss": 1.6741,
+      "step": 26383
+    },
+    {
+      "epoch": 0.5863111111111111,
+      "grad_norm": 1.6456549167633057,
+      "learning_rate": 8.275616803734163e-05,
+      "loss": 2.0613,
+      "step": 26384
+    },
+    {
+      "epoch": 0.5863333333333334,
+      "grad_norm": 1.6004148721694946,
+      "learning_rate": 8.275172260502334e-05,
+      "loss": 1.8647,
+      "step": 26385
+    },
+    {
+      "epoch": 0.5863555555555555,
+      "grad_norm": 1.6349995136260986,
+      "learning_rate": 8.274727717270505e-05,
+      "loss": 1.946,
+      "step": 26386
+    },
+    {
+      "epoch": 0.5863777777777778,
+      "grad_norm": 1.856724500656128,
+      "learning_rate": 8.274283174038676e-05,
+      "loss": 2.0633,
+      "step": 26387
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 1.818117618560791,
+      "learning_rate": 8.273838630806847e-05,
+      "loss": 2.2893,
+      "step": 26388
+    },
+    {
+      "epoch": 0.5864222222222222,
+      "grad_norm": 1.3961563110351562,
+      "learning_rate": 8.273394087575018e-05,
+      "loss": 1.608,
+      "step": 26389
+    },
+    {
+      "epoch": 0.5864444444444444,
+      "grad_norm": 1.90252685546875,
+      "learning_rate": 8.272949544343189e-05,
+      "loss": 2.1006,
+      "step": 26390
+    },
+    {
+      "epoch": 0.5864666666666667,
+      "grad_norm": 1.7041651010513306,
+      "learning_rate": 8.272505001111358e-05,
+      "loss": 1.7053,
+      "step": 26391
+    },
+    {
+      "epoch": 0.5864888888888888,
+      "grad_norm": 1.704955816268921,
+      "learning_rate": 8.272060457879529e-05,
+      "loss": 1.8423,
+      "step": 26392
+    },
+    {
+      "epoch": 0.5865111111111111,
+      "grad_norm": 2.031508445739746,
+      "learning_rate": 8.2716159146477e-05,
+      "loss": 1.7539,
+      "step": 26393
+    },
+    {
+      "epoch": 0.5865333333333334,
+      "grad_norm": 2.1782689094543457,
+      "learning_rate": 8.27117137141587e-05,
+      "loss": 1.8534,
+      "step": 26394
+    },
+    {
+      "epoch": 0.5865555555555556,
+      "grad_norm": 1.6237221956253052,
+      "learning_rate": 8.270726828184041e-05,
+      "loss": 1.8039,
+      "step": 26395
+    },
+    {
+      "epoch": 0.5865777777777778,
+      "grad_norm": 1.854526162147522,
+      "learning_rate": 8.270282284952212e-05,
+      "loss": 1.671,
+      "step": 26396
+    },
+    {
+      "epoch": 0.5866,
+      "grad_norm": 1.7956544160842896,
+      "learning_rate": 8.269837741720383e-05,
+      "loss": 1.6,
+      "step": 26397
+    },
+    {
+      "epoch": 0.5866222222222223,
+      "grad_norm": 2.6054320335388184,
+      "learning_rate": 8.269393198488554e-05,
+      "loss": 1.3538,
+      "step": 26398
+    },
+    {
+      "epoch": 0.5866444444444444,
+      "grad_norm": 1.5763986110687256,
+      "learning_rate": 8.268948655256725e-05,
+      "loss": 1.6026,
+      "step": 26399
+    },
+    {
+      "epoch": 0.5866666666666667,
+      "grad_norm": 1.8800048828125,
+      "learning_rate": 8.268504112024896e-05,
+      "loss": 1.4492,
+      "step": 26400
+    },
+    {
+      "epoch": 0.5866888888888889,
+      "grad_norm": 1.9422786235809326,
+      "learning_rate": 8.268059568793065e-05,
+      "loss": 2.3528,
+      "step": 26401
+    },
+    {
+      "epoch": 0.5867111111111111,
+      "grad_norm": 1.5031019449234009,
+      "learning_rate": 8.267615025561236e-05,
+      "loss": 2.304,
+      "step": 26402
+    },
+    {
+      "epoch": 0.5867333333333333,
+      "grad_norm": 1.8519642353057861,
+      "learning_rate": 8.267170482329407e-05,
+      "loss": 2.9398,
+      "step": 26403
+    },
+    {
+      "epoch": 0.5867555555555556,
+      "grad_norm": 1.6615465879440308,
+      "learning_rate": 8.266725939097577e-05,
+      "loss": 2.2684,
+      "step": 26404
+    },
+    {
+      "epoch": 0.5867777777777777,
+      "grad_norm": 1.7758432626724243,
+      "learning_rate": 8.266281395865748e-05,
+      "loss": 2.2158,
+      "step": 26405
+    },
+    {
+      "epoch": 0.5868,
+      "grad_norm": 1.4606879949569702,
+      "learning_rate": 8.265836852633919e-05,
+      "loss": 1.6692,
+      "step": 26406
+    },
+    {
+      "epoch": 0.5868222222222222,
+      "grad_norm": 1.423619270324707,
+      "learning_rate": 8.26539230940209e-05,
+      "loss": 2.2713,
+      "step": 26407
+    },
+    {
+      "epoch": 0.5868444444444444,
+      "grad_norm": 1.7291810512542725,
+      "learning_rate": 8.26494776617026e-05,
+      "loss": 2.4045,
+      "step": 26408
+    },
+    {
+      "epoch": 0.5868666666666666,
+      "grad_norm": 1.4429683685302734,
+      "learning_rate": 8.264503222938431e-05,
+      "loss": 1.9806,
+      "step": 26409
+    },
+    {
+      "epoch": 0.5868888888888889,
+      "grad_norm": 1.526571273803711,
+      "learning_rate": 8.264058679706602e-05,
+      "loss": 2.2861,
+      "step": 26410
+    },
+    {
+      "epoch": 0.5869111111111112,
+      "grad_norm": 1.5247364044189453,
+      "learning_rate": 8.263614136474772e-05,
+      "loss": 1.6332,
+      "step": 26411
+    },
+    {
+      "epoch": 0.5869333333333333,
+      "grad_norm": 1.7640910148620605,
+      "learning_rate": 8.263169593242943e-05,
+      "loss": 2.6262,
+      "step": 26412
+    },
+    {
+      "epoch": 0.5869555555555556,
+      "grad_norm": 1.8370826244354248,
+      "learning_rate": 8.262725050011114e-05,
+      "loss": 1.9565,
+      "step": 26413
+    },
+    {
+      "epoch": 0.5869777777777778,
+      "grad_norm": 1.6487253904342651,
+      "learning_rate": 8.262280506779285e-05,
+      "loss": 2.071,
+      "step": 26414
+    },
+    {
+      "epoch": 0.587,
+      "grad_norm": 1.5205276012420654,
+      "learning_rate": 8.261835963547455e-05,
+      "loss": 1.8026,
+      "step": 26415
+    },
+    {
+      "epoch": 0.5870222222222222,
+      "grad_norm": 1.7748128175735474,
+      "learning_rate": 8.261391420315627e-05,
+      "loss": 1.8529,
+      "step": 26416
+    },
+    {
+      "epoch": 0.5870444444444445,
+      "grad_norm": 1.5548124313354492,
+      "learning_rate": 8.260946877083798e-05,
+      "loss": 1.9147,
+      "step": 26417
+    },
+    {
+      "epoch": 0.5870666666666666,
+      "grad_norm": 1.6629937887191772,
+      "learning_rate": 8.260502333851967e-05,
+      "loss": 2.1774,
+      "step": 26418
+    },
+    {
+      "epoch": 0.5870888888888889,
+      "grad_norm": 1.7149839401245117,
+      "learning_rate": 8.260057790620138e-05,
+      "loss": 2.023,
+      "step": 26419
+    },
+    {
+      "epoch": 0.5871111111111111,
+      "grad_norm": 1.7205604314804077,
+      "learning_rate": 8.25961324738831e-05,
+      "loss": 2.1347,
+      "step": 26420
+    },
+    {
+      "epoch": 0.5871333333333333,
+      "grad_norm": 1.8264589309692383,
+      "learning_rate": 8.259168704156479e-05,
+      "loss": 2.0622,
+      "step": 26421
+    },
+    {
+      "epoch": 0.5871555555555555,
+      "grad_norm": 1.5748111009597778,
+      "learning_rate": 8.25872416092465e-05,
+      "loss": 1.7845,
+      "step": 26422
+    },
+    {
+      "epoch": 0.5871777777777778,
+      "grad_norm": 2.0303964614868164,
+      "learning_rate": 8.258279617692821e-05,
+      "loss": 2.1154,
+      "step": 26423
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 1.772826910018921,
+      "learning_rate": 8.257835074460992e-05,
+      "loss": 1.9938,
+      "step": 26424
+    },
+    {
+      "epoch": 0.5872222222222222,
+      "grad_norm": 1.7206312417984009,
+      "learning_rate": 8.257390531229163e-05,
+      "loss": 2.008,
+      "step": 26425
+    },
+    {
+      "epoch": 0.5872444444444445,
+      "grad_norm": 1.6457266807556152,
+      "learning_rate": 8.256945987997334e-05,
+      "loss": 2.0749,
+      "step": 26426
+    },
+    {
+      "epoch": 0.5872666666666667,
+      "grad_norm": 1.5992521047592163,
+      "learning_rate": 8.256501444765505e-05,
+      "loss": 2.2439,
+      "step": 26427
+    },
+    {
+      "epoch": 0.5872888888888889,
+      "grad_norm": 1.3555350303649902,
+      "learning_rate": 8.256056901533674e-05,
+      "loss": 1.2614,
+      "step": 26428
+    },
+    {
+      "epoch": 0.5873111111111111,
+      "grad_norm": 1.5201590061187744,
+      "learning_rate": 8.255612358301845e-05,
+      "loss": 1.8837,
+      "step": 26429
+    },
+    {
+      "epoch": 0.5873333333333334,
+      "grad_norm": 1.8117382526397705,
+      "learning_rate": 8.255167815070016e-05,
+      "loss": 2.3655,
+      "step": 26430
+    },
+    {
+      "epoch": 0.5873555555555555,
+      "grad_norm": 1.8975117206573486,
+      "learning_rate": 8.254723271838186e-05,
+      "loss": 2.1152,
+      "step": 26431
+    },
+    {
+      "epoch": 0.5873777777777778,
+      "grad_norm": 1.8291747570037842,
+      "learning_rate": 8.254278728606357e-05,
+      "loss": 1.8039,
+      "step": 26432
+    },
+    {
+      "epoch": 0.5874,
+      "grad_norm": 1.5239486694335938,
+      "learning_rate": 8.253834185374528e-05,
+      "loss": 1.3966,
+      "step": 26433
+    },
+    {
+      "epoch": 0.5874222222222222,
+      "grad_norm": 1.609764814376831,
+      "learning_rate": 8.253389642142699e-05,
+      "loss": 1.8746,
+      "step": 26434
+    },
+    {
+      "epoch": 0.5874444444444444,
+      "grad_norm": 1.4939262866973877,
+      "learning_rate": 8.25294509891087e-05,
+      "loss": 1.4485,
+      "step": 26435
+    },
+    {
+      "epoch": 0.5874666666666667,
+      "grad_norm": 1.7408115863800049,
+      "learning_rate": 8.25250055567904e-05,
+      "loss": 1.9704,
+      "step": 26436
+    },
+    {
+      "epoch": 0.5874888888888888,
+      "grad_norm": 1.6214052438735962,
+      "learning_rate": 8.252056012447212e-05,
+      "loss": 1.5611,
+      "step": 26437
+    },
+    {
+      "epoch": 0.5875111111111111,
+      "grad_norm": 1.5068247318267822,
+      "learning_rate": 8.251611469215381e-05,
+      "loss": 1.6896,
+      "step": 26438
+    },
+    {
+      "epoch": 0.5875333333333334,
+      "grad_norm": 1.753875494003296,
+      "learning_rate": 8.251166925983552e-05,
+      "loss": 1.7141,
+      "step": 26439
+    },
+    {
+      "epoch": 0.5875555555555556,
+      "grad_norm": 1.8102775812149048,
+      "learning_rate": 8.250722382751723e-05,
+      "loss": 1.956,
+      "step": 26440
+    },
+    {
+      "epoch": 0.5875777777777778,
+      "grad_norm": 1.626371145248413,
+      "learning_rate": 8.250277839519893e-05,
+      "loss": 1.7825,
+      "step": 26441
+    },
+    {
+      "epoch": 0.5876,
+      "grad_norm": 1.577762484550476,
+      "learning_rate": 8.249833296288064e-05,
+      "loss": 1.7609,
+      "step": 26442
+    },
+    {
+      "epoch": 0.5876222222222223,
+      "grad_norm": 1.9608583450317383,
+      "learning_rate": 8.249388753056235e-05,
+      "loss": 2.1549,
+      "step": 26443
+    },
+    {
+      "epoch": 0.5876444444444444,
+      "grad_norm": 1.8894906044006348,
+      "learning_rate": 8.248944209824406e-05,
+      "loss": 1.9205,
+      "step": 26444
+    },
+    {
+      "epoch": 0.5876666666666667,
+      "grad_norm": 1.5059431791305542,
+      "learning_rate": 8.248499666592577e-05,
+      "loss": 1.4438,
+      "step": 26445
+    },
+    {
+      "epoch": 0.5876888888888889,
+      "grad_norm": 1.8880411386489868,
+      "learning_rate": 8.248055123360748e-05,
+      "loss": 1.9452,
+      "step": 26446
+    },
+    {
+      "epoch": 0.5877111111111111,
+      "grad_norm": 1.5395166873931885,
+      "learning_rate": 8.247610580128919e-05,
+      "loss": 1.8412,
+      "step": 26447
+    },
+    {
+      "epoch": 0.5877333333333333,
+      "grad_norm": 1.722232460975647,
+      "learning_rate": 8.247166036897088e-05,
+      "loss": 1.7644,
+      "step": 26448
+    },
+    {
+      "epoch": 0.5877555555555556,
+      "grad_norm": 1.1945700645446777,
+      "learning_rate": 8.246721493665259e-05,
+      "loss": 0.5181,
+      "step": 26449
+    },
+    {
+      "epoch": 0.5877777777777777,
+      "grad_norm": 1.5664265155792236,
+      "learning_rate": 8.24627695043343e-05,
+      "loss": 0.8745,
+      "step": 26450
+    },
+    {
+      "epoch": 0.5878,
+      "grad_norm": 1.278299331665039,
+      "learning_rate": 8.2458324072016e-05,
+      "loss": 1.6046,
+      "step": 26451
+    },
+    {
+      "epoch": 0.5878222222222222,
+      "grad_norm": 1.794995665550232,
+      "learning_rate": 8.24538786396977e-05,
+      "loss": 2.6412,
+      "step": 26452
+    },
+    {
+      "epoch": 0.5878444444444444,
+      "grad_norm": 1.5944641828536987,
+      "learning_rate": 8.244943320737943e-05,
+      "loss": 2.638,
+      "step": 26453
+    },
+    {
+      "epoch": 0.5878666666666666,
+      "grad_norm": 1.603827953338623,
+      "learning_rate": 8.244498777506114e-05,
+      "loss": 1.2712,
+      "step": 26454
+    },
+    {
+      "epoch": 0.5878888888888889,
+      "grad_norm": 1.6544184684753418,
+      "learning_rate": 8.244054234274284e-05,
+      "loss": 1.7048,
+      "step": 26455
+    },
+    {
+      "epoch": 0.5879111111111112,
+      "grad_norm": 1.5918264389038086,
+      "learning_rate": 8.243609691042454e-05,
+      "loss": 2.0802,
+      "step": 26456
+    },
+    {
+      "epoch": 0.5879333333333333,
+      "grad_norm": 1.6266483068466187,
+      "learning_rate": 8.243165147810625e-05,
+      "loss": 1.6888,
+      "step": 26457
+    },
+    {
+      "epoch": 0.5879555555555556,
+      "grad_norm": 1.4718546867370605,
+      "learning_rate": 8.242720604578795e-05,
+      "loss": 1.9847,
+      "step": 26458
+    },
+    {
+      "epoch": 0.5879777777777778,
+      "grad_norm": 1.6215400695800781,
+      "learning_rate": 8.242276061346966e-05,
+      "loss": 2.5943,
+      "step": 26459
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 1.3390930891036987,
+      "learning_rate": 8.241831518115137e-05,
+      "loss": 0.9272,
+      "step": 26460
+    },
+    {
+      "epoch": 0.5880222222222222,
+      "grad_norm": 1.5605331659317017,
+      "learning_rate": 8.241386974883307e-05,
+      "loss": 2.1027,
+      "step": 26461
+    },
+    {
+      "epoch": 0.5880444444444445,
+      "grad_norm": 1.332210898399353,
+      "learning_rate": 8.240942431651479e-05,
+      "loss": 1.7529,
+      "step": 26462
+    },
+    {
+      "epoch": 0.5880666666666666,
+      "grad_norm": 1.414040207862854,
+      "learning_rate": 8.24049788841965e-05,
+      "loss": 1.9332,
+      "step": 26463
+    },
+    {
+      "epoch": 0.5880888888888889,
+      "grad_norm": 1.5289276838302612,
+      "learning_rate": 8.240053345187821e-05,
+      "loss": 2.1359,
+      "step": 26464
+    },
+    {
+      "epoch": 0.5881111111111111,
+      "grad_norm": 1.1083582639694214,
+      "learning_rate": 8.23960880195599e-05,
+      "loss": 1.1907,
+      "step": 26465
+    },
+    {
+      "epoch": 0.5881333333333333,
+      "grad_norm": 1.4909306764602661,
+      "learning_rate": 8.239164258724161e-05,
+      "loss": 1.8534,
+      "step": 26466
+    },
+    {
+      "epoch": 0.5881555555555555,
+      "grad_norm": 1.5297155380249023,
+      "learning_rate": 8.238719715492332e-05,
+      "loss": 1.7557,
+      "step": 26467
+    },
+    {
+      "epoch": 0.5881777777777778,
+      "grad_norm": 1.879301905632019,
+      "learning_rate": 8.238275172260502e-05,
+      "loss": 2.0541,
+      "step": 26468
+    },
+    {
+      "epoch": 0.5882,
+      "grad_norm": 1.8957713842391968,
+      "learning_rate": 8.237830629028673e-05,
+      "loss": 2.1396,
+      "step": 26469
+    },
+    {
+      "epoch": 0.5882222222222222,
+      "grad_norm": 1.3897382020950317,
+      "learning_rate": 8.237386085796844e-05,
+      "loss": 1.9988,
+      "step": 26470
+    },
+    {
+      "epoch": 0.5882444444444445,
+      "grad_norm": 1.6075379848480225,
+      "learning_rate": 8.236941542565015e-05,
+      "loss": 2.2945,
+      "step": 26471
+    },
+    {
+      "epoch": 0.5882666666666667,
+      "grad_norm": 1.8518240451812744,
+      "learning_rate": 8.236496999333186e-05,
+      "loss": 2.1176,
+      "step": 26472
+    },
+    {
+      "epoch": 0.5882888888888889,
+      "grad_norm": 1.6546710729599,
+      "learning_rate": 8.236052456101357e-05,
+      "loss": 1.8829,
+      "step": 26473
+    },
+    {
+      "epoch": 0.5883111111111111,
+      "grad_norm": 1.332216501235962,
+      "learning_rate": 8.235607912869528e-05,
+      "loss": 1.315,
+      "step": 26474
+    },
+    {
+      "epoch": 0.5883333333333334,
+      "grad_norm": 1.7014000415802002,
+      "learning_rate": 8.235163369637697e-05,
+      "loss": 2.2719,
+      "step": 26475
+    },
+    {
+      "epoch": 0.5883555555555555,
+      "grad_norm": 1.8507986068725586,
+      "learning_rate": 8.234718826405868e-05,
+      "loss": 1.9988,
+      "step": 26476
+    },
+    {
+      "epoch": 0.5883777777777778,
+      "grad_norm": 2.463884115219116,
+      "learning_rate": 8.234274283174039e-05,
+      "loss": 2.3464,
+      "step": 26477
+    },
+    {
+      "epoch": 0.5884,
+      "grad_norm": 3.2078611850738525,
+      "learning_rate": 8.233829739942209e-05,
+      "loss": 2.4075,
+      "step": 26478
+    },
+    {
+      "epoch": 0.5884222222222222,
+      "grad_norm": 1.769322156906128,
+      "learning_rate": 8.23338519671038e-05,
+      "loss": 1.9338,
+      "step": 26479
+    },
+    {
+      "epoch": 0.5884444444444444,
+      "grad_norm": 1.3190253973007202,
+      "learning_rate": 8.232940653478551e-05,
+      "loss": 1.2321,
+      "step": 26480
+    },
+    {
+      "epoch": 0.5884666666666667,
+      "grad_norm": 1.6068058013916016,
+      "learning_rate": 8.232496110246722e-05,
+      "loss": 1.9387,
+      "step": 26481
+    },
+    {
+      "epoch": 0.5884888888888888,
+      "grad_norm": 1.6955701112747192,
+      "learning_rate": 8.232051567014893e-05,
+      "loss": 2.0747,
+      "step": 26482
+    },
+    {
+      "epoch": 0.5885111111111111,
+      "grad_norm": 1.8967593908309937,
+      "learning_rate": 8.231607023783064e-05,
+      "loss": 1.7893,
+      "step": 26483
+    },
+    {
+      "epoch": 0.5885333333333334,
+      "grad_norm": 1.5220470428466797,
+      "learning_rate": 8.231162480551235e-05,
+      "loss": 1.8033,
+      "step": 26484
+    },
+    {
+      "epoch": 0.5885555555555556,
+      "grad_norm": 1.6682027578353882,
+      "learning_rate": 8.230717937319404e-05,
+      "loss": 2.1241,
+      "step": 26485
+    },
+    {
+      "epoch": 0.5885777777777778,
+      "grad_norm": 1.5411769151687622,
+      "learning_rate": 8.230273394087575e-05,
+      "loss": 1.5139,
+      "step": 26486
+    },
+    {
+      "epoch": 0.5886,
+      "grad_norm": 1.8869585990905762,
+      "learning_rate": 8.229828850855746e-05,
+      "loss": 2.0045,
+      "step": 26487
+    },
+    {
+      "epoch": 0.5886222222222223,
+      "grad_norm": 1.6759885549545288,
+      "learning_rate": 8.229384307623916e-05,
+      "loss": 1.6491,
+      "step": 26488
+    },
+    {
+      "epoch": 0.5886444444444444,
+      "grad_norm": 1.7548519372940063,
+      "learning_rate": 8.228939764392087e-05,
+      "loss": 2.0672,
+      "step": 26489
+    },
+    {
+      "epoch": 0.5886666666666667,
+      "grad_norm": 1.571402907371521,
+      "learning_rate": 8.228495221160259e-05,
+      "loss": 1.3744,
+      "step": 26490
+    },
+    {
+      "epoch": 0.5886888888888889,
+      "grad_norm": 1.9174574613571167,
+      "learning_rate": 8.22805067792843e-05,
+      "loss": 1.8788,
+      "step": 26491
+    },
+    {
+      "epoch": 0.5887111111111111,
+      "grad_norm": 2.084259033203125,
+      "learning_rate": 8.2276061346966e-05,
+      "loss": 2.3798,
+      "step": 26492
+    },
+    {
+      "epoch": 0.5887333333333333,
+      "grad_norm": 1.5270769596099854,
+      "learning_rate": 8.22716159146477e-05,
+      "loss": 1.8474,
+      "step": 26493
+    },
+    {
+      "epoch": 0.5887555555555556,
+      "grad_norm": 1.9949065446853638,
+      "learning_rate": 8.226717048232942e-05,
+      "loss": 2.2836,
+      "step": 26494
+    },
+    {
+      "epoch": 0.5887777777777777,
+      "grad_norm": 1.8067631721496582,
+      "learning_rate": 8.226272505001111e-05,
+      "loss": 1.8638,
+      "step": 26495
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 1.8312886953353882,
+      "learning_rate": 8.225827961769282e-05,
+      "loss": 2.3125,
+      "step": 26496
+    },
+    {
+      "epoch": 0.5888222222222222,
+      "grad_norm": 1.6324427127838135,
+      "learning_rate": 8.225383418537453e-05,
+      "loss": 1.4562,
+      "step": 26497
+    },
+    {
+      "epoch": 0.5888444444444444,
+      "grad_norm": 2.074530601501465,
+      "learning_rate": 8.224938875305623e-05,
+      "loss": 1.9811,
+      "step": 26498
+    },
+    {
+      "epoch": 0.5888666666666666,
+      "grad_norm": 1.7900502681732178,
+      "learning_rate": 8.224494332073795e-05,
+      "loss": 1.7443,
+      "step": 26499
+    },
+    {
+      "epoch": 0.5888888888888889,
+      "grad_norm": 1.8668699264526367,
+      "learning_rate": 8.224049788841966e-05,
+      "loss": 1.913,
+      "step": 26500
+    },
+    {
+      "epoch": 0.5889111111111112,
+      "grad_norm": 1.4245249032974243,
+      "learning_rate": 8.223605245610137e-05,
+      "loss": 2.6009,
+      "step": 26501
+    },
+    {
+      "epoch": 0.5889333333333333,
+      "grad_norm": 1.4777202606201172,
+      "learning_rate": 8.223160702378307e-05,
+      "loss": 2.6983,
+      "step": 26502
+    },
+    {
+      "epoch": 0.5889555555555556,
+      "grad_norm": 1.3658816814422607,
+      "learning_rate": 8.222716159146477e-05,
+      "loss": 2.0879,
+      "step": 26503
+    },
+    {
+      "epoch": 0.5889777777777778,
+      "grad_norm": 0.2192663699388504,
+      "learning_rate": 8.222271615914648e-05,
+      "loss": 0.0216,
+      "step": 26504
+    },
+    {
+      "epoch": 0.589,
+      "grad_norm": 1.0416237115859985,
+      "learning_rate": 8.221827072682818e-05,
+      "loss": 1.2148,
+      "step": 26505
+    },
+    {
+      "epoch": 0.5890222222222222,
+      "grad_norm": 1.6372934579849243,
+      "learning_rate": 8.221382529450989e-05,
+      "loss": 2.2565,
+      "step": 26506
+    },
+    {
+      "epoch": 0.5890444444444445,
+      "grad_norm": 1.6254594326019287,
+      "learning_rate": 8.22093798621916e-05,
+      "loss": 2.5999,
+      "step": 26507
+    },
+    {
+      "epoch": 0.5890666666666666,
+      "grad_norm": 1.6181801557540894,
+      "learning_rate": 8.220493442987331e-05,
+      "loss": 2.3094,
+      "step": 26508
+    },
+    {
+      "epoch": 0.5890888888888889,
+      "grad_norm": 1.6073154211044312,
+      "learning_rate": 8.220048899755502e-05,
+      "loss": 2.3929,
+      "step": 26509
+    },
+    {
+      "epoch": 0.5891111111111111,
+      "grad_norm": 1.4710923433303833,
+      "learning_rate": 8.219604356523673e-05,
+      "loss": 1.9685,
+      "step": 26510
+    },
+    {
+      "epoch": 0.5891333333333333,
+      "grad_norm": 1.5624597072601318,
+      "learning_rate": 8.219159813291844e-05,
+      "loss": 2.0818,
+      "step": 26511
+    },
+    {
+      "epoch": 0.5891555555555555,
+      "grad_norm": 1.6579220294952393,
+      "learning_rate": 8.218715270060013e-05,
+      "loss": 1.7105,
+      "step": 26512
+    },
+    {
+      "epoch": 0.5891777777777778,
+      "grad_norm": 1.7512640953063965,
+      "learning_rate": 8.218270726828184e-05,
+      "loss": 2.4301,
+      "step": 26513
+    },
+    {
+      "epoch": 0.5892,
+      "grad_norm": 1.403272271156311,
+      "learning_rate": 8.217826183596355e-05,
+      "loss": 1.5019,
+      "step": 26514
+    },
+    {
+      "epoch": 0.5892222222222222,
+      "grad_norm": 1.9031537771224976,
+      "learning_rate": 8.217381640364525e-05,
+      "loss": 2.4452,
+      "step": 26515
+    },
+    {
+      "epoch": 0.5892444444444445,
+      "grad_norm": 1.6179159879684448,
+      "learning_rate": 8.216937097132696e-05,
+      "loss": 2.1208,
+      "step": 26516
+    },
+    {
+      "epoch": 0.5892666666666667,
+      "grad_norm": 1.5489249229431152,
+      "learning_rate": 8.216492553900867e-05,
+      "loss": 1.6462,
+      "step": 26517
+    },
+    {
+      "epoch": 0.5892888888888889,
+      "grad_norm": 1.8130542039871216,
+      "learning_rate": 8.216048010669038e-05,
+      "loss": 1.6512,
+      "step": 26518
+    },
+    {
+      "epoch": 0.5893111111111111,
+      "grad_norm": 1.5840333700180054,
+      "learning_rate": 8.215603467437209e-05,
+      "loss": 1.6867,
+      "step": 26519
+    },
+    {
+      "epoch": 0.5893333333333334,
+      "grad_norm": 1.7145648002624512,
+      "learning_rate": 8.21515892420538e-05,
+      "loss": 2.2308,
+      "step": 26520
+    },
+    {
+      "epoch": 0.5893555555555555,
+      "grad_norm": 2.2711832523345947,
+      "learning_rate": 8.214714380973551e-05,
+      "loss": 2.3032,
+      "step": 26521
+    },
+    {
+      "epoch": 0.5893777777777778,
+      "grad_norm": 1.7569622993469238,
+      "learning_rate": 8.21426983774172e-05,
+      "loss": 1.9934,
+      "step": 26522
+    },
+    {
+      "epoch": 0.5894,
+      "grad_norm": 2.0704503059387207,
+      "learning_rate": 8.213825294509891e-05,
+      "loss": 2.0143,
+      "step": 26523
+    },
+    {
+      "epoch": 0.5894222222222222,
+      "grad_norm": 1.657915472984314,
+      "learning_rate": 8.213380751278062e-05,
+      "loss": 1.697,
+      "step": 26524
+    },
+    {
+      "epoch": 0.5894444444444444,
+      "grad_norm": 1.6642396450042725,
+      "learning_rate": 8.212936208046232e-05,
+      "loss": 1.6574,
+      "step": 26525
+    },
+    {
+      "epoch": 0.5894666666666667,
+      "grad_norm": 1.705404281616211,
+      "learning_rate": 8.212491664814403e-05,
+      "loss": 1.8843,
+      "step": 26526
+    },
+    {
+      "epoch": 0.5894888888888888,
+      "grad_norm": 1.951427936553955,
+      "learning_rate": 8.212047121582575e-05,
+      "loss": 2.2169,
+      "step": 26527
+    },
+    {
+      "epoch": 0.5895111111111111,
+      "grad_norm": 1.76736319065094,
+      "learning_rate": 8.211602578350746e-05,
+      "loss": 2.4436,
+      "step": 26528
+    },
+    {
+      "epoch": 0.5895333333333334,
+      "grad_norm": 1.6907325983047485,
+      "learning_rate": 8.211158035118916e-05,
+      "loss": 1.9543,
+      "step": 26529
+    },
+    {
+      "epoch": 0.5895555555555556,
+      "grad_norm": 1.6716514825820923,
+      "learning_rate": 8.210713491887087e-05,
+      "loss": 1.7508,
+      "step": 26530
+    },
+    {
+      "epoch": 0.5895777777777778,
+      "grad_norm": 1.727203130722046,
+      "learning_rate": 8.210268948655258e-05,
+      "loss": 2.0613,
+      "step": 26531
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 1.8019435405731201,
+      "learning_rate": 8.209824405423427e-05,
+      "loss": 1.9505,
+      "step": 26532
+    },
+    {
+      "epoch": 0.5896222222222223,
+      "grad_norm": 1.71755051612854,
+      "learning_rate": 8.209379862191598e-05,
+      "loss": 2.134,
+      "step": 26533
+    },
+    {
+      "epoch": 0.5896444444444444,
+      "grad_norm": 1.2752341032028198,
+      "learning_rate": 8.208935318959769e-05,
+      "loss": 0.8956,
+      "step": 26534
+    },
+    {
+      "epoch": 0.5896666666666667,
+      "grad_norm": 1.8847476243972778,
+      "learning_rate": 8.208490775727939e-05,
+      "loss": 2.0627,
+      "step": 26535
+    },
+    {
+      "epoch": 0.5896888888888889,
+      "grad_norm": 1.639291763305664,
+      "learning_rate": 8.208046232496111e-05,
+      "loss": 1.9175,
+      "step": 26536
+    },
+    {
+      "epoch": 0.5897111111111111,
+      "grad_norm": 2.0569376945495605,
+      "learning_rate": 8.207601689264282e-05,
+      "loss": 2.1619,
+      "step": 26537
+    },
+    {
+      "epoch": 0.5897333333333333,
+      "grad_norm": 1.775004267692566,
+      "learning_rate": 8.207157146032453e-05,
+      "loss": 1.8134,
+      "step": 26538
+    },
+    {
+      "epoch": 0.5897555555555556,
+      "grad_norm": 2.516251564025879,
+      "learning_rate": 8.206712602800623e-05,
+      "loss": 2.2565,
+      "step": 26539
+    },
+    {
+      "epoch": 0.5897777777777777,
+      "grad_norm": 1.5780636072158813,
+      "learning_rate": 8.206268059568794e-05,
+      "loss": 1.658,
+      "step": 26540
+    },
+    {
+      "epoch": 0.5898,
+      "grad_norm": 1.934320330619812,
+      "learning_rate": 8.205823516336965e-05,
+      "loss": 2.1937,
+      "step": 26541
+    },
+    {
+      "epoch": 0.5898222222222222,
+      "grad_norm": 1.937772274017334,
+      "learning_rate": 8.205378973105134e-05,
+      "loss": 2.0649,
+      "step": 26542
+    },
+    {
+      "epoch": 0.5898444444444444,
+      "grad_norm": 1.5901800394058228,
+      "learning_rate": 8.204934429873305e-05,
+      "loss": 1.6263,
+      "step": 26543
+    },
+    {
+      "epoch": 0.5898666666666667,
+      "grad_norm": 1.6580713987350464,
+      "learning_rate": 8.204489886641476e-05,
+      "loss": 1.7261,
+      "step": 26544
+    },
+    {
+      "epoch": 0.5898888888888889,
+      "grad_norm": 1.5570964813232422,
+      "learning_rate": 8.204045343409647e-05,
+      "loss": 1.4863,
+      "step": 26545
+    },
+    {
+      "epoch": 0.5899111111111112,
+      "grad_norm": 1.8327727317810059,
+      "learning_rate": 8.203600800177818e-05,
+      "loss": 1.8743,
+      "step": 26546
+    },
+    {
+      "epoch": 0.5899333333333333,
+      "grad_norm": 1.509567141532898,
+      "learning_rate": 8.203156256945989e-05,
+      "loss": 1.591,
+      "step": 26547
+    },
+    {
+      "epoch": 0.5899555555555556,
+      "grad_norm": 2.091273307800293,
+      "learning_rate": 8.20271171371416e-05,
+      "loss": 1.8144,
+      "step": 26548
+    },
+    {
+      "epoch": 0.5899777777777778,
+      "grad_norm": 1.8163344860076904,
+      "learning_rate": 8.20226717048233e-05,
+      "loss": 1.7362,
+      "step": 26549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0135276317596436,
+      "learning_rate": 8.2018226272505e-05,
+      "loss": 1.6102,
+      "step": 26550
+    },
+    {
+      "epoch": 0.5900222222222222,
+      "grad_norm": 1.781471610069275,
+      "learning_rate": 8.201378084018671e-05,
+      "loss": 2.6554,
+      "step": 26551
+    },
+    {
+      "epoch": 0.5900444444444445,
+      "grad_norm": 1.0032708644866943,
+      "learning_rate": 8.200933540786841e-05,
+      "loss": 0.8867,
+      "step": 26552
+    },
+    {
+      "epoch": 0.5900666666666666,
+      "grad_norm": 1.761935830116272,
+      "learning_rate": 8.200488997555012e-05,
+      "loss": 2.6627,
+      "step": 26553
+    },
+    {
+      "epoch": 0.5900888888888889,
+      "grad_norm": 1.5583018064498901,
+      "learning_rate": 8.200044454323183e-05,
+      "loss": 2.2719,
+      "step": 26554
+    },
+    {
+      "epoch": 0.5901111111111111,
+      "grad_norm": 1.6104395389556885,
+      "learning_rate": 8.199599911091354e-05,
+      "loss": 2.3873,
+      "step": 26555
+    },
+    {
+      "epoch": 0.5901333333333333,
+      "grad_norm": 1.6542582511901855,
+      "learning_rate": 8.199155367859525e-05,
+      "loss": 2.2377,
+      "step": 26556
+    },
+    {
+      "epoch": 0.5901555555555555,
+      "grad_norm": 1.507717251777649,
+      "learning_rate": 8.198710824627696e-05,
+      "loss": 1.8727,
+      "step": 26557
+    },
+    {
+      "epoch": 0.5901777777777778,
+      "grad_norm": 1.5798872709274292,
+      "learning_rate": 8.198266281395867e-05,
+      "loss": 2.4807,
+      "step": 26558
+    },
+    {
+      "epoch": 0.5902,
+      "grad_norm": 2.0253896713256836,
+      "learning_rate": 8.197821738164036e-05,
+      "loss": 2.1833,
+      "step": 26559
+    },
+    {
+      "epoch": 0.5902222222222222,
+      "grad_norm": 1.9096221923828125,
+      "learning_rate": 8.197377194932207e-05,
+      "loss": 2.0666,
+      "step": 26560
+    },
+    {
+      "epoch": 0.5902444444444445,
+      "grad_norm": 1.6258680820465088,
+      "learning_rate": 8.196932651700378e-05,
+      "loss": 1.9742,
+      "step": 26561
+    },
+    {
+      "epoch": 0.5902666666666667,
+      "grad_norm": 1.657233715057373,
+      "learning_rate": 8.196488108468548e-05,
+      "loss": 2.2751,
+      "step": 26562
+    },
+    {
+      "epoch": 0.5902888888888889,
+      "grad_norm": 2.256463050842285,
+      "learning_rate": 8.196043565236719e-05,
+      "loss": 2.3546,
+      "step": 26563
+    },
+    {
+      "epoch": 0.5903111111111111,
+      "grad_norm": 1.5345555543899536,
+      "learning_rate": 8.195599022004891e-05,
+      "loss": 2.0553,
+      "step": 26564
+    },
+    {
+      "epoch": 0.5903333333333334,
+      "grad_norm": 1.6566272974014282,
+      "learning_rate": 8.195154478773061e-05,
+      "loss": 1.8394,
+      "step": 26565
+    },
+    {
+      "epoch": 0.5903555555555555,
+      "grad_norm": 1.3703186511993408,
+      "learning_rate": 8.194709935541232e-05,
+      "loss": 1.2446,
+      "step": 26566
+    },
+    {
+      "epoch": 0.5903777777777778,
+      "grad_norm": 1.0013676881790161,
+      "learning_rate": 8.194265392309403e-05,
+      "loss": 0.0459,
+      "step": 26567
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 1.772273063659668,
+      "learning_rate": 8.193820849077574e-05,
+      "loss": 2.2504,
+      "step": 26568
+    },
+    {
+      "epoch": 0.5904222222222222,
+      "grad_norm": 1.7223542928695679,
+      "learning_rate": 8.193376305845743e-05,
+      "loss": 1.9782,
+      "step": 26569
+    },
+    {
+      "epoch": 0.5904444444444444,
+      "grad_norm": 1.4310003519058228,
+      "learning_rate": 8.192931762613914e-05,
+      "loss": 1.9249,
+      "step": 26570
+    },
+    {
+      "epoch": 0.5904666666666667,
+      "grad_norm": 1.986963152885437,
+      "learning_rate": 8.192487219382085e-05,
+      "loss": 2.1059,
+      "step": 26571
+    },
+    {
+      "epoch": 0.5904888888888888,
+      "grad_norm": 1.6062281131744385,
+      "learning_rate": 8.192042676150255e-05,
+      "loss": 1.9521,
+      "step": 26572
+    },
+    {
+      "epoch": 0.5905111111111111,
+      "grad_norm": 1.5529038906097412,
+      "learning_rate": 8.191598132918427e-05,
+      "loss": 1.7418,
+      "step": 26573
+    },
+    {
+      "epoch": 0.5905333333333334,
+      "grad_norm": 2.3296256065368652,
+      "learning_rate": 8.191153589686598e-05,
+      "loss": 2.0563,
+      "step": 26574
+    },
+    {
+      "epoch": 0.5905555555555555,
+      "grad_norm": 1.8755871057510376,
+      "learning_rate": 8.190709046454769e-05,
+      "loss": 2.1776,
+      "step": 26575
+    },
+    {
+      "epoch": 0.5905777777777778,
+      "grad_norm": 1.5734074115753174,
+      "learning_rate": 8.190264503222939e-05,
+      "loss": 1.8932,
+      "step": 26576
+    },
+    {
+      "epoch": 0.5906,
+      "grad_norm": 1.6150445938110352,
+      "learning_rate": 8.18981995999111e-05,
+      "loss": 1.8276,
+      "step": 26577
+    },
+    {
+      "epoch": 0.5906222222222223,
+      "grad_norm": 1.3324120044708252,
+      "learning_rate": 8.18937541675928e-05,
+      "loss": 1.8167,
+      "step": 26578
+    },
+    {
+      "epoch": 0.5906444444444444,
+      "grad_norm": 1.2015331983566284,
+      "learning_rate": 8.18893087352745e-05,
+      "loss": 1.0707,
+      "step": 26579
+    },
+    {
+      "epoch": 0.5906666666666667,
+      "grad_norm": 1.3903979063034058,
+      "learning_rate": 8.188486330295621e-05,
+      "loss": 1.3774,
+      "step": 26580
+    },
+    {
+      "epoch": 0.5906888888888889,
+      "grad_norm": 1.6445354223251343,
+      "learning_rate": 8.188041787063792e-05,
+      "loss": 1.4733,
+      "step": 26581
+    },
+    {
+      "epoch": 0.5907111111111111,
+      "grad_norm": 1.9705824851989746,
+      "learning_rate": 8.187597243831963e-05,
+      "loss": 1.829,
+      "step": 26582
+    },
+    {
+      "epoch": 0.5907333333333333,
+      "grad_norm": 1.145692229270935,
+      "learning_rate": 8.187152700600134e-05,
+      "loss": 0.9194,
+      "step": 26583
+    },
+    {
+      "epoch": 0.5907555555555556,
+      "grad_norm": 1.5932261943817139,
+      "learning_rate": 8.186708157368305e-05,
+      "loss": 1.5765,
+      "step": 26584
+    },
+    {
+      "epoch": 0.5907777777777777,
+      "grad_norm": 1.630245566368103,
+      "learning_rate": 8.186263614136476e-05,
+      "loss": 1.8371,
+      "step": 26585
+    },
+    {
+      "epoch": 0.5908,
+      "grad_norm": 1.6606966257095337,
+      "learning_rate": 8.185819070904646e-05,
+      "loss": 1.5915,
+      "step": 26586
+    },
+    {
+      "epoch": 0.5908222222222222,
+      "grad_norm": 1.1430854797363281,
+      "learning_rate": 8.185374527672817e-05,
+      "loss": 0.745,
+      "step": 26587
+    },
+    {
+      "epoch": 0.5908444444444444,
+      "grad_norm": 1.4599413871765137,
+      "learning_rate": 8.184929984440988e-05,
+      "loss": 1.1224,
+      "step": 26588
+    },
+    {
+      "epoch": 0.5908666666666667,
+      "grad_norm": 2.151262044906616,
+      "learning_rate": 8.184485441209157e-05,
+      "loss": 2.0447,
+      "step": 26589
+    },
+    {
+      "epoch": 0.5908888888888889,
+      "grad_norm": 1.7323678731918335,
+      "learning_rate": 8.184040897977328e-05,
+      "loss": 1.9831,
+      "step": 26590
+    },
+    {
+      "epoch": 0.5909111111111112,
+      "grad_norm": 1.604820728302002,
+      "learning_rate": 8.183596354745499e-05,
+      "loss": 1.4933,
+      "step": 26591
+    },
+    {
+      "epoch": 0.5909333333333333,
+      "grad_norm": 1.5943677425384521,
+      "learning_rate": 8.18315181151367e-05,
+      "loss": 1.6485,
+      "step": 26592
+    },
+    {
+      "epoch": 0.5909555555555556,
+      "grad_norm": 1.757006287574768,
+      "learning_rate": 8.182707268281841e-05,
+      "loss": 2.0849,
+      "step": 26593
+    },
+    {
+      "epoch": 0.5909777777777778,
+      "grad_norm": 1.4215257167816162,
+      "learning_rate": 8.182262725050012e-05,
+      "loss": 1.4631,
+      "step": 26594
+    },
+    {
+      "epoch": 0.591,
+      "grad_norm": 1.69182550907135,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 1.7487,
+      "step": 26595
+    },
+    {
+      "epoch": 0.5910222222222222,
+      "grad_norm": 2.1581408977508545,
+      "learning_rate": 8.181373638586353e-05,
+      "loss": 2.0966,
+      "step": 26596
+    },
+    {
+      "epoch": 0.5910444444444445,
+      "grad_norm": 2.17958927154541,
+      "learning_rate": 8.180929095354524e-05,
+      "loss": 1.7987,
+      "step": 26597
+    },
+    {
+      "epoch": 0.5910666666666666,
+      "grad_norm": 1.9482837915420532,
+      "learning_rate": 8.180484552122694e-05,
+      "loss": 1.793,
+      "step": 26598
+    },
+    {
+      "epoch": 0.5910888888888889,
+      "grad_norm": 1.6424596309661865,
+      "learning_rate": 8.180040008890864e-05,
+      "loss": 1.4073,
+      "step": 26599
+    },
+    {
+      "epoch": 0.5911111111111111,
+      "grad_norm": 2.3620777130126953,
+      "learning_rate": 8.179595465659035e-05,
+      "loss": 1.3749,
+      "step": 26600
+    },
+    {
+      "epoch": 0.5911333333333333,
+      "grad_norm": 1.8508235216140747,
+      "learning_rate": 8.179150922427207e-05,
+      "loss": 2.9108,
+      "step": 26601
+    },
+    {
+      "epoch": 0.5911555555555555,
+      "grad_norm": 1.1262704133987427,
+      "learning_rate": 8.178706379195377e-05,
+      "loss": 0.9763,
+      "step": 26602
+    },
+    {
+      "epoch": 0.5911777777777778,
+      "grad_norm": 1.5308245420455933,
+      "learning_rate": 8.178261835963548e-05,
+      "loss": 2.0178,
+      "step": 26603
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 1.6389973163604736,
+      "learning_rate": 8.177817292731719e-05,
+      "loss": 2.4368,
+      "step": 26604
+    },
+    {
+      "epoch": 0.5912222222222222,
+      "grad_norm": 1.7720253467559814,
+      "learning_rate": 8.17737274949989e-05,
+      "loss": 2.2833,
+      "step": 26605
+    },
+    {
+      "epoch": 0.5912444444444445,
+      "grad_norm": 1.401301383972168,
+      "learning_rate": 8.17692820626806e-05,
+      "loss": 2.1505,
+      "step": 26606
+    },
+    {
+      "epoch": 0.5912666666666667,
+      "grad_norm": 1.1152511835098267,
+      "learning_rate": 8.17648366303623e-05,
+      "loss": 1.1234,
+      "step": 26607
+    },
+    {
+      "epoch": 0.5912888888888889,
+      "grad_norm": 1.4615529775619507,
+      "learning_rate": 8.176039119804401e-05,
+      "loss": 1.8951,
+      "step": 26608
+    },
+    {
+      "epoch": 0.5913111111111111,
+      "grad_norm": 1.5637952089309692,
+      "learning_rate": 8.175594576572571e-05,
+      "loss": 2.4163,
+      "step": 26609
+    },
+    {
+      "epoch": 0.5913333333333334,
+      "grad_norm": 1.5488580465316772,
+      "learning_rate": 8.175150033340743e-05,
+      "loss": 2.0755,
+      "step": 26610
+    },
+    {
+      "epoch": 0.5913555555555555,
+      "grad_norm": 1.5902295112609863,
+      "learning_rate": 8.174705490108914e-05,
+      "loss": 1.8796,
+      "step": 26611
+    },
+    {
+      "epoch": 0.5913777777777778,
+      "grad_norm": 1.5014410018920898,
+      "learning_rate": 8.174260946877084e-05,
+      "loss": 2.2305,
+      "step": 26612
+    },
+    {
+      "epoch": 0.5914,
+      "grad_norm": 1.3349800109863281,
+      "learning_rate": 8.173816403645255e-05,
+      "loss": 1.4973,
+      "step": 26613
+    },
+    {
+      "epoch": 0.5914222222222222,
+      "grad_norm": 1.8313910961151123,
+      "learning_rate": 8.173371860413426e-05,
+      "loss": 2.1185,
+      "step": 26614
+    },
+    {
+      "epoch": 0.5914444444444444,
+      "grad_norm": 1.560365915298462,
+      "learning_rate": 8.172927317181597e-05,
+      "loss": 1.798,
+      "step": 26615
+    },
+    {
+      "epoch": 0.5914666666666667,
+      "grad_norm": 1.5687285661697388,
+      "learning_rate": 8.172482773949766e-05,
+      "loss": 1.802,
+      "step": 26616
+    },
+    {
+      "epoch": 0.5914888888888888,
+      "grad_norm": 1.6726188659667969,
+      "learning_rate": 8.172038230717937e-05,
+      "loss": 1.9302,
+      "step": 26617
+    },
+    {
+      "epoch": 0.5915111111111111,
+      "grad_norm": 2.1939918994903564,
+      "learning_rate": 8.171593687486108e-05,
+      "loss": 2.1833,
+      "step": 26618
+    },
+    {
+      "epoch": 0.5915333333333334,
+      "grad_norm": 1.5094565153121948,
+      "learning_rate": 8.171149144254279e-05,
+      "loss": 2.179,
+      "step": 26619
+    },
+    {
+      "epoch": 0.5915555555555555,
+      "grad_norm": 1.5371365547180176,
+      "learning_rate": 8.17070460102245e-05,
+      "loss": 1.798,
+      "step": 26620
+    },
+    {
+      "epoch": 0.5915777777777778,
+      "grad_norm": 1.4002341032028198,
+      "learning_rate": 8.170260057790621e-05,
+      "loss": 1.8906,
+      "step": 26621
+    },
+    {
+      "epoch": 0.5916,
+      "grad_norm": 1.7280828952789307,
+      "learning_rate": 8.169815514558791e-05,
+      "loss": 2.034,
+      "step": 26622
+    },
+    {
+      "epoch": 0.5916222222222223,
+      "grad_norm": 1.7399758100509644,
+      "learning_rate": 8.169370971326962e-05,
+      "loss": 2.2769,
+      "step": 26623
+    },
+    {
+      "epoch": 0.5916444444444444,
+      "grad_norm": 1.539891004562378,
+      "learning_rate": 8.168926428095133e-05,
+      "loss": 1.9422,
+      "step": 26624
+    },
+    {
+      "epoch": 0.5916666666666667,
+      "grad_norm": 1.6626595258712769,
+      "learning_rate": 8.168481884863304e-05,
+      "loss": 1.9142,
+      "step": 26625
+    },
+    {
+      "epoch": 0.5916888888888889,
+      "grad_norm": 1.5796253681182861,
+      "learning_rate": 8.168037341631473e-05,
+      "loss": 2.0004,
+      "step": 26626
+    },
+    {
+      "epoch": 0.5917111111111111,
+      "grad_norm": 1.6983789205551147,
+      "learning_rate": 8.167592798399644e-05,
+      "loss": 1.935,
+      "step": 26627
+    },
+    {
+      "epoch": 0.5917333333333333,
+      "grad_norm": 1.405887246131897,
+      "learning_rate": 8.167148255167815e-05,
+      "loss": 1.5105,
+      "step": 26628
+    },
+    {
+      "epoch": 0.5917555555555556,
+      "grad_norm": 1.676605463027954,
+      "learning_rate": 8.166703711935986e-05,
+      "loss": 1.7027,
+      "step": 26629
+    },
+    {
+      "epoch": 0.5917777777777777,
+      "grad_norm": 1.861316204071045,
+      "learning_rate": 8.166259168704157e-05,
+      "loss": 1.9841,
+      "step": 26630
+    },
+    {
+      "epoch": 0.5918,
+      "grad_norm": 1.7648253440856934,
+      "learning_rate": 8.165814625472328e-05,
+      "loss": 2.1883,
+      "step": 26631
+    },
+    {
+      "epoch": 0.5918222222222222,
+      "grad_norm": 1.637048363685608,
+      "learning_rate": 8.165370082240499e-05,
+      "loss": 1.5399,
+      "step": 26632
+    },
+    {
+      "epoch": 0.5918444444444444,
+      "grad_norm": 1.3285629749298096,
+      "learning_rate": 8.164925539008669e-05,
+      "loss": 1.1909,
+      "step": 26633
+    },
+    {
+      "epoch": 0.5918666666666667,
+      "grad_norm": 1.8250073194503784,
+      "learning_rate": 8.16448099577684e-05,
+      "loss": 2.0721,
+      "step": 26634
+    },
+    {
+      "epoch": 0.5918888888888889,
+      "grad_norm": 1.6213438510894775,
+      "learning_rate": 8.16403645254501e-05,
+      "loss": 1.9153,
+      "step": 26635
+    },
+    {
+      "epoch": 0.5919111111111112,
+      "grad_norm": 1.490524172782898,
+      "learning_rate": 8.16359190931318e-05,
+      "loss": 1.2641,
+      "step": 26636
+    },
+    {
+      "epoch": 0.5919333333333333,
+      "grad_norm": 1.869890570640564,
+      "learning_rate": 8.163147366081351e-05,
+      "loss": 1.3881,
+      "step": 26637
+    },
+    {
+      "epoch": 0.5919555555555556,
+      "grad_norm": 1.9931570291519165,
+      "learning_rate": 8.162702822849523e-05,
+      "loss": 1.5897,
+      "step": 26638
+    },
+    {
+      "epoch": 0.5919777777777778,
+      "grad_norm": 2.059723377227783,
+      "learning_rate": 8.162258279617693e-05,
+      "loss": 1.8027,
+      "step": 26639
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.8755348920822144,
+      "learning_rate": 8.161813736385864e-05,
+      "loss": 1.7455,
+      "step": 26640
+    },
+    {
+      "epoch": 0.5920222222222222,
+      "grad_norm": 1.8518706560134888,
+      "learning_rate": 8.161369193154035e-05,
+      "loss": 1.8116,
+      "step": 26641
+    },
+    {
+      "epoch": 0.5920444444444445,
+      "grad_norm": 2.008342981338501,
+      "learning_rate": 8.160924649922206e-05,
+      "loss": 2.3095,
+      "step": 26642
+    },
+    {
+      "epoch": 0.5920666666666666,
+      "grad_norm": 1.9992581605911255,
+      "learning_rate": 8.160480106690376e-05,
+      "loss": 1.831,
+      "step": 26643
+    },
+    {
+      "epoch": 0.5920888888888889,
+      "grad_norm": 1.7105638980865479,
+      "learning_rate": 8.160035563458547e-05,
+      "loss": 1.7077,
+      "step": 26644
+    },
+    {
+      "epoch": 0.5921111111111111,
+      "grad_norm": 1.5414973497390747,
+      "learning_rate": 8.159591020226717e-05,
+      "loss": 1.4228,
+      "step": 26645
+    },
+    {
+      "epoch": 0.5921333333333333,
+      "grad_norm": 1.8841001987457275,
+      "learning_rate": 8.159146476994887e-05,
+      "loss": 1.8394,
+      "step": 26646
+    },
+    {
+      "epoch": 0.5921555555555555,
+      "grad_norm": 1.8836796283721924,
+      "learning_rate": 8.15870193376306e-05,
+      "loss": 1.7217,
+      "step": 26647
+    },
+    {
+      "epoch": 0.5921777777777778,
+      "grad_norm": 2.0066640377044678,
+      "learning_rate": 8.15825739053123e-05,
+      "loss": 1.984,
+      "step": 26648
+    },
+    {
+      "epoch": 0.5922,
+      "grad_norm": 2.158877372741699,
+      "learning_rate": 8.1578128472994e-05,
+      "loss": 2.005,
+      "step": 26649
+    },
+    {
+      "epoch": 0.5922222222222222,
+      "grad_norm": 1.6916780471801758,
+      "learning_rate": 8.157368304067571e-05,
+      "loss": 1.3073,
+      "step": 26650
+    },
+    {
+      "epoch": 0.5922444444444445,
+      "grad_norm": 1.0765386819839478,
+      "learning_rate": 8.156923760835742e-05,
+      "loss": 1.078,
+      "step": 26651
+    },
+    {
+      "epoch": 0.5922666666666667,
+      "grad_norm": 1.3867547512054443,
+      "learning_rate": 8.156479217603913e-05,
+      "loss": 2.3633,
+      "step": 26652
+    },
+    {
+      "epoch": 0.5922888888888889,
+      "grad_norm": 1.541346788406372,
+      "learning_rate": 8.156034674372082e-05,
+      "loss": 2.5147,
+      "step": 26653
+    },
+    {
+      "epoch": 0.5923111111111111,
+      "grad_norm": 1.302173376083374,
+      "learning_rate": 8.155590131140253e-05,
+      "loss": 1.8573,
+      "step": 26654
+    },
+    {
+      "epoch": 0.5923333333333334,
+      "grad_norm": 1.4141172170639038,
+      "learning_rate": 8.155145587908424e-05,
+      "loss": 1.8953,
+      "step": 26655
+    },
+    {
+      "epoch": 0.5923555555555555,
+      "grad_norm": 1.8537484407424927,
+      "learning_rate": 8.154701044676595e-05,
+      "loss": 2.6232,
+      "step": 26656
+    },
+    {
+      "epoch": 0.5923777777777778,
+      "grad_norm": 1.457228660583496,
+      "learning_rate": 8.154256501444766e-05,
+      "loss": 2.0644,
+      "step": 26657
+    },
+    {
+      "epoch": 0.5924,
+      "grad_norm": 1.6981463432312012,
+      "learning_rate": 8.153811958212937e-05,
+      "loss": 1.7786,
+      "step": 26658
+    },
+    {
+      "epoch": 0.5924222222222222,
+      "grad_norm": 1.579250693321228,
+      "learning_rate": 8.153367414981107e-05,
+      "loss": 2.5667,
+      "step": 26659
+    },
+    {
+      "epoch": 0.5924444444444444,
+      "grad_norm": 1.561820387840271,
+      "learning_rate": 8.152922871749278e-05,
+      "loss": 2.0428,
+      "step": 26660
+    },
+    {
+      "epoch": 0.5924666666666667,
+      "grad_norm": 1.5794981718063354,
+      "learning_rate": 8.152478328517449e-05,
+      "loss": 2.0572,
+      "step": 26661
+    },
+    {
+      "epoch": 0.5924888888888888,
+      "grad_norm": 2.4824812412261963,
+      "learning_rate": 8.15203378528562e-05,
+      "loss": 1.0123,
+      "step": 26662
+    },
+    {
+      "epoch": 0.5925111111111111,
+      "grad_norm": 1.656442403793335,
+      "learning_rate": 8.15158924205379e-05,
+      "loss": 2.3156,
+      "step": 26663
+    },
+    {
+      "epoch": 0.5925333333333334,
+      "grad_norm": 1.7113794088363647,
+      "learning_rate": 8.15114469882196e-05,
+      "loss": 2.2904,
+      "step": 26664
+    },
+    {
+      "epoch": 0.5925555555555555,
+      "grad_norm": 1.4261407852172852,
+      "learning_rate": 8.150700155590131e-05,
+      "loss": 1.5333,
+      "step": 26665
+    },
+    {
+      "epoch": 0.5925777777777778,
+      "grad_norm": 1.7409135103225708,
+      "learning_rate": 8.150255612358302e-05,
+      "loss": 2.3401,
+      "step": 26666
+    },
+    {
+      "epoch": 0.5926,
+      "grad_norm": 1.390152096748352,
+      "learning_rate": 8.149811069126473e-05,
+      "loss": 1.2631,
+      "step": 26667
+    },
+    {
+      "epoch": 0.5926222222222223,
+      "grad_norm": 1.422816514968872,
+      "learning_rate": 8.149366525894644e-05,
+      "loss": 1.8948,
+      "step": 26668
+    },
+    {
+      "epoch": 0.5926444444444444,
+      "grad_norm": 1.5520660877227783,
+      "learning_rate": 8.148921982662814e-05,
+      "loss": 2.1724,
+      "step": 26669
+    },
+    {
+      "epoch": 0.5926666666666667,
+      "grad_norm": 1.649167776107788,
+      "learning_rate": 8.148477439430985e-05,
+      "loss": 2.0536,
+      "step": 26670
+    },
+    {
+      "epoch": 0.5926888888888889,
+      "grad_norm": 1.961836576461792,
+      "learning_rate": 8.148032896199156e-05,
+      "loss": 2.153,
+      "step": 26671
+    },
+    {
+      "epoch": 0.5927111111111111,
+      "grad_norm": 1.7347272634506226,
+      "learning_rate": 8.147588352967327e-05,
+      "loss": 2.1467,
+      "step": 26672
+    },
+    {
+      "epoch": 0.5927333333333333,
+      "grad_norm": 1.6931570768356323,
+      "learning_rate": 8.147143809735496e-05,
+      "loss": 1.6617,
+      "step": 26673
+    },
+    {
+      "epoch": 0.5927555555555556,
+      "grad_norm": 1.6291333436965942,
+      "learning_rate": 8.146699266503667e-05,
+      "loss": 1.7673,
+      "step": 26674
+    },
+    {
+      "epoch": 0.5927777777777777,
+      "grad_norm": 1.5067996978759766,
+      "learning_rate": 8.14625472327184e-05,
+      "loss": 1.8728,
+      "step": 26675
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 1.7630198001861572,
+      "learning_rate": 8.145810180040009e-05,
+      "loss": 1.6776,
+      "step": 26676
+    },
+    {
+      "epoch": 0.5928222222222223,
+      "grad_norm": 1.9861774444580078,
+      "learning_rate": 8.14536563680818e-05,
+      "loss": 1.7904,
+      "step": 26677
+    },
+    {
+      "epoch": 0.5928444444444444,
+      "grad_norm": 1.9535640478134155,
+      "learning_rate": 8.144921093576351e-05,
+      "loss": 2.577,
+      "step": 26678
+    },
+    {
+      "epoch": 0.5928666666666667,
+      "grad_norm": 1.636690616607666,
+      "learning_rate": 8.144476550344522e-05,
+      "loss": 2.0611,
+      "step": 26679
+    },
+    {
+      "epoch": 0.5928888888888889,
+      "grad_norm": 1.8471348285675049,
+      "learning_rate": 8.144032007112692e-05,
+      "loss": 1.5789,
+      "step": 26680
+    },
+    {
+      "epoch": 0.5929111111111112,
+      "grad_norm": 1.6389126777648926,
+      "learning_rate": 8.143587463880863e-05,
+      "loss": 2.0829,
+      "step": 26681
+    },
+    {
+      "epoch": 0.5929333333333333,
+      "grad_norm": 2.013946771621704,
+      "learning_rate": 8.143142920649034e-05,
+      "loss": 2.1446,
+      "step": 26682
+    },
+    {
+      "epoch": 0.5929555555555556,
+      "grad_norm": 1.6818547248840332,
+      "learning_rate": 8.142698377417203e-05,
+      "loss": 1.6905,
+      "step": 26683
+    },
+    {
+      "epoch": 0.5929777777777778,
+      "grad_norm": 1.9250750541687012,
+      "learning_rate": 8.142253834185376e-05,
+      "loss": 2.1326,
+      "step": 26684
+    },
+    {
+      "epoch": 0.593,
+      "grad_norm": 1.8171052932739258,
+      "learning_rate": 8.141809290953546e-05,
+      "loss": 1.8738,
+      "step": 26685
+    },
+    {
+      "epoch": 0.5930222222222222,
+      "grad_norm": 1.7382375001907349,
+      "learning_rate": 8.141364747721716e-05,
+      "loss": 1.8583,
+      "step": 26686
+    },
+    {
+      "epoch": 0.5930444444444445,
+      "grad_norm": 1.8500169515609741,
+      "learning_rate": 8.140920204489887e-05,
+      "loss": 1.7788,
+      "step": 26687
+    },
+    {
+      "epoch": 0.5930666666666666,
+      "grad_norm": 1.9859305620193481,
+      "learning_rate": 8.140475661258058e-05,
+      "loss": 1.9631,
+      "step": 26688
+    },
+    {
+      "epoch": 0.5930888888888889,
+      "grad_norm": 2.013277769088745,
+      "learning_rate": 8.140031118026229e-05,
+      "loss": 2.091,
+      "step": 26689
+    },
+    {
+      "epoch": 0.5931111111111111,
+      "grad_norm": 1.6553006172180176,
+      "learning_rate": 8.139586574794399e-05,
+      "loss": 2.0575,
+      "step": 26690
+    },
+    {
+      "epoch": 0.5931333333333333,
+      "grad_norm": 1.632489800453186,
+      "learning_rate": 8.13914203156257e-05,
+      "loss": 1.5623,
+      "step": 26691
+    },
+    {
+      "epoch": 0.5931555555555555,
+      "grad_norm": 1.8890914916992188,
+      "learning_rate": 8.13869748833074e-05,
+      "loss": 1.8982,
+      "step": 26692
+    },
+    {
+      "epoch": 0.5931777777777778,
+      "grad_norm": 1.7432690858840942,
+      "learning_rate": 8.138252945098911e-05,
+      "loss": 2.0591,
+      "step": 26693
+    },
+    {
+      "epoch": 0.5932,
+      "grad_norm": 1.7689228057861328,
+      "learning_rate": 8.137808401867082e-05,
+      "loss": 1.8778,
+      "step": 26694
+    },
+    {
+      "epoch": 0.5932222222222222,
+      "grad_norm": 1.571394920349121,
+      "learning_rate": 8.137363858635253e-05,
+      "loss": 1.9016,
+      "step": 26695
+    },
+    {
+      "epoch": 0.5932444444444445,
+      "grad_norm": 1.9821078777313232,
+      "learning_rate": 8.136919315403423e-05,
+      "loss": 2.078,
+      "step": 26696
+    },
+    {
+      "epoch": 0.5932666666666667,
+      "grad_norm": 1.533720850944519,
+      "learning_rate": 8.136474772171594e-05,
+      "loss": 1.6173,
+      "step": 26697
+    },
+    {
+      "epoch": 0.5932888888888889,
+      "grad_norm": 2.0955474376678467,
+      "learning_rate": 8.136030228939765e-05,
+      "loss": 1.8444,
+      "step": 26698
+    },
+    {
+      "epoch": 0.5933111111111111,
+      "grad_norm": 1.8464515209197998,
+      "learning_rate": 8.135585685707936e-05,
+      "loss": 1.9813,
+      "step": 26699
+    },
+    {
+      "epoch": 0.5933333333333334,
+      "grad_norm": 1.3876664638519287,
+      "learning_rate": 8.135141142476105e-05,
+      "loss": 1.0008,
+      "step": 26700
+    },
+    {
+      "epoch": 0.5933555555555555,
+      "grad_norm": 1.6724517345428467,
+      "learning_rate": 8.134696599244276e-05,
+      "loss": 2.8459,
+      "step": 26701
+    },
+    {
+      "epoch": 0.5933777777777778,
+      "grad_norm": 1.4604490995407104,
+      "learning_rate": 8.134252056012447e-05,
+      "loss": 2.2689,
+      "step": 26702
+    },
+    {
+      "epoch": 0.5934,
+      "grad_norm": 1.372445821762085,
+      "learning_rate": 8.133807512780618e-05,
+      "loss": 2.0997,
+      "step": 26703
+    },
+    {
+      "epoch": 0.5934222222222222,
+      "grad_norm": 1.5143563747406006,
+      "learning_rate": 8.13336296954879e-05,
+      "loss": 2.3801,
+      "step": 26704
+    },
+    {
+      "epoch": 0.5934444444444444,
+      "grad_norm": 1.5440425872802734,
+      "learning_rate": 8.13291842631696e-05,
+      "loss": 2.1438,
+      "step": 26705
+    },
+    {
+      "epoch": 0.5934666666666667,
+      "grad_norm": 1.5724350214004517,
+      "learning_rate": 8.13247388308513e-05,
+      "loss": 2.0209,
+      "step": 26706
+    },
+    {
+      "epoch": 0.5934888888888888,
+      "grad_norm": 1.6102594137191772,
+      "learning_rate": 8.132029339853301e-05,
+      "loss": 1.9863,
+      "step": 26707
+    },
+    {
+      "epoch": 0.5935111111111111,
+      "grad_norm": 1.4920109510421753,
+      "learning_rate": 8.131584796621472e-05,
+      "loss": 2.2177,
+      "step": 26708
+    },
+    {
+      "epoch": 0.5935333333333334,
+      "grad_norm": 1.2547584772109985,
+      "learning_rate": 8.131140253389643e-05,
+      "loss": 1.5696,
+      "step": 26709
+    },
+    {
+      "epoch": 0.5935555555555555,
+      "grad_norm": 1.7736376523971558,
+      "learning_rate": 8.130695710157812e-05,
+      "loss": 1.8298,
+      "step": 26710
+    },
+    {
+      "epoch": 0.5935777777777778,
+      "grad_norm": 1.727866530418396,
+      "learning_rate": 8.130251166925983e-05,
+      "loss": 2.2899,
+      "step": 26711
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 1.626160979270935,
+      "learning_rate": 8.129806623694156e-05,
+      "loss": 2.3531,
+      "step": 26712
+    },
+    {
+      "epoch": 0.5936222222222223,
+      "grad_norm": 1.6291509866714478,
+      "learning_rate": 8.129362080462325e-05,
+      "loss": 1.8856,
+      "step": 26713
+    },
+    {
+      "epoch": 0.5936444444444444,
+      "grad_norm": 1.7645621299743652,
+      "learning_rate": 8.128917537230496e-05,
+      "loss": 1.5955,
+      "step": 26714
+    },
+    {
+      "epoch": 0.5936666666666667,
+      "grad_norm": 1.6563748121261597,
+      "learning_rate": 8.128472993998667e-05,
+      "loss": 2.3929,
+      "step": 26715
+    },
+    {
+      "epoch": 0.5936888888888889,
+      "grad_norm": 1.5260820388793945,
+      "learning_rate": 8.128028450766837e-05,
+      "loss": 2.425,
+      "step": 26716
+    },
+    {
+      "epoch": 0.5937111111111111,
+      "grad_norm": 0.6744555830955505,
+      "learning_rate": 8.127583907535008e-05,
+      "loss": 0.0275,
+      "step": 26717
+    },
+    {
+      "epoch": 0.5937333333333333,
+      "grad_norm": 1.311850905418396,
+      "learning_rate": 8.127139364303179e-05,
+      "loss": 1.7537,
+      "step": 26718
+    },
+    {
+      "epoch": 0.5937555555555556,
+      "grad_norm": 1.5667086839675903,
+      "learning_rate": 8.12669482107135e-05,
+      "loss": 2.1051,
+      "step": 26719
+    },
+    {
+      "epoch": 0.5937777777777777,
+      "grad_norm": 1.6409789323806763,
+      "learning_rate": 8.126250277839519e-05,
+      "loss": 2.4806,
+      "step": 26720
+    },
+    {
+      "epoch": 0.5938,
+      "grad_norm": 1.585787057876587,
+      "learning_rate": 8.125805734607692e-05,
+      "loss": 2.4312,
+      "step": 26721
+    },
+    {
+      "epoch": 0.5938222222222223,
+      "grad_norm": 1.5881736278533936,
+      "learning_rate": 8.125361191375863e-05,
+      "loss": 2.2744,
+      "step": 26722
+    },
+    {
+      "epoch": 0.5938444444444444,
+      "grad_norm": 1.6588823795318604,
+      "learning_rate": 8.124916648144032e-05,
+      "loss": 1.711,
+      "step": 26723
+    },
+    {
+      "epoch": 0.5938666666666667,
+      "grad_norm": 1.717848300933838,
+      "learning_rate": 8.124472104912203e-05,
+      "loss": 2.1547,
+      "step": 26724
+    },
+    {
+      "epoch": 0.5938888888888889,
+      "grad_norm": 1.8647304773330688,
+      "learning_rate": 8.124027561680374e-05,
+      "loss": 2.0538,
+      "step": 26725
+    },
+    {
+      "epoch": 0.5939111111111111,
+      "grad_norm": 1.7777903079986572,
+      "learning_rate": 8.123583018448544e-05,
+      "loss": 2.1023,
+      "step": 26726
+    },
+    {
+      "epoch": 0.5939333333333333,
+      "grad_norm": 1.7060556411743164,
+      "learning_rate": 8.123138475216715e-05,
+      "loss": 1.4759,
+      "step": 26727
+    },
+    {
+      "epoch": 0.5939555555555556,
+      "grad_norm": 1.4728893041610718,
+      "learning_rate": 8.122693931984886e-05,
+      "loss": 1.4794,
+      "step": 26728
+    },
+    {
+      "epoch": 0.5939777777777778,
+      "grad_norm": 1.678420901298523,
+      "learning_rate": 8.122249388753057e-05,
+      "loss": 1.7817,
+      "step": 26729
+    },
+    {
+      "epoch": 0.594,
+      "grad_norm": 1.4201163053512573,
+      "learning_rate": 8.121804845521228e-05,
+      "loss": 1.6628,
+      "step": 26730
+    },
+    {
+      "epoch": 0.5940222222222222,
+      "grad_norm": 1.512225866317749,
+      "learning_rate": 8.121360302289399e-05,
+      "loss": 1.4771,
+      "step": 26731
+    },
+    {
+      "epoch": 0.5940444444444445,
+      "grad_norm": 1.8955464363098145,
+      "learning_rate": 8.12091575905757e-05,
+      "loss": 1.166,
+      "step": 26732
+    },
+    {
+      "epoch": 0.5940666666666666,
+      "grad_norm": 1.1601372957229614,
+      "learning_rate": 8.120471215825739e-05,
+      "loss": 1.0652,
+      "step": 26733
+    },
+    {
+      "epoch": 0.5940888888888889,
+      "grad_norm": 1.6053649187088013,
+      "learning_rate": 8.12002667259391e-05,
+      "loss": 1.7017,
+      "step": 26734
+    },
+    {
+      "epoch": 0.5941111111111111,
+      "grad_norm": 1.9037258625030518,
+      "learning_rate": 8.119582129362081e-05,
+      "loss": 2.0369,
+      "step": 26735
+    },
+    {
+      "epoch": 0.5941333333333333,
+      "grad_norm": 1.9601517915725708,
+      "learning_rate": 8.119137586130252e-05,
+      "loss": 1.8534,
+      "step": 26736
+    },
+    {
+      "epoch": 0.5941555555555555,
+      "grad_norm": 1.8852641582489014,
+      "learning_rate": 8.118693042898422e-05,
+      "loss": 1.8805,
+      "step": 26737
+    },
+    {
+      "epoch": 0.5941777777777778,
+      "grad_norm": 1.643731713294983,
+      "learning_rate": 8.118248499666593e-05,
+      "loss": 2.1002,
+      "step": 26738
+    },
+    {
+      "epoch": 0.5942,
+      "grad_norm": 1.9783738851547241,
+      "learning_rate": 8.117803956434764e-05,
+      "loss": 1.8242,
+      "step": 26739
+    },
+    {
+      "epoch": 0.5942222222222222,
+      "grad_norm": 1.477020502090454,
+      "learning_rate": 8.117359413202934e-05,
+      "loss": 1.3409,
+      "step": 26740
+    },
+    {
+      "epoch": 0.5942444444444445,
+      "grad_norm": 2.2099764347076416,
+      "learning_rate": 8.116914869971105e-05,
+      "loss": 1.9382,
+      "step": 26741
+    },
+    {
+      "epoch": 0.5942666666666667,
+      "grad_norm": 1.8193410634994507,
+      "learning_rate": 8.116470326739276e-05,
+      "loss": 1.6036,
+      "step": 26742
+    },
+    {
+      "epoch": 0.5942888888888889,
+      "grad_norm": 1.7083810567855835,
+      "learning_rate": 8.116025783507446e-05,
+      "loss": 1.7845,
+      "step": 26743
+    },
+    {
+      "epoch": 0.5943111111111111,
+      "grad_norm": 1.6544417142868042,
+      "learning_rate": 8.115581240275617e-05,
+      "loss": 1.7903,
+      "step": 26744
+    },
+    {
+      "epoch": 0.5943333333333334,
+      "grad_norm": 1.5880299806594849,
+      "learning_rate": 8.115136697043788e-05,
+      "loss": 1.7088,
+      "step": 26745
+    },
+    {
+      "epoch": 0.5943555555555555,
+      "grad_norm": 1.7833446264266968,
+      "learning_rate": 8.114692153811959e-05,
+      "loss": 2.0259,
+      "step": 26746
+    },
+    {
+      "epoch": 0.5943777777777778,
+      "grad_norm": 1.6049708127975464,
+      "learning_rate": 8.114247610580128e-05,
+      "loss": 1.682,
+      "step": 26747
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 2.0446813106536865,
+      "learning_rate": 8.1138030673483e-05,
+      "loss": 2.0802,
+      "step": 26748
+    },
+    {
+      "epoch": 0.5944222222222222,
+      "grad_norm": 2.010500192642212,
+      "learning_rate": 8.113358524116472e-05,
+      "loss": 2.2086,
+      "step": 26749
+    },
+    {
+      "epoch": 0.5944444444444444,
+      "grad_norm": 2.6268837451934814,
+      "learning_rate": 8.112913980884641e-05,
+      "loss": 1.9425,
+      "step": 26750
+    },
+    {
+      "epoch": 0.5944666666666667,
+      "grad_norm": 1.661972999572754,
+      "learning_rate": 8.112469437652812e-05,
+      "loss": 1.2295,
+      "step": 26751
+    },
+    {
+      "epoch": 0.5944888888888888,
+      "grad_norm": 1.4331251382827759,
+      "learning_rate": 8.112024894420983e-05,
+      "loss": 2.305,
+      "step": 26752
+    },
+    {
+      "epoch": 0.5945111111111111,
+      "grad_norm": 1.3672977685928345,
+      "learning_rate": 8.111580351189153e-05,
+      "loss": 2.128,
+      "step": 26753
+    },
+    {
+      "epoch": 0.5945333333333334,
+      "grad_norm": 1.5451977252960205,
+      "learning_rate": 8.111135807957324e-05,
+      "loss": 2.3535,
+      "step": 26754
+    },
+    {
+      "epoch": 0.5945555555555555,
+      "grad_norm": 1.458953619003296,
+      "learning_rate": 8.110691264725495e-05,
+      "loss": 2.1002,
+      "step": 26755
+    },
+    {
+      "epoch": 0.5945777777777778,
+      "grad_norm": 1.6705247163772583,
+      "learning_rate": 8.110246721493666e-05,
+      "loss": 2.0621,
+      "step": 26756
+    },
+    {
+      "epoch": 0.5946,
+      "grad_norm": 1.4237427711486816,
+      "learning_rate": 8.109802178261835e-05,
+      "loss": 2.0543,
+      "step": 26757
+    },
+    {
+      "epoch": 0.5946222222222223,
+      "grad_norm": 1.5021898746490479,
+      "learning_rate": 8.109357635030008e-05,
+      "loss": 1.9559,
+      "step": 26758
+    },
+    {
+      "epoch": 0.5946444444444444,
+      "grad_norm": 1.3429763317108154,
+      "learning_rate": 8.108913091798179e-05,
+      "loss": 1.5675,
+      "step": 26759
+    },
+    {
+      "epoch": 0.5946666666666667,
+      "grad_norm": 2.030367851257324,
+      "learning_rate": 8.108468548566348e-05,
+      "loss": 2.4005,
+      "step": 26760
+    },
+    {
+      "epoch": 0.5946888888888889,
+      "grad_norm": 1.536736011505127,
+      "learning_rate": 8.108024005334519e-05,
+      "loss": 1.945,
+      "step": 26761
+    },
+    {
+      "epoch": 0.5947111111111111,
+      "grad_norm": 1.4390614032745361,
+      "learning_rate": 8.10757946210269e-05,
+      "loss": 1.7824,
+      "step": 26762
+    },
+    {
+      "epoch": 0.5947333333333333,
+      "grad_norm": 1.8407021760940552,
+      "learning_rate": 8.10713491887086e-05,
+      "loss": 2.376,
+      "step": 26763
+    },
+    {
+      "epoch": 0.5947555555555556,
+      "grad_norm": 1.6480615139007568,
+      "learning_rate": 8.106690375639031e-05,
+      "loss": 2.0424,
+      "step": 26764
+    },
+    {
+      "epoch": 0.5947777777777777,
+      "grad_norm": 1.8025591373443604,
+      "learning_rate": 8.106245832407202e-05,
+      "loss": 1.9488,
+      "step": 26765
+    },
+    {
+      "epoch": 0.5948,
+      "grad_norm": 1.4991320371627808,
+      "learning_rate": 8.105801289175373e-05,
+      "loss": 0.7819,
+      "step": 26766
+    },
+    {
+      "epoch": 0.5948222222222223,
+      "grad_norm": 1.5261831283569336,
+      "learning_rate": 8.105356745943544e-05,
+      "loss": 2.0639,
+      "step": 26767
+    },
+    {
+      "epoch": 0.5948444444444444,
+      "grad_norm": 1.889927625656128,
+      "learning_rate": 8.104912202711715e-05,
+      "loss": 1.8238,
+      "step": 26768
+    },
+    {
+      "epoch": 0.5948666666666667,
+      "grad_norm": 1.794587254524231,
+      "learning_rate": 8.104467659479886e-05,
+      "loss": 2.3345,
+      "step": 26769
+    },
+    {
+      "epoch": 0.5948888888888889,
+      "grad_norm": 1.5195122957229614,
+      "learning_rate": 8.104023116248055e-05,
+      "loss": 1.3055,
+      "step": 26770
+    },
+    {
+      "epoch": 0.5949111111111111,
+      "grad_norm": 2.0479865074157715,
+      "learning_rate": 8.103578573016226e-05,
+      "loss": 2.0186,
+      "step": 26771
+    },
+    {
+      "epoch": 0.5949333333333333,
+      "grad_norm": 1.65352201461792,
+      "learning_rate": 8.103134029784397e-05,
+      "loss": 1.87,
+      "step": 26772
+    },
+    {
+      "epoch": 0.5949555555555556,
+      "grad_norm": 1.6952732801437378,
+      "learning_rate": 8.102689486552567e-05,
+      "loss": 1.7593,
+      "step": 26773
+    },
+    {
+      "epoch": 0.5949777777777778,
+      "grad_norm": 1.7520891427993774,
+      "learning_rate": 8.102244943320738e-05,
+      "loss": 2.2394,
+      "step": 26774
+    },
+    {
+      "epoch": 0.595,
+      "grad_norm": 1.7385798692703247,
+      "learning_rate": 8.101800400088909e-05,
+      "loss": 2.1476,
+      "step": 26775
+    },
+    {
+      "epoch": 0.5950222222222222,
+      "grad_norm": 1.6520934104919434,
+      "learning_rate": 8.10135585685708e-05,
+      "loss": 2.0333,
+      "step": 26776
+    },
+    {
+      "epoch": 0.5950444444444445,
+      "grad_norm": 1.8042887449264526,
+      "learning_rate": 8.10091131362525e-05,
+      "loss": 1.724,
+      "step": 26777
+    },
+    {
+      "epoch": 0.5950666666666666,
+      "grad_norm": 1.5326650142669678,
+      "learning_rate": 8.100466770393422e-05,
+      "loss": 1.7221,
+      "step": 26778
+    },
+    {
+      "epoch": 0.5950888888888889,
+      "grad_norm": 1.6813794374465942,
+      "learning_rate": 8.100022227161593e-05,
+      "loss": 1.9858,
+      "step": 26779
+    },
+    {
+      "epoch": 0.5951111111111111,
+      "grad_norm": 1.7051405906677246,
+      "learning_rate": 8.099577683929762e-05,
+      "loss": 1.801,
+      "step": 26780
+    },
+    {
+      "epoch": 0.5951333333333333,
+      "grad_norm": 1.1973179578781128,
+      "learning_rate": 8.099133140697933e-05,
+      "loss": 0.7489,
+      "step": 26781
+    },
+    {
+      "epoch": 0.5951555555555555,
+      "grad_norm": 1.8290960788726807,
+      "learning_rate": 8.098688597466104e-05,
+      "loss": 1.9373,
+      "step": 26782
+    },
+    {
+      "epoch": 0.5951777777777778,
+      "grad_norm": 1.7020169496536255,
+      "learning_rate": 8.098244054234274e-05,
+      "loss": 1.494,
+      "step": 26783
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 1.893183708190918,
+      "learning_rate": 8.097799511002445e-05,
+      "loss": 2.2692,
+      "step": 26784
+    },
+    {
+      "epoch": 0.5952222222222222,
+      "grad_norm": 1.7478630542755127,
+      "learning_rate": 8.097354967770616e-05,
+      "loss": 1.8236,
+      "step": 26785
+    },
+    {
+      "epoch": 0.5952444444444445,
+      "grad_norm": 1.9741427898406982,
+      "learning_rate": 8.096910424538788e-05,
+      "loss": 2.2287,
+      "step": 26786
+    },
+    {
+      "epoch": 0.5952666666666667,
+      "grad_norm": 1.6652030944824219,
+      "learning_rate": 8.096465881306957e-05,
+      "loss": 1.9211,
+      "step": 26787
+    },
+    {
+      "epoch": 0.5952888888888889,
+      "grad_norm": 1.9754842519760132,
+      "learning_rate": 8.096021338075128e-05,
+      "loss": 1.8296,
+      "step": 26788
+    },
+    {
+      "epoch": 0.5953111111111111,
+      "grad_norm": 1.5150765180587769,
+      "learning_rate": 8.0955767948433e-05,
+      "loss": 1.3967,
+      "step": 26789
+    },
+    {
+      "epoch": 0.5953333333333334,
+      "grad_norm": 1.212402582168579,
+      "learning_rate": 8.095132251611469e-05,
+      "loss": 0.8705,
+      "step": 26790
+    },
+    {
+      "epoch": 0.5953555555555555,
+      "grad_norm": 1.5794655084609985,
+      "learning_rate": 8.09468770837964e-05,
+      "loss": 1.8254,
+      "step": 26791
+    },
+    {
+      "epoch": 0.5953777777777778,
+      "grad_norm": 1.7709836959838867,
+      "learning_rate": 8.094243165147811e-05,
+      "loss": 1.8013,
+      "step": 26792
+    },
+    {
+      "epoch": 0.5954,
+      "grad_norm": 1.4090503454208374,
+      "learning_rate": 8.093798621915982e-05,
+      "loss": 1.6456,
+      "step": 26793
+    },
+    {
+      "epoch": 0.5954222222222222,
+      "grad_norm": 1.6370187997817993,
+      "learning_rate": 8.093354078684152e-05,
+      "loss": 1.549,
+      "step": 26794
+    },
+    {
+      "epoch": 0.5954444444444444,
+      "grad_norm": 1.7430535554885864,
+      "learning_rate": 8.092909535452324e-05,
+      "loss": 1.9142,
+      "step": 26795
+    },
+    {
+      "epoch": 0.5954666666666667,
+      "grad_norm": 1.836463212966919,
+      "learning_rate": 8.092464992220495e-05,
+      "loss": 1.9958,
+      "step": 26796
+    },
+    {
+      "epoch": 0.5954888888888888,
+      "grad_norm": 1.962188720703125,
+      "learning_rate": 8.092020448988664e-05,
+      "loss": 2.1326,
+      "step": 26797
+    },
+    {
+      "epoch": 0.5955111111111111,
+      "grad_norm": 1.878197431564331,
+      "learning_rate": 8.091575905756835e-05,
+      "loss": 2.121,
+      "step": 26798
+    },
+    {
+      "epoch": 0.5955333333333334,
+      "grad_norm": 1.8089852333068848,
+      "learning_rate": 8.091131362525006e-05,
+      "loss": 1.5793,
+      "step": 26799
+    },
+    {
+      "epoch": 0.5955555555555555,
+      "grad_norm": 1.736074686050415,
+      "learning_rate": 8.090686819293176e-05,
+      "loss": 0.9735,
+      "step": 26800
+    },
+    {
+      "epoch": 0.5955777777777778,
+      "grad_norm": 1.3884100914001465,
+      "learning_rate": 8.090242276061347e-05,
+      "loss": 1.0546,
+      "step": 26801
+    },
+    {
+      "epoch": 0.5956,
+      "grad_norm": 1.429738163948059,
+      "learning_rate": 8.089797732829518e-05,
+      "loss": 2.3998,
+      "step": 26802
+    },
+    {
+      "epoch": 0.5956222222222223,
+      "grad_norm": 1.3541944026947021,
+      "learning_rate": 8.089353189597689e-05,
+      "loss": 2.2969,
+      "step": 26803
+    },
+    {
+      "epoch": 0.5956444444444444,
+      "grad_norm": 1.5858944654464722,
+      "learning_rate": 8.08890864636586e-05,
+      "loss": 2.3351,
+      "step": 26804
+    },
+    {
+      "epoch": 0.5956666666666667,
+      "grad_norm": 1.5756072998046875,
+      "learning_rate": 8.088464103134031e-05,
+      "loss": 2.4645,
+      "step": 26805
+    },
+    {
+      "epoch": 0.5956888888888889,
+      "grad_norm": 1.969294548034668,
+      "learning_rate": 8.088019559902202e-05,
+      "loss": 2.1922,
+      "step": 26806
+    },
+    {
+      "epoch": 0.5957111111111111,
+      "grad_norm": 1.397650122642517,
+      "learning_rate": 8.087575016670371e-05,
+      "loss": 2.1963,
+      "step": 26807
+    },
+    {
+      "epoch": 0.5957333333333333,
+      "grad_norm": 1.8563101291656494,
+      "learning_rate": 8.087130473438542e-05,
+      "loss": 1.9459,
+      "step": 26808
+    },
+    {
+      "epoch": 0.5957555555555556,
+      "grad_norm": 1.721784234046936,
+      "learning_rate": 8.086685930206713e-05,
+      "loss": 2.3333,
+      "step": 26809
+    },
+    {
+      "epoch": 0.5957777777777777,
+      "grad_norm": 1.3764325380325317,
+      "learning_rate": 8.086241386974883e-05,
+      "loss": 2.0502,
+      "step": 26810
+    },
+    {
+      "epoch": 0.5958,
+      "grad_norm": 1.489363670349121,
+      "learning_rate": 8.085796843743054e-05,
+      "loss": 2.0739,
+      "step": 26811
+    },
+    {
+      "epoch": 0.5958222222222223,
+      "grad_norm": 1.897639274597168,
+      "learning_rate": 8.085352300511225e-05,
+      "loss": 2.2408,
+      "step": 26812
+    },
+    {
+      "epoch": 0.5958444444444444,
+      "grad_norm": 1.3819931745529175,
+      "learning_rate": 8.084907757279396e-05,
+      "loss": 1.6007,
+      "step": 26813
+    },
+    {
+      "epoch": 0.5958666666666667,
+      "grad_norm": 1.9608386754989624,
+      "learning_rate": 8.084463214047567e-05,
+      "loss": 1.6374,
+      "step": 26814
+    },
+    {
+      "epoch": 0.5958888888888889,
+      "grad_norm": 1.374138593673706,
+      "learning_rate": 8.084018670815738e-05,
+      "loss": 1.6545,
+      "step": 26815
+    },
+    {
+      "epoch": 0.5959111111111111,
+      "grad_norm": 1.7151923179626465,
+      "learning_rate": 8.083574127583909e-05,
+      "loss": 1.9196,
+      "step": 26816
+    },
+    {
+      "epoch": 0.5959333333333333,
+      "grad_norm": 1.099605917930603,
+      "learning_rate": 8.083129584352078e-05,
+      "loss": 1.0795,
+      "step": 26817
+    },
+    {
+      "epoch": 0.5959555555555556,
+      "grad_norm": 1.6276981830596924,
+      "learning_rate": 8.082685041120249e-05,
+      "loss": 1.5975,
+      "step": 26818
+    },
+    {
+      "epoch": 0.5959777777777778,
+      "grad_norm": 1.174397587776184,
+      "learning_rate": 8.08224049788842e-05,
+      "loss": 0.9732,
+      "step": 26819
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 1.9741249084472656,
+      "learning_rate": 8.08179595465659e-05,
+      "loss": 2.1333,
+      "step": 26820
+    },
+    {
+      "epoch": 0.5960222222222222,
+      "grad_norm": 1.6376827955245972,
+      "learning_rate": 8.081351411424761e-05,
+      "loss": 1.8163,
+      "step": 26821
+    },
+    {
+      "epoch": 0.5960444444444445,
+      "grad_norm": 1.5402346849441528,
+      "learning_rate": 8.080906868192932e-05,
+      "loss": 1.5948,
+      "step": 26822
+    },
+    {
+      "epoch": 0.5960666666666666,
+      "grad_norm": 1.6858675479888916,
+      "learning_rate": 8.080462324961104e-05,
+      "loss": 2.2483,
+      "step": 26823
+    },
+    {
+      "epoch": 0.5960888888888889,
+      "grad_norm": 1.6498875617980957,
+      "learning_rate": 8.080017781729274e-05,
+      "loss": 1.7901,
+      "step": 26824
+    },
+    {
+      "epoch": 0.5961111111111111,
+      "grad_norm": 1.6973052024841309,
+      "learning_rate": 8.079573238497445e-05,
+      "loss": 2.3223,
+      "step": 26825
+    },
+    {
+      "epoch": 0.5961333333333333,
+      "grad_norm": 1.4297351837158203,
+      "learning_rate": 8.079128695265616e-05,
+      "loss": 1.8346,
+      "step": 26826
+    },
+    {
+      "epoch": 0.5961555555555555,
+      "grad_norm": 1.4189152717590332,
+      "learning_rate": 8.078684152033785e-05,
+      "loss": 1.1534,
+      "step": 26827
+    },
+    {
+      "epoch": 0.5961777777777778,
+      "grad_norm": 1.5987457036972046,
+      "learning_rate": 8.078239608801956e-05,
+      "loss": 1.3904,
+      "step": 26828
+    },
+    {
+      "epoch": 0.5962,
+      "grad_norm": 1.690332055091858,
+      "learning_rate": 8.077795065570127e-05,
+      "loss": 1.9059,
+      "step": 26829
+    },
+    {
+      "epoch": 0.5962222222222222,
+      "grad_norm": 1.9026538133621216,
+      "learning_rate": 8.077350522338297e-05,
+      "loss": 1.976,
+      "step": 26830
+    },
+    {
+      "epoch": 0.5962444444444445,
+      "grad_norm": 1.623248815536499,
+      "learning_rate": 8.076905979106468e-05,
+      "loss": 1.9548,
+      "step": 26831
+    },
+    {
+      "epoch": 0.5962666666666666,
+      "grad_norm": 1.5890661478042603,
+      "learning_rate": 8.07646143587464e-05,
+      "loss": 1.2765,
+      "step": 26832
+    },
+    {
+      "epoch": 0.5962888888888889,
+      "grad_norm": 2.1992969512939453,
+      "learning_rate": 8.076016892642811e-05,
+      "loss": 2.1102,
+      "step": 26833
+    },
+    {
+      "epoch": 0.5963111111111111,
+      "grad_norm": 1.1912901401519775,
+      "learning_rate": 8.07557234941098e-05,
+      "loss": 0.7625,
+      "step": 26834
+    },
+    {
+      "epoch": 0.5963333333333334,
+      "grad_norm": 1.461975336074829,
+      "learning_rate": 8.075127806179151e-05,
+      "loss": 1.4908,
+      "step": 26835
+    },
+    {
+      "epoch": 0.5963555555555555,
+      "grad_norm": 2.027536630630493,
+      "learning_rate": 8.074683262947322e-05,
+      "loss": 2.1627,
+      "step": 26836
+    },
+    {
+      "epoch": 0.5963777777777778,
+      "grad_norm": 1.6185896396636963,
+      "learning_rate": 8.074238719715492e-05,
+      "loss": 1.3973,
+      "step": 26837
+    },
+    {
+      "epoch": 0.5964,
+      "grad_norm": 1.4064449071884155,
+      "learning_rate": 8.073794176483663e-05,
+      "loss": 0.8511,
+      "step": 26838
+    },
+    {
+      "epoch": 0.5964222222222222,
+      "grad_norm": 1.5895582437515259,
+      "learning_rate": 8.073349633251834e-05,
+      "loss": 1.9084,
+      "step": 26839
+    },
+    {
+      "epoch": 0.5964444444444444,
+      "grad_norm": 1.683510661125183,
+      "learning_rate": 8.072905090020005e-05,
+      "loss": 1.9935,
+      "step": 26840
+    },
+    {
+      "epoch": 0.5964666666666667,
+      "grad_norm": 1.5518105030059814,
+      "learning_rate": 8.072460546788176e-05,
+      "loss": 1.7772,
+      "step": 26841
+    },
+    {
+      "epoch": 0.5964888888888888,
+      "grad_norm": 1.6962943077087402,
+      "learning_rate": 8.072016003556347e-05,
+      "loss": 1.6512,
+      "step": 26842
+    },
+    {
+      "epoch": 0.5965111111111111,
+      "grad_norm": 1.9162324666976929,
+      "learning_rate": 8.071571460324518e-05,
+      "loss": 2.0387,
+      "step": 26843
+    },
+    {
+      "epoch": 0.5965333333333334,
+      "grad_norm": 1.843002438545227,
+      "learning_rate": 8.071126917092687e-05,
+      "loss": 1.7871,
+      "step": 26844
+    },
+    {
+      "epoch": 0.5965555555555555,
+      "grad_norm": 2.334334373474121,
+      "learning_rate": 8.070682373860858e-05,
+      "loss": 2.2448,
+      "step": 26845
+    },
+    {
+      "epoch": 0.5965777777777778,
+      "grad_norm": 1.9153010845184326,
+      "learning_rate": 8.07023783062903e-05,
+      "loss": 1.8737,
+      "step": 26846
+    },
+    {
+      "epoch": 0.5966,
+      "grad_norm": 1.7890713214874268,
+      "learning_rate": 8.069793287397199e-05,
+      "loss": 1.7194,
+      "step": 26847
+    },
+    {
+      "epoch": 0.5966222222222223,
+      "grad_norm": 2.0024993419647217,
+      "learning_rate": 8.06934874416537e-05,
+      "loss": 1.9438,
+      "step": 26848
+    },
+    {
+      "epoch": 0.5966444444444444,
+      "grad_norm": 1.8416032791137695,
+      "learning_rate": 8.068904200933541e-05,
+      "loss": 1.8522,
+      "step": 26849
+    },
+    {
+      "epoch": 0.5966666666666667,
+      "grad_norm": 1.50460946559906,
+      "learning_rate": 8.068459657701712e-05,
+      "loss": 1.2711,
+      "step": 26850
+    },
+    {
+      "epoch": 0.5966888888888889,
+      "grad_norm": 1.4480715990066528,
+      "learning_rate": 8.068015114469883e-05,
+      "loss": 2.5689,
+      "step": 26851
+    },
+    {
+      "epoch": 0.5967111111111111,
+      "grad_norm": 1.4781830310821533,
+      "learning_rate": 8.067570571238054e-05,
+      "loss": 2.4561,
+      "step": 26852
+    },
+    {
+      "epoch": 0.5967333333333333,
+      "grad_norm": 1.2978445291519165,
+      "learning_rate": 8.067126028006225e-05,
+      "loss": 1.1158,
+      "step": 26853
+    },
+    {
+      "epoch": 0.5967555555555556,
+      "grad_norm": 1.3282469511032104,
+      "learning_rate": 8.066681484774394e-05,
+      "loss": 2.0565,
+      "step": 26854
+    },
+    {
+      "epoch": 0.5967777777777777,
+      "grad_norm": 1.4978200197219849,
+      "learning_rate": 8.066236941542565e-05,
+      "loss": 1.8874,
+      "step": 26855
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 1.4156322479248047,
+      "learning_rate": 8.065792398310736e-05,
+      "loss": 2.0859,
+      "step": 26856
+    },
+    {
+      "epoch": 0.5968222222222223,
+      "grad_norm": 1.2539492845535278,
+      "learning_rate": 8.065347855078906e-05,
+      "loss": 1.0677,
+      "step": 26857
+    },
+    {
+      "epoch": 0.5968444444444444,
+      "grad_norm": 2.2956056594848633,
+      "learning_rate": 8.064903311847077e-05,
+      "loss": 2.5987,
+      "step": 26858
+    },
+    {
+      "epoch": 0.5968666666666667,
+      "grad_norm": 1.740151286125183,
+      "learning_rate": 8.064458768615248e-05,
+      "loss": 2.402,
+      "step": 26859
+    },
+    {
+      "epoch": 0.5968888888888889,
+      "grad_norm": 1.8559595346450806,
+      "learning_rate": 8.06401422538342e-05,
+      "loss": 1.8741,
+      "step": 26860
+    },
+    {
+      "epoch": 0.5969111111111111,
+      "grad_norm": 1.9488989114761353,
+      "learning_rate": 8.06356968215159e-05,
+      "loss": 2.3833,
+      "step": 26861
+    },
+    {
+      "epoch": 0.5969333333333333,
+      "grad_norm": 1.630929946899414,
+      "learning_rate": 8.06312513891976e-05,
+      "loss": 2.1018,
+      "step": 26862
+    },
+    {
+      "epoch": 0.5969555555555556,
+      "grad_norm": 1.4589755535125732,
+      "learning_rate": 8.062680595687932e-05,
+      "loss": 2.262,
+      "step": 26863
+    },
+    {
+      "epoch": 0.5969777777777778,
+      "grad_norm": 1.5186175107955933,
+      "learning_rate": 8.062236052456101e-05,
+      "loss": 2.0959,
+      "step": 26864
+    },
+    {
+      "epoch": 0.597,
+      "grad_norm": 1.5367112159729004,
+      "learning_rate": 8.061791509224272e-05,
+      "loss": 2.2553,
+      "step": 26865
+    },
+    {
+      "epoch": 0.5970222222222222,
+      "grad_norm": 1.5911413431167603,
+      "learning_rate": 8.061346965992443e-05,
+      "loss": 2.0367,
+      "step": 26866
+    },
+    {
+      "epoch": 0.5970444444444445,
+      "grad_norm": 1.7494949102401733,
+      "learning_rate": 8.060902422760613e-05,
+      "loss": 2.2267,
+      "step": 26867
+    },
+    {
+      "epoch": 0.5970666666666666,
+      "grad_norm": 1.4874684810638428,
+      "learning_rate": 8.060457879528784e-05,
+      "loss": 1.7659,
+      "step": 26868
+    },
+    {
+      "epoch": 0.5970888888888889,
+      "grad_norm": 1.7232106924057007,
+      "learning_rate": 8.060013336296956e-05,
+      "loss": 2.2392,
+      "step": 26869
+    },
+    {
+      "epoch": 0.5971111111111111,
+      "grad_norm": 1.575861930847168,
+      "learning_rate": 8.059568793065127e-05,
+      "loss": 1.9078,
+      "step": 26870
+    },
+    {
+      "epoch": 0.5971333333333333,
+      "grad_norm": 1.568634033203125,
+      "learning_rate": 8.059124249833297e-05,
+      "loss": 1.5753,
+      "step": 26871
+    },
+    {
+      "epoch": 0.5971555555555556,
+      "grad_norm": 1.902392864227295,
+      "learning_rate": 8.058679706601468e-05,
+      "loss": 2.2982,
+      "step": 26872
+    },
+    {
+      "epoch": 0.5971777777777778,
+      "grad_norm": 1.3850717544555664,
+      "learning_rate": 8.058235163369639e-05,
+      "loss": 1.8568,
+      "step": 26873
+    },
+    {
+      "epoch": 0.5972,
+      "grad_norm": 1.8104287385940552,
+      "learning_rate": 8.057790620137808e-05,
+      "loss": 1.9714,
+      "step": 26874
+    },
+    {
+      "epoch": 0.5972222222222222,
+      "grad_norm": 1.636537790298462,
+      "learning_rate": 8.057346076905979e-05,
+      "loss": 2.0433,
+      "step": 26875
+    },
+    {
+      "epoch": 0.5972444444444445,
+      "grad_norm": 1.733601689338684,
+      "learning_rate": 8.05690153367415e-05,
+      "loss": 1.7444,
+      "step": 26876
+    },
+    {
+      "epoch": 0.5972666666666666,
+      "grad_norm": 1.402464509010315,
+      "learning_rate": 8.056456990442321e-05,
+      "loss": 1.478,
+      "step": 26877
+    },
+    {
+      "epoch": 0.5972888888888889,
+      "grad_norm": 1.607865571975708,
+      "learning_rate": 8.056012447210492e-05,
+      "loss": 1.7318,
+      "step": 26878
+    },
+    {
+      "epoch": 0.5973111111111111,
+      "grad_norm": 1.7006738185882568,
+      "learning_rate": 8.055567903978663e-05,
+      "loss": 2.1914,
+      "step": 26879
+    },
+    {
+      "epoch": 0.5973333333333334,
+      "grad_norm": 1.8816903829574585,
+      "learning_rate": 8.055123360746834e-05,
+      "loss": 1.2202,
+      "step": 26880
+    },
+    {
+      "epoch": 0.5973555555555555,
+      "grad_norm": 1.9753090143203735,
+      "learning_rate": 8.054678817515004e-05,
+      "loss": 1.8315,
+      "step": 26881
+    },
+    {
+      "epoch": 0.5973777777777778,
+      "grad_norm": 1.8738881349563599,
+      "learning_rate": 8.054234274283174e-05,
+      "loss": 2.2523,
+      "step": 26882
+    },
+    {
+      "epoch": 0.5974,
+      "grad_norm": 1.477479338645935,
+      "learning_rate": 8.053789731051345e-05,
+      "loss": 1.6772,
+      "step": 26883
+    },
+    {
+      "epoch": 0.5974222222222222,
+      "grad_norm": 1.8067989349365234,
+      "learning_rate": 8.053345187819515e-05,
+      "loss": 1.8221,
+      "step": 26884
+    },
+    {
+      "epoch": 0.5974444444444444,
+      "grad_norm": 1.6162313222885132,
+      "learning_rate": 8.052900644587686e-05,
+      "loss": 1.7678,
+      "step": 26885
+    },
+    {
+      "epoch": 0.5974666666666667,
+      "grad_norm": 1.8906008005142212,
+      "learning_rate": 8.052456101355857e-05,
+      "loss": 1.9851,
+      "step": 26886
+    },
+    {
+      "epoch": 0.5974888888888888,
+      "grad_norm": 1.5765184164047241,
+      "learning_rate": 8.052011558124028e-05,
+      "loss": 1.7963,
+      "step": 26887
+    },
+    {
+      "epoch": 0.5975111111111111,
+      "grad_norm": 1.4867256879806519,
+      "learning_rate": 8.051567014892199e-05,
+      "loss": 1.7625,
+      "step": 26888
+    },
+    {
+      "epoch": 0.5975333333333334,
+      "grad_norm": 1.643934965133667,
+      "learning_rate": 8.05112247166037e-05,
+      "loss": 1.8958,
+      "step": 26889
+    },
+    {
+      "epoch": 0.5975555555555555,
+      "grad_norm": 1.6014454364776611,
+      "learning_rate": 8.050677928428541e-05,
+      "loss": 1.7788,
+      "step": 26890
+    },
+    {
+      "epoch": 0.5975777777777778,
+      "grad_norm": 3.079517126083374,
+      "learning_rate": 8.05023338519671e-05,
+      "loss": 1.5165,
+      "step": 26891
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 1.854453682899475,
+      "learning_rate": 8.049788841964881e-05,
+      "loss": 1.7971,
+      "step": 26892
+    },
+    {
+      "epoch": 0.5976222222222223,
+      "grad_norm": 1.4640413522720337,
+      "learning_rate": 8.049344298733052e-05,
+      "loss": 1.5388,
+      "step": 26893
+    },
+    {
+      "epoch": 0.5976444444444444,
+      "grad_norm": 1.8597694635391235,
+      "learning_rate": 8.048899755501222e-05,
+      "loss": 2.039,
+      "step": 26894
+    },
+    {
+      "epoch": 0.5976666666666667,
+      "grad_norm": 2.2400176525115967,
+      "learning_rate": 8.048455212269393e-05,
+      "loss": 2.0969,
+      "step": 26895
+    },
+    {
+      "epoch": 0.5976888888888889,
+      "grad_norm": 1.6783058643341064,
+      "learning_rate": 8.048010669037564e-05,
+      "loss": 1.7719,
+      "step": 26896
+    },
+    {
+      "epoch": 0.5977111111111111,
+      "grad_norm": 1.8262856006622314,
+      "learning_rate": 8.047566125805736e-05,
+      "loss": 1.6248,
+      "step": 26897
+    },
+    {
+      "epoch": 0.5977333333333333,
+      "grad_norm": 1.9277127981185913,
+      "learning_rate": 8.047121582573906e-05,
+      "loss": 1.6219,
+      "step": 26898
+    },
+    {
+      "epoch": 0.5977555555555556,
+      "grad_norm": 1.7251594066619873,
+      "learning_rate": 8.046677039342077e-05,
+      "loss": 1.7427,
+      "step": 26899
+    },
+    {
+      "epoch": 0.5977777777777777,
+      "grad_norm": 1.6436690092086792,
+      "learning_rate": 8.046232496110248e-05,
+      "loss": 1.4936,
+      "step": 26900
+    },
+    {
+      "epoch": 0.5978,
+      "grad_norm": 1.403838872909546,
+      "learning_rate": 8.045787952878417e-05,
+      "loss": 2.6362,
+      "step": 26901
+    },
+    {
+      "epoch": 0.5978222222222223,
+      "grad_norm": 1.383715271949768,
+      "learning_rate": 8.045343409646588e-05,
+      "loss": 2.1609,
+      "step": 26902
+    },
+    {
+      "epoch": 0.5978444444444444,
+      "grad_norm": 1.8453351259231567,
+      "learning_rate": 8.044898866414759e-05,
+      "loss": 2.2438,
+      "step": 26903
+    },
+    {
+      "epoch": 0.5978666666666667,
+      "grad_norm": 1.7187947034835815,
+      "learning_rate": 8.044454323182929e-05,
+      "loss": 2.2648,
+      "step": 26904
+    },
+    {
+      "epoch": 0.5978888888888889,
+      "grad_norm": 1.3632493019104004,
+      "learning_rate": 8.044009779951101e-05,
+      "loss": 2.1545,
+      "step": 26905
+    },
+    {
+      "epoch": 0.5979111111111111,
+      "grad_norm": 1.9141936302185059,
+      "learning_rate": 8.043565236719272e-05,
+      "loss": 1.6925,
+      "step": 26906
+    },
+    {
+      "epoch": 0.5979333333333333,
+      "grad_norm": 1.5372122526168823,
+      "learning_rate": 8.043120693487443e-05,
+      "loss": 2.392,
+      "step": 26907
+    },
+    {
+      "epoch": 0.5979555555555556,
+      "grad_norm": 1.2757940292358398,
+      "learning_rate": 8.042676150255613e-05,
+      "loss": 1.7866,
+      "step": 26908
+    },
+    {
+      "epoch": 0.5979777777777778,
+      "grad_norm": 1.9118685722351074,
+      "learning_rate": 8.042231607023784e-05,
+      "loss": 2.2627,
+      "step": 26909
+    },
+    {
+      "epoch": 0.598,
+      "grad_norm": 1.775957703590393,
+      "learning_rate": 8.041787063791955e-05,
+      "loss": 2.1612,
+      "step": 26910
+    },
+    {
+      "epoch": 0.5980222222222222,
+      "grad_norm": 1.5586014986038208,
+      "learning_rate": 8.041342520560124e-05,
+      "loss": 2.2685,
+      "step": 26911
+    },
+    {
+      "epoch": 0.5980444444444445,
+      "grad_norm": 1.7741633653640747,
+      "learning_rate": 8.040897977328295e-05,
+      "loss": 1.98,
+      "step": 26912
+    },
+    {
+      "epoch": 0.5980666666666666,
+      "grad_norm": 1.7802754640579224,
+      "learning_rate": 8.040453434096466e-05,
+      "loss": 2.5607,
+      "step": 26913
+    },
+    {
+      "epoch": 0.5980888888888889,
+      "grad_norm": 1.3747230768203735,
+      "learning_rate": 8.040008890864637e-05,
+      "loss": 1.3905,
+      "step": 26914
+    },
+    {
+      "epoch": 0.5981111111111111,
+      "grad_norm": 1.478218913078308,
+      "learning_rate": 8.039564347632808e-05,
+      "loss": 1.9997,
+      "step": 26915
+    },
+    {
+      "epoch": 0.5981333333333333,
+      "grad_norm": 2.119713544845581,
+      "learning_rate": 8.039119804400979e-05,
+      "loss": 1.7586,
+      "step": 26916
+    },
+    {
+      "epoch": 0.5981555555555556,
+      "grad_norm": 2.079786777496338,
+      "learning_rate": 8.03867526116915e-05,
+      "loss": 2.2577,
+      "step": 26917
+    },
+    {
+      "epoch": 0.5981777777777778,
+      "grad_norm": 1.048282504081726,
+      "learning_rate": 8.03823071793732e-05,
+      "loss": 0.9193,
+      "step": 26918
+    },
+    {
+      "epoch": 0.5982,
+      "grad_norm": 1.0226231813430786,
+      "learning_rate": 8.03778617470549e-05,
+      "loss": 1.0035,
+      "step": 26919
+    },
+    {
+      "epoch": 0.5982222222222222,
+      "grad_norm": 1.470382571220398,
+      "learning_rate": 8.037341631473662e-05,
+      "loss": 1.7619,
+      "step": 26920
+    },
+    {
+      "epoch": 0.5982444444444445,
+      "grad_norm": 1.536065697669983,
+      "learning_rate": 8.036897088241831e-05,
+      "loss": 1.8363,
+      "step": 26921
+    },
+    {
+      "epoch": 0.5982666666666666,
+      "grad_norm": 1.9655731916427612,
+      "learning_rate": 8.036452545010002e-05,
+      "loss": 2.2304,
+      "step": 26922
+    },
+    {
+      "epoch": 0.5982888888888889,
+      "grad_norm": 1.4935956001281738,
+      "learning_rate": 8.036008001778173e-05,
+      "loss": 1.5833,
+      "step": 26923
+    },
+    {
+      "epoch": 0.5983111111111111,
+      "grad_norm": 1.8283042907714844,
+      "learning_rate": 8.035563458546344e-05,
+      "loss": 2.0601,
+      "step": 26924
+    },
+    {
+      "epoch": 0.5983333333333334,
+      "grad_norm": 1.7895570993423462,
+      "learning_rate": 8.035118915314515e-05,
+      "loss": 2.0856,
+      "step": 26925
+    },
+    {
+      "epoch": 0.5983555555555555,
+      "grad_norm": 1.5136226415634155,
+      "learning_rate": 8.034674372082686e-05,
+      "loss": 2.1683,
+      "step": 26926
+    },
+    {
+      "epoch": 0.5983777777777778,
+      "grad_norm": 1.888890027999878,
+      "learning_rate": 8.034229828850857e-05,
+      "loss": 1.9483,
+      "step": 26927
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 1.9367108345031738,
+      "learning_rate": 8.033785285619027e-05,
+      "loss": 2.0916,
+      "step": 26928
+    },
+    {
+      "epoch": 0.5984222222222222,
+      "grad_norm": 1.6923370361328125,
+      "learning_rate": 8.033340742387197e-05,
+      "loss": 1.8858,
+      "step": 26929
+    },
+    {
+      "epoch": 0.5984444444444444,
+      "grad_norm": 1.7384182214736938,
+      "learning_rate": 8.032896199155368e-05,
+      "loss": 2.0751,
+      "step": 26930
+    },
+    {
+      "epoch": 0.5984666666666667,
+      "grad_norm": 1.5326626300811768,
+      "learning_rate": 8.032451655923538e-05,
+      "loss": 1.0639,
+      "step": 26931
+    },
+    {
+      "epoch": 0.5984888888888888,
+      "grad_norm": 1.4723471403121948,
+      "learning_rate": 8.032007112691709e-05,
+      "loss": 1.4927,
+      "step": 26932
+    },
+    {
+      "epoch": 0.5985111111111111,
+      "grad_norm": 1.6398212909698486,
+      "learning_rate": 8.03156256945988e-05,
+      "loss": 1.9547,
+      "step": 26933
+    },
+    {
+      "epoch": 0.5985333333333334,
+      "grad_norm": 1.6803056001663208,
+      "learning_rate": 8.031118026228051e-05,
+      "loss": 1.9571,
+      "step": 26934
+    },
+    {
+      "epoch": 0.5985555555555555,
+      "grad_norm": 1.5296698808670044,
+      "learning_rate": 8.030673482996222e-05,
+      "loss": 1.4227,
+      "step": 26935
+    },
+    {
+      "epoch": 0.5985777777777778,
+      "grad_norm": 1.6332881450653076,
+      "learning_rate": 8.030228939764393e-05,
+      "loss": 2.2374,
+      "step": 26936
+    },
+    {
+      "epoch": 0.5986,
+      "grad_norm": 1.9761298894882202,
+      "learning_rate": 8.029784396532564e-05,
+      "loss": 2.4395,
+      "step": 26937
+    },
+    {
+      "epoch": 0.5986222222222222,
+      "grad_norm": 1.8694782257080078,
+      "learning_rate": 8.029339853300733e-05,
+      "loss": 1.6662,
+      "step": 26938
+    },
+    {
+      "epoch": 0.5986444444444444,
+      "grad_norm": 1.7730891704559326,
+      "learning_rate": 8.028895310068904e-05,
+      "loss": 2.1371,
+      "step": 26939
+    },
+    {
+      "epoch": 0.5986666666666667,
+      "grad_norm": 1.6914303302764893,
+      "learning_rate": 8.028450766837075e-05,
+      "loss": 1.9322,
+      "step": 26940
+    },
+    {
+      "epoch": 0.5986888888888889,
+      "grad_norm": 1.6608036756515503,
+      "learning_rate": 8.028006223605245e-05,
+      "loss": 1.8645,
+      "step": 26941
+    },
+    {
+      "epoch": 0.5987111111111111,
+      "grad_norm": 1.7254583835601807,
+      "learning_rate": 8.027561680373417e-05,
+      "loss": 1.6468,
+      "step": 26942
+    },
+    {
+      "epoch": 0.5987333333333333,
+      "grad_norm": 1.714959979057312,
+      "learning_rate": 8.027117137141588e-05,
+      "loss": 1.8889,
+      "step": 26943
+    },
+    {
+      "epoch": 0.5987555555555556,
+      "grad_norm": 1.6050472259521484,
+      "learning_rate": 8.026672593909758e-05,
+      "loss": 1.8922,
+      "step": 26944
+    },
+    {
+      "epoch": 0.5987777777777777,
+      "grad_norm": 1.7325646877288818,
+      "learning_rate": 8.026228050677929e-05,
+      "loss": 1.9524,
+      "step": 26945
+    },
+    {
+      "epoch": 0.5988,
+      "grad_norm": 1.3277366161346436,
+      "learning_rate": 8.0257835074461e-05,
+      "loss": 0.8642,
+      "step": 26946
+    },
+    {
+      "epoch": 0.5988222222222223,
+      "grad_norm": 1.1544034481048584,
+      "learning_rate": 8.025338964214271e-05,
+      "loss": 0.9541,
+      "step": 26947
+    },
+    {
+      "epoch": 0.5988444444444444,
+      "grad_norm": 1.8778469562530518,
+      "learning_rate": 8.02489442098244e-05,
+      "loss": 2.0342,
+      "step": 26948
+    },
+    {
+      "epoch": 0.5988666666666667,
+      "grad_norm": 0.2326299399137497,
+      "learning_rate": 8.024449877750611e-05,
+      "loss": 0.0492,
+      "step": 26949
+    },
+    {
+      "epoch": 0.5988888888888889,
+      "grad_norm": 2.3978614807128906,
+      "learning_rate": 8.024005334518782e-05,
+      "loss": 2.062,
+      "step": 26950
+    },
+    {
+      "epoch": 0.5989111111111111,
+      "grad_norm": 1.5422933101654053,
+      "learning_rate": 8.023560791286953e-05,
+      "loss": 2.7222,
+      "step": 26951
+    },
+    {
+      "epoch": 0.5989333333333333,
+      "grad_norm": 1.4547905921936035,
+      "learning_rate": 8.023116248055124e-05,
+      "loss": 1.699,
+      "step": 26952
+    },
+    {
+      "epoch": 0.5989555555555556,
+      "grad_norm": 1.5786837339401245,
+      "learning_rate": 8.022671704823295e-05,
+      "loss": 3.1059,
+      "step": 26953
+    },
+    {
+      "epoch": 0.5989777777777778,
+      "grad_norm": 1.6623071432113647,
+      "learning_rate": 8.022227161591466e-05,
+      "loss": 2.5863,
+      "step": 26954
+    },
+    {
+      "epoch": 0.599,
+      "grad_norm": 1.663651704788208,
+      "learning_rate": 8.021782618359636e-05,
+      "loss": 1.9413,
+      "step": 26955
+    },
+    {
+      "epoch": 0.5990222222222222,
+      "grad_norm": 1.5665733814239502,
+      "learning_rate": 8.021338075127807e-05,
+      "loss": 2.3541,
+      "step": 26956
+    },
+    {
+      "epoch": 0.5990444444444445,
+      "grad_norm": 1.2850035429000854,
+      "learning_rate": 8.020893531895978e-05,
+      "loss": 2.0893,
+      "step": 26957
+    },
+    {
+      "epoch": 0.5990666666666666,
+      "grad_norm": 1.481261968612671,
+      "learning_rate": 8.020448988664147e-05,
+      "loss": 2.3503,
+      "step": 26958
+    },
+    {
+      "epoch": 0.5990888888888889,
+      "grad_norm": 1.7478474378585815,
+      "learning_rate": 8.020004445432318e-05,
+      "loss": 1.5484,
+      "step": 26959
+    },
+    {
+      "epoch": 0.5991111111111111,
+      "grad_norm": 1.60469651222229,
+      "learning_rate": 8.019559902200489e-05,
+      "loss": 2.0309,
+      "step": 26960
+    },
+    {
+      "epoch": 0.5991333333333333,
+      "grad_norm": 1.6925408840179443,
+      "learning_rate": 8.01911535896866e-05,
+      "loss": 0.9794,
+      "step": 26961
+    },
+    {
+      "epoch": 0.5991555555555556,
+      "grad_norm": 1.7684733867645264,
+      "learning_rate": 8.018670815736831e-05,
+      "loss": 2.2351,
+      "step": 26962
+    },
+    {
+      "epoch": 0.5991777777777778,
+      "grad_norm": 1.7226217985153198,
+      "learning_rate": 8.018226272505002e-05,
+      "loss": 1.963,
+      "step": 26963
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 1.6714963912963867,
+      "learning_rate": 8.017781729273173e-05,
+      "loss": 2.0087,
+      "step": 26964
+    },
+    {
+      "epoch": 0.5992222222222222,
+      "grad_norm": 1.5926263332366943,
+      "learning_rate": 8.017337186041343e-05,
+      "loss": 2.0484,
+      "step": 26965
+    },
+    {
+      "epoch": 0.5992444444444445,
+      "grad_norm": 1.4017750024795532,
+      "learning_rate": 8.016892642809514e-05,
+      "loss": 1.2894,
+      "step": 26966
+    },
+    {
+      "epoch": 0.5992666666666666,
+      "grad_norm": 1.0954105854034424,
+      "learning_rate": 8.016448099577685e-05,
+      "loss": 0.9788,
+      "step": 26967
+    },
+    {
+      "epoch": 0.5992888888888889,
+      "grad_norm": 1.7288366556167603,
+      "learning_rate": 8.016003556345854e-05,
+      "loss": 1.8245,
+      "step": 26968
+    },
+    {
+      "epoch": 0.5993111111111111,
+      "grad_norm": 1.6629772186279297,
+      "learning_rate": 8.015559013114025e-05,
+      "loss": 2.171,
+      "step": 26969
+    },
+    {
+      "epoch": 0.5993333333333334,
+      "grad_norm": 1.8875101804733276,
+      "learning_rate": 8.015114469882196e-05,
+      "loss": 2.114,
+      "step": 26970
+    },
+    {
+      "epoch": 0.5993555555555555,
+      "grad_norm": 1.5006781816482544,
+      "learning_rate": 8.014669926650367e-05,
+      "loss": 2.1082,
+      "step": 26971
+    },
+    {
+      "epoch": 0.5993777777777778,
+      "grad_norm": 1.7864024639129639,
+      "learning_rate": 8.014225383418538e-05,
+      "loss": 2.197,
+      "step": 26972
+    },
+    {
+      "epoch": 0.5994,
+      "grad_norm": 1.6741145849227905,
+      "learning_rate": 8.013780840186709e-05,
+      "loss": 1.6411,
+      "step": 26973
+    },
+    {
+      "epoch": 0.5994222222222222,
+      "grad_norm": 1.6694917678833008,
+      "learning_rate": 8.01333629695488e-05,
+      "loss": 1.8141,
+      "step": 26974
+    },
+    {
+      "epoch": 0.5994444444444444,
+      "grad_norm": 1.9536409378051758,
+      "learning_rate": 8.01289175372305e-05,
+      "loss": 2.1622,
+      "step": 26975
+    },
+    {
+      "epoch": 0.5994666666666667,
+      "grad_norm": 1.9600883722305298,
+      "learning_rate": 8.01244721049122e-05,
+      "loss": 1.9236,
+      "step": 26976
+    },
+    {
+      "epoch": 0.5994888888888888,
+      "grad_norm": 2.1006920337677,
+      "learning_rate": 8.012002667259391e-05,
+      "loss": 2.3782,
+      "step": 26977
+    },
+    {
+      "epoch": 0.5995111111111111,
+      "grad_norm": 1.7578033208847046,
+      "learning_rate": 8.011558124027561e-05,
+      "loss": 2.1157,
+      "step": 26978
+    },
+    {
+      "epoch": 0.5995333333333334,
+      "grad_norm": 1.7216506004333496,
+      "learning_rate": 8.011113580795733e-05,
+      "loss": 1.7657,
+      "step": 26979
+    },
+    {
+      "epoch": 0.5995555555555555,
+      "grad_norm": 1.7706249952316284,
+      "learning_rate": 8.010669037563904e-05,
+      "loss": 2.1157,
+      "step": 26980
+    },
+    {
+      "epoch": 0.5995777777777778,
+      "grad_norm": 1.936853051185608,
+      "learning_rate": 8.010224494332074e-05,
+      "loss": 1.9105,
+      "step": 26981
+    },
+    {
+      "epoch": 0.5996,
+      "grad_norm": 1.5771408081054688,
+      "learning_rate": 8.009779951100245e-05,
+      "loss": 1.8554,
+      "step": 26982
+    },
+    {
+      "epoch": 0.5996222222222222,
+      "grad_norm": 1.5390323400497437,
+      "learning_rate": 8.009335407868416e-05,
+      "loss": 1.734,
+      "step": 26983
+    },
+    {
+      "epoch": 0.5996444444444444,
+      "grad_norm": 1.6512367725372314,
+      "learning_rate": 8.008890864636587e-05,
+      "loss": 1.7134,
+      "step": 26984
+    },
+    {
+      "epoch": 0.5996666666666667,
+      "grad_norm": 2.0488808155059814,
+      "learning_rate": 8.008446321404756e-05,
+      "loss": 1.7957,
+      "step": 26985
+    },
+    {
+      "epoch": 0.5996888888888889,
+      "grad_norm": 1.9845051765441895,
+      "learning_rate": 8.008001778172927e-05,
+      "loss": 1.9532,
+      "step": 26986
+    },
+    {
+      "epoch": 0.5997111111111111,
+      "grad_norm": 1.6982089281082153,
+      "learning_rate": 8.007557234941098e-05,
+      "loss": 2.023,
+      "step": 26987
+    },
+    {
+      "epoch": 0.5997333333333333,
+      "grad_norm": 1.7800291776657104,
+      "learning_rate": 8.00711269170927e-05,
+      "loss": 1.9438,
+      "step": 26988
+    },
+    {
+      "epoch": 0.5997555555555556,
+      "grad_norm": 3.0561060905456543,
+      "learning_rate": 8.00666814847744e-05,
+      "loss": 2.0598,
+      "step": 26989
+    },
+    {
+      "epoch": 0.5997777777777777,
+      "grad_norm": 1.765212059020996,
+      "learning_rate": 8.006223605245611e-05,
+      "loss": 1.9058,
+      "step": 26990
+    },
+    {
+      "epoch": 0.5998,
+      "grad_norm": 1.9671481847763062,
+      "learning_rate": 8.005779062013781e-05,
+      "loss": 2.1272,
+      "step": 26991
+    },
+    {
+      "epoch": 0.5998222222222223,
+      "grad_norm": 1.9428707361221313,
+      "learning_rate": 8.005334518781952e-05,
+      "loss": 1.9046,
+      "step": 26992
+    },
+    {
+      "epoch": 0.5998444444444444,
+      "grad_norm": 1.8527716398239136,
+      "learning_rate": 8.004889975550123e-05,
+      "loss": 1.847,
+      "step": 26993
+    },
+    {
+      "epoch": 0.5998666666666667,
+      "grad_norm": 1.6317662000656128,
+      "learning_rate": 8.004445432318294e-05,
+      "loss": 1.5198,
+      "step": 26994
+    },
+    {
+      "epoch": 0.5998888888888889,
+      "grad_norm": 1.8692249059677124,
+      "learning_rate": 8.004000889086463e-05,
+      "loss": 1.895,
+      "step": 26995
+    },
+    {
+      "epoch": 0.5999111111111111,
+      "grad_norm": 1.947574496269226,
+      "learning_rate": 8.003556345854634e-05,
+      "loss": 2.2449,
+      "step": 26996
+    },
+    {
+      "epoch": 0.5999333333333333,
+      "grad_norm": 1.7928720712661743,
+      "learning_rate": 8.003111802622805e-05,
+      "loss": 1.9345,
+      "step": 26997
+    },
+    {
+      "epoch": 0.5999555555555556,
+      "grad_norm": 2.0767455101013184,
+      "learning_rate": 8.002667259390976e-05,
+      "loss": 2.2322,
+      "step": 26998
+    },
+    {
+      "epoch": 0.5999777777777778,
+      "grad_norm": 1.093999981880188,
+      "learning_rate": 8.002222716159147e-05,
+      "loss": 0.7272,
+      "step": 26999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6920832395553589,
+      "learning_rate": 8.001778172927318e-05,
+      "loss": 1.0741,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.8736392259597778,
+      "eval_runtime": 2159.761,
+      "eval_samples_per_second": 4.63,
+      "eval_steps_per_second": 4.63,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6000222222222222,
+      "grad_norm": 1.6577028036117554,
+      "learning_rate": 8.001333629695488e-05,
+      "loss": 2.6101,
+      "step": 27001
+    },
+    {
+      "epoch": 0.6000444444444445,
+      "grad_norm": 1.6142709255218506,
+      "learning_rate": 8.000889086463659e-05,
+      "loss": 2.748,
+      "step": 27002
+    },
+    {
+      "epoch": 0.6000666666666666,
+      "grad_norm": 1.508734941482544,
+      "learning_rate": 8.00044454323183e-05,
+      "loss": 2.6064,
+      "step": 27003
+    },
+    {
+      "epoch": 0.6000888888888889,
+      "grad_norm": 1.804945945739746,
+      "learning_rate": 8e-05,
+      "loss": 2.5162,
+      "step": 27004
+    },
+    {
+      "epoch": 0.6001111111111112,
+      "grad_norm": 1.4495036602020264,
+      "learning_rate": 7.99955545676817e-05,
+      "loss": 2.216,
+      "step": 27005
+    },
+    {
+      "epoch": 0.6001333333333333,
+      "grad_norm": 1.4844017028808594,
+      "learning_rate": 7.999110913536341e-05,
+      "loss": 2.0562,
+      "step": 27006
+    },
+    {
+      "epoch": 0.6001555555555556,
+      "grad_norm": 1.463944673538208,
+      "learning_rate": 7.998666370304512e-05,
+      "loss": 2.2244,
+      "step": 27007
+    },
+    {
+      "epoch": 0.6001777777777778,
+      "grad_norm": 1.755519151687622,
+      "learning_rate": 7.998221827072683e-05,
+      "loss": 2.3888,
+      "step": 27008
+    },
+    {
+      "epoch": 0.6002,
+      "grad_norm": 1.373860239982605,
+      "learning_rate": 7.997777283840854e-05,
+      "loss": 1.5492,
+      "step": 27009
+    },
+    {
+      "epoch": 0.6002222222222222,
+      "grad_norm": 1.4735991954803467,
+      "learning_rate": 7.997332740609025e-05,
+      "loss": 1.7861,
+      "step": 27010
+    },
+    {
+      "epoch": 0.6002444444444445,
+      "grad_norm": 1.7138311862945557,
+      "learning_rate": 7.996888197377196e-05,
+      "loss": 2.1563,
+      "step": 27011
+    },
+    {
+      "epoch": 0.6002666666666666,
+      "grad_norm": 1.6711690425872803,
+      "learning_rate": 7.996443654145366e-05,
+      "loss": 2.3283,
+      "step": 27012
+    },
+    {
+      "epoch": 0.6002888888888889,
+      "grad_norm": 2.0114669799804688,
+      "learning_rate": 7.995999110913537e-05,
+      "loss": 2.2527,
+      "step": 27013
+    },
+    {
+      "epoch": 0.6003111111111111,
+      "grad_norm": 1.8923662900924683,
+      "learning_rate": 7.995554567681708e-05,
+      "loss": 2.3929,
+      "step": 27014
+    },
+    {
+      "epoch": 0.6003333333333334,
+      "grad_norm": 1.4820717573165894,
+      "learning_rate": 7.995110024449877e-05,
+      "loss": 1.7815,
+      "step": 27015
+    },
+    {
+      "epoch": 0.6003555555555555,
+      "grad_norm": 1.8462886810302734,
+      "learning_rate": 7.99466548121805e-05,
+      "loss": 2.6735,
+      "step": 27016
+    },
+    {
+      "epoch": 0.6003777777777778,
+      "grad_norm": 1.7554908990859985,
+      "learning_rate": 7.99422093798622e-05,
+      "loss": 2.0213,
+      "step": 27017
+    },
+    {
+      "epoch": 0.6004,
+      "grad_norm": 2.3965280055999756,
+      "learning_rate": 7.99377639475439e-05,
+      "loss": 1.7916,
+      "step": 27018
+    },
+    {
+      "epoch": 0.6004222222222222,
+      "grad_norm": 1.4135266542434692,
+      "learning_rate": 7.993331851522561e-05,
+      "loss": 1.7435,
+      "step": 27019
+    },
+    {
+      "epoch": 0.6004444444444444,
+      "grad_norm": 1.4737776517868042,
+      "learning_rate": 7.992887308290732e-05,
+      "loss": 1.3768,
+      "step": 27020
+    },
+    {
+      "epoch": 0.6004666666666667,
+      "grad_norm": 1.8706743717193604,
+      "learning_rate": 7.992442765058903e-05,
+      "loss": 1.7092,
+      "step": 27021
+    },
+    {
+      "epoch": 0.6004888888888888,
+      "grad_norm": 1.7776674032211304,
+      "learning_rate": 7.991998221827073e-05,
+      "loss": 2.5811,
+      "step": 27022
+    },
+    {
+      "epoch": 0.6005111111111111,
+      "grad_norm": 1.6870381832122803,
+      "learning_rate": 7.991553678595244e-05,
+      "loss": 1.5549,
+      "step": 27023
+    },
+    {
+      "epoch": 0.6005333333333334,
+      "grad_norm": 1.9651492834091187,
+      "learning_rate": 7.991109135363414e-05,
+      "loss": 2.1999,
+      "step": 27024
+    },
+    {
+      "epoch": 0.6005555555555555,
+      "grad_norm": 1.5777024030685425,
+      "learning_rate": 7.990664592131585e-05,
+      "loss": 1.7351,
+      "step": 27025
+    },
+    {
+      "epoch": 0.6005777777777778,
+      "grad_norm": 1.9379805326461792,
+      "learning_rate": 7.990220048899756e-05,
+      "loss": 2.4563,
+      "step": 27026
+    },
+    {
+      "epoch": 0.6006,
+      "grad_norm": 1.9654110670089722,
+      "learning_rate": 7.989775505667927e-05,
+      "loss": 2.2118,
+      "step": 27027
+    },
+    {
+      "epoch": 0.6006222222222222,
+      "grad_norm": 1.7272353172302246,
+      "learning_rate": 7.989330962436097e-05,
+      "loss": 2.0969,
+      "step": 27028
+    },
+    {
+      "epoch": 0.6006444444444444,
+      "grad_norm": 1.3812412023544312,
+      "learning_rate": 7.988886419204268e-05,
+      "loss": 0.7072,
+      "step": 27029
+    },
+    {
+      "epoch": 0.6006666666666667,
+      "grad_norm": 1.810765266418457,
+      "learning_rate": 7.988441875972439e-05,
+      "loss": 1.5445,
+      "step": 27030
+    },
+    {
+      "epoch": 0.6006888888888889,
+      "grad_norm": 1.8113470077514648,
+      "learning_rate": 7.98799733274061e-05,
+      "loss": 1.9716,
+      "step": 27031
+    },
+    {
+      "epoch": 0.6007111111111111,
+      "grad_norm": 1.2004268169403076,
+      "learning_rate": 7.98755278950878e-05,
+      "loss": 0.9292,
+      "step": 27032
+    },
+    {
+      "epoch": 0.6007333333333333,
+      "grad_norm": 1.4961955547332764,
+      "learning_rate": 7.98710824627695e-05,
+      "loss": 1.6421,
+      "step": 27033
+    },
+    {
+      "epoch": 0.6007555555555556,
+      "grad_norm": 1.735058307647705,
+      "learning_rate": 7.986663703045121e-05,
+      "loss": 1.4196,
+      "step": 27034
+    },
+    {
+      "epoch": 0.6007777777777777,
+      "grad_norm": 1.5329277515411377,
+      "learning_rate": 7.986219159813292e-05,
+      "loss": 1.5782,
+      "step": 27035
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 1.7740753889083862,
+      "learning_rate": 7.985774616581463e-05,
+      "loss": 2.0045,
+      "step": 27036
+    },
+    {
+      "epoch": 0.6008222222222223,
+      "grad_norm": 1.8819478750228882,
+      "learning_rate": 7.985330073349634e-05,
+      "loss": 2.2053,
+      "step": 27037
+    },
+    {
+      "epoch": 0.6008444444444444,
+      "grad_norm": 2.052328586578369,
+      "learning_rate": 7.984885530117804e-05,
+      "loss": 2.061,
+      "step": 27038
+    },
+    {
+      "epoch": 0.6008666666666667,
+      "grad_norm": 1.8363018035888672,
+      "learning_rate": 7.984440986885975e-05,
+      "loss": 1.8221,
+      "step": 27039
+    },
+    {
+      "epoch": 0.6008888888888889,
+      "grad_norm": 1.8349599838256836,
+      "learning_rate": 7.983996443654146e-05,
+      "loss": 1.7493,
+      "step": 27040
+    },
+    {
+      "epoch": 0.6009111111111111,
+      "grad_norm": 1.7365055084228516,
+      "learning_rate": 7.983551900422317e-05,
+      "loss": 1.6845,
+      "step": 27041
+    },
+    {
+      "epoch": 0.6009333333333333,
+      "grad_norm": 1.5604342222213745,
+      "learning_rate": 7.983107357190486e-05,
+      "loss": 1.9154,
+      "step": 27042
+    },
+    {
+      "epoch": 0.6009555555555556,
+      "grad_norm": 1.7715667486190796,
+      "learning_rate": 7.982662813958657e-05,
+      "loss": 1.9098,
+      "step": 27043
+    },
+    {
+      "epoch": 0.6009777777777778,
+      "grad_norm": 1.7074742317199707,
+      "learning_rate": 7.982218270726828e-05,
+      "loss": 1.4472,
+      "step": 27044
+    },
+    {
+      "epoch": 0.601,
+      "grad_norm": 1.6556566953659058,
+      "learning_rate": 7.981773727494999e-05,
+      "loss": 1.8303,
+      "step": 27045
+    },
+    {
+      "epoch": 0.6010222222222222,
+      "grad_norm": 1.5229953527450562,
+      "learning_rate": 7.98132918426317e-05,
+      "loss": 1.9378,
+      "step": 27046
+    },
+    {
+      "epoch": 0.6010444444444445,
+      "grad_norm": 1.802293300628662,
+      "learning_rate": 7.980884641031341e-05,
+      "loss": 1.7766,
+      "step": 27047
+    },
+    {
+      "epoch": 0.6010666666666666,
+      "grad_norm": 1.222768783569336,
+      "learning_rate": 7.980440097799511e-05,
+      "loss": 0.7822,
+      "step": 27048
+    },
+    {
+      "epoch": 0.6010888888888889,
+      "grad_norm": 1.9860477447509766,
+      "learning_rate": 7.979995554567682e-05,
+      "loss": 1.8711,
+      "step": 27049
+    },
+    {
+      "epoch": 0.6011111111111112,
+      "grad_norm": 1.6315582990646362,
+      "learning_rate": 7.979551011335853e-05,
+      "loss": 0.8217,
+      "step": 27050
+    },
+    {
+      "epoch": 0.6011333333333333,
+      "grad_norm": 1.3168281316757202,
+      "learning_rate": 7.979106468104024e-05,
+      "loss": 2.2222,
+      "step": 27051
+    },
+    {
+      "epoch": 0.6011555555555556,
+      "grad_norm": 1.5810824632644653,
+      "learning_rate": 7.978661924872193e-05,
+      "loss": 2.4733,
+      "step": 27052
+    },
+    {
+      "epoch": 0.6011777777777778,
+      "grad_norm": 1.3996162414550781,
+      "learning_rate": 7.978217381640366e-05,
+      "loss": 2.0982,
+      "step": 27053
+    },
+    {
+      "epoch": 0.6012,
+      "grad_norm": 0.9520947933197021,
+      "learning_rate": 7.977772838408537e-05,
+      "loss": 1.1722,
+      "step": 27054
+    },
+    {
+      "epoch": 0.6012222222222222,
+      "grad_norm": 1.687771201133728,
+      "learning_rate": 7.977328295176706e-05,
+      "loss": 2.2697,
+      "step": 27055
+    },
+    {
+      "epoch": 0.6012444444444445,
+      "grad_norm": 1.4503023624420166,
+      "learning_rate": 7.976883751944877e-05,
+      "loss": 2.1093,
+      "step": 27056
+    },
+    {
+      "epoch": 0.6012666666666666,
+      "grad_norm": 1.4794197082519531,
+      "learning_rate": 7.976439208713048e-05,
+      "loss": 1.9424,
+      "step": 27057
+    },
+    {
+      "epoch": 0.6012888888888889,
+      "grad_norm": 1.83533775806427,
+      "learning_rate": 7.975994665481218e-05,
+      "loss": 2.3112,
+      "step": 27058
+    },
+    {
+      "epoch": 0.6013111111111111,
+      "grad_norm": 1.5439540147781372,
+      "learning_rate": 7.975550122249389e-05,
+      "loss": 1.4641,
+      "step": 27059
+    },
+    {
+      "epoch": 0.6013333333333334,
+      "grad_norm": 1.8201035261154175,
+      "learning_rate": 7.97510557901756e-05,
+      "loss": 1.9752,
+      "step": 27060
+    },
+    {
+      "epoch": 0.6013555555555555,
+      "grad_norm": 1.8052176237106323,
+      "learning_rate": 7.97466103578573e-05,
+      "loss": 2.1018,
+      "step": 27061
+    },
+    {
+      "epoch": 0.6013777777777778,
+      "grad_norm": 1.7336097955703735,
+      "learning_rate": 7.974216492553902e-05,
+      "loss": 1.8685,
+      "step": 27062
+    },
+    {
+      "epoch": 0.6014,
+      "grad_norm": 1.6656413078308105,
+      "learning_rate": 7.973771949322072e-05,
+      "loss": 1.6962,
+      "step": 27063
+    },
+    {
+      "epoch": 0.6014222222222222,
+      "grad_norm": 1.5584288835525513,
+      "learning_rate": 7.973327406090243e-05,
+      "loss": 1.8788,
+      "step": 27064
+    },
+    {
+      "epoch": 0.6014444444444444,
+      "grad_norm": 1.5663453340530396,
+      "learning_rate": 7.972882862858413e-05,
+      "loss": 2.4941,
+      "step": 27065
+    },
+    {
+      "epoch": 0.6014666666666667,
+      "grad_norm": 1.6916759014129639,
+      "learning_rate": 7.972438319626584e-05,
+      "loss": 2.1018,
+      "step": 27066
+    },
+    {
+      "epoch": 0.6014888888888889,
+      "grad_norm": 1.6306703090667725,
+      "learning_rate": 7.971993776394755e-05,
+      "loss": 1.6678,
+      "step": 27067
+    },
+    {
+      "epoch": 0.6015111111111111,
+      "grad_norm": 1.8184791803359985,
+      "learning_rate": 7.971549233162926e-05,
+      "loss": 0.8785,
+      "step": 27068
+    },
+    {
+      "epoch": 0.6015333333333334,
+      "grad_norm": 1.5695523023605347,
+      "learning_rate": 7.971104689931096e-05,
+      "loss": 2.0016,
+      "step": 27069
+    },
+    {
+      "epoch": 0.6015555555555555,
+      "grad_norm": 1.766474962234497,
+      "learning_rate": 7.970660146699267e-05,
+      "loss": 2.144,
+      "step": 27070
+    },
+    {
+      "epoch": 0.6015777777777778,
+      "grad_norm": 1.989862322807312,
+      "learning_rate": 7.970215603467437e-05,
+      "loss": 2.3518,
+      "step": 27071
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 1.598899483680725,
+      "learning_rate": 7.969771060235608e-05,
+      "loss": 1.8039,
+      "step": 27072
+    },
+    {
+      "epoch": 0.6016222222222222,
+      "grad_norm": 1.6440620422363281,
+      "learning_rate": 7.96932651700378e-05,
+      "loss": 1.9172,
+      "step": 27073
+    },
+    {
+      "epoch": 0.6016444444444444,
+      "grad_norm": 1.8292713165283203,
+      "learning_rate": 7.96888197377195e-05,
+      "loss": 1.6827,
+      "step": 27074
+    },
+    {
+      "epoch": 0.6016666666666667,
+      "grad_norm": 1.7788729667663574,
+      "learning_rate": 7.96843743054012e-05,
+      "loss": 2.0841,
+      "step": 27075
+    },
+    {
+      "epoch": 0.6016888888888889,
+      "grad_norm": 1.5874347686767578,
+      "learning_rate": 7.967992887308291e-05,
+      "loss": 1.6457,
+      "step": 27076
+    },
+    {
+      "epoch": 0.6017111111111111,
+      "grad_norm": 1.4802864789962769,
+      "learning_rate": 7.967548344076462e-05,
+      "loss": 2.049,
+      "step": 27077
+    },
+    {
+      "epoch": 0.6017333333333333,
+      "grad_norm": 1.557401180267334,
+      "learning_rate": 7.967103800844633e-05,
+      "loss": 0.8245,
+      "step": 27078
+    },
+    {
+      "epoch": 0.6017555555555556,
+      "grad_norm": 1.8262344598770142,
+      "learning_rate": 7.966659257612802e-05,
+      "loss": 2.1202,
+      "step": 27079
+    },
+    {
+      "epoch": 0.6017777777777777,
+      "grad_norm": 1.4381197690963745,
+      "learning_rate": 7.966214714380973e-05,
+      "loss": 1.3392,
+      "step": 27080
+    },
+    {
+      "epoch": 0.6018,
+      "grad_norm": 1.579521894454956,
+      "learning_rate": 7.965770171149144e-05,
+      "loss": 1.6266,
+      "step": 27081
+    },
+    {
+      "epoch": 0.6018222222222223,
+      "grad_norm": 1.4592291116714478,
+      "learning_rate": 7.965325627917315e-05,
+      "loss": 1.678,
+      "step": 27082
+    },
+    {
+      "epoch": 0.6018444444444444,
+      "grad_norm": 1.5992571115493774,
+      "learning_rate": 7.964881084685486e-05,
+      "loss": 1.6441,
+      "step": 27083
+    },
+    {
+      "epoch": 0.6018666666666667,
+      "grad_norm": 1.9749088287353516,
+      "learning_rate": 7.964436541453657e-05,
+      "loss": 2.295,
+      "step": 27084
+    },
+    {
+      "epoch": 0.6018888888888889,
+      "grad_norm": 1.8254121541976929,
+      "learning_rate": 7.963991998221827e-05,
+      "loss": 2.1561,
+      "step": 27085
+    },
+    {
+      "epoch": 0.6019111111111111,
+      "grad_norm": 2.735177993774414,
+      "learning_rate": 7.963547454989998e-05,
+      "loss": 2.914,
+      "step": 27086
+    },
+    {
+      "epoch": 0.6019333333333333,
+      "grad_norm": 1.5838193893432617,
+      "learning_rate": 7.963102911758169e-05,
+      "loss": 1.8465,
+      "step": 27087
+    },
+    {
+      "epoch": 0.6019555555555556,
+      "grad_norm": 1.8101228475570679,
+      "learning_rate": 7.96265836852634e-05,
+      "loss": 1.9933,
+      "step": 27088
+    },
+    {
+      "epoch": 0.6019777777777777,
+      "grad_norm": 1.8943954706192017,
+      "learning_rate": 7.96221382529451e-05,
+      "loss": 1.4846,
+      "step": 27089
+    },
+    {
+      "epoch": 0.602,
+      "grad_norm": 1.3854761123657227,
+      "learning_rate": 7.961769282062682e-05,
+      "loss": 0.8323,
+      "step": 27090
+    },
+    {
+      "epoch": 0.6020222222222222,
+      "grad_norm": 1.6953483819961548,
+      "learning_rate": 7.961324738830853e-05,
+      "loss": 1.8218,
+      "step": 27091
+    },
+    {
+      "epoch": 0.6020444444444445,
+      "grad_norm": 1.7189913988113403,
+      "learning_rate": 7.960880195599022e-05,
+      "loss": 1.8765,
+      "step": 27092
+    },
+    {
+      "epoch": 0.6020666666666666,
+      "grad_norm": 1.7521131038665771,
+      "learning_rate": 7.960435652367193e-05,
+      "loss": 2.0528,
+      "step": 27093
+    },
+    {
+      "epoch": 0.6020888888888889,
+      "grad_norm": 1.5336365699768066,
+      "learning_rate": 7.959991109135364e-05,
+      "loss": 1.6545,
+      "step": 27094
+    },
+    {
+      "epoch": 0.6021111111111112,
+      "grad_norm": 1.950052261352539,
+      "learning_rate": 7.959546565903534e-05,
+      "loss": 2.2977,
+      "step": 27095
+    },
+    {
+      "epoch": 0.6021333333333333,
+      "grad_norm": 2.1121826171875,
+      "learning_rate": 7.959102022671705e-05,
+      "loss": 2.1558,
+      "step": 27096
+    },
+    {
+      "epoch": 0.6021555555555556,
+      "grad_norm": 1.7638791799545288,
+      "learning_rate": 7.958657479439876e-05,
+      "loss": 1.8762,
+      "step": 27097
+    },
+    {
+      "epoch": 0.6021777777777778,
+      "grad_norm": 1.2896921634674072,
+      "learning_rate": 7.958212936208047e-05,
+      "loss": 0.9417,
+      "step": 27098
+    },
+    {
+      "epoch": 0.6022,
+      "grad_norm": 1.7924249172210693,
+      "learning_rate": 7.957768392976218e-05,
+      "loss": 1.9258,
+      "step": 27099
+    },
+    {
+      "epoch": 0.6022222222222222,
+      "grad_norm": 1.7163163423538208,
+      "learning_rate": 7.957323849744389e-05,
+      "loss": 1.5863,
+      "step": 27100
+    },
+    {
+      "epoch": 0.6022444444444445,
+      "grad_norm": 1.3778804540634155,
+      "learning_rate": 7.95687930651256e-05,
+      "loss": 2.2595,
+      "step": 27101
+    },
+    {
+      "epoch": 0.6022666666666666,
+      "grad_norm": 1.197900652885437,
+      "learning_rate": 7.956434763280729e-05,
+      "loss": 1.1102,
+      "step": 27102
+    },
+    {
+      "epoch": 0.6022888888888889,
+      "grad_norm": 1.3239737749099731,
+      "learning_rate": 7.9559902200489e-05,
+      "loss": 2.09,
+      "step": 27103
+    },
+    {
+      "epoch": 0.6023111111111111,
+      "grad_norm": 1.4392409324645996,
+      "learning_rate": 7.955545676817071e-05,
+      "loss": 1.5999,
+      "step": 27104
+    },
+    {
+      "epoch": 0.6023333333333334,
+      "grad_norm": 1.487004041671753,
+      "learning_rate": 7.955101133585241e-05,
+      "loss": 2.2757,
+      "step": 27105
+    },
+    {
+      "epoch": 0.6023555555555555,
+      "grad_norm": 2.519434690475464,
+      "learning_rate": 7.954656590353412e-05,
+      "loss": 1.6574,
+      "step": 27106
+    },
+    {
+      "epoch": 0.6023777777777778,
+      "grad_norm": 1.5450921058654785,
+      "learning_rate": 7.954212047121583e-05,
+      "loss": 2.3597,
+      "step": 27107
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 1.5056664943695068,
+      "learning_rate": 7.953767503889754e-05,
+      "loss": 1.8347,
+      "step": 27108
+    },
+    {
+      "epoch": 0.6024222222222222,
+      "grad_norm": 1.5468626022338867,
+      "learning_rate": 7.953322960657925e-05,
+      "loss": 1.9519,
+      "step": 27109
+    },
+    {
+      "epoch": 0.6024444444444444,
+      "grad_norm": 1.6627495288848877,
+      "learning_rate": 7.952878417426096e-05,
+      "loss": 1.8685,
+      "step": 27110
+    },
+    {
+      "epoch": 0.6024666666666667,
+      "grad_norm": 1.7398810386657715,
+      "learning_rate": 7.952433874194266e-05,
+      "loss": 2.2111,
+      "step": 27111
+    },
+    {
+      "epoch": 0.6024888888888889,
+      "grad_norm": 1.4979416131973267,
+      "learning_rate": 7.951989330962436e-05,
+      "loss": 2.2772,
+      "step": 27112
+    },
+    {
+      "epoch": 0.6025111111111111,
+      "grad_norm": 1.7348592281341553,
+      "learning_rate": 7.951544787730607e-05,
+      "loss": 2.2954,
+      "step": 27113
+    },
+    {
+      "epoch": 0.6025333333333334,
+      "grad_norm": 1.4830193519592285,
+      "learning_rate": 7.951100244498778e-05,
+      "loss": 2.1515,
+      "step": 27114
+    },
+    {
+      "epoch": 0.6025555555555555,
+      "grad_norm": 1.6449952125549316,
+      "learning_rate": 7.950655701266949e-05,
+      "loss": 2.1656,
+      "step": 27115
+    },
+    {
+      "epoch": 0.6025777777777778,
+      "grad_norm": 1.628627061843872,
+      "learning_rate": 7.950211158035119e-05,
+      "loss": 1.8167,
+      "step": 27116
+    },
+    {
+      "epoch": 0.6026,
+      "grad_norm": 1.6549086570739746,
+      "learning_rate": 7.94976661480329e-05,
+      "loss": 2.0722,
+      "step": 27117
+    },
+    {
+      "epoch": 0.6026222222222222,
+      "grad_norm": 1.1302205324172974,
+      "learning_rate": 7.94932207157146e-05,
+      "loss": 1.0459,
+      "step": 27118
+    },
+    {
+      "epoch": 0.6026444444444444,
+      "grad_norm": 1.723571538925171,
+      "learning_rate": 7.948877528339631e-05,
+      "loss": 1.9982,
+      "step": 27119
+    },
+    {
+      "epoch": 0.6026666666666667,
+      "grad_norm": 1.8001817464828491,
+      "learning_rate": 7.948432985107802e-05,
+      "loss": 2.4451,
+      "step": 27120
+    },
+    {
+      "epoch": 0.6026888888888889,
+      "grad_norm": 1.7819477319717407,
+      "learning_rate": 7.947988441875973e-05,
+      "loss": 2.1226,
+      "step": 27121
+    },
+    {
+      "epoch": 0.6027111111111111,
+      "grad_norm": 1.7987935543060303,
+      "learning_rate": 7.947543898644143e-05,
+      "loss": 2.1471,
+      "step": 27122
+    },
+    {
+      "epoch": 0.6027333333333333,
+      "grad_norm": 1.8866928815841675,
+      "learning_rate": 7.947099355412314e-05,
+      "loss": 2.2982,
+      "step": 27123
+    },
+    {
+      "epoch": 0.6027555555555556,
+      "grad_norm": 2.157236337661743,
+      "learning_rate": 7.946654812180485e-05,
+      "loss": 2.5947,
+      "step": 27124
+    },
+    {
+      "epoch": 0.6027777777777777,
+      "grad_norm": 1.7541967630386353,
+      "learning_rate": 7.946210268948656e-05,
+      "loss": 2.1122,
+      "step": 27125
+    },
+    {
+      "epoch": 0.6028,
+      "grad_norm": 1.5409049987792969,
+      "learning_rate": 7.945765725716825e-05,
+      "loss": 2.2893,
+      "step": 27126
+    },
+    {
+      "epoch": 0.6028222222222223,
+      "grad_norm": 1.9576064348220825,
+      "learning_rate": 7.945321182484998e-05,
+      "loss": 1.741,
+      "step": 27127
+    },
+    {
+      "epoch": 0.6028444444444444,
+      "grad_norm": 1.5584945678710938,
+      "learning_rate": 7.944876639253169e-05,
+      "loss": 2.0153,
+      "step": 27128
+    },
+    {
+      "epoch": 0.6028666666666667,
+      "grad_norm": 2.0886170864105225,
+      "learning_rate": 7.944432096021338e-05,
+      "loss": 2.3039,
+      "step": 27129
+    },
+    {
+      "epoch": 0.6028888888888889,
+      "grad_norm": 1.5287795066833496,
+      "learning_rate": 7.94398755278951e-05,
+      "loss": 1.5705,
+      "step": 27130
+    },
+    {
+      "epoch": 0.6029111111111111,
+      "grad_norm": 1.9877333641052246,
+      "learning_rate": 7.94354300955768e-05,
+      "loss": 2.0677,
+      "step": 27131
+    },
+    {
+      "epoch": 0.6029333333333333,
+      "grad_norm": 1.8227648735046387,
+      "learning_rate": 7.94309846632585e-05,
+      "loss": 2.0896,
+      "step": 27132
+    },
+    {
+      "epoch": 0.6029555555555556,
+      "grad_norm": 1.9447449445724487,
+      "learning_rate": 7.942653923094021e-05,
+      "loss": 1.9537,
+      "step": 27133
+    },
+    {
+      "epoch": 0.6029777777777777,
+      "grad_norm": 0.20771636068820953,
+      "learning_rate": 7.942209379862192e-05,
+      "loss": 0.0279,
+      "step": 27134
+    },
+    {
+      "epoch": 0.603,
+      "grad_norm": 2.1986615657806396,
+      "learning_rate": 7.941764836630363e-05,
+      "loss": 2.0665,
+      "step": 27135
+    },
+    {
+      "epoch": 0.6030222222222222,
+      "grad_norm": 1.7483482360839844,
+      "learning_rate": 7.941320293398534e-05,
+      "loss": 1.8281,
+      "step": 27136
+    },
+    {
+      "epoch": 0.6030444444444445,
+      "grad_norm": 1.5220845937728882,
+      "learning_rate": 7.940875750166705e-05,
+      "loss": 1.44,
+      "step": 27137
+    },
+    {
+      "epoch": 0.6030666666666666,
+      "grad_norm": 1.4676389694213867,
+      "learning_rate": 7.940431206934876e-05,
+      "loss": 1.396,
+      "step": 27138
+    },
+    {
+      "epoch": 0.6030888888888889,
+      "grad_norm": 1.8141095638275146,
+      "learning_rate": 7.939986663703045e-05,
+      "loss": 1.8311,
+      "step": 27139
+    },
+    {
+      "epoch": 0.6031111111111112,
+      "grad_norm": 2.3061838150024414,
+      "learning_rate": 7.939542120471216e-05,
+      "loss": 2.5424,
+      "step": 27140
+    },
+    {
+      "epoch": 0.6031333333333333,
+      "grad_norm": 1.7188256978988647,
+      "learning_rate": 7.939097577239387e-05,
+      "loss": 2.0499,
+      "step": 27141
+    },
+    {
+      "epoch": 0.6031555555555556,
+      "grad_norm": 1.8286235332489014,
+      "learning_rate": 7.938653034007557e-05,
+      "loss": 1.9255,
+      "step": 27142
+    },
+    {
+      "epoch": 0.6031777777777778,
+      "grad_norm": 1.7270867824554443,
+      "learning_rate": 7.938208490775728e-05,
+      "loss": 1.5199,
+      "step": 27143
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 1.7218559980392456,
+      "learning_rate": 7.937763947543899e-05,
+      "loss": 1.9342,
+      "step": 27144
+    },
+    {
+      "epoch": 0.6032222222222222,
+      "grad_norm": 1.8925360441207886,
+      "learning_rate": 7.93731940431207e-05,
+      "loss": 1.7951,
+      "step": 27145
+    },
+    {
+      "epoch": 0.6032444444444445,
+      "grad_norm": 1.7069977521896362,
+      "learning_rate": 7.93687486108024e-05,
+      "loss": 1.7802,
+      "step": 27146
+    },
+    {
+      "epoch": 0.6032666666666666,
+      "grad_norm": 1.6345694065093994,
+      "learning_rate": 7.936430317848412e-05,
+      "loss": 1.7784,
+      "step": 27147
+    },
+    {
+      "epoch": 0.6032888888888889,
+      "grad_norm": 1.5925861597061157,
+      "learning_rate": 7.935985774616583e-05,
+      "loss": 1.5868,
+      "step": 27148
+    },
+    {
+      "epoch": 0.6033111111111111,
+      "grad_norm": 1.817430019378662,
+      "learning_rate": 7.935541231384752e-05,
+      "loss": 1.6037,
+      "step": 27149
+    },
+    {
+      "epoch": 0.6033333333333334,
+      "grad_norm": 1.9700731039047241,
+      "learning_rate": 7.935096688152923e-05,
+      "loss": 1.9079,
+      "step": 27150
+    },
+    {
+      "epoch": 0.6033555555555555,
+      "grad_norm": 1.0064431428909302,
+      "learning_rate": 7.934652144921094e-05,
+      "loss": 1.0072,
+      "step": 27151
+    },
+    {
+      "epoch": 0.6033777777777778,
+      "grad_norm": 1.370635986328125,
+      "learning_rate": 7.934207601689264e-05,
+      "loss": 1.9392,
+      "step": 27152
+    },
+    {
+      "epoch": 0.6034,
+      "grad_norm": 0.9711331129074097,
+      "learning_rate": 7.933763058457435e-05,
+      "loss": 1.1252,
+      "step": 27153
+    },
+    {
+      "epoch": 0.6034222222222222,
+      "grad_norm": 1.2324848175048828,
+      "learning_rate": 7.933318515225606e-05,
+      "loss": 1.3029,
+      "step": 27154
+    },
+    {
+      "epoch": 0.6034444444444444,
+      "grad_norm": 1.7871129512786865,
+      "learning_rate": 7.932873971993777e-05,
+      "loss": 2.4237,
+      "step": 27155
+    },
+    {
+      "epoch": 0.6034666666666667,
+      "grad_norm": 1.6308362483978271,
+      "learning_rate": 7.932429428761948e-05,
+      "loss": 2.4098,
+      "step": 27156
+    },
+    {
+      "epoch": 0.6034888888888889,
+      "grad_norm": 1.7439035177230835,
+      "learning_rate": 7.931984885530119e-05,
+      "loss": 2.3384,
+      "step": 27157
+    },
+    {
+      "epoch": 0.6035111111111111,
+      "grad_norm": 1.5378587245941162,
+      "learning_rate": 7.93154034229829e-05,
+      "loss": 2.3004,
+      "step": 27158
+    },
+    {
+      "epoch": 0.6035333333333334,
+      "grad_norm": 1.8610527515411377,
+      "learning_rate": 7.931095799066459e-05,
+      "loss": 2.3126,
+      "step": 27159
+    },
+    {
+      "epoch": 0.6035555555555555,
+      "grad_norm": 1.5449259281158447,
+      "learning_rate": 7.93065125583463e-05,
+      "loss": 2.3821,
+      "step": 27160
+    },
+    {
+      "epoch": 0.6035777777777778,
+      "grad_norm": 1.38969886302948,
+      "learning_rate": 7.930206712602801e-05,
+      "loss": 1.047,
+      "step": 27161
+    },
+    {
+      "epoch": 0.6036,
+      "grad_norm": 1.143798828125,
+      "learning_rate": 7.92976216937097e-05,
+      "loss": 0.9785,
+      "step": 27162
+    },
+    {
+      "epoch": 0.6036222222222222,
+      "grad_norm": 1.6547763347625732,
+      "learning_rate": 7.929317626139142e-05,
+      "loss": 1.6648,
+      "step": 27163
+    },
+    {
+      "epoch": 0.6036444444444444,
+      "grad_norm": 1.617087960243225,
+      "learning_rate": 7.928873082907314e-05,
+      "loss": 1.9899,
+      "step": 27164
+    },
+    {
+      "epoch": 0.6036666666666667,
+      "grad_norm": 1.8198072910308838,
+      "learning_rate": 7.928428539675485e-05,
+      "loss": 1.7478,
+      "step": 27165
+    },
+    {
+      "epoch": 0.6036888888888889,
+      "grad_norm": 1.6863237619400024,
+      "learning_rate": 7.927983996443654e-05,
+      "loss": 1.818,
+      "step": 27166
+    },
+    {
+      "epoch": 0.6037111111111111,
+      "grad_norm": 1.8025907278060913,
+      "learning_rate": 7.927539453211825e-05,
+      "loss": 2.3106,
+      "step": 27167
+    },
+    {
+      "epoch": 0.6037333333333333,
+      "grad_norm": 1.6756824254989624,
+      "learning_rate": 7.927094909979996e-05,
+      "loss": 1.7787,
+      "step": 27168
+    },
+    {
+      "epoch": 0.6037555555555556,
+      "grad_norm": 1.456647276878357,
+      "learning_rate": 7.926650366748166e-05,
+      "loss": 2.0248,
+      "step": 27169
+    },
+    {
+      "epoch": 0.6037777777777777,
+      "grad_norm": 1.53804612159729,
+      "learning_rate": 7.926205823516337e-05,
+      "loss": 1.9316,
+      "step": 27170
+    },
+    {
+      "epoch": 0.6038,
+      "grad_norm": 1.5869379043579102,
+      "learning_rate": 7.925761280284508e-05,
+      "loss": 1.799,
+      "step": 27171
+    },
+    {
+      "epoch": 0.6038222222222223,
+      "grad_norm": 1.6954032182693481,
+      "learning_rate": 7.925316737052679e-05,
+      "loss": 2.29,
+      "step": 27172
+    },
+    {
+      "epoch": 0.6038444444444444,
+      "grad_norm": 1.6798923015594482,
+      "learning_rate": 7.92487219382085e-05,
+      "loss": 1.4173,
+      "step": 27173
+    },
+    {
+      "epoch": 0.6038666666666667,
+      "grad_norm": 1.5482616424560547,
+      "learning_rate": 7.924427650589021e-05,
+      "loss": 2.2306,
+      "step": 27174
+    },
+    {
+      "epoch": 0.6038888888888889,
+      "grad_norm": 1.8107613325119019,
+      "learning_rate": 7.923983107357192e-05,
+      "loss": 1.8407,
+      "step": 27175
+    },
+    {
+      "epoch": 0.6039111111111111,
+      "grad_norm": 1.5991325378417969,
+      "learning_rate": 7.923538564125361e-05,
+      "loss": 1.8254,
+      "step": 27176
+    },
+    {
+      "epoch": 0.6039333333333333,
+      "grad_norm": 1.5964834690093994,
+      "learning_rate": 7.923094020893532e-05,
+      "loss": 1.9868,
+      "step": 27177
+    },
+    {
+      "epoch": 0.6039555555555556,
+      "grad_norm": 1.3201031684875488,
+      "learning_rate": 7.922649477661703e-05,
+      "loss": 1.4255,
+      "step": 27178
+    },
+    {
+      "epoch": 0.6039777777777777,
+      "grad_norm": 1.61896550655365,
+      "learning_rate": 7.922204934429873e-05,
+      "loss": 1.9333,
+      "step": 27179
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 1.6375930309295654,
+      "learning_rate": 7.921760391198044e-05,
+      "loss": 1.5589,
+      "step": 27180
+    },
+    {
+      "epoch": 0.6040222222222222,
+      "grad_norm": 1.830711841583252,
+      "learning_rate": 7.921315847966215e-05,
+      "loss": 1.9801,
+      "step": 27181
+    },
+    {
+      "epoch": 0.6040444444444445,
+      "grad_norm": 1.8794705867767334,
+      "learning_rate": 7.920871304734386e-05,
+      "loss": 2.5178,
+      "step": 27182
+    },
+    {
+      "epoch": 0.6040666666666666,
+      "grad_norm": 2.103665590286255,
+      "learning_rate": 7.920426761502557e-05,
+      "loss": 2.5723,
+      "step": 27183
+    },
+    {
+      "epoch": 0.6040888888888889,
+      "grad_norm": 1.7538213729858398,
+      "learning_rate": 7.919982218270728e-05,
+      "loss": 1.7259,
+      "step": 27184
+    },
+    {
+      "epoch": 0.6041111111111112,
+      "grad_norm": 1.6424554586410522,
+      "learning_rate": 7.919537675038899e-05,
+      "loss": 1.8007,
+      "step": 27185
+    },
+    {
+      "epoch": 0.6041333333333333,
+      "grad_norm": 1.4428913593292236,
+      "learning_rate": 7.919093131807068e-05,
+      "loss": 1.6839,
+      "step": 27186
+    },
+    {
+      "epoch": 0.6041555555555556,
+      "grad_norm": 0.2574564516544342,
+      "learning_rate": 7.918648588575239e-05,
+      "loss": 0.0344,
+      "step": 27187
+    },
+    {
+      "epoch": 0.6041777777777778,
+      "grad_norm": 1.7294193506240845,
+      "learning_rate": 7.91820404534341e-05,
+      "loss": 1.8667,
+      "step": 27188
+    },
+    {
+      "epoch": 0.6042,
+      "grad_norm": 1.966111660003662,
+      "learning_rate": 7.91775950211158e-05,
+      "loss": 1.6948,
+      "step": 27189
+    },
+    {
+      "epoch": 0.6042222222222222,
+      "grad_norm": 1.6431033611297607,
+      "learning_rate": 7.917314958879751e-05,
+      "loss": 1.8165,
+      "step": 27190
+    },
+    {
+      "epoch": 0.6042444444444445,
+      "grad_norm": 1.6426076889038086,
+      "learning_rate": 7.916870415647922e-05,
+      "loss": 1.5127,
+      "step": 27191
+    },
+    {
+      "epoch": 0.6042666666666666,
+      "grad_norm": 1.80479896068573,
+      "learning_rate": 7.916425872416093e-05,
+      "loss": 1.668,
+      "step": 27192
+    },
+    {
+      "epoch": 0.6042888888888889,
+      "grad_norm": 1.8984283208847046,
+      "learning_rate": 7.915981329184264e-05,
+      "loss": 1.628,
+      "step": 27193
+    },
+    {
+      "epoch": 0.6043111111111111,
+      "grad_norm": 1.841007947921753,
+      "learning_rate": 7.915536785952435e-05,
+      "loss": 1.7181,
+      "step": 27194
+    },
+    {
+      "epoch": 0.6043333333333333,
+      "grad_norm": 1.8353080749511719,
+      "learning_rate": 7.915092242720606e-05,
+      "loss": 1.7911,
+      "step": 27195
+    },
+    {
+      "epoch": 0.6043555555555555,
+      "grad_norm": 2.1403040885925293,
+      "learning_rate": 7.914647699488775e-05,
+      "loss": 2.1268,
+      "step": 27196
+    },
+    {
+      "epoch": 0.6043777777777778,
+      "grad_norm": 1.801777958869934,
+      "learning_rate": 7.914203156256946e-05,
+      "loss": 1.7257,
+      "step": 27197
+    },
+    {
+      "epoch": 0.6044,
+      "grad_norm": 1.9252780675888062,
+      "learning_rate": 7.913758613025117e-05,
+      "loss": 1.7415,
+      "step": 27198
+    },
+    {
+      "epoch": 0.6044222222222222,
+      "grad_norm": 2.0512959957122803,
+      "learning_rate": 7.913314069793287e-05,
+      "loss": 1.5827,
+      "step": 27199
+    },
+    {
+      "epoch": 0.6044444444444445,
+      "grad_norm": 2.0335724353790283,
+      "learning_rate": 7.912869526561458e-05,
+      "loss": 1.6761,
+      "step": 27200
+    },
+    {
+      "epoch": 0.6044666666666667,
+      "grad_norm": 1.3120962381362915,
+      "learning_rate": 7.91242498332963e-05,
+      "loss": 2.4573,
+      "step": 27201
+    },
+    {
+      "epoch": 0.6044888888888889,
+      "grad_norm": 1.1846799850463867,
+      "learning_rate": 7.911980440097801e-05,
+      "loss": 1.0285,
+      "step": 27202
+    },
+    {
+      "epoch": 0.6045111111111111,
+      "grad_norm": 1.609234094619751,
+      "learning_rate": 7.91153589686597e-05,
+      "loss": 2.4472,
+      "step": 27203
+    },
+    {
+      "epoch": 0.6045333333333334,
+      "grad_norm": 1.491317868232727,
+      "learning_rate": 7.911091353634142e-05,
+      "loss": 2.3505,
+      "step": 27204
+    },
+    {
+      "epoch": 0.6045555555555555,
+      "grad_norm": 1.569449782371521,
+      "learning_rate": 7.910646810402312e-05,
+      "loss": 2.0256,
+      "step": 27205
+    },
+    {
+      "epoch": 0.6045777777777778,
+      "grad_norm": 1.4014590978622437,
+      "learning_rate": 7.910202267170482e-05,
+      "loss": 2.0003,
+      "step": 27206
+    },
+    {
+      "epoch": 0.6046,
+      "grad_norm": 1.7583537101745605,
+      "learning_rate": 7.909757723938653e-05,
+      "loss": 2.4809,
+      "step": 27207
+    },
+    {
+      "epoch": 0.6046222222222222,
+      "grad_norm": 1.4294236898422241,
+      "learning_rate": 7.909313180706824e-05,
+      "loss": 2.1107,
+      "step": 27208
+    },
+    {
+      "epoch": 0.6046444444444444,
+      "grad_norm": 1.4733035564422607,
+      "learning_rate": 7.908868637474994e-05,
+      "loss": 1.9086,
+      "step": 27209
+    },
+    {
+      "epoch": 0.6046666666666667,
+      "grad_norm": 2.2566213607788086,
+      "learning_rate": 7.908424094243166e-05,
+      "loss": 2.483,
+      "step": 27210
+    },
+    {
+      "epoch": 0.6046888888888889,
+      "grad_norm": 1.5477021932601929,
+      "learning_rate": 7.907979551011337e-05,
+      "loss": 1.8564,
+      "step": 27211
+    },
+    {
+      "epoch": 0.6047111111111111,
+      "grad_norm": 1.6028445959091187,
+      "learning_rate": 7.907535007779508e-05,
+      "loss": 2.3341,
+      "step": 27212
+    },
+    {
+      "epoch": 0.6047333333333333,
+      "grad_norm": 1.5999501943588257,
+      "learning_rate": 7.907090464547677e-05,
+      "loss": 2.1941,
+      "step": 27213
+    },
+    {
+      "epoch": 0.6047555555555556,
+      "grad_norm": 1.3347578048706055,
+      "learning_rate": 7.906645921315848e-05,
+      "loss": 1.9165,
+      "step": 27214
+    },
+    {
+      "epoch": 0.6047777777777777,
+      "grad_norm": 1.7065314054489136,
+      "learning_rate": 7.90620137808402e-05,
+      "loss": 1.665,
+      "step": 27215
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 1.4490126371383667,
+      "learning_rate": 7.905756834852189e-05,
+      "loss": 1.9074,
+      "step": 27216
+    },
+    {
+      "epoch": 0.6048222222222223,
+      "grad_norm": 1.6703357696533203,
+      "learning_rate": 7.90531229162036e-05,
+      "loss": 2.252,
+      "step": 27217
+    },
+    {
+      "epoch": 0.6048444444444444,
+      "grad_norm": 1.8422366380691528,
+      "learning_rate": 7.904867748388531e-05,
+      "loss": 2.2362,
+      "step": 27218
+    },
+    {
+      "epoch": 0.6048666666666667,
+      "grad_norm": 1.5340640544891357,
+      "learning_rate": 7.904423205156702e-05,
+      "loss": 1.6657,
+      "step": 27219
+    },
+    {
+      "epoch": 0.6048888888888889,
+      "grad_norm": 1.7700345516204834,
+      "learning_rate": 7.903978661924873e-05,
+      "loss": 2.0036,
+      "step": 27220
+    },
+    {
+      "epoch": 0.6049111111111111,
+      "grad_norm": 1.6051772832870483,
+      "learning_rate": 7.903534118693044e-05,
+      "loss": 1.916,
+      "step": 27221
+    },
+    {
+      "epoch": 0.6049333333333333,
+      "grad_norm": 1.0932059288024902,
+      "learning_rate": 7.903089575461215e-05,
+      "loss": 0.9623,
+      "step": 27222
+    },
+    {
+      "epoch": 0.6049555555555556,
+      "grad_norm": 1.5146722793579102,
+      "learning_rate": 7.902645032229384e-05,
+      "loss": 1.7762,
+      "step": 27223
+    },
+    {
+      "epoch": 0.6049777777777777,
+      "grad_norm": 1.7188857793807983,
+      "learning_rate": 7.902200488997555e-05,
+      "loss": 2.0552,
+      "step": 27224
+    },
+    {
+      "epoch": 0.605,
+      "grad_norm": 1.9077479839324951,
+      "learning_rate": 7.901755945765726e-05,
+      "loss": 2.2053,
+      "step": 27225
+    },
+    {
+      "epoch": 0.6050222222222222,
+      "grad_norm": 1.4720269441604614,
+      "learning_rate": 7.901311402533896e-05,
+      "loss": 1.8231,
+      "step": 27226
+    },
+    {
+      "epoch": 0.6050444444444445,
+      "grad_norm": 2.066772222518921,
+      "learning_rate": 7.900866859302067e-05,
+      "loss": 1.6994,
+      "step": 27227
+    },
+    {
+      "epoch": 0.6050666666666666,
+      "grad_norm": 1.4600918292999268,
+      "learning_rate": 7.900422316070238e-05,
+      "loss": 1.5678,
+      "step": 27228
+    },
+    {
+      "epoch": 0.6050888888888889,
+      "grad_norm": 1.7550400495529175,
+      "learning_rate": 7.899977772838409e-05,
+      "loss": 1.6418,
+      "step": 27229
+    },
+    {
+      "epoch": 0.6051111111111112,
+      "grad_norm": 1.729282021522522,
+      "learning_rate": 7.89953322960658e-05,
+      "loss": 2.0124,
+      "step": 27230
+    },
+    {
+      "epoch": 0.6051333333333333,
+      "grad_norm": 1.740505576133728,
+      "learning_rate": 7.899088686374751e-05,
+      "loss": 1.8965,
+      "step": 27231
+    },
+    {
+      "epoch": 0.6051555555555556,
+      "grad_norm": 1.786176085472107,
+      "learning_rate": 7.898644143142922e-05,
+      "loss": 2.0738,
+      "step": 27232
+    },
+    {
+      "epoch": 0.6051777777777778,
+      "grad_norm": 1.7551746368408203,
+      "learning_rate": 7.898199599911091e-05,
+      "loss": 2.156,
+      "step": 27233
+    },
+    {
+      "epoch": 0.6052,
+      "grad_norm": 1.7817493677139282,
+      "learning_rate": 7.897755056679262e-05,
+      "loss": 2.104,
+      "step": 27234
+    },
+    {
+      "epoch": 0.6052222222222222,
+      "grad_norm": 1.8015249967575073,
+      "learning_rate": 7.897310513447433e-05,
+      "loss": 1.7312,
+      "step": 27235
+    },
+    {
+      "epoch": 0.6052444444444445,
+      "grad_norm": 1.0835497379302979,
+      "learning_rate": 7.896865970215603e-05,
+      "loss": 0.8679,
+      "step": 27236
+    },
+    {
+      "epoch": 0.6052666666666666,
+      "grad_norm": 1.620829701423645,
+      "learning_rate": 7.896421426983774e-05,
+      "loss": 1.9263,
+      "step": 27237
+    },
+    {
+      "epoch": 0.6052888888888889,
+      "grad_norm": 1.481078863143921,
+      "learning_rate": 7.895976883751946e-05,
+      "loss": 1.5824,
+      "step": 27238
+    },
+    {
+      "epoch": 0.6053111111111111,
+      "grad_norm": 2.2054390907287598,
+      "learning_rate": 7.895532340520117e-05,
+      "loss": 2.0811,
+      "step": 27239
+    },
+    {
+      "epoch": 0.6053333333333333,
+      "grad_norm": 1.6682418584823608,
+      "learning_rate": 7.895087797288287e-05,
+      "loss": 1.8911,
+      "step": 27240
+    },
+    {
+      "epoch": 0.6053555555555555,
+      "grad_norm": 1.5624669790267944,
+      "learning_rate": 7.894643254056458e-05,
+      "loss": 1.5574,
+      "step": 27241
+    },
+    {
+      "epoch": 0.6053777777777778,
+      "grad_norm": 1.5307351350784302,
+      "learning_rate": 7.894198710824629e-05,
+      "loss": 1.6266,
+      "step": 27242
+    },
+    {
+      "epoch": 0.6054,
+      "grad_norm": 1.8009110689163208,
+      "learning_rate": 7.893754167592798e-05,
+      "loss": 1.8157,
+      "step": 27243
+    },
+    {
+      "epoch": 0.6054222222222222,
+      "grad_norm": 2.319335460662842,
+      "learning_rate": 7.893309624360969e-05,
+      "loss": 1.9652,
+      "step": 27244
+    },
+    {
+      "epoch": 0.6054444444444445,
+      "grad_norm": 1.7203480005264282,
+      "learning_rate": 7.89286508112914e-05,
+      "loss": 1.9449,
+      "step": 27245
+    },
+    {
+      "epoch": 0.6054666666666667,
+      "grad_norm": 1.9069678783416748,
+      "learning_rate": 7.89242053789731e-05,
+      "loss": 1.9355,
+      "step": 27246
+    },
+    {
+      "epoch": 0.6054888888888889,
+      "grad_norm": 2.3656368255615234,
+      "learning_rate": 7.891975994665482e-05,
+      "loss": 2.183,
+      "step": 27247
+    },
+    {
+      "epoch": 0.6055111111111111,
+      "grad_norm": 1.825256109237671,
+      "learning_rate": 7.891531451433653e-05,
+      "loss": 1.6806,
+      "step": 27248
+    },
+    {
+      "epoch": 0.6055333333333334,
+      "grad_norm": 1.5662660598754883,
+      "learning_rate": 7.891086908201824e-05,
+      "loss": 1.3733,
+      "step": 27249
+    },
+    {
+      "epoch": 0.6055555555555555,
+      "grad_norm": 1.6754958629608154,
+      "learning_rate": 7.890642364969994e-05,
+      "loss": 1.415,
+      "step": 27250
+    },
+    {
+      "epoch": 0.6055777777777778,
+      "grad_norm": 1.008105754852295,
+      "learning_rate": 7.890197821738165e-05,
+      "loss": 1.1714,
+      "step": 27251
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 1.5311484336853027,
+      "learning_rate": 7.889753278506336e-05,
+      "loss": 2.2801,
+      "step": 27252
+    },
+    {
+      "epoch": 0.6056222222222222,
+      "grad_norm": 1.440260648727417,
+      "learning_rate": 7.889308735274505e-05,
+      "loss": 1.8973,
+      "step": 27253
+    },
+    {
+      "epoch": 0.6056444444444444,
+      "grad_norm": 1.7567195892333984,
+      "learning_rate": 7.888864192042676e-05,
+      "loss": 2.3299,
+      "step": 27254
+    },
+    {
+      "epoch": 0.6056666666666667,
+      "grad_norm": 1.4583789110183716,
+      "learning_rate": 7.888419648810847e-05,
+      "loss": 2.1668,
+      "step": 27255
+    },
+    {
+      "epoch": 0.6056888888888889,
+      "grad_norm": 1.5148966312408447,
+      "learning_rate": 7.887975105579018e-05,
+      "loss": 2.1548,
+      "step": 27256
+    },
+    {
+      "epoch": 0.6057111111111111,
+      "grad_norm": 1.9175471067428589,
+      "learning_rate": 7.887530562347189e-05,
+      "loss": 2.3519,
+      "step": 27257
+    },
+    {
+      "epoch": 0.6057333333333333,
+      "grad_norm": 1.5939949750900269,
+      "learning_rate": 7.88708601911536e-05,
+      "loss": 2.2127,
+      "step": 27258
+    },
+    {
+      "epoch": 0.6057555555555556,
+      "grad_norm": 0.9889760613441467,
+      "learning_rate": 7.886641475883531e-05,
+      "loss": 1.006,
+      "step": 27259
+    },
+    {
+      "epoch": 0.6057777777777777,
+      "grad_norm": 1.8088592290878296,
+      "learning_rate": 7.8861969326517e-05,
+      "loss": 1.9756,
+      "step": 27260
+    },
+    {
+      "epoch": 0.6058,
+      "grad_norm": 1.4391329288482666,
+      "learning_rate": 7.885752389419871e-05,
+      "loss": 1.9602,
+      "step": 27261
+    },
+    {
+      "epoch": 0.6058222222222223,
+      "grad_norm": 1.4422990083694458,
+      "learning_rate": 7.885307846188042e-05,
+      "loss": 1.9709,
+      "step": 27262
+    },
+    {
+      "epoch": 0.6058444444444444,
+      "grad_norm": 1.6840225458145142,
+      "learning_rate": 7.884863302956212e-05,
+      "loss": 2.2694,
+      "step": 27263
+    },
+    {
+      "epoch": 0.6058666666666667,
+      "grad_norm": 1.5895969867706299,
+      "learning_rate": 7.884418759724383e-05,
+      "loss": 1.5241,
+      "step": 27264
+    },
+    {
+      "epoch": 0.6058888888888889,
+      "grad_norm": 1.3911423683166504,
+      "learning_rate": 7.883974216492554e-05,
+      "loss": 1.8146,
+      "step": 27265
+    },
+    {
+      "epoch": 0.6059111111111111,
+      "grad_norm": 1.4521762132644653,
+      "learning_rate": 7.883529673260725e-05,
+      "loss": 1.9944,
+      "step": 27266
+    },
+    {
+      "epoch": 0.6059333333333333,
+      "grad_norm": 1.5626418590545654,
+      "learning_rate": 7.883085130028896e-05,
+      "loss": 1.9758,
+      "step": 27267
+    },
+    {
+      "epoch": 0.6059555555555556,
+      "grad_norm": 1.6539586782455444,
+      "learning_rate": 7.882640586797067e-05,
+      "loss": 1.8796,
+      "step": 27268
+    },
+    {
+      "epoch": 0.6059777777777777,
+      "grad_norm": 1.6141746044158936,
+      "learning_rate": 7.882196043565238e-05,
+      "loss": 1.9365,
+      "step": 27269
+    },
+    {
+      "epoch": 0.606,
+      "grad_norm": 1.7146342992782593,
+      "learning_rate": 7.881751500333407e-05,
+      "loss": 1.8579,
+      "step": 27270
+    },
+    {
+      "epoch": 0.6060222222222222,
+      "grad_norm": 1.6857411861419678,
+      "learning_rate": 7.881306957101578e-05,
+      "loss": 1.7205,
+      "step": 27271
+    },
+    {
+      "epoch": 0.6060444444444445,
+      "grad_norm": 1.569933295249939,
+      "learning_rate": 7.88086241386975e-05,
+      "loss": 1.335,
+      "step": 27272
+    },
+    {
+      "epoch": 0.6060666666666666,
+      "grad_norm": 1.507858395576477,
+      "learning_rate": 7.880417870637919e-05,
+      "loss": 1.7434,
+      "step": 27273
+    },
+    {
+      "epoch": 0.6060888888888889,
+      "grad_norm": 1.8040244579315186,
+      "learning_rate": 7.87997332740609e-05,
+      "loss": 2.213,
+      "step": 27274
+    },
+    {
+      "epoch": 0.6061111111111112,
+      "grad_norm": 1.3667738437652588,
+      "learning_rate": 7.879528784174262e-05,
+      "loss": 1.572,
+      "step": 27275
+    },
+    {
+      "epoch": 0.6061333333333333,
+      "grad_norm": 1.5162640810012817,
+      "learning_rate": 7.879084240942433e-05,
+      "loss": 1.8487,
+      "step": 27276
+    },
+    {
+      "epoch": 0.6061555555555556,
+      "grad_norm": 1.6777446269989014,
+      "learning_rate": 7.878639697710603e-05,
+      "loss": 1.6855,
+      "step": 27277
+    },
+    {
+      "epoch": 0.6061777777777778,
+      "grad_norm": 1.7804574966430664,
+      "learning_rate": 7.878195154478774e-05,
+      "loss": 2.2795,
+      "step": 27278
+    },
+    {
+      "epoch": 0.6062,
+      "grad_norm": 1.5175312757492065,
+      "learning_rate": 7.877750611246945e-05,
+      "loss": 0.8556,
+      "step": 27279
+    },
+    {
+      "epoch": 0.6062222222222222,
+      "grad_norm": 2.4243533611297607,
+      "learning_rate": 7.877306068015114e-05,
+      "loss": 2.2855,
+      "step": 27280
+    },
+    {
+      "epoch": 0.6062444444444445,
+      "grad_norm": 1.8361667394638062,
+      "learning_rate": 7.876861524783285e-05,
+      "loss": 1.5862,
+      "step": 27281
+    },
+    {
+      "epoch": 0.6062666666666666,
+      "grad_norm": 1.846785068511963,
+      "learning_rate": 7.876416981551456e-05,
+      "loss": 1.634,
+      "step": 27282
+    },
+    {
+      "epoch": 0.6062888888888889,
+      "grad_norm": 1.6902236938476562,
+      "learning_rate": 7.875972438319626e-05,
+      "loss": 1.8415,
+      "step": 27283
+    },
+    {
+      "epoch": 0.6063111111111111,
+      "grad_norm": 1.7941762208938599,
+      "learning_rate": 7.875527895087798e-05,
+      "loss": 2.0218,
+      "step": 27284
+    },
+    {
+      "epoch": 0.6063333333333333,
+      "grad_norm": 2.0850422382354736,
+      "learning_rate": 7.875083351855969e-05,
+      "loss": 2.2714,
+      "step": 27285
+    },
+    {
+      "epoch": 0.6063555555555555,
+      "grad_norm": 1.5774202346801758,
+      "learning_rate": 7.87463880862414e-05,
+      "loss": 1.7006,
+      "step": 27286
+    },
+    {
+      "epoch": 0.6063777777777778,
+      "grad_norm": 1.839981198310852,
+      "learning_rate": 7.87419426539231e-05,
+      "loss": 1.9587,
+      "step": 27287
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 2.1352789402008057,
+      "learning_rate": 7.87374972216048e-05,
+      "loss": 1.5797,
+      "step": 27288
+    },
+    {
+      "epoch": 0.6064222222222222,
+      "grad_norm": 1.870865821838379,
+      "learning_rate": 7.873305178928652e-05,
+      "loss": 2.1551,
+      "step": 27289
+    },
+    {
+      "epoch": 0.6064444444444445,
+      "grad_norm": 1.6629317998886108,
+      "learning_rate": 7.872860635696821e-05,
+      "loss": 1.7822,
+      "step": 27290
+    },
+    {
+      "epoch": 0.6064666666666667,
+      "grad_norm": 1.589384913444519,
+      "learning_rate": 7.872416092464992e-05,
+      "loss": 1.4413,
+      "step": 27291
+    },
+    {
+      "epoch": 0.6064888888888889,
+      "grad_norm": 2.1463875770568848,
+      "learning_rate": 7.871971549233163e-05,
+      "loss": 1.874,
+      "step": 27292
+    },
+    {
+      "epoch": 0.6065111111111111,
+      "grad_norm": 1.9512488842010498,
+      "learning_rate": 7.871527006001334e-05,
+      "loss": 1.79,
+      "step": 27293
+    },
+    {
+      "epoch": 0.6065333333333334,
+      "grad_norm": 1.5353994369506836,
+      "learning_rate": 7.871082462769505e-05,
+      "loss": 1.5228,
+      "step": 27294
+    },
+    {
+      "epoch": 0.6065555555555555,
+      "grad_norm": 1.8420060873031616,
+      "learning_rate": 7.870637919537676e-05,
+      "loss": 2.0819,
+      "step": 27295
+    },
+    {
+      "epoch": 0.6065777777777778,
+      "grad_norm": 1.686511754989624,
+      "learning_rate": 7.870193376305847e-05,
+      "loss": 1.87,
+      "step": 27296
+    },
+    {
+      "epoch": 0.6066,
+      "grad_norm": 1.0739492177963257,
+      "learning_rate": 7.869748833074017e-05,
+      "loss": 0.8274,
+      "step": 27297
+    },
+    {
+      "epoch": 0.6066222222222222,
+      "grad_norm": 1.7083632946014404,
+      "learning_rate": 7.869304289842188e-05,
+      "loss": 1.4793,
+      "step": 27298
+    },
+    {
+      "epoch": 0.6066444444444444,
+      "grad_norm": 1.871471643447876,
+      "learning_rate": 7.868859746610359e-05,
+      "loss": 1.7419,
+      "step": 27299
+    },
+    {
+      "epoch": 0.6066666666666667,
+      "grad_norm": 2.1125330924987793,
+      "learning_rate": 7.868415203378528e-05,
+      "loss": 2.0606,
+      "step": 27300
+    },
+    {
+      "epoch": 0.6066888888888889,
+      "grad_norm": 1.3407315015792847,
+      "learning_rate": 7.867970660146699e-05,
+      "loss": 2.1311,
+      "step": 27301
+    },
+    {
+      "epoch": 0.6067111111111111,
+      "grad_norm": 1.8985822200775146,
+      "learning_rate": 7.86752611691487e-05,
+      "loss": 2.2525,
+      "step": 27302
+    },
+    {
+      "epoch": 0.6067333333333333,
+      "grad_norm": 1.5596638917922974,
+      "learning_rate": 7.867081573683041e-05,
+      "loss": 2.4506,
+      "step": 27303
+    },
+    {
+      "epoch": 0.6067555555555556,
+      "grad_norm": 2.5087380409240723,
+      "learning_rate": 7.866637030451212e-05,
+      "loss": 2.2414,
+      "step": 27304
+    },
+    {
+      "epoch": 0.6067777777777777,
+      "grad_norm": 1.667316198348999,
+      "learning_rate": 7.866192487219383e-05,
+      "loss": 2.4115,
+      "step": 27305
+    },
+    {
+      "epoch": 0.6068,
+      "grad_norm": 1.6813075542449951,
+      "learning_rate": 7.865747943987554e-05,
+      "loss": 2.1599,
+      "step": 27306
+    },
+    {
+      "epoch": 0.6068222222222223,
+      "grad_norm": 1.5888385772705078,
+      "learning_rate": 7.865303400755724e-05,
+      "loss": 2.2666,
+      "step": 27307
+    },
+    {
+      "epoch": 0.6068444444444444,
+      "grad_norm": 1.5145535469055176,
+      "learning_rate": 7.864858857523894e-05,
+      "loss": 2.2976,
+      "step": 27308
+    },
+    {
+      "epoch": 0.6068666666666667,
+      "grad_norm": 1.3751749992370605,
+      "learning_rate": 7.864414314292065e-05,
+      "loss": 1.326,
+      "step": 27309
+    },
+    {
+      "epoch": 0.6068888888888889,
+      "grad_norm": 1.3869388103485107,
+      "learning_rate": 7.863969771060235e-05,
+      "loss": 1.7573,
+      "step": 27310
+    },
+    {
+      "epoch": 0.6069111111111111,
+      "grad_norm": 1.439608097076416,
+      "learning_rate": 7.863525227828406e-05,
+      "loss": 1.6373,
+      "step": 27311
+    },
+    {
+      "epoch": 0.6069333333333333,
+      "grad_norm": 1.6022894382476807,
+      "learning_rate": 7.863080684596578e-05,
+      "loss": 1.9089,
+      "step": 27312
+    },
+    {
+      "epoch": 0.6069555555555556,
+      "grad_norm": 1.7709941864013672,
+      "learning_rate": 7.862636141364748e-05,
+      "loss": 2.3322,
+      "step": 27313
+    },
+    {
+      "epoch": 0.6069777777777777,
+      "grad_norm": 1.721430778503418,
+      "learning_rate": 7.862191598132919e-05,
+      "loss": 2.0007,
+      "step": 27314
+    },
+    {
+      "epoch": 0.607,
+      "grad_norm": 1.7934560775756836,
+      "learning_rate": 7.86174705490109e-05,
+      "loss": 2.5062,
+      "step": 27315
+    },
+    {
+      "epoch": 0.6070222222222222,
+      "grad_norm": 1.1028038263320923,
+      "learning_rate": 7.861302511669261e-05,
+      "loss": 0.6188,
+      "step": 27316
+    },
+    {
+      "epoch": 0.6070444444444445,
+      "grad_norm": 1.4227097034454346,
+      "learning_rate": 7.86085796843743e-05,
+      "loss": 1.8313,
+      "step": 27317
+    },
+    {
+      "epoch": 0.6070666666666666,
+      "grad_norm": 1.8850692510604858,
+      "learning_rate": 7.860413425205601e-05,
+      "loss": 2.3957,
+      "step": 27318
+    },
+    {
+      "epoch": 0.6070888888888889,
+      "grad_norm": 1.5272574424743652,
+      "learning_rate": 7.859968881973772e-05,
+      "loss": 1.9499,
+      "step": 27319
+    },
+    {
+      "epoch": 0.6071111111111112,
+      "grad_norm": 1.9343817234039307,
+      "learning_rate": 7.859524338741942e-05,
+      "loss": 2.4507,
+      "step": 27320
+    },
+    {
+      "epoch": 0.6071333333333333,
+      "grad_norm": 2.05100679397583,
+      "learning_rate": 7.859079795510114e-05,
+      "loss": 2.0323,
+      "step": 27321
+    },
+    {
+      "epoch": 0.6071555555555556,
+      "grad_norm": 1.5332660675048828,
+      "learning_rate": 7.858635252278285e-05,
+      "loss": 1.4606,
+      "step": 27322
+    },
+    {
+      "epoch": 0.6071777777777778,
+      "grad_norm": 1.5789637565612793,
+      "learning_rate": 7.858190709046455e-05,
+      "loss": 2.0889,
+      "step": 27323
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 1.5830676555633545,
+      "learning_rate": 7.857746165814626e-05,
+      "loss": 1.7474,
+      "step": 27324
+    },
+    {
+      "epoch": 0.6072222222222222,
+      "grad_norm": 1.554005742073059,
+      "learning_rate": 7.857301622582797e-05,
+      "loss": 1.7042,
+      "step": 27325
+    },
+    {
+      "epoch": 0.6072444444444445,
+      "grad_norm": 1.6211838722229004,
+      "learning_rate": 7.856857079350968e-05,
+      "loss": 2.1875,
+      "step": 27326
+    },
+    {
+      "epoch": 0.6072666666666666,
+      "grad_norm": 1.5243436098098755,
+      "learning_rate": 7.856412536119137e-05,
+      "loss": 1.3351,
+      "step": 27327
+    },
+    {
+      "epoch": 0.6072888888888889,
+      "grad_norm": 1.7410411834716797,
+      "learning_rate": 7.855967992887308e-05,
+      "loss": 2.1272,
+      "step": 27328
+    },
+    {
+      "epoch": 0.6073111111111111,
+      "grad_norm": 1.3875596523284912,
+      "learning_rate": 7.855523449655479e-05,
+      "loss": 1.3673,
+      "step": 27329
+    },
+    {
+      "epoch": 0.6073333333333333,
+      "grad_norm": 1.7309584617614746,
+      "learning_rate": 7.85507890642365e-05,
+      "loss": 1.8283,
+      "step": 27330
+    },
+    {
+      "epoch": 0.6073555555555555,
+      "grad_norm": 1.6569769382476807,
+      "learning_rate": 7.854634363191821e-05,
+      "loss": 1.8252,
+      "step": 27331
+    },
+    {
+      "epoch": 0.6073777777777778,
+      "grad_norm": 1.6704628467559814,
+      "learning_rate": 7.854189819959992e-05,
+      "loss": 1.3384,
+      "step": 27332
+    },
+    {
+      "epoch": 0.6074,
+      "grad_norm": 2.1140072345733643,
+      "learning_rate": 7.853745276728163e-05,
+      "loss": 2.4161,
+      "step": 27333
+    },
+    {
+      "epoch": 0.6074222222222222,
+      "grad_norm": 1.9944144487380981,
+      "learning_rate": 7.853300733496333e-05,
+      "loss": 1.5231,
+      "step": 27334
+    },
+    {
+      "epoch": 0.6074444444444445,
+      "grad_norm": 0.5336458086967468,
+      "learning_rate": 7.852856190264504e-05,
+      "loss": 0.0307,
+      "step": 27335
+    },
+    {
+      "epoch": 0.6074666666666667,
+      "grad_norm": 1.9870555400848389,
+      "learning_rate": 7.852411647032675e-05,
+      "loss": 2.0026,
+      "step": 27336
+    },
+    {
+      "epoch": 0.6074888888888889,
+      "grad_norm": 1.7697476148605347,
+      "learning_rate": 7.851967103800844e-05,
+      "loss": 1.9533,
+      "step": 27337
+    },
+    {
+      "epoch": 0.6075111111111111,
+      "grad_norm": 1.5129780769348145,
+      "learning_rate": 7.851522560569015e-05,
+      "loss": 1.7916,
+      "step": 27338
+    },
+    {
+      "epoch": 0.6075333333333334,
+      "grad_norm": 1.7161614894866943,
+      "learning_rate": 7.851078017337186e-05,
+      "loss": 1.9484,
+      "step": 27339
+    },
+    {
+      "epoch": 0.6075555555555555,
+      "grad_norm": 2.0948965549468994,
+      "learning_rate": 7.850633474105357e-05,
+      "loss": 1.9074,
+      "step": 27340
+    },
+    {
+      "epoch": 0.6075777777777778,
+      "grad_norm": 1.6959728002548218,
+      "learning_rate": 7.850188930873528e-05,
+      "loss": 1.0733,
+      "step": 27341
+    },
+    {
+      "epoch": 0.6076,
+      "grad_norm": 2.467322826385498,
+      "learning_rate": 7.849744387641699e-05,
+      "loss": 2.4551,
+      "step": 27342
+    },
+    {
+      "epoch": 0.6076222222222222,
+      "grad_norm": 2.0165164470672607,
+      "learning_rate": 7.84929984440987e-05,
+      "loss": 2.0362,
+      "step": 27343
+    },
+    {
+      "epoch": 0.6076444444444444,
+      "grad_norm": 1.6614750623703003,
+      "learning_rate": 7.84885530117804e-05,
+      "loss": 1.601,
+      "step": 27344
+    },
+    {
+      "epoch": 0.6076666666666667,
+      "grad_norm": 1.7933533191680908,
+      "learning_rate": 7.84841075794621e-05,
+      "loss": 1.6952,
+      "step": 27345
+    },
+    {
+      "epoch": 0.6076888888888888,
+      "grad_norm": 1.201366901397705,
+      "learning_rate": 7.847966214714382e-05,
+      "loss": 0.7091,
+      "step": 27346
+    },
+    {
+      "epoch": 0.6077111111111111,
+      "grad_norm": 1.7066757678985596,
+      "learning_rate": 7.847521671482551e-05,
+      "loss": 1.773,
+      "step": 27347
+    },
+    {
+      "epoch": 0.6077333333333333,
+      "grad_norm": 2.016390800476074,
+      "learning_rate": 7.847077128250722e-05,
+      "loss": 1.7361,
+      "step": 27348
+    },
+    {
+      "epoch": 0.6077555555555556,
+      "grad_norm": 2.0306246280670166,
+      "learning_rate": 7.846632585018894e-05,
+      "loss": 2.0121,
+      "step": 27349
+    },
+    {
+      "epoch": 0.6077777777777778,
+      "grad_norm": 2.100975751876831,
+      "learning_rate": 7.846188041787064e-05,
+      "loss": 1.523,
+      "step": 27350
+    },
+    {
+      "epoch": 0.6078,
+      "grad_norm": 1.7427887916564941,
+      "learning_rate": 7.845743498555235e-05,
+      "loss": 3.027,
+      "step": 27351
+    },
+    {
+      "epoch": 0.6078222222222223,
+      "grad_norm": 1.467203974723816,
+      "learning_rate": 7.845298955323406e-05,
+      "loss": 2.4911,
+      "step": 27352
+    },
+    {
+      "epoch": 0.6078444444444444,
+      "grad_norm": 1.4917911291122437,
+      "learning_rate": 7.844854412091577e-05,
+      "loss": 1.8796,
+      "step": 27353
+    },
+    {
+      "epoch": 0.6078666666666667,
+      "grad_norm": 1.814010500907898,
+      "learning_rate": 7.844409868859747e-05,
+      "loss": 2.5444,
+      "step": 27354
+    },
+    {
+      "epoch": 0.6078888888888889,
+      "grad_norm": 1.3394798040390015,
+      "learning_rate": 7.843965325627917e-05,
+      "loss": 2.0518,
+      "step": 27355
+    },
+    {
+      "epoch": 0.6079111111111111,
+      "grad_norm": 1.4140102863311768,
+      "learning_rate": 7.843520782396088e-05,
+      "loss": 1.8226,
+      "step": 27356
+    },
+    {
+      "epoch": 0.6079333333333333,
+      "grad_norm": 1.5619611740112305,
+      "learning_rate": 7.843076239164258e-05,
+      "loss": 2.2057,
+      "step": 27357
+    },
+    {
+      "epoch": 0.6079555555555556,
+      "grad_norm": 1.7596157789230347,
+      "learning_rate": 7.84263169593243e-05,
+      "loss": 2.0465,
+      "step": 27358
+    },
+    {
+      "epoch": 0.6079777777777777,
+      "grad_norm": 1.662670612335205,
+      "learning_rate": 7.842187152700601e-05,
+      "loss": 2.4204,
+      "step": 27359
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.7003495693206787,
+      "learning_rate": 7.841742609468771e-05,
+      "loss": 2.2705,
+      "step": 27360
+    },
+    {
+      "epoch": 0.6080222222222222,
+      "grad_norm": 1.5685595273971558,
+      "learning_rate": 7.841298066236942e-05,
+      "loss": 1.9704,
+      "step": 27361
+    },
+    {
+      "epoch": 0.6080444444444445,
+      "grad_norm": 1.3184891939163208,
+      "learning_rate": 7.840853523005113e-05,
+      "loss": 1.5464,
+      "step": 27362
+    },
+    {
+      "epoch": 0.6080666666666666,
+      "grad_norm": 1.4261938333511353,
+      "learning_rate": 7.840408979773284e-05,
+      "loss": 1.9069,
+      "step": 27363
+    },
+    {
+      "epoch": 0.6080888888888889,
+      "grad_norm": 1.6896491050720215,
+      "learning_rate": 7.839964436541453e-05,
+      "loss": 2.0931,
+      "step": 27364
+    },
+    {
+      "epoch": 0.6081111111111112,
+      "grad_norm": 1.500627040863037,
+      "learning_rate": 7.839519893309624e-05,
+      "loss": 2.434,
+      "step": 27365
+    },
+    {
+      "epoch": 0.6081333333333333,
+      "grad_norm": 1.5913405418395996,
+      "learning_rate": 7.839075350077795e-05,
+      "loss": 1.9811,
+      "step": 27366
+    },
+    {
+      "epoch": 0.6081555555555556,
+      "grad_norm": 1.6660000085830688,
+      "learning_rate": 7.838630806845966e-05,
+      "loss": 2.097,
+      "step": 27367
+    },
+    {
+      "epoch": 0.6081777777777778,
+      "grad_norm": 1.4471626281738281,
+      "learning_rate": 7.838186263614137e-05,
+      "loss": 1.7771,
+      "step": 27368
+    },
+    {
+      "epoch": 0.6082,
+      "grad_norm": 1.5671677589416504,
+      "learning_rate": 7.837741720382308e-05,
+      "loss": 1.8492,
+      "step": 27369
+    },
+    {
+      "epoch": 0.6082222222222222,
+      "grad_norm": 1.9340373277664185,
+      "learning_rate": 7.837297177150478e-05,
+      "loss": 2.2269,
+      "step": 27370
+    },
+    {
+      "epoch": 0.6082444444444445,
+      "grad_norm": 1.7883458137512207,
+      "learning_rate": 7.836852633918649e-05,
+      "loss": 2.0126,
+      "step": 27371
+    },
+    {
+      "epoch": 0.6082666666666666,
+      "grad_norm": 1.9143633842468262,
+      "learning_rate": 7.83640809068682e-05,
+      "loss": 1.8424,
+      "step": 27372
+    },
+    {
+      "epoch": 0.6082888888888889,
+      "grad_norm": 1.63086998462677,
+      "learning_rate": 7.835963547454991e-05,
+      "loss": 1.5828,
+      "step": 27373
+    },
+    {
+      "epoch": 0.6083111111111111,
+      "grad_norm": 1.1815743446350098,
+      "learning_rate": 7.83551900422316e-05,
+      "loss": 1.0505,
+      "step": 27374
+    },
+    {
+      "epoch": 0.6083333333333333,
+      "grad_norm": 1.5303187370300293,
+      "learning_rate": 7.835074460991331e-05,
+      "loss": 1.6238,
+      "step": 27375
+    },
+    {
+      "epoch": 0.6083555555555555,
+      "grad_norm": 1.7506102323532104,
+      "learning_rate": 7.834629917759502e-05,
+      "loss": 1.8319,
+      "step": 27376
+    },
+    {
+      "epoch": 0.6083777777777778,
+      "grad_norm": 1.247265338897705,
+      "learning_rate": 7.834185374527673e-05,
+      "loss": 0.7261,
+      "step": 27377
+    },
+    {
+      "epoch": 0.6084,
+      "grad_norm": 1.6127877235412598,
+      "learning_rate": 7.833740831295844e-05,
+      "loss": 1.714,
+      "step": 27378
+    },
+    {
+      "epoch": 0.6084222222222222,
+      "grad_norm": 1.9505119323730469,
+      "learning_rate": 7.833296288064015e-05,
+      "loss": 1.7232,
+      "step": 27379
+    },
+    {
+      "epoch": 0.6084444444444445,
+      "grad_norm": 1.985364317893982,
+      "learning_rate": 7.832851744832185e-05,
+      "loss": 2.0324,
+      "step": 27380
+    },
+    {
+      "epoch": 0.6084666666666667,
+      "grad_norm": 1.7504750490188599,
+      "learning_rate": 7.832407201600356e-05,
+      "loss": 1.4993,
+      "step": 27381
+    },
+    {
+      "epoch": 0.6084888888888889,
+      "grad_norm": 1.5885565280914307,
+      "learning_rate": 7.831962658368527e-05,
+      "loss": 1.5342,
+      "step": 27382
+    },
+    {
+      "epoch": 0.6085111111111111,
+      "grad_norm": 1.6928197145462036,
+      "learning_rate": 7.831518115136698e-05,
+      "loss": 1.4337,
+      "step": 27383
+    },
+    {
+      "epoch": 0.6085333333333334,
+      "grad_norm": 1.6309256553649902,
+      "learning_rate": 7.831073571904867e-05,
+      "loss": 2.0033,
+      "step": 27384
+    },
+    {
+      "epoch": 0.6085555555555555,
+      "grad_norm": 1.6423968076705933,
+      "learning_rate": 7.830629028673038e-05,
+      "loss": 1.3243,
+      "step": 27385
+    },
+    {
+      "epoch": 0.6085777777777778,
+      "grad_norm": 1.1410245895385742,
+      "learning_rate": 7.83018448544121e-05,
+      "loss": 0.9618,
+      "step": 27386
+    },
+    {
+      "epoch": 0.6086,
+      "grad_norm": 1.119278073310852,
+      "learning_rate": 7.82973994220938e-05,
+      "loss": 0.7651,
+      "step": 27387
+    },
+    {
+      "epoch": 0.6086222222222222,
+      "grad_norm": 1.5896695852279663,
+      "learning_rate": 7.829295398977551e-05,
+      "loss": 1.9563,
+      "step": 27388
+    },
+    {
+      "epoch": 0.6086444444444444,
+      "grad_norm": 1.5537821054458618,
+      "learning_rate": 7.828850855745722e-05,
+      "loss": 1.7107,
+      "step": 27389
+    },
+    {
+      "epoch": 0.6086666666666667,
+      "grad_norm": 1.7940185070037842,
+      "learning_rate": 7.828406312513893e-05,
+      "loss": 1.8262,
+      "step": 27390
+    },
+    {
+      "epoch": 0.6086888888888888,
+      "grad_norm": 1.3464748859405518,
+      "learning_rate": 7.827961769282063e-05,
+      "loss": 1.406,
+      "step": 27391
+    },
+    {
+      "epoch": 0.6087111111111111,
+      "grad_norm": 1.7492023706436157,
+      "learning_rate": 7.827517226050234e-05,
+      "loss": 1.7909,
+      "step": 27392
+    },
+    {
+      "epoch": 0.6087333333333333,
+      "grad_norm": 1.7525969743728638,
+      "learning_rate": 7.827072682818405e-05,
+      "loss": 1.8682,
+      "step": 27393
+    },
+    {
+      "epoch": 0.6087555555555556,
+      "grad_norm": 1.7507290840148926,
+      "learning_rate": 7.826628139586574e-05,
+      "loss": 1.6276,
+      "step": 27394
+    },
+    {
+      "epoch": 0.6087777777777778,
+      "grad_norm": 1.8546593189239502,
+      "learning_rate": 7.826183596354746e-05,
+      "loss": 1.8663,
+      "step": 27395
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 1.7273077964782715,
+      "learning_rate": 7.825739053122917e-05,
+      "loss": 1.669,
+      "step": 27396
+    },
+    {
+      "epoch": 0.6088222222222223,
+      "grad_norm": 1.8696497678756714,
+      "learning_rate": 7.825294509891087e-05,
+      "loss": 1.6732,
+      "step": 27397
+    },
+    {
+      "epoch": 0.6088444444444444,
+      "grad_norm": 1.6298273801803589,
+      "learning_rate": 7.824849966659258e-05,
+      "loss": 1.4251,
+      "step": 27398
+    },
+    {
+      "epoch": 0.6088666666666667,
+      "grad_norm": 1.6738899946212769,
+      "learning_rate": 7.824405423427429e-05,
+      "loss": 1.6784,
+      "step": 27399
+    },
+    {
+      "epoch": 0.6088888888888889,
+      "grad_norm": 1.6951320171356201,
+      "learning_rate": 7.8239608801956e-05,
+      "loss": 1.3885,
+      "step": 27400
+    },
+    {
+      "epoch": 0.6089111111111111,
+      "grad_norm": 1.6318590641021729,
+      "learning_rate": 7.82351633696377e-05,
+      "loss": 2.4896,
+      "step": 27401
+    },
+    {
+      "epoch": 0.6089333333333333,
+      "grad_norm": 0.9586099982261658,
+      "learning_rate": 7.82307179373194e-05,
+      "loss": 0.0219,
+      "step": 27402
+    },
+    {
+      "epoch": 0.6089555555555556,
+      "grad_norm": 1.6465784311294556,
+      "learning_rate": 7.822627250500111e-05,
+      "loss": 2.6767,
+      "step": 27403
+    },
+    {
+      "epoch": 0.6089777777777777,
+      "grad_norm": 1.5100815296173096,
+      "learning_rate": 7.822182707268282e-05,
+      "loss": 2.0998,
+      "step": 27404
+    },
+    {
+      "epoch": 0.609,
+      "grad_norm": 1.8771125078201294,
+      "learning_rate": 7.821738164036453e-05,
+      "loss": 1.9324,
+      "step": 27405
+    },
+    {
+      "epoch": 0.6090222222222222,
+      "grad_norm": 1.4821887016296387,
+      "learning_rate": 7.821293620804624e-05,
+      "loss": 2.3447,
+      "step": 27406
+    },
+    {
+      "epoch": 0.6090444444444445,
+      "grad_norm": 1.6751375198364258,
+      "learning_rate": 7.820849077572794e-05,
+      "loss": 2.2993,
+      "step": 27407
+    },
+    {
+      "epoch": 0.6090666666666666,
+      "grad_norm": 1.3872592449188232,
+      "learning_rate": 7.820404534340965e-05,
+      "loss": 1.4902,
+      "step": 27408
+    },
+    {
+      "epoch": 0.6090888888888889,
+      "grad_norm": 1.9620487689971924,
+      "learning_rate": 7.819959991109136e-05,
+      "loss": 2.845,
+      "step": 27409
+    },
+    {
+      "epoch": 0.6091111111111112,
+      "grad_norm": 2.082174777984619,
+      "learning_rate": 7.819515447877307e-05,
+      "loss": 2.0938,
+      "step": 27410
+    },
+    {
+      "epoch": 0.6091333333333333,
+      "grad_norm": 1.4058952331542969,
+      "learning_rate": 7.819070904645476e-05,
+      "loss": 2.0692,
+      "step": 27411
+    },
+    {
+      "epoch": 0.6091555555555556,
+      "grad_norm": 1.7376422882080078,
+      "learning_rate": 7.818626361413647e-05,
+      "loss": 2.0961,
+      "step": 27412
+    },
+    {
+      "epoch": 0.6091777777777778,
+      "grad_norm": 1.646878719329834,
+      "learning_rate": 7.818181818181818e-05,
+      "loss": 1.6845,
+      "step": 27413
+    },
+    {
+      "epoch": 0.6092,
+      "grad_norm": 1.623456358909607,
+      "learning_rate": 7.817737274949989e-05,
+      "loss": 2.2751,
+      "step": 27414
+    },
+    {
+      "epoch": 0.6092222222222222,
+      "grad_norm": 1.3822822570800781,
+      "learning_rate": 7.81729273171816e-05,
+      "loss": 1.9871,
+      "step": 27415
+    },
+    {
+      "epoch": 0.6092444444444445,
+      "grad_norm": 1.3027005195617676,
+      "learning_rate": 7.816848188486331e-05,
+      "loss": 1.0903,
+      "step": 27416
+    },
+    {
+      "epoch": 0.6092666666666666,
+      "grad_norm": 1.6102577447891235,
+      "learning_rate": 7.816403645254501e-05,
+      "loss": 2.0865,
+      "step": 27417
+    },
+    {
+      "epoch": 0.6092888888888889,
+      "grad_norm": 1.9533787965774536,
+      "learning_rate": 7.815959102022672e-05,
+      "loss": 1.7968,
+      "step": 27418
+    },
+    {
+      "epoch": 0.6093111111111111,
+      "grad_norm": 1.5359073877334595,
+      "learning_rate": 7.815514558790843e-05,
+      "loss": 2.1051,
+      "step": 27419
+    },
+    {
+      "epoch": 0.6093333333333333,
+      "grad_norm": 1.6860942840576172,
+      "learning_rate": 7.815070015559014e-05,
+      "loss": 2.169,
+      "step": 27420
+    },
+    {
+      "epoch": 0.6093555555555555,
+      "grad_norm": 1.7165453433990479,
+      "learning_rate": 7.814625472327183e-05,
+      "loss": 0.0347,
+      "step": 27421
+    },
+    {
+      "epoch": 0.6093777777777778,
+      "grad_norm": 1.7087652683258057,
+      "learning_rate": 7.814180929095354e-05,
+      "loss": 1.7233,
+      "step": 27422
+    },
+    {
+      "epoch": 0.6094,
+      "grad_norm": 1.8145941495895386,
+      "learning_rate": 7.813736385863527e-05,
+      "loss": 2.054,
+      "step": 27423
+    },
+    {
+      "epoch": 0.6094222222222222,
+      "grad_norm": 1.6878859996795654,
+      "learning_rate": 7.813291842631696e-05,
+      "loss": 2.2226,
+      "step": 27424
+    },
+    {
+      "epoch": 0.6094444444444445,
+      "grad_norm": 1.7352796792984009,
+      "learning_rate": 7.812847299399867e-05,
+      "loss": 1.7145,
+      "step": 27425
+    },
+    {
+      "epoch": 0.6094666666666667,
+      "grad_norm": 1.5544723272323608,
+      "learning_rate": 7.812402756168038e-05,
+      "loss": 1.967,
+      "step": 27426
+    },
+    {
+      "epoch": 0.6094888888888889,
+      "grad_norm": 1.5665216445922852,
+      "learning_rate": 7.811958212936208e-05,
+      "loss": 1.8513,
+      "step": 27427
+    },
+    {
+      "epoch": 0.6095111111111111,
+      "grad_norm": 1.6698379516601562,
+      "learning_rate": 7.811513669704379e-05,
+      "loss": 1.9505,
+      "step": 27428
+    },
+    {
+      "epoch": 0.6095333333333334,
+      "grad_norm": 1.8726260662078857,
+      "learning_rate": 7.81106912647255e-05,
+      "loss": 2.0823,
+      "step": 27429
+    },
+    {
+      "epoch": 0.6095555555555555,
+      "grad_norm": 1.5356457233428955,
+      "learning_rate": 7.81062458324072e-05,
+      "loss": 1.6717,
+      "step": 27430
+    },
+    {
+      "epoch": 0.6095777777777778,
+      "grad_norm": 1.660161018371582,
+      "learning_rate": 7.81018004000889e-05,
+      "loss": 2.0189,
+      "step": 27431
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 1.7959243059158325,
+      "learning_rate": 7.809735496777063e-05,
+      "loss": 2.1529,
+      "step": 27432
+    },
+    {
+      "epoch": 0.6096222222222222,
+      "grad_norm": 1.4932820796966553,
+      "learning_rate": 7.809290953545234e-05,
+      "loss": 1.0458,
+      "step": 27433
+    },
+    {
+      "epoch": 0.6096444444444444,
+      "grad_norm": 1.8300398588180542,
+      "learning_rate": 7.808846410313403e-05,
+      "loss": 1.8499,
+      "step": 27434
+    },
+    {
+      "epoch": 0.6096666666666667,
+      "grad_norm": 1.5831258296966553,
+      "learning_rate": 7.808401867081574e-05,
+      "loss": 1.7654,
+      "step": 27435
+    },
+    {
+      "epoch": 0.6096888888888888,
+      "grad_norm": 0.13884931802749634,
+      "learning_rate": 7.807957323849745e-05,
+      "loss": 0.0269,
+      "step": 27436
+    },
+    {
+      "epoch": 0.6097111111111111,
+      "grad_norm": 1.4457045793533325,
+      "learning_rate": 7.807512780617915e-05,
+      "loss": 1.692,
+      "step": 27437
+    },
+    {
+      "epoch": 0.6097333333333333,
+      "grad_norm": 1.3845202922821045,
+      "learning_rate": 7.807068237386086e-05,
+      "loss": 1.7402,
+      "step": 27438
+    },
+    {
+      "epoch": 0.6097555555555556,
+      "grad_norm": 1.6479185819625854,
+      "learning_rate": 7.806623694154257e-05,
+      "loss": 1.9473,
+      "step": 27439
+    },
+    {
+      "epoch": 0.6097777777777778,
+      "grad_norm": 2.167131185531616,
+      "learning_rate": 7.806179150922428e-05,
+      "loss": 1.6456,
+      "step": 27440
+    },
+    {
+      "epoch": 0.6098,
+      "grad_norm": 1.7285823822021484,
+      "learning_rate": 7.805734607690599e-05,
+      "loss": 1.9564,
+      "step": 27441
+    },
+    {
+      "epoch": 0.6098222222222223,
+      "grad_norm": 1.8743493556976318,
+      "learning_rate": 7.80529006445877e-05,
+      "loss": 1.9648,
+      "step": 27442
+    },
+    {
+      "epoch": 0.6098444444444444,
+      "grad_norm": 1.6637998819351196,
+      "learning_rate": 7.80484552122694e-05,
+      "loss": 1.7091,
+      "step": 27443
+    },
+    {
+      "epoch": 0.6098666666666667,
+      "grad_norm": 1.7053593397140503,
+      "learning_rate": 7.80440097799511e-05,
+      "loss": 1.613,
+      "step": 27444
+    },
+    {
+      "epoch": 0.6098888888888889,
+      "grad_norm": 1.8503854274749756,
+      "learning_rate": 7.803956434763281e-05,
+      "loss": 1.8533,
+      "step": 27445
+    },
+    {
+      "epoch": 0.6099111111111111,
+      "grad_norm": 1.8011876344680786,
+      "learning_rate": 7.803511891531452e-05,
+      "loss": 1.6717,
+      "step": 27446
+    },
+    {
+      "epoch": 0.6099333333333333,
+      "grad_norm": 1.5371845960617065,
+      "learning_rate": 7.803067348299623e-05,
+      "loss": 1.4231,
+      "step": 27447
+    },
+    {
+      "epoch": 0.6099555555555556,
+      "grad_norm": 2.444550037384033,
+      "learning_rate": 7.802622805067793e-05,
+      "loss": 2.5735,
+      "step": 27448
+    },
+    {
+      "epoch": 0.6099777777777777,
+      "grad_norm": 0.2376312017440796,
+      "learning_rate": 7.802178261835964e-05,
+      "loss": 0.0471,
+      "step": 27449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6536933183670044,
+      "learning_rate": 7.801733718604134e-05,
+      "loss": 0.9774,
+      "step": 27450
+    },
+    {
+      "epoch": 0.6100222222222222,
+      "grad_norm": 1.3775696754455566,
+      "learning_rate": 7.801289175372305e-05,
+      "loss": 1.3143,
+      "step": 27451
+    },
+    {
+      "epoch": 0.6100444444444444,
+      "grad_norm": 1.582527995109558,
+      "learning_rate": 7.800844632140476e-05,
+      "loss": 2.2777,
+      "step": 27452
+    },
+    {
+      "epoch": 0.6100666666666666,
+      "grad_norm": 1.4803003072738647,
+      "learning_rate": 7.800400088908647e-05,
+      "loss": 1.9213,
+      "step": 27453
+    },
+    {
+      "epoch": 0.6100888888888889,
+      "grad_norm": 1.5033972263336182,
+      "learning_rate": 7.799955545676817e-05,
+      "loss": 2.2959,
+      "step": 27454
+    },
+    {
+      "epoch": 0.6101111111111112,
+      "grad_norm": 1.5518680810928345,
+      "learning_rate": 7.799511002444988e-05,
+      "loss": 2.2657,
+      "step": 27455
+    },
+    {
+      "epoch": 0.6101333333333333,
+      "grad_norm": 1.8300862312316895,
+      "learning_rate": 7.799066459213159e-05,
+      "loss": 2.2406,
+      "step": 27456
+    },
+    {
+      "epoch": 0.6101555555555556,
+      "grad_norm": 1.669783353805542,
+      "learning_rate": 7.79862191598133e-05,
+      "loss": 2.5407,
+      "step": 27457
+    },
+    {
+      "epoch": 0.6101777777777778,
+      "grad_norm": 1.7596453428268433,
+      "learning_rate": 7.7981773727495e-05,
+      "loss": 2.4514,
+      "step": 27458
+    },
+    {
+      "epoch": 0.6102,
+      "grad_norm": 1.6170973777770996,
+      "learning_rate": 7.79773282951767e-05,
+      "loss": 1.728,
+      "step": 27459
+    },
+    {
+      "epoch": 0.6102222222222222,
+      "grad_norm": 1.6608338356018066,
+      "learning_rate": 7.797288286285843e-05,
+      "loss": 2.3238,
+      "step": 27460
+    },
+    {
+      "epoch": 0.6102444444444445,
+      "grad_norm": 1.5813875198364258,
+      "learning_rate": 7.796843743054012e-05,
+      "loss": 1.9771,
+      "step": 27461
+    },
+    {
+      "epoch": 0.6102666666666666,
+      "grad_norm": 1.5008454322814941,
+      "learning_rate": 7.796399199822183e-05,
+      "loss": 1.8478,
+      "step": 27462
+    },
+    {
+      "epoch": 0.6102888888888889,
+      "grad_norm": 1.8809629678726196,
+      "learning_rate": 7.795954656590354e-05,
+      "loss": 2.0884,
+      "step": 27463
+    },
+    {
+      "epoch": 0.6103111111111111,
+      "grad_norm": 1.7680113315582275,
+      "learning_rate": 7.795510113358524e-05,
+      "loss": 2.0763,
+      "step": 27464
+    },
+    {
+      "epoch": 0.6103333333333333,
+      "grad_norm": 1.5036847591400146,
+      "learning_rate": 7.795065570126695e-05,
+      "loss": 2.0254,
+      "step": 27465
+    },
+    {
+      "epoch": 0.6103555555555555,
+      "grad_norm": 2.1038553714752197,
+      "learning_rate": 7.794621026894866e-05,
+      "loss": 2.4336,
+      "step": 27466
+    },
+    {
+      "epoch": 0.6103777777777778,
+      "grad_norm": 1.8677781820297241,
+      "learning_rate": 7.794176483663037e-05,
+      "loss": 2.2394,
+      "step": 27467
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 1.5848768949508667,
+      "learning_rate": 7.793731940431206e-05,
+      "loss": 2.1178,
+      "step": 27468
+    },
+    {
+      "epoch": 0.6104222222222222,
+      "grad_norm": 1.632435917854309,
+      "learning_rate": 7.793287397199379e-05,
+      "loss": 2.0393,
+      "step": 27469
+    },
+    {
+      "epoch": 0.6104444444444445,
+      "grad_norm": 2.08906626701355,
+      "learning_rate": 7.79284285396755e-05,
+      "loss": 1.511,
+      "step": 27470
+    },
+    {
+      "epoch": 0.6104666666666667,
+      "grad_norm": 1.5188167095184326,
+      "learning_rate": 7.792398310735719e-05,
+      "loss": 1.4935,
+      "step": 27471
+    },
+    {
+      "epoch": 0.6104888888888889,
+      "grad_norm": 1.579158067703247,
+      "learning_rate": 7.79195376750389e-05,
+      "loss": 1.81,
+      "step": 27472
+    },
+    {
+      "epoch": 0.6105111111111111,
+      "grad_norm": 1.6517826318740845,
+      "learning_rate": 7.791509224272061e-05,
+      "loss": 1.4328,
+      "step": 27473
+    },
+    {
+      "epoch": 0.6105333333333334,
+      "grad_norm": 1.608399510383606,
+      "learning_rate": 7.791064681040231e-05,
+      "loss": 1.9586,
+      "step": 27474
+    },
+    {
+      "epoch": 0.6105555555555555,
+      "grad_norm": 1.4978665113449097,
+      "learning_rate": 7.790620137808402e-05,
+      "loss": 1.7627,
+      "step": 27475
+    },
+    {
+      "epoch": 0.6105777777777778,
+      "grad_norm": 1.7245503664016724,
+      "learning_rate": 7.790175594576573e-05,
+      "loss": 2.1325,
+      "step": 27476
+    },
+    {
+      "epoch": 0.6106,
+      "grad_norm": 1.6567022800445557,
+      "learning_rate": 7.789731051344744e-05,
+      "loss": 1.58,
+      "step": 27477
+    },
+    {
+      "epoch": 0.6106222222222222,
+      "grad_norm": 1.8098033666610718,
+      "learning_rate": 7.789286508112915e-05,
+      "loss": 1.4043,
+      "step": 27478
+    },
+    {
+      "epoch": 0.6106444444444444,
+      "grad_norm": 1.8955461978912354,
+      "learning_rate": 7.788841964881086e-05,
+      "loss": 2.1242,
+      "step": 27479
+    },
+    {
+      "epoch": 0.6106666666666667,
+      "grad_norm": 1.7299959659576416,
+      "learning_rate": 7.788397421649257e-05,
+      "loss": 1.9613,
+      "step": 27480
+    },
+    {
+      "epoch": 0.6106888888888888,
+      "grad_norm": 1.5184121131896973,
+      "learning_rate": 7.787952878417426e-05,
+      "loss": 1.9766,
+      "step": 27481
+    },
+    {
+      "epoch": 0.6107111111111111,
+      "grad_norm": 1.6224559545516968,
+      "learning_rate": 7.787508335185597e-05,
+      "loss": 1.9477,
+      "step": 27482
+    },
+    {
+      "epoch": 0.6107333333333334,
+      "grad_norm": 1.9605573415756226,
+      "learning_rate": 7.787063791953768e-05,
+      "loss": 1.6935,
+      "step": 27483
+    },
+    {
+      "epoch": 0.6107555555555556,
+      "grad_norm": 1.8318862915039062,
+      "learning_rate": 7.786619248721938e-05,
+      "loss": 2.1745,
+      "step": 27484
+    },
+    {
+      "epoch": 0.6107777777777778,
+      "grad_norm": 1.8009825944900513,
+      "learning_rate": 7.786174705490109e-05,
+      "loss": 1.6737,
+      "step": 27485
+    },
+    {
+      "epoch": 0.6108,
+      "grad_norm": 1.9569668769836426,
+      "learning_rate": 7.78573016225828e-05,
+      "loss": 1.6701,
+      "step": 27486
+    },
+    {
+      "epoch": 0.6108222222222223,
+      "grad_norm": 1.9783457517623901,
+      "learning_rate": 7.78528561902645e-05,
+      "loss": 2.1003,
+      "step": 27487
+    },
+    {
+      "epoch": 0.6108444444444444,
+      "grad_norm": 1.848065972328186,
+      "learning_rate": 7.784841075794622e-05,
+      "loss": 1.8746,
+      "step": 27488
+    },
+    {
+      "epoch": 0.6108666666666667,
+      "grad_norm": 1.9032918214797974,
+      "learning_rate": 7.784396532562792e-05,
+      "loss": 1.8572,
+      "step": 27489
+    },
+    {
+      "epoch": 0.6108888888888889,
+      "grad_norm": 1.7915080785751343,
+      "learning_rate": 7.783951989330963e-05,
+      "loss": 1.771,
+      "step": 27490
+    },
+    {
+      "epoch": 0.6109111111111111,
+      "grad_norm": 1.5801305770874023,
+      "learning_rate": 7.783507446099133e-05,
+      "loss": 1.8012,
+      "step": 27491
+    },
+    {
+      "epoch": 0.6109333333333333,
+      "grad_norm": 1.9359170198440552,
+      "learning_rate": 7.783062902867304e-05,
+      "loss": 2.2787,
+      "step": 27492
+    },
+    {
+      "epoch": 0.6109555555555556,
+      "grad_norm": 1.6912680864334106,
+      "learning_rate": 7.782618359635475e-05,
+      "loss": 1.6391,
+      "step": 27493
+    },
+    {
+      "epoch": 0.6109777777777777,
+      "grad_norm": 1.9229602813720703,
+      "learning_rate": 7.782173816403646e-05,
+      "loss": 1.5608,
+      "step": 27494
+    },
+    {
+      "epoch": 0.611,
+      "grad_norm": 1.6610921621322632,
+      "learning_rate": 7.781729273171816e-05,
+      "loss": 1.9686,
+      "step": 27495
+    },
+    {
+      "epoch": 0.6110222222222222,
+      "grad_norm": 1.874182939529419,
+      "learning_rate": 7.781284729939987e-05,
+      "loss": 2.2633,
+      "step": 27496
+    },
+    {
+      "epoch": 0.6110444444444444,
+      "grad_norm": 1.746493935585022,
+      "learning_rate": 7.780840186708159e-05,
+      "loss": 1.7766,
+      "step": 27497
+    },
+    {
+      "epoch": 0.6110666666666666,
+      "grad_norm": 1.7898156642913818,
+      "learning_rate": 7.780395643476328e-05,
+      "loss": 1.9386,
+      "step": 27498
+    },
+    {
+      "epoch": 0.6110888888888889,
+      "grad_norm": 1.5614951848983765,
+      "learning_rate": 7.7799511002445e-05,
+      "loss": 1.7405,
+      "step": 27499
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 1.543195366859436,
+      "learning_rate": 7.77950655701267e-05,
+      "loss": 1.1225,
+      "step": 27500
+    },
+    {
+      "epoch": 0.6111333333333333,
+      "grad_norm": 1.5423308610916138,
+      "learning_rate": 7.77906201378084e-05,
+      "loss": 2.1288,
+      "step": 27501
+    },
+    {
+      "epoch": 0.6111555555555556,
+      "grad_norm": 1.339261531829834,
+      "learning_rate": 7.778617470549011e-05,
+      "loss": 2.4123,
+      "step": 27502
+    },
+    {
+      "epoch": 0.6111777777777778,
+      "grad_norm": 1.3870769739151,
+      "learning_rate": 7.778172927317182e-05,
+      "loss": 2.0975,
+      "step": 27503
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 1.0712703466415405,
+      "learning_rate": 7.777728384085353e-05,
+      "loss": 1.2014,
+      "step": 27504
+    },
+    {
+      "epoch": 0.6112222222222222,
+      "grad_norm": 1.5670156478881836,
+      "learning_rate": 7.777283840853522e-05,
+      "loss": 1.9654,
+      "step": 27505
+    },
+    {
+      "epoch": 0.6112444444444445,
+      "grad_norm": 1.445992350578308,
+      "learning_rate": 7.776839297621695e-05,
+      "loss": 2.1042,
+      "step": 27506
+    },
+    {
+      "epoch": 0.6112666666666666,
+      "grad_norm": 1.7465412616729736,
+      "learning_rate": 7.776394754389866e-05,
+      "loss": 2.0741,
+      "step": 27507
+    },
+    {
+      "epoch": 0.6112888888888889,
+      "grad_norm": 1.6744312047958374,
+      "learning_rate": 7.775950211158035e-05,
+      "loss": 2.7767,
+      "step": 27508
+    },
+    {
+      "epoch": 0.6113111111111111,
+      "grad_norm": 1.4504873752593994,
+      "learning_rate": 7.775505667926206e-05,
+      "loss": 2.3511,
+      "step": 27509
+    },
+    {
+      "epoch": 0.6113333333333333,
+      "grad_norm": 1.8150758743286133,
+      "learning_rate": 7.775061124694377e-05,
+      "loss": 2.4931,
+      "step": 27510
+    },
+    {
+      "epoch": 0.6113555555555555,
+      "grad_norm": 1.5724668502807617,
+      "learning_rate": 7.774616581462547e-05,
+      "loss": 2.251,
+      "step": 27511
+    },
+    {
+      "epoch": 0.6113777777777778,
+      "grad_norm": 1.3290016651153564,
+      "learning_rate": 7.774172038230718e-05,
+      "loss": 1.1945,
+      "step": 27512
+    },
+    {
+      "epoch": 0.6114,
+      "grad_norm": 1.7260475158691406,
+      "learning_rate": 7.773727494998889e-05,
+      "loss": 1.8431,
+      "step": 27513
+    },
+    {
+      "epoch": 0.6114222222222222,
+      "grad_norm": 1.6018413305282593,
+      "learning_rate": 7.77328295176706e-05,
+      "loss": 2.1677,
+      "step": 27514
+    },
+    {
+      "epoch": 0.6114444444444445,
+      "grad_norm": 1.3727264404296875,
+      "learning_rate": 7.772838408535231e-05,
+      "loss": 1.2625,
+      "step": 27515
+    },
+    {
+      "epoch": 0.6114666666666667,
+      "grad_norm": 1.6395866870880127,
+      "learning_rate": 7.772393865303402e-05,
+      "loss": 2.2097,
+      "step": 27516
+    },
+    {
+      "epoch": 0.6114888888888889,
+      "grad_norm": 1.7023524045944214,
+      "learning_rate": 7.771949322071573e-05,
+      "loss": 2.1454,
+      "step": 27517
+    },
+    {
+      "epoch": 0.6115111111111111,
+      "grad_norm": 1.579649567604065,
+      "learning_rate": 7.771504778839742e-05,
+      "loss": 1.9694,
+      "step": 27518
+    },
+    {
+      "epoch": 0.6115333333333334,
+      "grad_norm": 1.4795308113098145,
+      "learning_rate": 7.771060235607913e-05,
+      "loss": 2.1181,
+      "step": 27519
+    },
+    {
+      "epoch": 0.6115555555555555,
+      "grad_norm": 1.7093156576156616,
+      "learning_rate": 7.770615692376084e-05,
+      "loss": 2.3267,
+      "step": 27520
+    },
+    {
+      "epoch": 0.6115777777777778,
+      "grad_norm": 1.8103622198104858,
+      "learning_rate": 7.770171149144254e-05,
+      "loss": 1.9654,
+      "step": 27521
+    },
+    {
+      "epoch": 0.6116,
+      "grad_norm": 1.6107460260391235,
+      "learning_rate": 7.769726605912425e-05,
+      "loss": 1.3434,
+      "step": 27522
+    },
+    {
+      "epoch": 0.6116222222222222,
+      "grad_norm": 1.6847106218338013,
+      "learning_rate": 7.769282062680596e-05,
+      "loss": 2.5195,
+      "step": 27523
+    },
+    {
+      "epoch": 0.6116444444444444,
+      "grad_norm": 1.7250392436981201,
+      "learning_rate": 7.768837519448767e-05,
+      "loss": 1.5947,
+      "step": 27524
+    },
+    {
+      "epoch": 0.6116666666666667,
+      "grad_norm": 1.9823029041290283,
+      "learning_rate": 7.768392976216938e-05,
+      "loss": 2.2627,
+      "step": 27525
+    },
+    {
+      "epoch": 0.6116888888888888,
+      "grad_norm": 1.8412489891052246,
+      "learning_rate": 7.767948432985109e-05,
+      "loss": 2.0382,
+      "step": 27526
+    },
+    {
+      "epoch": 0.6117111111111111,
+      "grad_norm": 1.7352039813995361,
+      "learning_rate": 7.76750388975328e-05,
+      "loss": 2.2732,
+      "step": 27527
+    },
+    {
+      "epoch": 0.6117333333333334,
+      "grad_norm": 1.6988086700439453,
+      "learning_rate": 7.767059346521449e-05,
+      "loss": 2.2386,
+      "step": 27528
+    },
+    {
+      "epoch": 0.6117555555555556,
+      "grad_norm": 1.6477222442626953,
+      "learning_rate": 7.76661480328962e-05,
+      "loss": 1.4523,
+      "step": 27529
+    },
+    {
+      "epoch": 0.6117777777777778,
+      "grad_norm": 2.1043894290924072,
+      "learning_rate": 7.766170260057791e-05,
+      "loss": 2.5293,
+      "step": 27530
+    },
+    {
+      "epoch": 0.6118,
+      "grad_norm": 1.773444652557373,
+      "learning_rate": 7.765725716825961e-05,
+      "loss": 2.1363,
+      "step": 27531
+    },
+    {
+      "epoch": 0.6118222222222223,
+      "grad_norm": 1.4800810813903809,
+      "learning_rate": 7.765281173594132e-05,
+      "loss": 1.8682,
+      "step": 27532
+    },
+    {
+      "epoch": 0.6118444444444444,
+      "grad_norm": 1.6080175638198853,
+      "learning_rate": 7.764836630362303e-05,
+      "loss": 1.6381,
+      "step": 27533
+    },
+    {
+      "epoch": 0.6118666666666667,
+      "grad_norm": 2.55741548538208,
+      "learning_rate": 7.764392087130475e-05,
+      "loss": 1.8665,
+      "step": 27534
+    },
+    {
+      "epoch": 0.6118888888888889,
+      "grad_norm": 1.5693540573120117,
+      "learning_rate": 7.763947543898645e-05,
+      "loss": 1.8233,
+      "step": 27535
+    },
+    {
+      "epoch": 0.6119111111111111,
+      "grad_norm": 1.696226954460144,
+      "learning_rate": 7.763503000666816e-05,
+      "loss": 1.6646,
+      "step": 27536
+    },
+    {
+      "epoch": 0.6119333333333333,
+      "grad_norm": 2.2691214084625244,
+      "learning_rate": 7.763058457434986e-05,
+      "loss": 2.2198,
+      "step": 27537
+    },
+    {
+      "epoch": 0.6119555555555556,
+      "grad_norm": 1.8994804620742798,
+      "learning_rate": 7.762613914203156e-05,
+      "loss": 2.2349,
+      "step": 27538
+    },
+    {
+      "epoch": 0.6119777777777777,
+      "grad_norm": 1.870898962020874,
+      "learning_rate": 7.762169370971327e-05,
+      "loss": 1.89,
+      "step": 27539
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 1.6274160146713257,
+      "learning_rate": 7.761724827739498e-05,
+      "loss": 1.7772,
+      "step": 27540
+    },
+    {
+      "epoch": 0.6120222222222222,
+      "grad_norm": 1.788335919380188,
+      "learning_rate": 7.761280284507668e-05,
+      "loss": 1.9265,
+      "step": 27541
+    },
+    {
+      "epoch": 0.6120444444444444,
+      "grad_norm": 1.7671217918395996,
+      "learning_rate": 7.760835741275839e-05,
+      "loss": 1.5955,
+      "step": 27542
+    },
+    {
+      "epoch": 0.6120666666666666,
+      "grad_norm": 1.9399702548980713,
+      "learning_rate": 7.760391198044011e-05,
+      "loss": 1.886,
+      "step": 27543
+    },
+    {
+      "epoch": 0.6120888888888889,
+      "grad_norm": 1.8233189582824707,
+      "learning_rate": 7.759946654812182e-05,
+      "loss": 1.7791,
+      "step": 27544
+    },
+    {
+      "epoch": 0.6121111111111112,
+      "grad_norm": 1.925369143486023,
+      "learning_rate": 7.759502111580351e-05,
+      "loss": 1.9089,
+      "step": 27545
+    },
+    {
+      "epoch": 0.6121333333333333,
+      "grad_norm": 1.8397341966629028,
+      "learning_rate": 7.759057568348522e-05,
+      "loss": 1.9789,
+      "step": 27546
+    },
+    {
+      "epoch": 0.6121555555555556,
+      "grad_norm": 1.8775615692138672,
+      "learning_rate": 7.758613025116693e-05,
+      "loss": 1.6122,
+      "step": 27547
+    },
+    {
+      "epoch": 0.6121777777777778,
+      "grad_norm": 2.0925540924072266,
+      "learning_rate": 7.758168481884863e-05,
+      "loss": 2.0339,
+      "step": 27548
+    },
+    {
+      "epoch": 0.6122,
+      "grad_norm": 2.579909086227417,
+      "learning_rate": 7.757723938653034e-05,
+      "loss": 1.9227,
+      "step": 27549
+    },
+    {
+      "epoch": 0.6122222222222222,
+      "grad_norm": 1.542546033859253,
+      "learning_rate": 7.757279395421205e-05,
+      "loss": 1.2056,
+      "step": 27550
+    },
+    {
+      "epoch": 0.6122444444444445,
+      "grad_norm": 1.1781761646270752,
+      "learning_rate": 7.756834852189376e-05,
+      "loss": 1.4263,
+      "step": 27551
+    },
+    {
+      "epoch": 0.6122666666666666,
+      "grad_norm": 1.4291971921920776,
+      "learning_rate": 7.756390308957547e-05,
+      "loss": 2.5189,
+      "step": 27552
+    },
+    {
+      "epoch": 0.6122888888888889,
+      "grad_norm": 1.0243903398513794,
+      "learning_rate": 7.755945765725718e-05,
+      "loss": 1.1948,
+      "step": 27553
+    },
+    {
+      "epoch": 0.6123111111111111,
+      "grad_norm": 1.0803135633468628,
+      "learning_rate": 7.755501222493889e-05,
+      "loss": 1.1821,
+      "step": 27554
+    },
+    {
+      "epoch": 0.6123333333333333,
+      "grad_norm": 1.4516526460647583,
+      "learning_rate": 7.755056679262058e-05,
+      "loss": 2.0488,
+      "step": 27555
+    },
+    {
+      "epoch": 0.6123555555555555,
+      "grad_norm": 1.5359574556350708,
+      "learning_rate": 7.754612136030229e-05,
+      "loss": 2.3512,
+      "step": 27556
+    },
+    {
+      "epoch": 0.6123777777777778,
+      "grad_norm": 1.5224683284759521,
+      "learning_rate": 7.7541675927984e-05,
+      "loss": 2.1224,
+      "step": 27557
+    },
+    {
+      "epoch": 0.6124,
+      "grad_norm": 1.4551442861557007,
+      "learning_rate": 7.75372304956657e-05,
+      "loss": 1.8269,
+      "step": 27558
+    },
+    {
+      "epoch": 0.6124222222222222,
+      "grad_norm": 1.4387001991271973,
+      "learning_rate": 7.753278506334741e-05,
+      "loss": 2.0894,
+      "step": 27559
+    },
+    {
+      "epoch": 0.6124444444444445,
+      "grad_norm": 2.2286064624786377,
+      "learning_rate": 7.752833963102912e-05,
+      "loss": 2.4909,
+      "step": 27560
+    },
+    {
+      "epoch": 0.6124666666666667,
+      "grad_norm": 1.8037059307098389,
+      "learning_rate": 7.752389419871083e-05,
+      "loss": 2.1049,
+      "step": 27561
+    },
+    {
+      "epoch": 0.6124888888888889,
+      "grad_norm": 1.8678205013275146,
+      "learning_rate": 7.751944876639254e-05,
+      "loss": 2.0448,
+      "step": 27562
+    },
+    {
+      "epoch": 0.6125111111111111,
+      "grad_norm": 1.6603857278823853,
+      "learning_rate": 7.751500333407425e-05,
+      "loss": 1.2267,
+      "step": 27563
+    },
+    {
+      "epoch": 0.6125333333333334,
+      "grad_norm": 1.5285018682479858,
+      "learning_rate": 7.751055790175596e-05,
+      "loss": 2.072,
+      "step": 27564
+    },
+    {
+      "epoch": 0.6125555555555555,
+      "grad_norm": 1.6152604818344116,
+      "learning_rate": 7.750611246943765e-05,
+      "loss": 2.062,
+      "step": 27565
+    },
+    {
+      "epoch": 0.6125777777777778,
+      "grad_norm": 1.51948082447052,
+      "learning_rate": 7.750166703711936e-05,
+      "loss": 1.5071,
+      "step": 27566
+    },
+    {
+      "epoch": 0.6126,
+      "grad_norm": 1.1879931688308716,
+      "learning_rate": 7.749722160480107e-05,
+      "loss": 1.1506,
+      "step": 27567
+    },
+    {
+      "epoch": 0.6126222222222222,
+      "grad_norm": 1.9524271488189697,
+      "learning_rate": 7.749277617248277e-05,
+      "loss": 2.4532,
+      "step": 27568
+    },
+    {
+      "epoch": 0.6126444444444444,
+      "grad_norm": 1.6999812126159668,
+      "learning_rate": 7.748833074016448e-05,
+      "loss": 1.6778,
+      "step": 27569
+    },
+    {
+      "epoch": 0.6126666666666667,
+      "grad_norm": 1.7853630781173706,
+      "learning_rate": 7.748388530784619e-05,
+      "loss": 2.3016,
+      "step": 27570
+    },
+    {
+      "epoch": 0.6126888888888888,
+      "grad_norm": 1.782258152961731,
+      "learning_rate": 7.747943987552791e-05,
+      "loss": 1.65,
+      "step": 27571
+    },
+    {
+      "epoch": 0.6127111111111111,
+      "grad_norm": 1.5772422552108765,
+      "learning_rate": 7.74749944432096e-05,
+      "loss": 2.018,
+      "step": 27572
+    },
+    {
+      "epoch": 0.6127333333333334,
+      "grad_norm": 2.0711746215820312,
+      "learning_rate": 7.747054901089132e-05,
+      "loss": 2.2871,
+      "step": 27573
+    },
+    {
+      "epoch": 0.6127555555555556,
+      "grad_norm": 1.7900344133377075,
+      "learning_rate": 7.746610357857303e-05,
+      "loss": 1.9562,
+      "step": 27574
+    },
+    {
+      "epoch": 0.6127777777777778,
+      "grad_norm": 1.6933283805847168,
+      "learning_rate": 7.746165814625472e-05,
+      "loss": 1.9733,
+      "step": 27575
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 1.7561581134796143,
+      "learning_rate": 7.745721271393643e-05,
+      "loss": 2.1765,
+      "step": 27576
+    },
+    {
+      "epoch": 0.6128222222222223,
+      "grad_norm": 1.5848567485809326,
+      "learning_rate": 7.745276728161814e-05,
+      "loss": 1.8434,
+      "step": 27577
+    },
+    {
+      "epoch": 0.6128444444444444,
+      "grad_norm": 1.4822124242782593,
+      "learning_rate": 7.744832184929984e-05,
+      "loss": 2.0103,
+      "step": 27578
+    },
+    {
+      "epoch": 0.6128666666666667,
+      "grad_norm": 1.7740105390548706,
+      "learning_rate": 7.744387641698155e-05,
+      "loss": 2.0244,
+      "step": 27579
+    },
+    {
+      "epoch": 0.6128888888888889,
+      "grad_norm": 1.7334976196289062,
+      "learning_rate": 7.743943098466327e-05,
+      "loss": 1.9153,
+      "step": 27580
+    },
+    {
+      "epoch": 0.6129111111111111,
+      "grad_norm": 1.635332465171814,
+      "learning_rate": 7.743498555234498e-05,
+      "loss": 1.9374,
+      "step": 27581
+    },
+    {
+      "epoch": 0.6129333333333333,
+      "grad_norm": 1.4835823774337769,
+      "learning_rate": 7.743054012002668e-05,
+      "loss": 2.0732,
+      "step": 27582
+    },
+    {
+      "epoch": 0.6129555555555556,
+      "grad_norm": 1.4527803659439087,
+      "learning_rate": 7.742609468770839e-05,
+      "loss": 1.6363,
+      "step": 27583
+    },
+    {
+      "epoch": 0.6129777777777777,
+      "grad_norm": 1.216811180114746,
+      "learning_rate": 7.74216492553901e-05,
+      "loss": 0.8762,
+      "step": 27584
+    },
+    {
+      "epoch": 0.613,
+      "grad_norm": 1.8603177070617676,
+      "learning_rate": 7.741720382307179e-05,
+      "loss": 2.3087,
+      "step": 27585
+    },
+    {
+      "epoch": 0.6130222222222222,
+      "grad_norm": 1.8066385984420776,
+      "learning_rate": 7.74127583907535e-05,
+      "loss": 1.8244,
+      "step": 27586
+    },
+    {
+      "epoch": 0.6130444444444444,
+      "grad_norm": 1.5782830715179443,
+      "learning_rate": 7.740831295843521e-05,
+      "loss": 1.5668,
+      "step": 27587
+    },
+    {
+      "epoch": 0.6130666666666666,
+      "grad_norm": 1.8840564489364624,
+      "learning_rate": 7.740386752611692e-05,
+      "loss": 1.7731,
+      "step": 27588
+    },
+    {
+      "epoch": 0.6130888888888889,
+      "grad_norm": 2.1401329040527344,
+      "learning_rate": 7.739942209379863e-05,
+      "loss": 2.0632,
+      "step": 27589
+    },
+    {
+      "epoch": 0.6131111111111112,
+      "grad_norm": 1.8816437721252441,
+      "learning_rate": 7.739497666148034e-05,
+      "loss": 1.6173,
+      "step": 27590
+    },
+    {
+      "epoch": 0.6131333333333333,
+      "grad_norm": 2.0451958179473877,
+      "learning_rate": 7.739053122916205e-05,
+      "loss": 2.0848,
+      "step": 27591
+    },
+    {
+      "epoch": 0.6131555555555556,
+      "grad_norm": 1.6593228578567505,
+      "learning_rate": 7.738608579684374e-05,
+      "loss": 1.7136,
+      "step": 27592
+    },
+    {
+      "epoch": 0.6131777777777778,
+      "grad_norm": 1.5360110998153687,
+      "learning_rate": 7.738164036452545e-05,
+      "loss": 1.747,
+      "step": 27593
+    },
+    {
+      "epoch": 0.6132,
+      "grad_norm": 2.1586971282958984,
+      "learning_rate": 7.737719493220716e-05,
+      "loss": 1.9396,
+      "step": 27594
+    },
+    {
+      "epoch": 0.6132222222222222,
+      "grad_norm": 1.9175387620925903,
+      "learning_rate": 7.737274949988886e-05,
+      "loss": 1.7093,
+      "step": 27595
+    },
+    {
+      "epoch": 0.6132444444444445,
+      "grad_norm": 1.982151985168457,
+      "learning_rate": 7.736830406757057e-05,
+      "loss": 2.1863,
+      "step": 27596
+    },
+    {
+      "epoch": 0.6132666666666666,
+      "grad_norm": 3.3169479370117188,
+      "learning_rate": 7.736385863525228e-05,
+      "loss": 1.985,
+      "step": 27597
+    },
+    {
+      "epoch": 0.6132888888888889,
+      "grad_norm": 1.7234865427017212,
+      "learning_rate": 7.735941320293399e-05,
+      "loss": 1.9525,
+      "step": 27598
+    },
+    {
+      "epoch": 0.6133111111111111,
+      "grad_norm": 1.8913743495941162,
+      "learning_rate": 7.73549677706157e-05,
+      "loss": 2.1799,
+      "step": 27599
+    },
+    {
+      "epoch": 0.6133333333333333,
+      "grad_norm": 1.910362720489502,
+      "learning_rate": 7.735052233829741e-05,
+      "loss": 1.761,
+      "step": 27600
+    },
+    {
+      "epoch": 0.6133555555555555,
+      "grad_norm": 1.485051155090332,
+      "learning_rate": 7.734607690597912e-05,
+      "loss": 2.3293,
+      "step": 27601
+    },
+    {
+      "epoch": 0.6133777777777778,
+      "grad_norm": 1.4898614883422852,
+      "learning_rate": 7.734163147366081e-05,
+      "loss": 2.4178,
+      "step": 27602
+    },
+    {
+      "epoch": 0.6134,
+      "grad_norm": 1.735337734222412,
+      "learning_rate": 7.733718604134252e-05,
+      "loss": 2.3771,
+      "step": 27603
+    },
+    {
+      "epoch": 0.6134222222222222,
+      "grad_norm": 0.31236639618873596,
+      "learning_rate": 7.733274060902423e-05,
+      "loss": 0.0177,
+      "step": 27604
+    },
+    {
+      "epoch": 0.6134444444444445,
+      "grad_norm": 1.4100191593170166,
+      "learning_rate": 7.732829517670593e-05,
+      "loss": 2.1479,
+      "step": 27605
+    },
+    {
+      "epoch": 0.6134666666666667,
+      "grad_norm": 1.3722394704818726,
+      "learning_rate": 7.732384974438764e-05,
+      "loss": 2.2028,
+      "step": 27606
+    },
+    {
+      "epoch": 0.6134888888888889,
+      "grad_norm": 1.4756319522857666,
+      "learning_rate": 7.731940431206935e-05,
+      "loss": 2.1144,
+      "step": 27607
+    },
+    {
+      "epoch": 0.6135111111111111,
+      "grad_norm": 1.1815156936645508,
+      "learning_rate": 7.731495887975107e-05,
+      "loss": 1.5594,
+      "step": 27608
+    },
+    {
+      "epoch": 0.6135333333333334,
+      "grad_norm": 1.2476978302001953,
+      "learning_rate": 7.731051344743277e-05,
+      "loss": 1.2661,
+      "step": 27609
+    },
+    {
+      "epoch": 0.6135555555555555,
+      "grad_norm": 1.892336368560791,
+      "learning_rate": 7.730606801511448e-05,
+      "loss": 2.3488,
+      "step": 27610
+    },
+    {
+      "epoch": 0.6135777777777778,
+      "grad_norm": 1.6618032455444336,
+      "learning_rate": 7.730162258279619e-05,
+      "loss": 1.6773,
+      "step": 27611
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 1.2995035648345947,
+      "learning_rate": 7.729717715047788e-05,
+      "loss": 1.4888,
+      "step": 27612
+    },
+    {
+      "epoch": 0.6136222222222222,
+      "grad_norm": 1.5355890989303589,
+      "learning_rate": 7.729273171815959e-05,
+      "loss": 1.7641,
+      "step": 27613
+    },
+    {
+      "epoch": 0.6136444444444444,
+      "grad_norm": 1.7926889657974243,
+      "learning_rate": 7.72882862858413e-05,
+      "loss": 1.903,
+      "step": 27614
+    },
+    {
+      "epoch": 0.6136666666666667,
+      "grad_norm": 1.98298978805542,
+      "learning_rate": 7.7283840853523e-05,
+      "loss": 2.0607,
+      "step": 27615
+    },
+    {
+      "epoch": 0.6136888888888888,
+      "grad_norm": 1.5406136512756348,
+      "learning_rate": 7.727939542120471e-05,
+      "loss": 1.96,
+      "step": 27616
+    },
+    {
+      "epoch": 0.6137111111111111,
+      "grad_norm": 1.7651652097702026,
+      "learning_rate": 7.727494998888643e-05,
+      "loss": 1.8261,
+      "step": 27617
+    },
+    {
+      "epoch": 0.6137333333333334,
+      "grad_norm": 1.607982873916626,
+      "learning_rate": 7.727050455656814e-05,
+      "loss": 1.7859,
+      "step": 27618
+    },
+    {
+      "epoch": 0.6137555555555556,
+      "grad_norm": 1.7538888454437256,
+      "learning_rate": 7.726605912424984e-05,
+      "loss": 2.2275,
+      "step": 27619
+    },
+    {
+      "epoch": 0.6137777777777778,
+      "grad_norm": 2.276658773422241,
+      "learning_rate": 7.726161369193155e-05,
+      "loss": 2.1569,
+      "step": 27620
+    },
+    {
+      "epoch": 0.6138,
+      "grad_norm": 1.7184829711914062,
+      "learning_rate": 7.725716825961326e-05,
+      "loss": 2.0479,
+      "step": 27621
+    },
+    {
+      "epoch": 0.6138222222222223,
+      "grad_norm": 1.8018218278884888,
+      "learning_rate": 7.725272282729495e-05,
+      "loss": 1.9534,
+      "step": 27622
+    },
+    {
+      "epoch": 0.6138444444444444,
+      "grad_norm": 1.7157447338104248,
+      "learning_rate": 7.724827739497666e-05,
+      "loss": 1.936,
+      "step": 27623
+    },
+    {
+      "epoch": 0.6138666666666667,
+      "grad_norm": 1.4542107582092285,
+      "learning_rate": 7.724383196265837e-05,
+      "loss": 1.5425,
+      "step": 27624
+    },
+    {
+      "epoch": 0.6138888888888889,
+      "grad_norm": 2.096475124359131,
+      "learning_rate": 7.723938653034008e-05,
+      "loss": 2.4732,
+      "step": 27625
+    },
+    {
+      "epoch": 0.6139111111111111,
+      "grad_norm": 1.5585896968841553,
+      "learning_rate": 7.723494109802179e-05,
+      "loss": 2.1209,
+      "step": 27626
+    },
+    {
+      "epoch": 0.6139333333333333,
+      "grad_norm": 2.434840679168701,
+      "learning_rate": 7.72304956657035e-05,
+      "loss": 2.603,
+      "step": 27627
+    },
+    {
+      "epoch": 0.6139555555555556,
+      "grad_norm": 1.888709545135498,
+      "learning_rate": 7.722605023338521e-05,
+      "loss": 2.3049,
+      "step": 27628
+    },
+    {
+      "epoch": 0.6139777777777777,
+      "grad_norm": 1.717698335647583,
+      "learning_rate": 7.72216048010669e-05,
+      "loss": 2.0233,
+      "step": 27629
+    },
+    {
+      "epoch": 0.614,
+      "grad_norm": 1.5587272644042969,
+      "learning_rate": 7.721715936874862e-05,
+      "loss": 1.8363,
+      "step": 27630
+    },
+    {
+      "epoch": 0.6140222222222222,
+      "grad_norm": 1.546492099761963,
+      "learning_rate": 7.721271393643032e-05,
+      "loss": 2.0981,
+      "step": 27631
+    },
+    {
+      "epoch": 0.6140444444444444,
+      "grad_norm": 1.110864520072937,
+      "learning_rate": 7.720826850411202e-05,
+      "loss": 1.0081,
+      "step": 27632
+    },
+    {
+      "epoch": 0.6140666666666666,
+      "grad_norm": 1.1011080741882324,
+      "learning_rate": 7.720382307179373e-05,
+      "loss": 0.9227,
+      "step": 27633
+    },
+    {
+      "epoch": 0.6140888888888889,
+      "grad_norm": 1.4774861335754395,
+      "learning_rate": 7.719937763947544e-05,
+      "loss": 1.7885,
+      "step": 27634
+    },
+    {
+      "epoch": 0.6141111111111112,
+      "grad_norm": 1.8963218927383423,
+      "learning_rate": 7.719493220715715e-05,
+      "loss": 1.8461,
+      "step": 27635
+    },
+    {
+      "epoch": 0.6141333333333333,
+      "grad_norm": 1.8257931470870972,
+      "learning_rate": 7.719048677483886e-05,
+      "loss": 1.7892,
+      "step": 27636
+    },
+    {
+      "epoch": 0.6141555555555556,
+      "grad_norm": 1.488949179649353,
+      "learning_rate": 7.718604134252057e-05,
+      "loss": 1.2848,
+      "step": 27637
+    },
+    {
+      "epoch": 0.6141777777777778,
+      "grad_norm": 1.8301564455032349,
+      "learning_rate": 7.718159591020228e-05,
+      "loss": 2.1069,
+      "step": 27638
+    },
+    {
+      "epoch": 0.6142,
+      "grad_norm": 1.6307073831558228,
+      "learning_rate": 7.717715047788397e-05,
+      "loss": 1.7145,
+      "step": 27639
+    },
+    {
+      "epoch": 0.6142222222222222,
+      "grad_norm": 1.733677864074707,
+      "learning_rate": 7.717270504556568e-05,
+      "loss": 2.2545,
+      "step": 27640
+    },
+    {
+      "epoch": 0.6142444444444445,
+      "grad_norm": 1.5401047468185425,
+      "learning_rate": 7.71682596132474e-05,
+      "loss": 1.225,
+      "step": 27641
+    },
+    {
+      "epoch": 0.6142666666666666,
+      "grad_norm": 0.9527256488800049,
+      "learning_rate": 7.716381418092909e-05,
+      "loss": 0.762,
+      "step": 27642
+    },
+    {
+      "epoch": 0.6142888888888889,
+      "grad_norm": 1.9015424251556396,
+      "learning_rate": 7.71593687486108e-05,
+      "loss": 1.8785,
+      "step": 27643
+    },
+    {
+      "epoch": 0.6143111111111111,
+      "grad_norm": 1.754401445388794,
+      "learning_rate": 7.715492331629251e-05,
+      "loss": 2.1387,
+      "step": 27644
+    },
+    {
+      "epoch": 0.6143333333333333,
+      "grad_norm": 1.8856477737426758,
+      "learning_rate": 7.715047788397422e-05,
+      "loss": 1.8633,
+      "step": 27645
+    },
+    {
+      "epoch": 0.6143555555555555,
+      "grad_norm": 1.6553912162780762,
+      "learning_rate": 7.714603245165593e-05,
+      "loss": 1.6771,
+      "step": 27646
+    },
+    {
+      "epoch": 0.6143777777777778,
+      "grad_norm": 1.5054494142532349,
+      "learning_rate": 7.714158701933764e-05,
+      "loss": 1.3938,
+      "step": 27647
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.715761661529541,
+      "learning_rate": 7.713714158701935e-05,
+      "loss": 2.0282,
+      "step": 27648
+    },
+    {
+      "epoch": 0.6144222222222222,
+      "grad_norm": 1.7670443058013916,
+      "learning_rate": 7.713269615470104e-05,
+      "loss": 0.9393,
+      "step": 27649
+    },
+    {
+      "epoch": 0.6144444444444445,
+      "grad_norm": 1.943627119064331,
+      "learning_rate": 7.712825072238275e-05,
+      "loss": 1.3714,
+      "step": 27650
+    },
+    {
+      "epoch": 0.6144666666666667,
+      "grad_norm": 1.4573408365249634,
+      "learning_rate": 7.712380529006446e-05,
+      "loss": 2.2397,
+      "step": 27651
+    },
+    {
+      "epoch": 0.6144888888888889,
+      "grad_norm": 1.8467220067977905,
+      "learning_rate": 7.711935985774616e-05,
+      "loss": 2.8679,
+      "step": 27652
+    },
+    {
+      "epoch": 0.6145111111111111,
+      "grad_norm": 0.8519179224967957,
+      "learning_rate": 7.711491442542787e-05,
+      "loss": 0.9065,
+      "step": 27653
+    },
+    {
+      "epoch": 0.6145333333333334,
+      "grad_norm": 1.5563408136367798,
+      "learning_rate": 7.711046899310959e-05,
+      "loss": 2.1656,
+      "step": 27654
+    },
+    {
+      "epoch": 0.6145555555555555,
+      "grad_norm": 2.1301660537719727,
+      "learning_rate": 7.71060235607913e-05,
+      "loss": 2.4985,
+      "step": 27655
+    },
+    {
+      "epoch": 0.6145777777777778,
+      "grad_norm": 1.8298200368881226,
+      "learning_rate": 7.7101578128473e-05,
+      "loss": 2.1938,
+      "step": 27656
+    },
+    {
+      "epoch": 0.6146,
+      "grad_norm": 1.7975974082946777,
+      "learning_rate": 7.709713269615471e-05,
+      "loss": 2.0496,
+      "step": 27657
+    },
+    {
+      "epoch": 0.6146222222222222,
+      "grad_norm": 1.5617163181304932,
+      "learning_rate": 7.709268726383642e-05,
+      "loss": 2.2042,
+      "step": 27658
+    },
+    {
+      "epoch": 0.6146444444444444,
+      "grad_norm": 1.6176177263259888,
+      "learning_rate": 7.708824183151811e-05,
+      "loss": 1.6636,
+      "step": 27659
+    },
+    {
+      "epoch": 0.6146666666666667,
+      "grad_norm": 1.4682356119155884,
+      "learning_rate": 7.708379639919982e-05,
+      "loss": 1.7942,
+      "step": 27660
+    },
+    {
+      "epoch": 0.6146888888888888,
+      "grad_norm": 1.9298056364059448,
+      "learning_rate": 7.707935096688153e-05,
+      "loss": 2.0287,
+      "step": 27661
+    },
+    {
+      "epoch": 0.6147111111111111,
+      "grad_norm": 1.5962305068969727,
+      "learning_rate": 7.707490553456324e-05,
+      "loss": 1.9764,
+      "step": 27662
+    },
+    {
+      "epoch": 0.6147333333333334,
+      "grad_norm": 1.8374494314193726,
+      "learning_rate": 7.707046010224495e-05,
+      "loss": 2.0601,
+      "step": 27663
+    },
+    {
+      "epoch": 0.6147555555555556,
+      "grad_norm": 1.7612316608428955,
+      "learning_rate": 7.706601466992666e-05,
+      "loss": 2.5086,
+      "step": 27664
+    },
+    {
+      "epoch": 0.6147777777777778,
+      "grad_norm": 1.8300474882125854,
+      "learning_rate": 7.706156923760837e-05,
+      "loss": 1.891,
+      "step": 27665
+    },
+    {
+      "epoch": 0.6148,
+      "grad_norm": 1.481309175491333,
+      "learning_rate": 7.705712380529007e-05,
+      "loss": 1.8192,
+      "step": 27666
+    },
+    {
+      "epoch": 0.6148222222222223,
+      "grad_norm": 1.5958552360534668,
+      "learning_rate": 7.705267837297178e-05,
+      "loss": 2.1367,
+      "step": 27667
+    },
+    {
+      "epoch": 0.6148444444444444,
+      "grad_norm": 1.4381794929504395,
+      "learning_rate": 7.704823294065349e-05,
+      "loss": 1.667,
+      "step": 27668
+    },
+    {
+      "epoch": 0.6148666666666667,
+      "grad_norm": 1.7751753330230713,
+      "learning_rate": 7.704378750833518e-05,
+      "loss": 2.3798,
+      "step": 27669
+    },
+    {
+      "epoch": 0.6148888888888889,
+      "grad_norm": 1.361076831817627,
+      "learning_rate": 7.703934207601689e-05,
+      "loss": 1.3767,
+      "step": 27670
+    },
+    {
+      "epoch": 0.6149111111111111,
+      "grad_norm": 1.8791605234146118,
+      "learning_rate": 7.70348966436986e-05,
+      "loss": 2.2393,
+      "step": 27671
+    },
+    {
+      "epoch": 0.6149333333333333,
+      "grad_norm": 1.0605690479278564,
+      "learning_rate": 7.703045121138031e-05,
+      "loss": 0.9758,
+      "step": 27672
+    },
+    {
+      "epoch": 0.6149555555555556,
+      "grad_norm": 1.5713154077529907,
+      "learning_rate": 7.702600577906202e-05,
+      "loss": 1.5473,
+      "step": 27673
+    },
+    {
+      "epoch": 0.6149777777777777,
+      "grad_norm": 1.6930948495864868,
+      "learning_rate": 7.702156034674373e-05,
+      "loss": 2.4817,
+      "step": 27674
+    },
+    {
+      "epoch": 0.615,
+      "grad_norm": 1.694022297859192,
+      "learning_rate": 7.701711491442544e-05,
+      "loss": 2.0426,
+      "step": 27675
+    },
+    {
+      "epoch": 0.6150222222222222,
+      "grad_norm": 1.4905400276184082,
+      "learning_rate": 7.701266948210714e-05,
+      "loss": 1.7319,
+      "step": 27676
+    },
+    {
+      "epoch": 0.6150444444444444,
+      "grad_norm": 1.6683911085128784,
+      "learning_rate": 7.700822404978885e-05,
+      "loss": 2.1189,
+      "step": 27677
+    },
+    {
+      "epoch": 0.6150666666666667,
+      "grad_norm": 1.7028203010559082,
+      "learning_rate": 7.700377861747055e-05,
+      "loss": 2.313,
+      "step": 27678
+    },
+    {
+      "epoch": 0.6150888888888889,
+      "grad_norm": 1.4560718536376953,
+      "learning_rate": 7.699933318515225e-05,
+      "loss": 1.8408,
+      "step": 27679
+    },
+    {
+      "epoch": 0.6151111111111112,
+      "grad_norm": 1.5160682201385498,
+      "learning_rate": 7.699488775283396e-05,
+      "loss": 2.0621,
+      "step": 27680
+    },
+    {
+      "epoch": 0.6151333333333333,
+      "grad_norm": 1.6277612447738647,
+      "learning_rate": 7.699044232051567e-05,
+      "loss": 1.9481,
+      "step": 27681
+    },
+    {
+      "epoch": 0.6151555555555556,
+      "grad_norm": 1.7742987871170044,
+      "learning_rate": 7.698599688819738e-05,
+      "loss": 2.1653,
+      "step": 27682
+    },
+    {
+      "epoch": 0.6151777777777778,
+      "grad_norm": 1.5646719932556152,
+      "learning_rate": 7.698155145587909e-05,
+      "loss": 1.8406,
+      "step": 27683
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 1.6861202716827393,
+      "learning_rate": 7.69771060235608e-05,
+      "loss": 2.6977,
+      "step": 27684
+    },
+    {
+      "epoch": 0.6152222222222222,
+      "grad_norm": 2.004915475845337,
+      "learning_rate": 7.697266059124251e-05,
+      "loss": 2.252,
+      "step": 27685
+    },
+    {
+      "epoch": 0.6152444444444445,
+      "grad_norm": 0.9852021336555481,
+      "learning_rate": 7.69682151589242e-05,
+      "loss": 0.8603,
+      "step": 27686
+    },
+    {
+      "epoch": 0.6152666666666666,
+      "grad_norm": 1.5224124193191528,
+      "learning_rate": 7.696376972660591e-05,
+      "loss": 1.9327,
+      "step": 27687
+    },
+    {
+      "epoch": 0.6152888888888889,
+      "grad_norm": 1.6800435781478882,
+      "learning_rate": 7.695932429428762e-05,
+      "loss": 1.7287,
+      "step": 27688
+    },
+    {
+      "epoch": 0.6153111111111111,
+      "grad_norm": 1.6324905157089233,
+      "learning_rate": 7.695487886196932e-05,
+      "loss": 1.5948,
+      "step": 27689
+    },
+    {
+      "epoch": 0.6153333333333333,
+      "grad_norm": 1.5132930278778076,
+      "learning_rate": 7.695043342965103e-05,
+      "loss": 1.4836,
+      "step": 27690
+    },
+    {
+      "epoch": 0.6153555555555555,
+      "grad_norm": 2.3151564598083496,
+      "learning_rate": 7.694598799733275e-05,
+      "loss": 1.8424,
+      "step": 27691
+    },
+    {
+      "epoch": 0.6153777777777778,
+      "grad_norm": 1.6328611373901367,
+      "learning_rate": 7.694154256501445e-05,
+      "loss": 1.5843,
+      "step": 27692
+    },
+    {
+      "epoch": 0.6154,
+      "grad_norm": 2.0141165256500244,
+      "learning_rate": 7.693709713269616e-05,
+      "loss": 1.8026,
+      "step": 27693
+    },
+    {
+      "epoch": 0.6154222222222222,
+      "grad_norm": 1.8666143417358398,
+      "learning_rate": 7.693265170037787e-05,
+      "loss": 2.0635,
+      "step": 27694
+    },
+    {
+      "epoch": 0.6154444444444445,
+      "grad_norm": 2.37695574760437,
+      "learning_rate": 7.692820626805958e-05,
+      "loss": 1.9471,
+      "step": 27695
+    },
+    {
+      "epoch": 0.6154666666666667,
+      "grad_norm": 1.7559750080108643,
+      "learning_rate": 7.692376083574127e-05,
+      "loss": 1.5826,
+      "step": 27696
+    },
+    {
+      "epoch": 0.6154888888888889,
+      "grad_norm": 1.5847703218460083,
+      "learning_rate": 7.691931540342298e-05,
+      "loss": 1.3723,
+      "step": 27697
+    },
+    {
+      "epoch": 0.6155111111111111,
+      "grad_norm": 1.7543034553527832,
+      "learning_rate": 7.691486997110469e-05,
+      "loss": 1.6881,
+      "step": 27698
+    },
+    {
+      "epoch": 0.6155333333333334,
+      "grad_norm": 1.7109777927398682,
+      "learning_rate": 7.69104245387864e-05,
+      "loss": 1.6384,
+      "step": 27699
+    },
+    {
+      "epoch": 0.6155555555555555,
+      "grad_norm": 1.8934999704360962,
+      "learning_rate": 7.690597910646811e-05,
+      "loss": 1.4734,
+      "step": 27700
+    },
+    {
+      "epoch": 0.6155777777777778,
+      "grad_norm": 1.2288439273834229,
+      "learning_rate": 7.690153367414982e-05,
+      "loss": 1.3923,
+      "step": 27701
+    },
+    {
+      "epoch": 0.6156,
+      "grad_norm": 1.6553082466125488,
+      "learning_rate": 7.689708824183152e-05,
+      "loss": 1.8585,
+      "step": 27702
+    },
+    {
+      "epoch": 0.6156222222222222,
+      "grad_norm": 1.480249285697937,
+      "learning_rate": 7.689264280951323e-05,
+      "loss": 2.0241,
+      "step": 27703
+    },
+    {
+      "epoch": 0.6156444444444444,
+      "grad_norm": 1.4849505424499512,
+      "learning_rate": 7.688819737719494e-05,
+      "loss": 2.1186,
+      "step": 27704
+    },
+    {
+      "epoch": 0.6156666666666667,
+      "grad_norm": 1.8371179103851318,
+      "learning_rate": 7.688375194487665e-05,
+      "loss": 2.0136,
+      "step": 27705
+    },
+    {
+      "epoch": 0.6156888888888888,
+      "grad_norm": 2.0982840061187744,
+      "learning_rate": 7.687930651255834e-05,
+      "loss": 2.2286,
+      "step": 27706
+    },
+    {
+      "epoch": 0.6157111111111111,
+      "grad_norm": 1.4270573854446411,
+      "learning_rate": 7.687486108024005e-05,
+      "loss": 2.0153,
+      "step": 27707
+    },
+    {
+      "epoch": 0.6157333333333334,
+      "grad_norm": 1.5784318447113037,
+      "learning_rate": 7.687041564792176e-05,
+      "loss": 1.8786,
+      "step": 27708
+    },
+    {
+      "epoch": 0.6157555555555555,
+      "grad_norm": 1.4389392137527466,
+      "learning_rate": 7.686597021560347e-05,
+      "loss": 1.6763,
+      "step": 27709
+    },
+    {
+      "epoch": 0.6157777777777778,
+      "grad_norm": 1.6586129665374756,
+      "learning_rate": 7.686152478328518e-05,
+      "loss": 2.0527,
+      "step": 27710
+    },
+    {
+      "epoch": 0.6158,
+      "grad_norm": 1.6005334854125977,
+      "learning_rate": 7.685707935096689e-05,
+      "loss": 2.1591,
+      "step": 27711
+    },
+    {
+      "epoch": 0.6158222222222223,
+      "grad_norm": 1.971131443977356,
+      "learning_rate": 7.68526339186486e-05,
+      "loss": 1.7389,
+      "step": 27712
+    },
+    {
+      "epoch": 0.6158444444444444,
+      "grad_norm": 1.5543437004089355,
+      "learning_rate": 7.68481884863303e-05,
+      "loss": 1.9799,
+      "step": 27713
+    },
+    {
+      "epoch": 0.6158666666666667,
+      "grad_norm": 1.8680613040924072,
+      "learning_rate": 7.6843743054012e-05,
+      "loss": 1.5808,
+      "step": 27714
+    },
+    {
+      "epoch": 0.6158888888888889,
+      "grad_norm": 1.8699873685836792,
+      "learning_rate": 7.683929762169372e-05,
+      "loss": 2.1983,
+      "step": 27715
+    },
+    {
+      "epoch": 0.6159111111111111,
+      "grad_norm": 1.506440281867981,
+      "learning_rate": 7.683485218937541e-05,
+      "loss": 2.2206,
+      "step": 27716
+    },
+    {
+      "epoch": 0.6159333333333333,
+      "grad_norm": 1.7313804626464844,
+      "learning_rate": 7.683040675705712e-05,
+      "loss": 1.9819,
+      "step": 27717
+    },
+    {
+      "epoch": 0.6159555555555556,
+      "grad_norm": 1.4775224924087524,
+      "learning_rate": 7.682596132473883e-05,
+      "loss": 2.0975,
+      "step": 27718
+    },
+    {
+      "epoch": 0.6159777777777777,
+      "grad_norm": 1.656072974205017,
+      "learning_rate": 7.682151589242054e-05,
+      "loss": 2.4082,
+      "step": 27719
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 1.63306725025177,
+      "learning_rate": 7.681707046010225e-05,
+      "loss": 1.7959,
+      "step": 27720
+    },
+    {
+      "epoch": 0.6160222222222222,
+      "grad_norm": 2.13395357131958,
+      "learning_rate": 7.681262502778396e-05,
+      "loss": 0.0452,
+      "step": 27721
+    },
+    {
+      "epoch": 0.6160444444444444,
+      "grad_norm": 1.8045676946640015,
+      "learning_rate": 7.680817959546567e-05,
+      "loss": 2.08,
+      "step": 27722
+    },
+    {
+      "epoch": 0.6160666666666667,
+      "grad_norm": 1.8881685733795166,
+      "learning_rate": 7.680373416314737e-05,
+      "loss": 2.011,
+      "step": 27723
+    },
+    {
+      "epoch": 0.6160888888888889,
+      "grad_norm": 1.9771424531936646,
+      "learning_rate": 7.679928873082908e-05,
+      "loss": 2.2206,
+      "step": 27724
+    },
+    {
+      "epoch": 0.6161111111111112,
+      "grad_norm": 1.520660161972046,
+      "learning_rate": 7.679484329851079e-05,
+      "loss": 2.0039,
+      "step": 27725
+    },
+    {
+      "epoch": 0.6161333333333333,
+      "grad_norm": 1.7060950994491577,
+      "learning_rate": 7.679039786619248e-05,
+      "loss": 2.3537,
+      "step": 27726
+    },
+    {
+      "epoch": 0.6161555555555556,
+      "grad_norm": 2.0422251224517822,
+      "learning_rate": 7.67859524338742e-05,
+      "loss": 1.8996,
+      "step": 27727
+    },
+    {
+      "epoch": 0.6161777777777778,
+      "grad_norm": 1.5210211277008057,
+      "learning_rate": 7.678150700155591e-05,
+      "loss": 1.9079,
+      "step": 27728
+    },
+    {
+      "epoch": 0.6162,
+      "grad_norm": 1.5915676355361938,
+      "learning_rate": 7.677706156923761e-05,
+      "loss": 2.0076,
+      "step": 27729
+    },
+    {
+      "epoch": 0.6162222222222222,
+      "grad_norm": 1.6983424425125122,
+      "learning_rate": 7.677261613691932e-05,
+      "loss": 2.2466,
+      "step": 27730
+    },
+    {
+      "epoch": 0.6162444444444445,
+      "grad_norm": 1.7953312397003174,
+      "learning_rate": 7.676817070460103e-05,
+      "loss": 2.1523,
+      "step": 27731
+    },
+    {
+      "epoch": 0.6162666666666666,
+      "grad_norm": 1.9210249185562134,
+      "learning_rate": 7.676372527228274e-05,
+      "loss": 2.0222,
+      "step": 27732
+    },
+    {
+      "epoch": 0.6162888888888889,
+      "grad_norm": 1.5903412103652954,
+      "learning_rate": 7.675927983996443e-05,
+      "loss": 1.6526,
+      "step": 27733
+    },
+    {
+      "epoch": 0.6163111111111111,
+      "grad_norm": 1.300310730934143,
+      "learning_rate": 7.675483440764614e-05,
+      "loss": 0.0304,
+      "step": 27734
+    },
+    {
+      "epoch": 0.6163333333333333,
+      "grad_norm": 1.4345909357070923,
+      "learning_rate": 7.675038897532785e-05,
+      "loss": 1.624,
+      "step": 27735
+    },
+    {
+      "epoch": 0.6163555555555555,
+      "grad_norm": 1.777416467666626,
+      "learning_rate": 7.674594354300956e-05,
+      "loss": 1.6793,
+      "step": 27736
+    },
+    {
+      "epoch": 0.6163777777777778,
+      "grad_norm": 1.7176923751831055,
+      "learning_rate": 7.674149811069127e-05,
+      "loss": 1.6987,
+      "step": 27737
+    },
+    {
+      "epoch": 0.6164,
+      "grad_norm": 1.9084837436676025,
+      "learning_rate": 7.673705267837298e-05,
+      "loss": 1.9548,
+      "step": 27738
+    },
+    {
+      "epoch": 0.6164222222222222,
+      "grad_norm": 1.5448215007781982,
+      "learning_rate": 7.673260724605468e-05,
+      "loss": 1.2958,
+      "step": 27739
+    },
+    {
+      "epoch": 0.6164444444444445,
+      "grad_norm": 1.9638711214065552,
+      "learning_rate": 7.672816181373639e-05,
+      "loss": 1.3227,
+      "step": 27740
+    },
+    {
+      "epoch": 0.6164666666666667,
+      "grad_norm": 1.9058576822280884,
+      "learning_rate": 7.67237163814181e-05,
+      "loss": 1.6331,
+      "step": 27741
+    },
+    {
+      "epoch": 0.6164888888888889,
+      "grad_norm": 1.6180567741394043,
+      "learning_rate": 7.671927094909981e-05,
+      "loss": 1.9669,
+      "step": 27742
+    },
+    {
+      "epoch": 0.6165111111111111,
+      "grad_norm": 1.7356921434402466,
+      "learning_rate": 7.67148255167815e-05,
+      "loss": 2.1797,
+      "step": 27743
+    },
+    {
+      "epoch": 0.6165333333333334,
+      "grad_norm": 1.7215243577957153,
+      "learning_rate": 7.671038008446321e-05,
+      "loss": 1.6665,
+      "step": 27744
+    },
+    {
+      "epoch": 0.6165555555555555,
+      "grad_norm": 2.344080924987793,
+      "learning_rate": 7.670593465214492e-05,
+      "loss": 2.1785,
+      "step": 27745
+    },
+    {
+      "epoch": 0.6165777777777778,
+      "grad_norm": 1.594612956047058,
+      "learning_rate": 7.670148921982663e-05,
+      "loss": 1.5843,
+      "step": 27746
+    },
+    {
+      "epoch": 0.6166,
+      "grad_norm": 2.3588497638702393,
+      "learning_rate": 7.669704378750834e-05,
+      "loss": 1.8844,
+      "step": 27747
+    },
+    {
+      "epoch": 0.6166222222222222,
+      "grad_norm": 1.6239612102508545,
+      "learning_rate": 7.669259835519005e-05,
+      "loss": 1.6552,
+      "step": 27748
+    },
+    {
+      "epoch": 0.6166444444444444,
+      "grad_norm": 1.8449560403823853,
+      "learning_rate": 7.668815292287175e-05,
+      "loss": 1.7764,
+      "step": 27749
+    },
+    {
+      "epoch": 0.6166666666666667,
+      "grad_norm": 1.4692752361297607,
+      "learning_rate": 7.668370749055346e-05,
+      "loss": 1.2563,
+      "step": 27750
+    },
+    {
+      "epoch": 0.6166888888888888,
+      "grad_norm": 1.3470090627670288,
+      "learning_rate": 7.667926205823517e-05,
+      "loss": 2.1952,
+      "step": 27751
+    },
+    {
+      "epoch": 0.6167111111111111,
+      "grad_norm": 1.4804136753082275,
+      "learning_rate": 7.667481662591688e-05,
+      "loss": 2.3699,
+      "step": 27752
+    },
+    {
+      "epoch": 0.6167333333333334,
+      "grad_norm": 1.5747419595718384,
+      "learning_rate": 7.667037119359857e-05,
+      "loss": 2.253,
+      "step": 27753
+    },
+    {
+      "epoch": 0.6167555555555555,
+      "grad_norm": 1.3297574520111084,
+      "learning_rate": 7.666592576128028e-05,
+      "loss": 1.1228,
+      "step": 27754
+    },
+    {
+      "epoch": 0.6167777777777778,
+      "grad_norm": 1.3217922449111938,
+      "learning_rate": 7.666148032896199e-05,
+      "loss": 1.8382,
+      "step": 27755
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 1.4520697593688965,
+      "learning_rate": 7.66570348966437e-05,
+      "loss": 2.1111,
+      "step": 27756
+    },
+    {
+      "epoch": 0.6168222222222223,
+      "grad_norm": 1.4051185846328735,
+      "learning_rate": 7.665258946432541e-05,
+      "loss": 1.8979,
+      "step": 27757
+    },
+    {
+      "epoch": 0.6168444444444444,
+      "grad_norm": 1.55694580078125,
+      "learning_rate": 7.664814403200712e-05,
+      "loss": 2.0987,
+      "step": 27758
+    },
+    {
+      "epoch": 0.6168666666666667,
+      "grad_norm": 1.452420711517334,
+      "learning_rate": 7.664369859968882e-05,
+      "loss": 1.866,
+      "step": 27759
+    },
+    {
+      "epoch": 0.6168888888888889,
+      "grad_norm": 1.4672553539276123,
+      "learning_rate": 7.663925316737053e-05,
+      "loss": 1.8999,
+      "step": 27760
+    },
+    {
+      "epoch": 0.6169111111111111,
+      "grad_norm": 1.9965301752090454,
+      "learning_rate": 7.663480773505224e-05,
+      "loss": 2.7316,
+      "step": 27761
+    },
+    {
+      "epoch": 0.6169333333333333,
+      "grad_norm": 1.617195725440979,
+      "learning_rate": 7.663036230273395e-05,
+      "loss": 2.4018,
+      "step": 27762
+    },
+    {
+      "epoch": 0.6169555555555556,
+      "grad_norm": 1.7019859552383423,
+      "learning_rate": 7.662591687041564e-05,
+      "loss": 2.1988,
+      "step": 27763
+    },
+    {
+      "epoch": 0.6169777777777777,
+      "grad_norm": 1.8571925163269043,
+      "learning_rate": 7.662147143809737e-05,
+      "loss": 2.2452,
+      "step": 27764
+    },
+    {
+      "epoch": 0.617,
+      "grad_norm": 1.922149658203125,
+      "learning_rate": 7.661702600577907e-05,
+      "loss": 2.1675,
+      "step": 27765
+    },
+    {
+      "epoch": 0.6170222222222222,
+      "grad_norm": 1.533893346786499,
+      "learning_rate": 7.661258057346077e-05,
+      "loss": 1.4528,
+      "step": 27766
+    },
+    {
+      "epoch": 0.6170444444444444,
+      "grad_norm": 1.5106412172317505,
+      "learning_rate": 7.660813514114248e-05,
+      "loss": 1.8997,
+      "step": 27767
+    },
+    {
+      "epoch": 0.6170666666666667,
+      "grad_norm": 1.6652741432189941,
+      "learning_rate": 7.660368970882419e-05,
+      "loss": 2.0417,
+      "step": 27768
+    },
+    {
+      "epoch": 0.6170888888888889,
+      "grad_norm": 1.8946317434310913,
+      "learning_rate": 7.65992442765059e-05,
+      "loss": 2.3318,
+      "step": 27769
+    },
+    {
+      "epoch": 0.6171111111111112,
+      "grad_norm": 2.1442677974700928,
+      "learning_rate": 7.65947988441876e-05,
+      "loss": 1.679,
+      "step": 27770
+    },
+    {
+      "epoch": 0.6171333333333333,
+      "grad_norm": 1.441885232925415,
+      "learning_rate": 7.65903534118693e-05,
+      "loss": 1.6786,
+      "step": 27771
+    },
+    {
+      "epoch": 0.6171555555555556,
+      "grad_norm": 1.5308915376663208,
+      "learning_rate": 7.658590797955102e-05,
+      "loss": 1.8118,
+      "step": 27772
+    },
+    {
+      "epoch": 0.6171777777777778,
+      "grad_norm": 2.0713369846343994,
+      "learning_rate": 7.658146254723272e-05,
+      "loss": 2.2014,
+      "step": 27773
+    },
+    {
+      "epoch": 0.6172,
+      "grad_norm": 1.5334278345108032,
+      "learning_rate": 7.657701711491443e-05,
+      "loss": 2.0471,
+      "step": 27774
+    },
+    {
+      "epoch": 0.6172222222222222,
+      "grad_norm": 1.7738068103790283,
+      "learning_rate": 7.657257168259614e-05,
+      "loss": 1.888,
+      "step": 27775
+    },
+    {
+      "epoch": 0.6172444444444445,
+      "grad_norm": 2.077033519744873,
+      "learning_rate": 7.656812625027784e-05,
+      "loss": 1.7194,
+      "step": 27776
+    },
+    {
+      "epoch": 0.6172666666666666,
+      "grad_norm": 1.8221032619476318,
+      "learning_rate": 7.656368081795955e-05,
+      "loss": 1.9039,
+      "step": 27777
+    },
+    {
+      "epoch": 0.6172888888888889,
+      "grad_norm": 1.6100168228149414,
+      "learning_rate": 7.655923538564126e-05,
+      "loss": 1.8672,
+      "step": 27778
+    },
+    {
+      "epoch": 0.6173111111111111,
+      "grad_norm": 2.5112087726593018,
+      "learning_rate": 7.655478995332297e-05,
+      "loss": 2.4041,
+      "step": 27779
+    },
+    {
+      "epoch": 0.6173333333333333,
+      "grad_norm": 1.9133623838424683,
+      "learning_rate": 7.655034452100467e-05,
+      "loss": 1.8831,
+      "step": 27780
+    },
+    {
+      "epoch": 0.6173555555555555,
+      "grad_norm": 1.3843563795089722,
+      "learning_rate": 7.654589908868637e-05,
+      "loss": 1.1407,
+      "step": 27781
+    },
+    {
+      "epoch": 0.6173777777777778,
+      "grad_norm": 1.566394567489624,
+      "learning_rate": 7.654145365636808e-05,
+      "loss": 1.4861,
+      "step": 27782
+    },
+    {
+      "epoch": 0.6174,
+      "grad_norm": 1.7080655097961426,
+      "learning_rate": 7.65370082240498e-05,
+      "loss": 1.7517,
+      "step": 27783
+    },
+    {
+      "epoch": 0.6174222222222222,
+      "grad_norm": 1.7616134881973267,
+      "learning_rate": 7.65325627917315e-05,
+      "loss": 1.8535,
+      "step": 27784
+    },
+    {
+      "epoch": 0.6174444444444445,
+      "grad_norm": 1.7613402605056763,
+      "learning_rate": 7.652811735941321e-05,
+      "loss": 1.4388,
+      "step": 27785
+    },
+    {
+      "epoch": 0.6174666666666667,
+      "grad_norm": 1.4484586715698242,
+      "learning_rate": 7.652367192709491e-05,
+      "loss": 1.7956,
+      "step": 27786
+    },
+    {
+      "epoch": 0.6174888888888889,
+      "grad_norm": 1.606410264968872,
+      "learning_rate": 7.651922649477662e-05,
+      "loss": 1.9883,
+      "step": 27787
+    },
+    {
+      "epoch": 0.6175111111111111,
+      "grad_norm": 2.0317225456237793,
+      "learning_rate": 7.651478106245833e-05,
+      "loss": 1.5763,
+      "step": 27788
+    },
+    {
+      "epoch": 0.6175333333333334,
+      "grad_norm": 1.7747282981872559,
+      "learning_rate": 7.651033563014004e-05,
+      "loss": 1.9427,
+      "step": 27789
+    },
+    {
+      "epoch": 0.6175555555555555,
+      "grad_norm": 1.4215421676635742,
+      "learning_rate": 7.650589019782173e-05,
+      "loss": 1.4721,
+      "step": 27790
+    },
+    {
+      "epoch": 0.6175777777777778,
+      "grad_norm": 1.934670090675354,
+      "learning_rate": 7.650144476550344e-05,
+      "loss": 1.6565,
+      "step": 27791
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 2.0426065921783447,
+      "learning_rate": 7.649699933318515e-05,
+      "loss": 1.6775,
+      "step": 27792
+    },
+    {
+      "epoch": 0.6176222222222222,
+      "grad_norm": 1.7949936389923096,
+      "learning_rate": 7.649255390086686e-05,
+      "loss": 1.7326,
+      "step": 27793
+    },
+    {
+      "epoch": 0.6176444444444444,
+      "grad_norm": 1.9961637258529663,
+      "learning_rate": 7.648810846854857e-05,
+      "loss": 1.7102,
+      "step": 27794
+    },
+    {
+      "epoch": 0.6176666666666667,
+      "grad_norm": 1.7532471418380737,
+      "learning_rate": 7.648366303623028e-05,
+      "loss": 1.7731,
+      "step": 27795
+    },
+    {
+      "epoch": 0.6176888888888888,
+      "grad_norm": 2.2291626930236816,
+      "learning_rate": 7.647921760391198e-05,
+      "loss": 2.2091,
+      "step": 27796
+    },
+    {
+      "epoch": 0.6177111111111111,
+      "grad_norm": 2.012691020965576,
+      "learning_rate": 7.647477217159369e-05,
+      "loss": 1.9455,
+      "step": 27797
+    },
+    {
+      "epoch": 0.6177333333333334,
+      "grad_norm": 1.7636815309524536,
+      "learning_rate": 7.64703267392754e-05,
+      "loss": 1.5035,
+      "step": 27798
+    },
+    {
+      "epoch": 0.6177555555555555,
+      "grad_norm": 1.8447431325912476,
+      "learning_rate": 7.646588130695711e-05,
+      "loss": 1.3523,
+      "step": 27799
+    },
+    {
+      "epoch": 0.6177777777777778,
+      "grad_norm": 1.5177545547485352,
+      "learning_rate": 7.64614358746388e-05,
+      "loss": 1.147,
+      "step": 27800
+    },
+    {
+      "epoch": 0.6178,
+      "grad_norm": 1.4865962266921997,
+      "learning_rate": 7.645699044232053e-05,
+      "loss": 2.593,
+      "step": 27801
+    },
+    {
+      "epoch": 0.6178222222222223,
+      "grad_norm": 1.5389937162399292,
+      "learning_rate": 7.645254501000224e-05,
+      "loss": 2.2661,
+      "step": 27802
+    },
+    {
+      "epoch": 0.6178444444444444,
+      "grad_norm": 1.3967429399490356,
+      "learning_rate": 7.644809957768393e-05,
+      "loss": 2.368,
+      "step": 27803
+    },
+    {
+      "epoch": 0.6178666666666667,
+      "grad_norm": 1.512739896774292,
+      "learning_rate": 7.644365414536564e-05,
+      "loss": 2.3682,
+      "step": 27804
+    },
+    {
+      "epoch": 0.6178888888888889,
+      "grad_norm": 1.7297879457473755,
+      "learning_rate": 7.643920871304735e-05,
+      "loss": 2.2775,
+      "step": 27805
+    },
+    {
+      "epoch": 0.6179111111111111,
+      "grad_norm": 1.5569535493850708,
+      "learning_rate": 7.643476328072905e-05,
+      "loss": 1.8969,
+      "step": 27806
+    },
+    {
+      "epoch": 0.6179333333333333,
+      "grad_norm": 2.3133952617645264,
+      "learning_rate": 7.643031784841076e-05,
+      "loss": 2.1335,
+      "step": 27807
+    },
+    {
+      "epoch": 0.6179555555555556,
+      "grad_norm": 1.477544903755188,
+      "learning_rate": 7.642587241609247e-05,
+      "loss": 1.9267,
+      "step": 27808
+    },
+    {
+      "epoch": 0.6179777777777777,
+      "grad_norm": 1.5673898458480835,
+      "learning_rate": 7.642142698377418e-05,
+      "loss": 2.1177,
+      "step": 27809
+    },
+    {
+      "epoch": 0.618,
+      "grad_norm": 1.7615025043487549,
+      "learning_rate": 7.641698155145589e-05,
+      "loss": 2.2152,
+      "step": 27810
+    },
+    {
+      "epoch": 0.6180222222222223,
+      "grad_norm": 2.3260927200317383,
+      "learning_rate": 7.64125361191376e-05,
+      "loss": 2.2751,
+      "step": 27811
+    },
+    {
+      "epoch": 0.6180444444444444,
+      "grad_norm": 1.756048321723938,
+      "learning_rate": 7.64080906868193e-05,
+      "loss": 2.406,
+      "step": 27812
+    },
+    {
+      "epoch": 0.6180666666666667,
+      "grad_norm": 2.2361507415771484,
+      "learning_rate": 7.6403645254501e-05,
+      "loss": 1.9301,
+      "step": 27813
+    },
+    {
+      "epoch": 0.6180888888888889,
+      "grad_norm": 1.5776044130325317,
+      "learning_rate": 7.639919982218271e-05,
+      "loss": 2.2465,
+      "step": 27814
+    },
+    {
+      "epoch": 0.6181111111111111,
+      "grad_norm": 1.456405520439148,
+      "learning_rate": 7.639475438986442e-05,
+      "loss": 2.1656,
+      "step": 27815
+    },
+    {
+      "epoch": 0.6181333333333333,
+      "grad_norm": 1.5151910781860352,
+      "learning_rate": 7.639030895754612e-05,
+      "loss": 1.9655,
+      "step": 27816
+    },
+    {
+      "epoch": 0.6181555555555556,
+      "grad_norm": 1.819975733757019,
+      "learning_rate": 7.638586352522783e-05,
+      "loss": 1.7451,
+      "step": 27817
+    },
+    {
+      "epoch": 0.6181777777777778,
+      "grad_norm": 1.6116926670074463,
+      "learning_rate": 7.638141809290954e-05,
+      "loss": 1.9346,
+      "step": 27818
+    },
+    {
+      "epoch": 0.6182,
+      "grad_norm": 1.3958470821380615,
+      "learning_rate": 7.637697266059125e-05,
+      "loss": 1.57,
+      "step": 27819
+    },
+    {
+      "epoch": 0.6182222222222222,
+      "grad_norm": 1.6095985174179077,
+      "learning_rate": 7.637252722827295e-05,
+      "loss": 2.1294,
+      "step": 27820
+    },
+    {
+      "epoch": 0.6182444444444445,
+      "grad_norm": 1.5986542701721191,
+      "learning_rate": 7.636808179595466e-05,
+      "loss": 2.2341,
+      "step": 27821
+    },
+    {
+      "epoch": 0.6182666666666666,
+      "grad_norm": 1.6214473247528076,
+      "learning_rate": 7.636363636363637e-05,
+      "loss": 2.133,
+      "step": 27822
+    },
+    {
+      "epoch": 0.6182888888888889,
+      "grad_norm": 1.593881368637085,
+      "learning_rate": 7.635919093131807e-05,
+      "loss": 1.7351,
+      "step": 27823
+    },
+    {
+      "epoch": 0.6183111111111111,
+      "grad_norm": 1.630756139755249,
+      "learning_rate": 7.635474549899978e-05,
+      "loss": 1.7785,
+      "step": 27824
+    },
+    {
+      "epoch": 0.6183333333333333,
+      "grad_norm": 1.658084750175476,
+      "learning_rate": 7.635030006668149e-05,
+      "loss": 1.5986,
+      "step": 27825
+    },
+    {
+      "epoch": 0.6183555555555555,
+      "grad_norm": 1.6406036615371704,
+      "learning_rate": 7.63458546343632e-05,
+      "loss": 1.8687,
+      "step": 27826
+    },
+    {
+      "epoch": 0.6183777777777778,
+      "grad_norm": 1.6786532402038574,
+      "learning_rate": 7.63414092020449e-05,
+      "loss": 1.6849,
+      "step": 27827
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 1.719092607498169,
+      "learning_rate": 7.63369637697266e-05,
+      "loss": 1.7264,
+      "step": 27828
+    },
+    {
+      "epoch": 0.6184222222222222,
+      "grad_norm": 1.5456290245056152,
+      "learning_rate": 7.633251833740831e-05,
+      "loss": 1.8574,
+      "step": 27829
+    },
+    {
+      "epoch": 0.6184444444444445,
+      "grad_norm": 1.4805980920791626,
+      "learning_rate": 7.632807290509002e-05,
+      "loss": 1.6378,
+      "step": 27830
+    },
+    {
+      "epoch": 0.6184666666666667,
+      "grad_norm": 1.7630740404129028,
+      "learning_rate": 7.632362747277173e-05,
+      "loss": 1.7688,
+      "step": 27831
+    },
+    {
+      "epoch": 0.6184888888888889,
+      "grad_norm": 1.7586926221847534,
+      "learning_rate": 7.631918204045344e-05,
+      "loss": 2.1503,
+      "step": 27832
+    },
+    {
+      "epoch": 0.6185111111111111,
+      "grad_norm": 1.1839295625686646,
+      "learning_rate": 7.631473660813514e-05,
+      "loss": 0.9218,
+      "step": 27833
+    },
+    {
+      "epoch": 0.6185333333333334,
+      "grad_norm": 1.8422901630401611,
+      "learning_rate": 7.631029117581685e-05,
+      "loss": 2.0849,
+      "step": 27834
+    },
+    {
+      "epoch": 0.6185555555555555,
+      "grad_norm": 2.327899217605591,
+      "learning_rate": 7.630584574349856e-05,
+      "loss": 2.2619,
+      "step": 27835
+    },
+    {
+      "epoch": 0.6185777777777778,
+      "grad_norm": 1.7383606433868408,
+      "learning_rate": 7.630140031118027e-05,
+      "loss": 1.9783,
+      "step": 27836
+    },
+    {
+      "epoch": 0.6186,
+      "grad_norm": 1.705854892730713,
+      "learning_rate": 7.629695487886196e-05,
+      "loss": 1.8867,
+      "step": 27837
+    },
+    {
+      "epoch": 0.6186222222222222,
+      "grad_norm": 1.2072322368621826,
+      "learning_rate": 7.629250944654369e-05,
+      "loss": 1.3474,
+      "step": 27838
+    },
+    {
+      "epoch": 0.6186444444444444,
+      "grad_norm": 1.545650601387024,
+      "learning_rate": 7.62880640142254e-05,
+      "loss": 1.4679,
+      "step": 27839
+    },
+    {
+      "epoch": 0.6186666666666667,
+      "grad_norm": 1.7425636053085327,
+      "learning_rate": 7.628361858190709e-05,
+      "loss": 2.004,
+      "step": 27840
+    },
+    {
+      "epoch": 0.6186888888888888,
+      "grad_norm": 1.946216344833374,
+      "learning_rate": 7.62791731495888e-05,
+      "loss": 1.6654,
+      "step": 27841
+    },
+    {
+      "epoch": 0.6187111111111111,
+      "grad_norm": 1.860919713973999,
+      "learning_rate": 7.627472771727051e-05,
+      "loss": 1.8677,
+      "step": 27842
+    },
+    {
+      "epoch": 0.6187333333333334,
+      "grad_norm": 1.6177150011062622,
+      "learning_rate": 7.627028228495221e-05,
+      "loss": 1.694,
+      "step": 27843
+    },
+    {
+      "epoch": 0.6187555555555555,
+      "grad_norm": 1.7288682460784912,
+      "learning_rate": 7.626583685263392e-05,
+      "loss": 1.6698,
+      "step": 27844
+    },
+    {
+      "epoch": 0.6187777777777778,
+      "grad_norm": 2.123058795928955,
+      "learning_rate": 7.626139142031563e-05,
+      "loss": 1.6972,
+      "step": 27845
+    },
+    {
+      "epoch": 0.6188,
+      "grad_norm": 1.999112844467163,
+      "learning_rate": 7.625694598799734e-05,
+      "loss": 1.6694,
+      "step": 27846
+    },
+    {
+      "epoch": 0.6188222222222223,
+      "grad_norm": 1.8956773281097412,
+      "learning_rate": 7.625250055567905e-05,
+      "loss": 1.946,
+      "step": 27847
+    },
+    {
+      "epoch": 0.6188444444444444,
+      "grad_norm": 2.0588977336883545,
+      "learning_rate": 7.624805512336076e-05,
+      "loss": 1.6527,
+      "step": 27848
+    },
+    {
+      "epoch": 0.6188666666666667,
+      "grad_norm": 1.6914948225021362,
+      "learning_rate": 7.624360969104247e-05,
+      "loss": 1.3385,
+      "step": 27849
+    },
+    {
+      "epoch": 0.6188888888888889,
+      "grad_norm": 1.8900210857391357,
+      "learning_rate": 7.623916425872416e-05,
+      "loss": 1.3768,
+      "step": 27850
+    },
+    {
+      "epoch": 0.6189111111111111,
+      "grad_norm": 1.0682233572006226,
+      "learning_rate": 7.623471882640587e-05,
+      "loss": 1.192,
+      "step": 27851
+    },
+    {
+      "epoch": 0.6189333333333333,
+      "grad_norm": 1.622100591659546,
+      "learning_rate": 7.623027339408758e-05,
+      "loss": 2.828,
+      "step": 27852
+    },
+    {
+      "epoch": 0.6189555555555556,
+      "grad_norm": 1.6287745237350464,
+      "learning_rate": 7.622582796176928e-05,
+      "loss": 2.5527,
+      "step": 27853
+    },
+    {
+      "epoch": 0.6189777777777777,
+      "grad_norm": 1.8082754611968994,
+      "learning_rate": 7.622138252945099e-05,
+      "loss": 2.5409,
+      "step": 27854
+    },
+    {
+      "epoch": 0.619,
+      "grad_norm": 1.6269252300262451,
+      "learning_rate": 7.62169370971327e-05,
+      "loss": 2.1211,
+      "step": 27855
+    },
+    {
+      "epoch": 0.6190222222222223,
+      "grad_norm": 1.4222098588943481,
+      "learning_rate": 7.62124916648144e-05,
+      "loss": 2.0527,
+      "step": 27856
+    },
+    {
+      "epoch": 0.6190444444444444,
+      "grad_norm": 1.8875646591186523,
+      "learning_rate": 7.620804623249612e-05,
+      "loss": 2.3583,
+      "step": 27857
+    },
+    {
+      "epoch": 0.6190666666666667,
+      "grad_norm": 1.4680495262145996,
+      "learning_rate": 7.620360080017783e-05,
+      "loss": 2.3106,
+      "step": 27858
+    },
+    {
+      "epoch": 0.6190888888888889,
+      "grad_norm": 1.6367533206939697,
+      "learning_rate": 7.619915536785954e-05,
+      "loss": 2.1393,
+      "step": 27859
+    },
+    {
+      "epoch": 0.6191111111111111,
+      "grad_norm": 1.9394482374191284,
+      "learning_rate": 7.619470993554123e-05,
+      "loss": 1.9793,
+      "step": 27860
+    },
+    {
+      "epoch": 0.6191333333333333,
+      "grad_norm": 1.4326545000076294,
+      "learning_rate": 7.619026450322294e-05,
+      "loss": 2.1417,
+      "step": 27861
+    },
+    {
+      "epoch": 0.6191555555555556,
+      "grad_norm": 1.5940648317337036,
+      "learning_rate": 7.618581907090465e-05,
+      "loss": 2.3789,
+      "step": 27862
+    },
+    {
+      "epoch": 0.6191777777777778,
+      "grad_norm": 1.5996965169906616,
+      "learning_rate": 7.618137363858635e-05,
+      "loss": 1.6626,
+      "step": 27863
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 1.4322491884231567,
+      "learning_rate": 7.617692820626806e-05,
+      "loss": 2.0181,
+      "step": 27864
+    },
+    {
+      "epoch": 0.6192222222222222,
+      "grad_norm": 1.4913889169692993,
+      "learning_rate": 7.617248277394977e-05,
+      "loss": 2.2868,
+      "step": 27865
+    },
+    {
+      "epoch": 0.6192444444444445,
+      "grad_norm": 1.576112151145935,
+      "learning_rate": 7.616803734163148e-05,
+      "loss": 2.0001,
+      "step": 27866
+    },
+    {
+      "epoch": 0.6192666666666666,
+      "grad_norm": 2.043811082839966,
+      "learning_rate": 7.616359190931319e-05,
+      "loss": 2.1336,
+      "step": 27867
+    },
+    {
+      "epoch": 0.6192888888888889,
+      "grad_norm": 1.5992106199264526,
+      "learning_rate": 7.61591464769949e-05,
+      "loss": 1.9645,
+      "step": 27868
+    },
+    {
+      "epoch": 0.6193111111111111,
+      "grad_norm": 1.496248483657837,
+      "learning_rate": 7.61547010446766e-05,
+      "loss": 1.5838,
+      "step": 27869
+    },
+    {
+      "epoch": 0.6193333333333333,
+      "grad_norm": 2.008216381072998,
+      "learning_rate": 7.61502556123583e-05,
+      "loss": 2.4949,
+      "step": 27870
+    },
+    {
+      "epoch": 0.6193555555555555,
+      "grad_norm": 1.7011308670043945,
+      "learning_rate": 7.614581018004001e-05,
+      "loss": 2.3068,
+      "step": 27871
+    },
+    {
+      "epoch": 0.6193777777777778,
+      "grad_norm": 1.6385869979858398,
+      "learning_rate": 7.614136474772172e-05,
+      "loss": 1.1203,
+      "step": 27872
+    },
+    {
+      "epoch": 0.6194,
+      "grad_norm": 1.4843419790267944,
+      "learning_rate": 7.613691931540343e-05,
+      "loss": 1.6449,
+      "step": 27873
+    },
+    {
+      "epoch": 0.6194222222222222,
+      "grad_norm": 1.7454240322113037,
+      "learning_rate": 7.613247388308513e-05,
+      "loss": 1.6117,
+      "step": 27874
+    },
+    {
+      "epoch": 0.6194444444444445,
+      "grad_norm": 2.1531989574432373,
+      "learning_rate": 7.612802845076685e-05,
+      "loss": 2.2732,
+      "step": 27875
+    },
+    {
+      "epoch": 0.6194666666666667,
+      "grad_norm": 1.6188414096832275,
+      "learning_rate": 7.612358301844856e-05,
+      "loss": 1.614,
+      "step": 27876
+    },
+    {
+      "epoch": 0.6194888888888889,
+      "grad_norm": 1.970890760421753,
+      "learning_rate": 7.611913758613025e-05,
+      "loss": 2.3413,
+      "step": 27877
+    },
+    {
+      "epoch": 0.6195111111111111,
+      "grad_norm": 1.5219156742095947,
+      "learning_rate": 7.611469215381196e-05,
+      "loss": 1.6374,
+      "step": 27878
+    },
+    {
+      "epoch": 0.6195333333333334,
+      "grad_norm": 1.6971971988677979,
+      "learning_rate": 7.611024672149367e-05,
+      "loss": 1.5241,
+      "step": 27879
+    },
+    {
+      "epoch": 0.6195555555555555,
+      "grad_norm": 1.7241084575653076,
+      "learning_rate": 7.610580128917537e-05,
+      "loss": 2.0711,
+      "step": 27880
+    },
+    {
+      "epoch": 0.6195777777777778,
+      "grad_norm": 1.8422069549560547,
+      "learning_rate": 7.610135585685708e-05,
+      "loss": 2.0101,
+      "step": 27881
+    },
+    {
+      "epoch": 0.6196,
+      "grad_norm": 1.7453213930130005,
+      "learning_rate": 7.609691042453879e-05,
+      "loss": 1.7256,
+      "step": 27882
+    },
+    {
+      "epoch": 0.6196222222222222,
+      "grad_norm": 1.6272207498550415,
+      "learning_rate": 7.60924649922205e-05,
+      "loss": 1.8167,
+      "step": 27883
+    },
+    {
+      "epoch": 0.6196444444444444,
+      "grad_norm": 2.016037940979004,
+      "learning_rate": 7.608801955990221e-05,
+      "loss": 1.7058,
+      "step": 27884
+    },
+    {
+      "epoch": 0.6196666666666667,
+      "grad_norm": 2.0873775482177734,
+      "learning_rate": 7.608357412758392e-05,
+      "loss": 1.9221,
+      "step": 27885
+    },
+    {
+      "epoch": 0.6196888888888888,
+      "grad_norm": 1.8145103454589844,
+      "learning_rate": 7.607912869526563e-05,
+      "loss": 1.9094,
+      "step": 27886
+    },
+    {
+      "epoch": 0.6197111111111111,
+      "grad_norm": 2.046666145324707,
+      "learning_rate": 7.607468326294732e-05,
+      "loss": 2.0466,
+      "step": 27887
+    },
+    {
+      "epoch": 0.6197333333333334,
+      "grad_norm": 1.5662391185760498,
+      "learning_rate": 7.607023783062903e-05,
+      "loss": 2.0985,
+      "step": 27888
+    },
+    {
+      "epoch": 0.6197555555555555,
+      "grad_norm": 1.8663203716278076,
+      "learning_rate": 7.606579239831074e-05,
+      "loss": 1.8951,
+      "step": 27889
+    },
+    {
+      "epoch": 0.6197777777777778,
+      "grad_norm": 1.8117085695266724,
+      "learning_rate": 7.606134696599244e-05,
+      "loss": 1.6704,
+      "step": 27890
+    },
+    {
+      "epoch": 0.6198,
+      "grad_norm": 1.5927486419677734,
+      "learning_rate": 7.605690153367415e-05,
+      "loss": 1.6549,
+      "step": 27891
+    },
+    {
+      "epoch": 0.6198222222222223,
+      "grad_norm": 1.8937994241714478,
+      "learning_rate": 7.605245610135586e-05,
+      "loss": 2.0085,
+      "step": 27892
+    },
+    {
+      "epoch": 0.6198444444444444,
+      "grad_norm": 1.9662067890167236,
+      "learning_rate": 7.604801066903757e-05,
+      "loss": 2.1213,
+      "step": 27893
+    },
+    {
+      "epoch": 0.6198666666666667,
+      "grad_norm": 2.182756185531616,
+      "learning_rate": 7.604356523671928e-05,
+      "loss": 1.8149,
+      "step": 27894
+    },
+    {
+      "epoch": 0.6198888888888889,
+      "grad_norm": 1.598151683807373,
+      "learning_rate": 7.603911980440099e-05,
+      "loss": 1.779,
+      "step": 27895
+    },
+    {
+      "epoch": 0.6199111111111111,
+      "grad_norm": 1.8498104810714722,
+      "learning_rate": 7.60346743720827e-05,
+      "loss": 1.5005,
+      "step": 27896
+    },
+    {
+      "epoch": 0.6199333333333333,
+      "grad_norm": 1.8257795572280884,
+      "learning_rate": 7.603022893976439e-05,
+      "loss": 1.6445,
+      "step": 27897
+    },
+    {
+      "epoch": 0.6199555555555556,
+      "grad_norm": 1.6474151611328125,
+      "learning_rate": 7.60257835074461e-05,
+      "loss": 1.7728,
+      "step": 27898
+    },
+    {
+      "epoch": 0.6199777777777777,
+      "grad_norm": 1.864945888519287,
+      "learning_rate": 7.602133807512781e-05,
+      "loss": 1.8965,
+      "step": 27899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7285076379776,
+      "learning_rate": 7.601689264280951e-05,
+      "loss": 1.7502,
+      "step": 27900
+    },
+    {
+      "epoch": 0.6200222222222223,
+      "grad_norm": 1.3599979877471924,
+      "learning_rate": 7.601244721049122e-05,
+      "loss": 2.4592,
+      "step": 27901
+    },
+    {
+      "epoch": 0.6200444444444444,
+      "grad_norm": 1.5312752723693848,
+      "learning_rate": 7.600800177817293e-05,
+      "loss": 2.6304,
+      "step": 27902
+    },
+    {
+      "epoch": 0.6200666666666667,
+      "grad_norm": 1.5203238725662231,
+      "learning_rate": 7.600355634585464e-05,
+      "loss": 2.3161,
+      "step": 27903
+    },
+    {
+      "epoch": 0.6200888888888889,
+      "grad_norm": 1.6090893745422363,
+      "learning_rate": 7.599911091353635e-05,
+      "loss": 2.5759,
+      "step": 27904
+    },
+    {
+      "epoch": 0.6201111111111111,
+      "grad_norm": 1.3716859817504883,
+      "learning_rate": 7.599466548121806e-05,
+      "loss": 2.0005,
+      "step": 27905
+    },
+    {
+      "epoch": 0.6201333333333333,
+      "grad_norm": 1.4314205646514893,
+      "learning_rate": 7.599022004889977e-05,
+      "loss": 1.8578,
+      "step": 27906
+    },
+    {
+      "epoch": 0.6201555555555556,
+      "grad_norm": 1.3883862495422363,
+      "learning_rate": 7.598577461658146e-05,
+      "loss": 1.7013,
+      "step": 27907
+    },
+    {
+      "epoch": 0.6201777777777778,
+      "grad_norm": 1.5877196788787842,
+      "learning_rate": 7.598132918426317e-05,
+      "loss": 1.9532,
+      "step": 27908
+    },
+    {
+      "epoch": 0.6202,
+      "grad_norm": 1.5918232202529907,
+      "learning_rate": 7.597688375194488e-05,
+      "loss": 1.0095,
+      "step": 27909
+    },
+    {
+      "epoch": 0.6202222222222222,
+      "grad_norm": 1.5913398265838623,
+      "learning_rate": 7.597243831962658e-05,
+      "loss": 2.2015,
+      "step": 27910
+    },
+    {
+      "epoch": 0.6202444444444445,
+      "grad_norm": 1.7465524673461914,
+      "learning_rate": 7.596799288730829e-05,
+      "loss": 2.2779,
+      "step": 27911
+    },
+    {
+      "epoch": 0.6202666666666666,
+      "grad_norm": 1.722979187965393,
+      "learning_rate": 7.596354745499001e-05,
+      "loss": 2.114,
+      "step": 27912
+    },
+    {
+      "epoch": 0.6202888888888889,
+      "grad_norm": 1.6596341133117676,
+      "learning_rate": 7.595910202267172e-05,
+      "loss": 2.0469,
+      "step": 27913
+    },
+    {
+      "epoch": 0.6203111111111111,
+      "grad_norm": 1.748468041419983,
+      "learning_rate": 7.595465659035342e-05,
+      "loss": 1.9282,
+      "step": 27914
+    },
+    {
+      "epoch": 0.6203333333333333,
+      "grad_norm": 1.206313967704773,
+      "learning_rate": 7.595021115803512e-05,
+      "loss": 1.0193,
+      "step": 27915
+    },
+    {
+      "epoch": 0.6203555555555555,
+      "grad_norm": 2.0730957984924316,
+      "learning_rate": 7.594576572571683e-05,
+      "loss": 1.5116,
+      "step": 27916
+    },
+    {
+      "epoch": 0.6203777777777778,
+      "grad_norm": 1.7398715019226074,
+      "learning_rate": 7.594132029339853e-05,
+      "loss": 1.719,
+      "step": 27917
+    },
+    {
+      "epoch": 0.6204,
+      "grad_norm": 1.5374395847320557,
+      "learning_rate": 7.593687486108024e-05,
+      "loss": 1.5994,
+      "step": 27918
+    },
+    {
+      "epoch": 0.6204222222222222,
+      "grad_norm": 1.6957192420959473,
+      "learning_rate": 7.593242942876195e-05,
+      "loss": 2.3119,
+      "step": 27919
+    },
+    {
+      "epoch": 0.6204444444444445,
+      "grad_norm": 1.5393242835998535,
+      "learning_rate": 7.592798399644365e-05,
+      "loss": 2.1632,
+      "step": 27920
+    },
+    {
+      "epoch": 0.6204666666666667,
+      "grad_norm": 1.851560354232788,
+      "learning_rate": 7.592353856412537e-05,
+      "loss": 1.865,
+      "step": 27921
+    },
+    {
+      "epoch": 0.6204888888888889,
+      "grad_norm": 1.640065312385559,
+      "learning_rate": 7.591909313180708e-05,
+      "loss": 1.5166,
+      "step": 27922
+    },
+    {
+      "epoch": 0.6205111111111111,
+      "grad_norm": 1.5717978477478027,
+      "learning_rate": 7.591464769948879e-05,
+      "loss": 1.854,
+      "step": 27923
+    },
+    {
+      "epoch": 0.6205333333333334,
+      "grad_norm": 1.7012500762939453,
+      "learning_rate": 7.591020226717048e-05,
+      "loss": 1.856,
+      "step": 27924
+    },
+    {
+      "epoch": 0.6205555555555555,
+      "grad_norm": 2.284024477005005,
+      "learning_rate": 7.59057568348522e-05,
+      "loss": 1.8496,
+      "step": 27925
+    },
+    {
+      "epoch": 0.6205777777777778,
+      "grad_norm": 1.928219199180603,
+      "learning_rate": 7.59013114025339e-05,
+      "loss": 1.7534,
+      "step": 27926
+    },
+    {
+      "epoch": 0.6206,
+      "grad_norm": 1.737140417098999,
+      "learning_rate": 7.58968659702156e-05,
+      "loss": 1.5632,
+      "step": 27927
+    },
+    {
+      "epoch": 0.6206222222222222,
+      "grad_norm": 1.9052971601486206,
+      "learning_rate": 7.589242053789731e-05,
+      "loss": 2.1908,
+      "step": 27928
+    },
+    {
+      "epoch": 0.6206444444444444,
+      "grad_norm": 1.4619766473770142,
+      "learning_rate": 7.588797510557902e-05,
+      "loss": 1.838,
+      "step": 27929
+    },
+    {
+      "epoch": 0.6206666666666667,
+      "grad_norm": 1.62483811378479,
+      "learning_rate": 7.588352967326073e-05,
+      "loss": 1.7586,
+      "step": 27930
+    },
+    {
+      "epoch": 0.6206888888888888,
+      "grad_norm": 1.580270767211914,
+      "learning_rate": 7.587908424094244e-05,
+      "loss": 1.9294,
+      "step": 27931
+    },
+    {
+      "epoch": 0.6207111111111111,
+      "grad_norm": 1.4798740148544312,
+      "learning_rate": 7.587463880862415e-05,
+      "loss": 1.0451,
+      "step": 27932
+    },
+    {
+      "epoch": 0.6207333333333334,
+      "grad_norm": 1.5176098346710205,
+      "learning_rate": 7.587019337630586e-05,
+      "loss": 1.2598,
+      "step": 27933
+    },
+    {
+      "epoch": 0.6207555555555555,
+      "grad_norm": 1.3974021673202515,
+      "learning_rate": 7.586574794398755e-05,
+      "loss": 1.5877,
+      "step": 27934
+    },
+    {
+      "epoch": 0.6207777777777778,
+      "grad_norm": 1.6418516635894775,
+      "learning_rate": 7.586130251166926e-05,
+      "loss": 1.785,
+      "step": 27935
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 1.3830448389053345,
+      "learning_rate": 7.585685707935097e-05,
+      "loss": 1.246,
+      "step": 27936
+    },
+    {
+      "epoch": 0.6208222222222223,
+      "grad_norm": 1.5359796285629272,
+      "learning_rate": 7.585241164703267e-05,
+      "loss": 1.7877,
+      "step": 27937
+    },
+    {
+      "epoch": 0.6208444444444444,
+      "grad_norm": 2.0273587703704834,
+      "learning_rate": 7.584796621471438e-05,
+      "loss": 2.0674,
+      "step": 27938
+    },
+    {
+      "epoch": 0.6208666666666667,
+      "grad_norm": 1.653097152709961,
+      "learning_rate": 7.584352078239609e-05,
+      "loss": 2.0481,
+      "step": 27939
+    },
+    {
+      "epoch": 0.6208888888888889,
+      "grad_norm": 1.8791024684906006,
+      "learning_rate": 7.58390753500778e-05,
+      "loss": 1.8427,
+      "step": 27940
+    },
+    {
+      "epoch": 0.6209111111111111,
+      "grad_norm": 1.7230693101882935,
+      "learning_rate": 7.583462991775951e-05,
+      "loss": 1.6818,
+      "step": 27941
+    },
+    {
+      "epoch": 0.6209333333333333,
+      "grad_norm": 1.7749496698379517,
+      "learning_rate": 7.583018448544122e-05,
+      "loss": 1.9271,
+      "step": 27942
+    },
+    {
+      "epoch": 0.6209555555555556,
+      "grad_norm": 1.8635108470916748,
+      "learning_rate": 7.582573905312293e-05,
+      "loss": 1.7799,
+      "step": 27943
+    },
+    {
+      "epoch": 0.6209777777777777,
+      "grad_norm": 2.007655620574951,
+      "learning_rate": 7.582129362080462e-05,
+      "loss": 1.8257,
+      "step": 27944
+    },
+    {
+      "epoch": 0.621,
+      "grad_norm": 1.7758183479309082,
+      "learning_rate": 7.581684818848633e-05,
+      "loss": 2.0091,
+      "step": 27945
+    },
+    {
+      "epoch": 0.6210222222222223,
+      "grad_norm": 2.0555312633514404,
+      "learning_rate": 7.581240275616804e-05,
+      "loss": 2.1853,
+      "step": 27946
+    },
+    {
+      "epoch": 0.6210444444444444,
+      "grad_norm": 1.8315494060516357,
+      "learning_rate": 7.580795732384974e-05,
+      "loss": 1.8743,
+      "step": 27947
+    },
+    {
+      "epoch": 0.6210666666666667,
+      "grad_norm": 1.6850496530532837,
+      "learning_rate": 7.580351189153145e-05,
+      "loss": 1.6656,
+      "step": 27948
+    },
+    {
+      "epoch": 0.6210888888888889,
+      "grad_norm": 1.6388370990753174,
+      "learning_rate": 7.579906645921317e-05,
+      "loss": 1.2549,
+      "step": 27949
+    },
+    {
+      "epoch": 0.6211111111111111,
+      "grad_norm": 1.9473472833633423,
+      "learning_rate": 7.579462102689488e-05,
+      "loss": 1.1684,
+      "step": 27950
+    },
+    {
+      "epoch": 0.6211333333333333,
+      "grad_norm": 1.6398197412490845,
+      "learning_rate": 7.579017559457658e-05,
+      "loss": 2.7279,
+      "step": 27951
+    },
+    {
+      "epoch": 0.6211555555555556,
+      "grad_norm": 1.6046913862228394,
+      "learning_rate": 7.578573016225829e-05,
+      "loss": 2.8082,
+      "step": 27952
+    },
+    {
+      "epoch": 0.6211777777777778,
+      "grad_norm": 1.714526891708374,
+      "learning_rate": 7.578128472994e-05,
+      "loss": 3.4,
+      "step": 27953
+    },
+    {
+      "epoch": 0.6212,
+      "grad_norm": 1.4914047718048096,
+      "learning_rate": 7.577683929762169e-05,
+      "loss": 1.0971,
+      "step": 27954
+    },
+    {
+      "epoch": 0.6212222222222222,
+      "grad_norm": 1.3853058815002441,
+      "learning_rate": 7.57723938653034e-05,
+      "loss": 1.858,
+      "step": 27955
+    },
+    {
+      "epoch": 0.6212444444444445,
+      "grad_norm": 1.5049077272415161,
+      "learning_rate": 7.576794843298511e-05,
+      "loss": 2.2232,
+      "step": 27956
+    },
+    {
+      "epoch": 0.6212666666666666,
+      "grad_norm": 1.389599323272705,
+      "learning_rate": 7.576350300066681e-05,
+      "loss": 2.0723,
+      "step": 27957
+    },
+    {
+      "epoch": 0.6212888888888889,
+      "grad_norm": 1.340166449546814,
+      "learning_rate": 7.575905756834853e-05,
+      "loss": 1.3373,
+      "step": 27958
+    },
+    {
+      "epoch": 0.6213111111111111,
+      "grad_norm": 1.8471781015396118,
+      "learning_rate": 7.575461213603024e-05,
+      "loss": 2.4835,
+      "step": 27959
+    },
+    {
+      "epoch": 0.6213333333333333,
+      "grad_norm": 1.6483877897262573,
+      "learning_rate": 7.575016670371195e-05,
+      "loss": 2.3488,
+      "step": 27960
+    },
+    {
+      "epoch": 0.6213555555555555,
+      "grad_norm": 1.5012586116790771,
+      "learning_rate": 7.574572127139365e-05,
+      "loss": 1.88,
+      "step": 27961
+    },
+    {
+      "epoch": 0.6213777777777778,
+      "grad_norm": 1.6846944093704224,
+      "learning_rate": 7.574127583907535e-05,
+      "loss": 1.9666,
+      "step": 27962
+    },
+    {
+      "epoch": 0.6214,
+      "grad_norm": 1.699049711227417,
+      "learning_rate": 7.573683040675706e-05,
+      "loss": 2.4547,
+      "step": 27963
+    },
+    {
+      "epoch": 0.6214222222222222,
+      "grad_norm": 1.53487229347229,
+      "learning_rate": 7.573238497443876e-05,
+      "loss": 1.9296,
+      "step": 27964
+    },
+    {
+      "epoch": 0.6214444444444445,
+      "grad_norm": 1.4856736660003662,
+      "learning_rate": 7.572793954212047e-05,
+      "loss": 1.4113,
+      "step": 27965
+    },
+    {
+      "epoch": 0.6214666666666666,
+      "grad_norm": 1.408342719078064,
+      "learning_rate": 7.572349410980218e-05,
+      "loss": 1.6205,
+      "step": 27966
+    },
+    {
+      "epoch": 0.6214888888888889,
+      "grad_norm": 1.648368239402771,
+      "learning_rate": 7.571904867748389e-05,
+      "loss": 2.2225,
+      "step": 27967
+    },
+    {
+      "epoch": 0.6215111111111111,
+      "grad_norm": 1.4867284297943115,
+      "learning_rate": 7.57146032451656e-05,
+      "loss": 2.0689,
+      "step": 27968
+    },
+    {
+      "epoch": 0.6215333333333334,
+      "grad_norm": 1.564137578010559,
+      "learning_rate": 7.571015781284731e-05,
+      "loss": 2.3272,
+      "step": 27969
+    },
+    {
+      "epoch": 0.6215555555555555,
+      "grad_norm": 1.6785781383514404,
+      "learning_rate": 7.570571238052902e-05,
+      "loss": 1.8413,
+      "step": 27970
+    },
+    {
+      "epoch": 0.6215777777777778,
+      "grad_norm": 1.7516751289367676,
+      "learning_rate": 7.570126694821071e-05,
+      "loss": 1.9945,
+      "step": 27971
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 1.6698040962219238,
+      "learning_rate": 7.569682151589242e-05,
+      "loss": 2.0592,
+      "step": 27972
+    },
+    {
+      "epoch": 0.6216222222222222,
+      "grad_norm": 1.8508039712905884,
+      "learning_rate": 7.569237608357413e-05,
+      "loss": 1.9574,
+      "step": 27973
+    },
+    {
+      "epoch": 0.6216444444444444,
+      "grad_norm": 1.6818801164627075,
+      "learning_rate": 7.568793065125583e-05,
+      "loss": 1.4262,
+      "step": 27974
+    },
+    {
+      "epoch": 0.6216666666666667,
+      "grad_norm": 1.641690969467163,
+      "learning_rate": 7.568348521893754e-05,
+      "loss": 1.9255,
+      "step": 27975
+    },
+    {
+      "epoch": 0.6216888888888888,
+      "grad_norm": 1.6074247360229492,
+      "learning_rate": 7.567903978661925e-05,
+      "loss": 2.1551,
+      "step": 27976
+    },
+    {
+      "epoch": 0.6217111111111111,
+      "grad_norm": 1.5521165132522583,
+      "learning_rate": 7.567459435430096e-05,
+      "loss": 1.7597,
+      "step": 27977
+    },
+    {
+      "epoch": 0.6217333333333334,
+      "grad_norm": 1.5441497564315796,
+      "learning_rate": 7.567014892198267e-05,
+      "loss": 1.7756,
+      "step": 27978
+    },
+    {
+      "epoch": 0.6217555555555555,
+      "grad_norm": 1.7832576036453247,
+      "learning_rate": 7.566570348966438e-05,
+      "loss": 1.7453,
+      "step": 27979
+    },
+    {
+      "epoch": 0.6217777777777778,
+      "grad_norm": 1.5918539762496948,
+      "learning_rate": 7.566125805734609e-05,
+      "loss": 1.6821,
+      "step": 27980
+    },
+    {
+      "epoch": 0.6218,
+      "grad_norm": 1.6350274085998535,
+      "learning_rate": 7.565681262502778e-05,
+      "loss": 2.1195,
+      "step": 27981
+    },
+    {
+      "epoch": 0.6218222222222223,
+      "grad_norm": 1.7357590198516846,
+      "learning_rate": 7.565236719270949e-05,
+      "loss": 1.8812,
+      "step": 27982
+    },
+    {
+      "epoch": 0.6218444444444444,
+      "grad_norm": 1.4452167749404907,
+      "learning_rate": 7.56479217603912e-05,
+      "loss": 0.903,
+      "step": 27983
+    },
+    {
+      "epoch": 0.6218666666666667,
+      "grad_norm": 1.1117844581604004,
+      "learning_rate": 7.56434763280729e-05,
+      "loss": 0.8629,
+      "step": 27984
+    },
+    {
+      "epoch": 0.6218888888888889,
+      "grad_norm": 1.4665799140930176,
+      "learning_rate": 7.563903089575461e-05,
+      "loss": 1.6197,
+      "step": 27985
+    },
+    {
+      "epoch": 0.6219111111111111,
+      "grad_norm": 1.592340111732483,
+      "learning_rate": 7.563458546343633e-05,
+      "loss": 1.5545,
+      "step": 27986
+    },
+    {
+      "epoch": 0.6219333333333333,
+      "grad_norm": 1.722834587097168,
+      "learning_rate": 7.563014003111804e-05,
+      "loss": 1.8558,
+      "step": 27987
+    },
+    {
+      "epoch": 0.6219555555555556,
+      "grad_norm": 2.186478614807129,
+      "learning_rate": 7.562569459879974e-05,
+      "loss": 1.8553,
+      "step": 27988
+    },
+    {
+      "epoch": 0.6219777777777777,
+      "grad_norm": 1.7810604572296143,
+      "learning_rate": 7.562124916648145e-05,
+      "loss": 1.935,
+      "step": 27989
+    },
+    {
+      "epoch": 0.622,
+      "grad_norm": 1.6212259531021118,
+      "learning_rate": 7.561680373416316e-05,
+      "loss": 1.8521,
+      "step": 27990
+    },
+    {
+      "epoch": 0.6220222222222223,
+      "grad_norm": 2.047581672668457,
+      "learning_rate": 7.561235830184485e-05,
+      "loss": 2.0427,
+      "step": 27991
+    },
+    {
+      "epoch": 0.6220444444444444,
+      "grad_norm": 1.82438063621521,
+      "learning_rate": 7.560791286952656e-05,
+      "loss": 1.4275,
+      "step": 27992
+    },
+    {
+      "epoch": 0.6220666666666667,
+      "grad_norm": 1.806922197341919,
+      "learning_rate": 7.560346743720827e-05,
+      "loss": 1.6137,
+      "step": 27993
+    },
+    {
+      "epoch": 0.6220888888888889,
+      "grad_norm": 1.7153993844985962,
+      "learning_rate": 7.559902200488997e-05,
+      "loss": 2.0712,
+      "step": 27994
+    },
+    {
+      "epoch": 0.6221111111111111,
+      "grad_norm": 1.5733906030654907,
+      "learning_rate": 7.559457657257169e-05,
+      "loss": 1.3682,
+      "step": 27995
+    },
+    {
+      "epoch": 0.6221333333333333,
+      "grad_norm": 2.5927658081054688,
+      "learning_rate": 7.55901311402534e-05,
+      "loss": 2.0729,
+      "step": 27996
+    },
+    {
+      "epoch": 0.6221555555555556,
+      "grad_norm": 1.656896948814392,
+      "learning_rate": 7.558568570793511e-05,
+      "loss": 1.3535,
+      "step": 27997
+    },
+    {
+      "epoch": 0.6221777777777778,
+      "grad_norm": 1.611059308052063,
+      "learning_rate": 7.55812402756168e-05,
+      "loss": 1.6851,
+      "step": 27998
+    },
+    {
+      "epoch": 0.6222,
+      "grad_norm": 1.9701584577560425,
+      "learning_rate": 7.557679484329852e-05,
+      "loss": 2.047,
+      "step": 27999
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 1.6851133108139038,
+      "learning_rate": 7.557234941098023e-05,
+      "loss": 1.1879,
+      "step": 28000
+    },
+    {
+      "epoch": 0.6222444444444445,
+      "grad_norm": 1.6479599475860596,
+      "learning_rate": 7.556790397866192e-05,
+      "loss": 2.4432,
+      "step": 28001
+    },
+    {
+      "epoch": 0.6222666666666666,
+      "grad_norm": 1.5716572999954224,
+      "learning_rate": 7.556345854634363e-05,
+      "loss": 2.1938,
+      "step": 28002
+    },
+    {
+      "epoch": 0.6222888888888889,
+      "grad_norm": 2.1492650508880615,
+      "learning_rate": 7.555901311402534e-05,
+      "loss": 2.1888,
+      "step": 28003
+    },
+    {
+      "epoch": 0.6223111111111111,
+      "grad_norm": 1.7056307792663574,
+      "learning_rate": 7.555456768170705e-05,
+      "loss": 2.339,
+      "step": 28004
+    },
+    {
+      "epoch": 0.6223333333333333,
+      "grad_norm": 1.5766496658325195,
+      "learning_rate": 7.555012224938876e-05,
+      "loss": 2.9654,
+      "step": 28005
+    },
+    {
+      "epoch": 0.6223555555555556,
+      "grad_norm": 1.76160728931427,
+      "learning_rate": 7.554567681707047e-05,
+      "loss": 2.2328,
+      "step": 28006
+    },
+    {
+      "epoch": 0.6223777777777778,
+      "grad_norm": 1.5594091415405273,
+      "learning_rate": 7.554123138475218e-05,
+      "loss": 2.0131,
+      "step": 28007
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 1.8020453453063965,
+      "learning_rate": 7.553678595243388e-05,
+      "loss": 1.6641,
+      "step": 28008
+    },
+    {
+      "epoch": 0.6224222222222222,
+      "grad_norm": 1.7662761211395264,
+      "learning_rate": 7.553234052011559e-05,
+      "loss": 2.189,
+      "step": 28009
+    },
+    {
+      "epoch": 0.6224444444444445,
+      "grad_norm": 1.743079423904419,
+      "learning_rate": 7.55278950877973e-05,
+      "loss": 2.2716,
+      "step": 28010
+    },
+    {
+      "epoch": 0.6224666666666666,
+      "grad_norm": 1.5527716875076294,
+      "learning_rate": 7.552344965547899e-05,
+      "loss": 1.6795,
+      "step": 28011
+    },
+    {
+      "epoch": 0.6224888888888889,
+      "grad_norm": 1.5938984155654907,
+      "learning_rate": 7.55190042231607e-05,
+      "loss": 1.5725,
+      "step": 28012
+    },
+    {
+      "epoch": 0.6225111111111111,
+      "grad_norm": 1.3984183073043823,
+      "learning_rate": 7.551455879084241e-05,
+      "loss": 1.5888,
+      "step": 28013
+    },
+    {
+      "epoch": 0.6225333333333334,
+      "grad_norm": 1.4453216791152954,
+      "learning_rate": 7.551011335852412e-05,
+      "loss": 1.8982,
+      "step": 28014
+    },
+    {
+      "epoch": 0.6225555555555555,
+      "grad_norm": 1.5529475212097168,
+      "learning_rate": 7.550566792620583e-05,
+      "loss": 2.0408,
+      "step": 28015
+    },
+    {
+      "epoch": 0.6225777777777778,
+      "grad_norm": 1.4605592489242554,
+      "learning_rate": 7.550122249388754e-05,
+      "loss": 1.8066,
+      "step": 28016
+    },
+    {
+      "epoch": 0.6226,
+      "grad_norm": 1.8840243816375732,
+      "learning_rate": 7.549677706156925e-05,
+      "loss": 2.7807,
+      "step": 28017
+    },
+    {
+      "epoch": 0.6226222222222222,
+      "grad_norm": 1.7677241563796997,
+      "learning_rate": 7.549233162925094e-05,
+      "loss": 1.994,
+      "step": 28018
+    },
+    {
+      "epoch": 0.6226444444444444,
+      "grad_norm": 1.7343677282333374,
+      "learning_rate": 7.548788619693265e-05,
+      "loss": 2.0937,
+      "step": 28019
+    },
+    {
+      "epoch": 0.6226666666666667,
+      "grad_norm": 1.4782507419586182,
+      "learning_rate": 7.548344076461436e-05,
+      "loss": 1.6051,
+      "step": 28020
+    },
+    {
+      "epoch": 0.6226888888888888,
+      "grad_norm": 1.741336703300476,
+      "learning_rate": 7.547899533229606e-05,
+      "loss": 2.2404,
+      "step": 28021
+    },
+    {
+      "epoch": 0.6227111111111111,
+      "grad_norm": 1.725960612297058,
+      "learning_rate": 7.547454989997777e-05,
+      "loss": 1.658,
+      "step": 28022
+    },
+    {
+      "epoch": 0.6227333333333334,
+      "grad_norm": 1.852859377861023,
+      "learning_rate": 7.547010446765949e-05,
+      "loss": 2.0704,
+      "step": 28023
+    },
+    {
+      "epoch": 0.6227555555555555,
+      "grad_norm": 1.7915359735488892,
+      "learning_rate": 7.546565903534119e-05,
+      "loss": 2.1554,
+      "step": 28024
+    },
+    {
+      "epoch": 0.6227777777777778,
+      "grad_norm": 1.321905255317688,
+      "learning_rate": 7.54612136030229e-05,
+      "loss": 1.337,
+      "step": 28025
+    },
+    {
+      "epoch": 0.6228,
+      "grad_norm": 1.7915064096450806,
+      "learning_rate": 7.545676817070461e-05,
+      "loss": 1.6202,
+      "step": 28026
+    },
+    {
+      "epoch": 0.6228222222222223,
+      "grad_norm": 1.7633399963378906,
+      "learning_rate": 7.545232273838632e-05,
+      "loss": 1.8875,
+      "step": 28027
+    },
+    {
+      "epoch": 0.6228444444444444,
+      "grad_norm": 1.0724031925201416,
+      "learning_rate": 7.544787730606801e-05,
+      "loss": 0.8833,
+      "step": 28028
+    },
+    {
+      "epoch": 0.6228666666666667,
+      "grad_norm": 1.6533682346343994,
+      "learning_rate": 7.544343187374972e-05,
+      "loss": 1.6894,
+      "step": 28029
+    },
+    {
+      "epoch": 0.6228888888888889,
+      "grad_norm": 1.649261236190796,
+      "learning_rate": 7.543898644143143e-05,
+      "loss": 2.0422,
+      "step": 28030
+    },
+    {
+      "epoch": 0.6229111111111111,
+      "grad_norm": 1.9536653757095337,
+      "learning_rate": 7.543454100911313e-05,
+      "loss": 2.3201,
+      "step": 28031
+    },
+    {
+      "epoch": 0.6229333333333333,
+      "grad_norm": 1.7542853355407715,
+      "learning_rate": 7.543009557679485e-05,
+      "loss": 1.7455,
+      "step": 28032
+    },
+    {
+      "epoch": 0.6229555555555556,
+      "grad_norm": 1.3951656818389893,
+      "learning_rate": 7.542565014447656e-05,
+      "loss": 1.2113,
+      "step": 28033
+    },
+    {
+      "epoch": 0.6229777777777777,
+      "grad_norm": 1.873515009880066,
+      "learning_rate": 7.542120471215827e-05,
+      "loss": 1.6083,
+      "step": 28034
+    },
+    {
+      "epoch": 0.623,
+      "grad_norm": 1.6619421243667603,
+      "learning_rate": 7.541675927983997e-05,
+      "loss": 1.9856,
+      "step": 28035
+    },
+    {
+      "epoch": 0.6230222222222223,
+      "grad_norm": 1.208182692527771,
+      "learning_rate": 7.541231384752168e-05,
+      "loss": 0.8777,
+      "step": 28036
+    },
+    {
+      "epoch": 0.6230444444444444,
+      "grad_norm": 1.8139747381210327,
+      "learning_rate": 7.540786841520339e-05,
+      "loss": 1.6417,
+      "step": 28037
+    },
+    {
+      "epoch": 0.6230666666666667,
+      "grad_norm": 1.5409775972366333,
+      "learning_rate": 7.540342298288508e-05,
+      "loss": 1.7358,
+      "step": 28038
+    },
+    {
+      "epoch": 0.6230888888888889,
+      "grad_norm": 1.7688947916030884,
+      "learning_rate": 7.539897755056679e-05,
+      "loss": 2.0103,
+      "step": 28039
+    },
+    {
+      "epoch": 0.6231111111111111,
+      "grad_norm": 1.628895878791809,
+      "learning_rate": 7.53945321182485e-05,
+      "loss": 1.6595,
+      "step": 28040
+    },
+    {
+      "epoch": 0.6231333333333333,
+      "grad_norm": 1.785688877105713,
+      "learning_rate": 7.539008668593021e-05,
+      "loss": 1.721,
+      "step": 28041
+    },
+    {
+      "epoch": 0.6231555555555556,
+      "grad_norm": 1.9412195682525635,
+      "learning_rate": 7.538564125361192e-05,
+      "loss": 1.8939,
+      "step": 28042
+    },
+    {
+      "epoch": 0.6231777777777778,
+      "grad_norm": 2.1074423789978027,
+      "learning_rate": 7.538119582129363e-05,
+      "loss": 1.5976,
+      "step": 28043
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 2.227391242980957,
+      "learning_rate": 7.537675038897534e-05,
+      "loss": 1.9225,
+      "step": 28044
+    },
+    {
+      "epoch": 0.6232222222222222,
+      "grad_norm": 1.847145915031433,
+      "learning_rate": 7.537230495665704e-05,
+      "loss": 1.5274,
+      "step": 28045
+    },
+    {
+      "epoch": 0.6232444444444445,
+      "grad_norm": 2.542794942855835,
+      "learning_rate": 7.536785952433875e-05,
+      "loss": 2.4347,
+      "step": 28046
+    },
+    {
+      "epoch": 0.6232666666666666,
+      "grad_norm": 1.9294025897979736,
+      "learning_rate": 7.536341409202046e-05,
+      "loss": 1.9662,
+      "step": 28047
+    },
+    {
+      "epoch": 0.6232888888888889,
+      "grad_norm": 1.1842116117477417,
+      "learning_rate": 7.535896865970215e-05,
+      "loss": 0.8846,
+      "step": 28048
+    },
+    {
+      "epoch": 0.6233111111111111,
+      "grad_norm": 1.9983824491500854,
+      "learning_rate": 7.535452322738386e-05,
+      "loss": 1.8756,
+      "step": 28049
+    },
+    {
+      "epoch": 0.6233333333333333,
+      "grad_norm": 1.9489094018936157,
+      "learning_rate": 7.535007779506557e-05,
+      "loss": 1.6686,
+      "step": 28050
+    },
+    {
+      "epoch": 0.6233555555555556,
+      "grad_norm": 1.8719598054885864,
+      "learning_rate": 7.534563236274728e-05,
+      "loss": 2.0377,
+      "step": 28051
+    },
+    {
+      "epoch": 0.6233777777777778,
+      "grad_norm": 1.7876286506652832,
+      "learning_rate": 7.534118693042899e-05,
+      "loss": 2.6911,
+      "step": 28052
+    },
+    {
+      "epoch": 0.6234,
+      "grad_norm": 1.4274696111679077,
+      "learning_rate": 7.53367414981107e-05,
+      "loss": 1.9744,
+      "step": 28053
+    },
+    {
+      "epoch": 0.6234222222222222,
+      "grad_norm": 1.7839244604110718,
+      "learning_rate": 7.533229606579241e-05,
+      "loss": 2.4189,
+      "step": 28054
+    },
+    {
+      "epoch": 0.6234444444444445,
+      "grad_norm": 1.5739293098449707,
+      "learning_rate": 7.53278506334741e-05,
+      "loss": 1.8991,
+      "step": 28055
+    },
+    {
+      "epoch": 0.6234666666666666,
+      "grad_norm": 2.45786190032959,
+      "learning_rate": 7.532340520115582e-05,
+      "loss": 2.428,
+      "step": 28056
+    },
+    {
+      "epoch": 0.6234888888888889,
+      "grad_norm": 2.522510290145874,
+      "learning_rate": 7.531895976883752e-05,
+      "loss": 2.4247,
+      "step": 28057
+    },
+    {
+      "epoch": 0.6235111111111111,
+      "grad_norm": 1.6587966680526733,
+      "learning_rate": 7.531451433651922e-05,
+      "loss": 2.0945,
+      "step": 28058
+    },
+    {
+      "epoch": 0.6235333333333334,
+      "grad_norm": 1.7723037004470825,
+      "learning_rate": 7.531006890420093e-05,
+      "loss": 2.1222,
+      "step": 28059
+    },
+    {
+      "epoch": 0.6235555555555555,
+      "grad_norm": 1.5230958461761475,
+      "learning_rate": 7.530562347188265e-05,
+      "loss": 2.4104,
+      "step": 28060
+    },
+    {
+      "epoch": 0.6235777777777778,
+      "grad_norm": 1.460243582725525,
+      "learning_rate": 7.530117803956435e-05,
+      "loss": 0.8193,
+      "step": 28061
+    },
+    {
+      "epoch": 0.6236,
+      "grad_norm": 1.7510052919387817,
+      "learning_rate": 7.529673260724606e-05,
+      "loss": 1.968,
+      "step": 28062
+    },
+    {
+      "epoch": 0.6236222222222222,
+      "grad_norm": 1.5463873147964478,
+      "learning_rate": 7.529228717492777e-05,
+      "loss": 2.3151,
+      "step": 28063
+    },
+    {
+      "epoch": 0.6236444444444444,
+      "grad_norm": 1.510756492614746,
+      "learning_rate": 7.528784174260948e-05,
+      "loss": 1.5273,
+      "step": 28064
+    },
+    {
+      "epoch": 0.6236666666666667,
+      "grad_norm": 1.7299911975860596,
+      "learning_rate": 7.528339631029117e-05,
+      "loss": 2.1592,
+      "step": 28065
+    },
+    {
+      "epoch": 0.6236888888888888,
+      "grad_norm": 1.775102972984314,
+      "learning_rate": 7.527895087797288e-05,
+      "loss": 2.5766,
+      "step": 28066
+    },
+    {
+      "epoch": 0.6237111111111111,
+      "grad_norm": 1.8978391885757446,
+      "learning_rate": 7.52745054456546e-05,
+      "loss": 2.8868,
+      "step": 28067
+    },
+    {
+      "epoch": 0.6237333333333334,
+      "grad_norm": 1.4233916997909546,
+      "learning_rate": 7.527006001333629e-05,
+      "loss": 1.7611,
+      "step": 28068
+    },
+    {
+      "epoch": 0.6237555555555555,
+      "grad_norm": 1.5011509656906128,
+      "learning_rate": 7.526561458101801e-05,
+      "loss": 2.0248,
+      "step": 28069
+    },
+    {
+      "epoch": 0.6237777777777778,
+      "grad_norm": 1.5596365928649902,
+      "learning_rate": 7.526116914869972e-05,
+      "loss": 2.1135,
+      "step": 28070
+    },
+    {
+      "epoch": 0.6238,
+      "grad_norm": 0.19666370749473572,
+      "learning_rate": 7.525672371638142e-05,
+      "loss": 0.0258,
+      "step": 28071
+    },
+    {
+      "epoch": 0.6238222222222222,
+      "grad_norm": 1.6305148601531982,
+      "learning_rate": 7.525227828406313e-05,
+      "loss": 1.9894,
+      "step": 28072
+    },
+    {
+      "epoch": 0.6238444444444444,
+      "grad_norm": 1.5677214860916138,
+      "learning_rate": 7.524783285174484e-05,
+      "loss": 1.9228,
+      "step": 28073
+    },
+    {
+      "epoch": 0.6238666666666667,
+      "grad_norm": 1.7453768253326416,
+      "learning_rate": 7.524338741942655e-05,
+      "loss": 2.0331,
+      "step": 28074
+    },
+    {
+      "epoch": 0.6238888888888889,
+      "grad_norm": 1.7649633884429932,
+      "learning_rate": 7.523894198710824e-05,
+      "loss": 2.5632,
+      "step": 28075
+    },
+    {
+      "epoch": 0.6239111111111111,
+      "grad_norm": 1.564621925354004,
+      "learning_rate": 7.523449655478995e-05,
+      "loss": 1.9003,
+      "step": 28076
+    },
+    {
+      "epoch": 0.6239333333333333,
+      "grad_norm": 1.7973852157592773,
+      "learning_rate": 7.523005112247166e-05,
+      "loss": 2.2091,
+      "step": 28077
+    },
+    {
+      "epoch": 0.6239555555555556,
+      "grad_norm": 1.7107487916946411,
+      "learning_rate": 7.522560569015337e-05,
+      "loss": 1.9508,
+      "step": 28078
+    },
+    {
+      "epoch": 0.6239777777777777,
+      "grad_norm": 1.6050406694412231,
+      "learning_rate": 7.522116025783508e-05,
+      "loss": 1.8418,
+      "step": 28079
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.5532035827636719,
+      "learning_rate": 7.521671482551679e-05,
+      "loss": 1.7462,
+      "step": 28080
+    },
+    {
+      "epoch": 0.6240222222222223,
+      "grad_norm": 1.7981599569320679,
+      "learning_rate": 7.521226939319849e-05,
+      "loss": 2.0927,
+      "step": 28081
+    },
+    {
+      "epoch": 0.6240444444444444,
+      "grad_norm": 2.0693376064300537,
+      "learning_rate": 7.52078239608802e-05,
+      "loss": 2.339,
+      "step": 28082
+    },
+    {
+      "epoch": 0.6240666666666667,
+      "grad_norm": 2.317495822906494,
+      "learning_rate": 7.520337852856191e-05,
+      "loss": 2.6896,
+      "step": 28083
+    },
+    {
+      "epoch": 0.6240888888888889,
+      "grad_norm": 1.6972509622573853,
+      "learning_rate": 7.519893309624362e-05,
+      "loss": 2.0378,
+      "step": 28084
+    },
+    {
+      "epoch": 0.6241111111111111,
+      "grad_norm": 1.830173373222351,
+      "learning_rate": 7.519448766392531e-05,
+      "loss": 2.1773,
+      "step": 28085
+    },
+    {
+      "epoch": 0.6241333333333333,
+      "grad_norm": 2.4986438751220703,
+      "learning_rate": 7.519004223160702e-05,
+      "loss": 1.8763,
+      "step": 28086
+    },
+    {
+      "epoch": 0.6241555555555556,
+      "grad_norm": 1.9697855710983276,
+      "learning_rate": 7.518559679928873e-05,
+      "loss": 1.5322,
+      "step": 28087
+    },
+    {
+      "epoch": 0.6241777777777778,
+      "grad_norm": 1.905541181564331,
+      "learning_rate": 7.518115136697044e-05,
+      "loss": 1.6298,
+      "step": 28088
+    },
+    {
+      "epoch": 0.6242,
+      "grad_norm": 1.5465525388717651,
+      "learning_rate": 7.517670593465215e-05,
+      "loss": 1.4808,
+      "step": 28089
+    },
+    {
+      "epoch": 0.6242222222222222,
+      "grad_norm": 1.9321972131729126,
+      "learning_rate": 7.517226050233386e-05,
+      "loss": 1.6778,
+      "step": 28090
+    },
+    {
+      "epoch": 0.6242444444444445,
+      "grad_norm": 1.919901728630066,
+      "learning_rate": 7.516781507001557e-05,
+      "loss": 2.01,
+      "step": 28091
+    },
+    {
+      "epoch": 0.6242666666666666,
+      "grad_norm": 1.8274459838867188,
+      "learning_rate": 7.516336963769727e-05,
+      "loss": 2.2137,
+      "step": 28092
+    },
+    {
+      "epoch": 0.6242888888888889,
+      "grad_norm": 1.757928729057312,
+      "learning_rate": 7.515892420537898e-05,
+      "loss": 1.6233,
+      "step": 28093
+    },
+    {
+      "epoch": 0.6243111111111111,
+      "grad_norm": 1.7088501453399658,
+      "learning_rate": 7.515447877306069e-05,
+      "loss": 1.7571,
+      "step": 28094
+    },
+    {
+      "epoch": 0.6243333333333333,
+      "grad_norm": 2.1138648986816406,
+      "learning_rate": 7.515003334074238e-05,
+      "loss": 2.3092,
+      "step": 28095
+    },
+    {
+      "epoch": 0.6243555555555556,
+      "grad_norm": 1.675551414489746,
+      "learning_rate": 7.514558790842409e-05,
+      "loss": 1.7107,
+      "step": 28096
+    },
+    {
+      "epoch": 0.6243777777777778,
+      "grad_norm": 1.9008680582046509,
+      "learning_rate": 7.514114247610581e-05,
+      "loss": 2.0442,
+      "step": 28097
+    },
+    {
+      "epoch": 0.6244,
+      "grad_norm": 2.154712438583374,
+      "learning_rate": 7.513669704378751e-05,
+      "loss": 1.9904,
+      "step": 28098
+    },
+    {
+      "epoch": 0.6244222222222222,
+      "grad_norm": 1.4223718643188477,
+      "learning_rate": 7.513225161146922e-05,
+      "loss": 0.9587,
+      "step": 28099
+    },
+    {
+      "epoch": 0.6244444444444445,
+      "grad_norm": 2.5552480220794678,
+      "learning_rate": 7.512780617915093e-05,
+      "loss": 1.6455,
+      "step": 28100
+    },
+    {
+      "epoch": 0.6244666666666666,
+      "grad_norm": 1.4165558815002441,
+      "learning_rate": 7.512336074683264e-05,
+      "loss": 2.0838,
+      "step": 28101
+    },
+    {
+      "epoch": 0.6244888888888889,
+      "grad_norm": 1.561410665512085,
+      "learning_rate": 7.511891531451434e-05,
+      "loss": 2.2894,
+      "step": 28102
+    },
+    {
+      "epoch": 0.6245111111111111,
+      "grad_norm": 1.31568443775177,
+      "learning_rate": 7.511446988219605e-05,
+      "loss": 2.087,
+      "step": 28103
+    },
+    {
+      "epoch": 0.6245333333333334,
+      "grad_norm": 1.6914708614349365,
+      "learning_rate": 7.511002444987775e-05,
+      "loss": 1.9268,
+      "step": 28104
+    },
+    {
+      "epoch": 0.6245555555555555,
+      "grad_norm": 1.6146295070648193,
+      "learning_rate": 7.510557901755945e-05,
+      "loss": 2.0463,
+      "step": 28105
+    },
+    {
+      "epoch": 0.6245777777777778,
+      "grad_norm": 1.5655505657196045,
+      "learning_rate": 7.510113358524117e-05,
+      "loss": 2.1122,
+      "step": 28106
+    },
+    {
+      "epoch": 0.6246,
+      "grad_norm": 1.5095285177230835,
+      "learning_rate": 7.509668815292288e-05,
+      "loss": 2.0841,
+      "step": 28107
+    },
+    {
+      "epoch": 0.6246222222222222,
+      "grad_norm": 1.6935125589370728,
+      "learning_rate": 7.509224272060458e-05,
+      "loss": 1.5109,
+      "step": 28108
+    },
+    {
+      "epoch": 0.6246444444444444,
+      "grad_norm": 1.5695096254348755,
+      "learning_rate": 7.508779728828629e-05,
+      "loss": 2.0201,
+      "step": 28109
+    },
+    {
+      "epoch": 0.6246666666666667,
+      "grad_norm": 1.5828744173049927,
+      "learning_rate": 7.5083351855968e-05,
+      "loss": 2.3864,
+      "step": 28110
+    },
+    {
+      "epoch": 0.6246888888888888,
+      "grad_norm": 1.9256597757339478,
+      "learning_rate": 7.507890642364971e-05,
+      "loss": 1.3124,
+      "step": 28111
+    },
+    {
+      "epoch": 0.6247111111111111,
+      "grad_norm": 1.8132530450820923,
+      "learning_rate": 7.50744609913314e-05,
+      "loss": 1.9949,
+      "step": 28112
+    },
+    {
+      "epoch": 0.6247333333333334,
+      "grad_norm": 1.9781066179275513,
+      "learning_rate": 7.507001555901311e-05,
+      "loss": 2.2829,
+      "step": 28113
+    },
+    {
+      "epoch": 0.6247555555555555,
+      "grad_norm": 1.0637400150299072,
+      "learning_rate": 7.506557012669482e-05,
+      "loss": 1.1418,
+      "step": 28114
+    },
+    {
+      "epoch": 0.6247777777777778,
+      "grad_norm": 1.4971493482589722,
+      "learning_rate": 7.506112469437653e-05,
+      "loss": 1.9301,
+      "step": 28115
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 1.9875372648239136,
+      "learning_rate": 7.505667926205824e-05,
+      "loss": 1.9572,
+      "step": 28116
+    },
+    {
+      "epoch": 0.6248222222222222,
+      "grad_norm": 1.5765823125839233,
+      "learning_rate": 7.505223382973995e-05,
+      "loss": 2.1686,
+      "step": 28117
+    },
+    {
+      "epoch": 0.6248444444444444,
+      "grad_norm": 1.5952030420303345,
+      "learning_rate": 7.504778839742165e-05,
+      "loss": 1.7775,
+      "step": 28118
+    },
+    {
+      "epoch": 0.6248666666666667,
+      "grad_norm": 1.9303405284881592,
+      "learning_rate": 7.504334296510336e-05,
+      "loss": 2.2577,
+      "step": 28119
+    },
+    {
+      "epoch": 0.6248888888888889,
+      "grad_norm": 1.6546903848648071,
+      "learning_rate": 7.503889753278507e-05,
+      "loss": 1.9196,
+      "step": 28120
+    },
+    {
+      "epoch": 0.6249111111111111,
+      "grad_norm": 1.4748766422271729,
+      "learning_rate": 7.503445210046678e-05,
+      "loss": 1.9852,
+      "step": 28121
+    },
+    {
+      "epoch": 0.6249333333333333,
+      "grad_norm": 1.8407598733901978,
+      "learning_rate": 7.503000666814847e-05,
+      "loss": 1.7293,
+      "step": 28122
+    },
+    {
+      "epoch": 0.6249555555555556,
+      "grad_norm": 1.8829072713851929,
+      "learning_rate": 7.502556123583018e-05,
+      "loss": 2.2661,
+      "step": 28123
+    },
+    {
+      "epoch": 0.6249777777777777,
+      "grad_norm": 1.5948944091796875,
+      "learning_rate": 7.502111580351189e-05,
+      "loss": 1.7127,
+      "step": 28124
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 1.6878756284713745,
+      "learning_rate": 7.50166703711936e-05,
+      "loss": 2.1244,
+      "step": 28125
+    },
+    {
+      "epoch": 0.6250222222222223,
+      "grad_norm": 1.5194244384765625,
+      "learning_rate": 7.501222493887531e-05,
+      "loss": 1.5466,
+      "step": 28126
+    },
+    {
+      "epoch": 0.6250444444444444,
+      "grad_norm": 1.6358795166015625,
+      "learning_rate": 7.500777950655702e-05,
+      "loss": 1.8983,
+      "step": 28127
+    },
+    {
+      "epoch": 0.6250666666666667,
+      "grad_norm": 1.9146581888198853,
+      "learning_rate": 7.500333407423872e-05,
+      "loss": 2.1941,
+      "step": 28128
+    },
+    {
+      "epoch": 0.6250888888888889,
+      "grad_norm": 1.477559208869934,
+      "learning_rate": 7.499888864192043e-05,
+      "loss": 1.1678,
+      "step": 28129
+    },
+    {
+      "epoch": 0.6251111111111111,
+      "grad_norm": 1.3219830989837646,
+      "learning_rate": 7.499444320960214e-05,
+      "loss": 1.3111,
+      "step": 28130
+    },
+    {
+      "epoch": 0.6251333333333333,
+      "grad_norm": 1.3379400968551636,
+      "learning_rate": 7.498999777728385e-05,
+      "loss": 1.2058,
+      "step": 28131
+    },
+    {
+      "epoch": 0.6251555555555556,
+      "grad_norm": 1.565895676612854,
+      "learning_rate": 7.498555234496554e-05,
+      "loss": 2.0447,
+      "step": 28132
+    },
+    {
+      "epoch": 0.6251777777777778,
+      "grad_norm": 1.7458685636520386,
+      "learning_rate": 7.498110691264725e-05,
+      "loss": 1.6086,
+      "step": 28133
+    },
+    {
+      "epoch": 0.6252,
+      "grad_norm": 1.5337018966674805,
+      "learning_rate": 7.497666148032898e-05,
+      "loss": 1.7856,
+      "step": 28134
+    },
+    {
+      "epoch": 0.6252222222222222,
+      "grad_norm": 2.4415087699890137,
+      "learning_rate": 7.497221604801067e-05,
+      "loss": 2.4248,
+      "step": 28135
+    },
+    {
+      "epoch": 0.6252444444444445,
+      "grad_norm": 1.7598092555999756,
+      "learning_rate": 7.496777061569238e-05,
+      "loss": 2.0031,
+      "step": 28136
+    },
+    {
+      "epoch": 0.6252666666666666,
+      "grad_norm": 1.657050609588623,
+      "learning_rate": 7.496332518337409e-05,
+      "loss": 1.6827,
+      "step": 28137
+    },
+    {
+      "epoch": 0.6252888888888889,
+      "grad_norm": 1.4472593069076538,
+      "learning_rate": 7.495887975105579e-05,
+      "loss": 1.5198,
+      "step": 28138
+    },
+    {
+      "epoch": 0.6253111111111112,
+      "grad_norm": 0.7550368309020996,
+      "learning_rate": 7.49544343187375e-05,
+      "loss": 0.033,
+      "step": 28139
+    },
+    {
+      "epoch": 0.6253333333333333,
+      "grad_norm": 1.9193103313446045,
+      "learning_rate": 7.49499888864192e-05,
+      "loss": 1.9755,
+      "step": 28140
+    },
+    {
+      "epoch": 0.6253555555555556,
+      "grad_norm": 2.672001361846924,
+      "learning_rate": 7.494554345410092e-05,
+      "loss": 1.7431,
+      "step": 28141
+    },
+    {
+      "epoch": 0.6253777777777778,
+      "grad_norm": 1.5714386701583862,
+      "learning_rate": 7.494109802178261e-05,
+      "loss": 1.9451,
+      "step": 28142
+    },
+    {
+      "epoch": 0.6254,
+      "grad_norm": 1.6924715042114258,
+      "learning_rate": 7.493665258946434e-05,
+      "loss": 1.7028,
+      "step": 28143
+    },
+    {
+      "epoch": 0.6254222222222222,
+      "grad_norm": 1.9398430585861206,
+      "learning_rate": 7.493220715714604e-05,
+      "loss": 2.1162,
+      "step": 28144
+    },
+    {
+      "epoch": 0.6254444444444445,
+      "grad_norm": 1.5882272720336914,
+      "learning_rate": 7.492776172482774e-05,
+      "loss": 1.4288,
+      "step": 28145
+    },
+    {
+      "epoch": 0.6254666666666666,
+      "grad_norm": 1.8803850412368774,
+      "learning_rate": 7.492331629250945e-05,
+      "loss": 1.7391,
+      "step": 28146
+    },
+    {
+      "epoch": 0.6254888888888889,
+      "grad_norm": 1.6644755601882935,
+      "learning_rate": 7.491887086019116e-05,
+      "loss": 1.9229,
+      "step": 28147
+    },
+    {
+      "epoch": 0.6255111111111111,
+      "grad_norm": 2.1229488849639893,
+      "learning_rate": 7.491442542787287e-05,
+      "loss": 1.6049,
+      "step": 28148
+    },
+    {
+      "epoch": 0.6255333333333334,
+      "grad_norm": 1.5672968626022339,
+      "learning_rate": 7.490997999555457e-05,
+      "loss": 1.3858,
+      "step": 28149
+    },
+    {
+      "epoch": 0.6255555555555555,
+      "grad_norm": 1.2970843315124512,
+      "learning_rate": 7.490553456323628e-05,
+      "loss": 0.5406,
+      "step": 28150
+    },
+    {
+      "epoch": 0.6255777777777778,
+      "grad_norm": 1.9396559000015259,
+      "learning_rate": 7.490108913091799e-05,
+      "loss": 2.4668,
+      "step": 28151
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 1.4108716249465942,
+      "learning_rate": 7.48966436985997e-05,
+      "loss": 1.8329,
+      "step": 28152
+    },
+    {
+      "epoch": 0.6256222222222222,
+      "grad_norm": 1.3884369134902954,
+      "learning_rate": 7.48921982662814e-05,
+      "loss": 1.9971,
+      "step": 28153
+    },
+    {
+      "epoch": 0.6256444444444444,
+      "grad_norm": 1.5661218166351318,
+      "learning_rate": 7.488775283396311e-05,
+      "loss": 2.4888,
+      "step": 28154
+    },
+    {
+      "epoch": 0.6256666666666667,
+      "grad_norm": 1.4409959316253662,
+      "learning_rate": 7.488330740164481e-05,
+      "loss": 1.4596,
+      "step": 28155
+    },
+    {
+      "epoch": 0.6256888888888889,
+      "grad_norm": 1.6241331100463867,
+      "learning_rate": 7.487886196932652e-05,
+      "loss": 2.219,
+      "step": 28156
+    },
+    {
+      "epoch": 0.6257111111111111,
+      "grad_norm": 1.4938559532165527,
+      "learning_rate": 7.487441653700823e-05,
+      "loss": 2.295,
+      "step": 28157
+    },
+    {
+      "epoch": 0.6257333333333334,
+      "grad_norm": 1.578965187072754,
+      "learning_rate": 7.486997110468994e-05,
+      "loss": 2.276,
+      "step": 28158
+    },
+    {
+      "epoch": 0.6257555555555555,
+      "grad_norm": 1.4421511888504028,
+      "learning_rate": 7.486552567237163e-05,
+      "loss": 1.905,
+      "step": 28159
+    },
+    {
+      "epoch": 0.6257777777777778,
+      "grad_norm": 1.6099575757980347,
+      "learning_rate": 7.486108024005334e-05,
+      "loss": 1.9368,
+      "step": 28160
+    },
+    {
+      "epoch": 0.6258,
+      "grad_norm": 1.6032575368881226,
+      "learning_rate": 7.485663480773505e-05,
+      "loss": 1.7127,
+      "step": 28161
+    },
+    {
+      "epoch": 0.6258222222222222,
+      "grad_norm": 1.719819188117981,
+      "learning_rate": 7.485218937541676e-05,
+      "loss": 2.3936,
+      "step": 28162
+    },
+    {
+      "epoch": 0.6258444444444444,
+      "grad_norm": 1.682454228401184,
+      "learning_rate": 7.484774394309847e-05,
+      "loss": 2.2743,
+      "step": 28163
+    },
+    {
+      "epoch": 0.6258666666666667,
+      "grad_norm": 1.3777753114700317,
+      "learning_rate": 7.484329851078018e-05,
+      "loss": 0.9647,
+      "step": 28164
+    },
+    {
+      "epoch": 0.6258888888888889,
+      "grad_norm": 1.7213762998580933,
+      "learning_rate": 7.483885307846188e-05,
+      "loss": 2.2361,
+      "step": 28165
+    },
+    {
+      "epoch": 0.6259111111111111,
+      "grad_norm": 1.4945802688598633,
+      "learning_rate": 7.483440764614359e-05,
+      "loss": 1.7347,
+      "step": 28166
+    },
+    {
+      "epoch": 0.6259333333333333,
+      "grad_norm": 1.432998538017273,
+      "learning_rate": 7.48299622138253e-05,
+      "loss": 1.5117,
+      "step": 28167
+    },
+    {
+      "epoch": 0.6259555555555556,
+      "grad_norm": 1.8089045286178589,
+      "learning_rate": 7.482551678150701e-05,
+      "loss": 2.1483,
+      "step": 28168
+    },
+    {
+      "epoch": 0.6259777777777777,
+      "grad_norm": 1.4634894132614136,
+      "learning_rate": 7.48210713491887e-05,
+      "loss": 1.7531,
+      "step": 28169
+    },
+    {
+      "epoch": 0.626,
+      "grad_norm": 1.5178799629211426,
+      "learning_rate": 7.481662591687041e-05,
+      "loss": 1.5721,
+      "step": 28170
+    },
+    {
+      "epoch": 0.6260222222222223,
+      "grad_norm": 1.5224968194961548,
+      "learning_rate": 7.481218048455214e-05,
+      "loss": 1.9887,
+      "step": 28171
+    },
+    {
+      "epoch": 0.6260444444444444,
+      "grad_norm": 1.3402128219604492,
+      "learning_rate": 7.480773505223383e-05,
+      "loss": 1.6201,
+      "step": 28172
+    },
+    {
+      "epoch": 0.6260666666666667,
+      "grad_norm": 1.4766627550125122,
+      "learning_rate": 7.480328961991554e-05,
+      "loss": 1.7079,
+      "step": 28173
+    },
+    {
+      "epoch": 0.6260888888888889,
+      "grad_norm": 1.459479570388794,
+      "learning_rate": 7.479884418759725e-05,
+      "loss": 1.4734,
+      "step": 28174
+    },
+    {
+      "epoch": 0.6261111111111111,
+      "grad_norm": 1.8193565607070923,
+      "learning_rate": 7.479439875527895e-05,
+      "loss": 2.2156,
+      "step": 28175
+    },
+    {
+      "epoch": 0.6261333333333333,
+      "grad_norm": 1.553693413734436,
+      "learning_rate": 7.478995332296066e-05,
+      "loss": 1.5702,
+      "step": 28176
+    },
+    {
+      "epoch": 0.6261555555555556,
+      "grad_norm": 2.1361544132232666,
+      "learning_rate": 7.478550789064237e-05,
+      "loss": 1.8119,
+      "step": 28177
+    },
+    {
+      "epoch": 0.6261777777777778,
+      "grad_norm": 1.7945271730422974,
+      "learning_rate": 7.478106245832408e-05,
+      "loss": 2.1368,
+      "step": 28178
+    },
+    {
+      "epoch": 0.6262,
+      "grad_norm": 1.1177501678466797,
+      "learning_rate": 7.477661702600577e-05,
+      "loss": 0.7338,
+      "step": 28179
+    },
+    {
+      "epoch": 0.6262222222222222,
+      "grad_norm": 1.8593310117721558,
+      "learning_rate": 7.47721715936875e-05,
+      "loss": 2.0467,
+      "step": 28180
+    },
+    {
+      "epoch": 0.6262444444444445,
+      "grad_norm": 2.0093624591827393,
+      "learning_rate": 7.47677261613692e-05,
+      "loss": 1.9046,
+      "step": 28181
+    },
+    {
+      "epoch": 0.6262666666666666,
+      "grad_norm": 2.9138686656951904,
+      "learning_rate": 7.47632807290509e-05,
+      "loss": 2.1785,
+      "step": 28182
+    },
+    {
+      "epoch": 0.6262888888888889,
+      "grad_norm": 1.8873893022537231,
+      "learning_rate": 7.475883529673261e-05,
+      "loss": 1.7869,
+      "step": 28183
+    },
+    {
+      "epoch": 0.6263111111111112,
+      "grad_norm": 1.9957176446914673,
+      "learning_rate": 7.475438986441432e-05,
+      "loss": 1.9079,
+      "step": 28184
+    },
+    {
+      "epoch": 0.6263333333333333,
+      "grad_norm": 1.9087276458740234,
+      "learning_rate": 7.474994443209602e-05,
+      "loss": 2.3609,
+      "step": 28185
+    },
+    {
+      "epoch": 0.6263555555555556,
+      "grad_norm": 1.6395312547683716,
+      "learning_rate": 7.474549899977773e-05,
+      "loss": 1.5422,
+      "step": 28186
+    },
+    {
+      "epoch": 0.6263777777777778,
+      "grad_norm": 1.694113850593567,
+      "learning_rate": 7.474105356745944e-05,
+      "loss": 1.4948,
+      "step": 28187
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 2.090855121612549,
+      "learning_rate": 7.473660813514115e-05,
+      "loss": 2.3313,
+      "step": 28188
+    },
+    {
+      "epoch": 0.6264222222222222,
+      "grad_norm": 1.4489983320236206,
+      "learning_rate": 7.473216270282286e-05,
+      "loss": 1.4332,
+      "step": 28189
+    },
+    {
+      "epoch": 0.6264444444444445,
+      "grad_norm": 1.713574767112732,
+      "learning_rate": 7.472771727050457e-05,
+      "loss": 1.8765,
+      "step": 28190
+    },
+    {
+      "epoch": 0.6264666666666666,
+      "grad_norm": 1.9305528402328491,
+      "learning_rate": 7.472327183818627e-05,
+      "loss": 1.434,
+      "step": 28191
+    },
+    {
+      "epoch": 0.6264888888888889,
+      "grad_norm": 2.201230764389038,
+      "learning_rate": 7.471882640586797e-05,
+      "loss": 1.412,
+      "step": 28192
+    },
+    {
+      "epoch": 0.6265111111111111,
+      "grad_norm": 1.8815728425979614,
+      "learning_rate": 7.471438097354968e-05,
+      "loss": 1.5491,
+      "step": 28193
+    },
+    {
+      "epoch": 0.6265333333333334,
+      "grad_norm": 1.3930691480636597,
+      "learning_rate": 7.470993554123139e-05,
+      "loss": 1.5035,
+      "step": 28194
+    },
+    {
+      "epoch": 0.6265555555555555,
+      "grad_norm": 1.7491741180419922,
+      "learning_rate": 7.470549010891309e-05,
+      "loss": 1.6875,
+      "step": 28195
+    },
+    {
+      "epoch": 0.6265777777777778,
+      "grad_norm": 2.281614303588867,
+      "learning_rate": 7.47010446765948e-05,
+      "loss": 2.1702,
+      "step": 28196
+    },
+    {
+      "epoch": 0.6266,
+      "grad_norm": 1.478087306022644,
+      "learning_rate": 7.46965992442765e-05,
+      "loss": 1.4802,
+      "step": 28197
+    },
+    {
+      "epoch": 0.6266222222222222,
+      "grad_norm": 2.5004618167877197,
+      "learning_rate": 7.469215381195822e-05,
+      "loss": 1.9388,
+      "step": 28198
+    },
+    {
+      "epoch": 0.6266444444444444,
+      "grad_norm": 3.1494030952453613,
+      "learning_rate": 7.468770837963992e-05,
+      "loss": 2.192,
+      "step": 28199
+    },
+    {
+      "epoch": 0.6266666666666667,
+      "grad_norm": 1.9602372646331787,
+      "learning_rate": 7.468326294732163e-05,
+      "loss": 1.2015,
+      "step": 28200
+    },
+    {
+      "epoch": 0.6266888888888889,
+      "grad_norm": 1.803371548652649,
+      "learning_rate": 7.467881751500334e-05,
+      "loss": 2.1237,
+      "step": 28201
+    },
+    {
+      "epoch": 0.6267111111111111,
+      "grad_norm": 1.682560682296753,
+      "learning_rate": 7.467437208268504e-05,
+      "loss": 2.3588,
+      "step": 28202
+    },
+    {
+      "epoch": 0.6267333333333334,
+      "grad_norm": 1.5700569152832031,
+      "learning_rate": 7.466992665036675e-05,
+      "loss": 2.4099,
+      "step": 28203
+    },
+    {
+      "epoch": 0.6267555555555555,
+      "grad_norm": 1.806165099143982,
+      "learning_rate": 7.466548121804846e-05,
+      "loss": 2.2097,
+      "step": 28204
+    },
+    {
+      "epoch": 0.6267777777777778,
+      "grad_norm": 1.451766014099121,
+      "learning_rate": 7.466103578573017e-05,
+      "loss": 1.657,
+      "step": 28205
+    },
+    {
+      "epoch": 0.6268,
+      "grad_norm": 1.5212419033050537,
+      "learning_rate": 7.465659035341186e-05,
+      "loss": 2.3481,
+      "step": 28206
+    },
+    {
+      "epoch": 0.6268222222222222,
+      "grad_norm": 1.485065221786499,
+      "learning_rate": 7.465214492109357e-05,
+      "loss": 2.3113,
+      "step": 28207
+    },
+    {
+      "epoch": 0.6268444444444444,
+      "grad_norm": 1.5031648874282837,
+      "learning_rate": 7.46476994887753e-05,
+      "loss": 1.9426,
+      "step": 28208
+    },
+    {
+      "epoch": 0.6268666666666667,
+      "grad_norm": 1.8476406335830688,
+      "learning_rate": 7.4643254056457e-05,
+      "loss": 1.9221,
+      "step": 28209
+    },
+    {
+      "epoch": 0.6268888888888889,
+      "grad_norm": 1.5286202430725098,
+      "learning_rate": 7.46388086241387e-05,
+      "loss": 1.9807,
+      "step": 28210
+    },
+    {
+      "epoch": 0.6269111111111111,
+      "grad_norm": 1.6897189617156982,
+      "learning_rate": 7.463436319182041e-05,
+      "loss": 1.8853,
+      "step": 28211
+    },
+    {
+      "epoch": 0.6269333333333333,
+      "grad_norm": 1.8951705694198608,
+      "learning_rate": 7.462991775950211e-05,
+      "loss": 1.9416,
+      "step": 28212
+    },
+    {
+      "epoch": 0.6269555555555556,
+      "grad_norm": 1.5698425769805908,
+      "learning_rate": 7.462547232718382e-05,
+      "loss": 1.5634,
+      "step": 28213
+    },
+    {
+      "epoch": 0.6269777777777777,
+      "grad_norm": 1.63398277759552,
+      "learning_rate": 7.462102689486553e-05,
+      "loss": 2.0327,
+      "step": 28214
+    },
+    {
+      "epoch": 0.627,
+      "grad_norm": 1.5755960941314697,
+      "learning_rate": 7.461658146254724e-05,
+      "loss": 1.6802,
+      "step": 28215
+    },
+    {
+      "epoch": 0.6270222222222223,
+      "grad_norm": 0.6174606084823608,
+      "learning_rate": 7.461213603022893e-05,
+      "loss": 0.0305,
+      "step": 28216
+    },
+    {
+      "epoch": 0.6270444444444444,
+      "grad_norm": 1.692030429840088,
+      "learning_rate": 7.460769059791066e-05,
+      "loss": 2.2251,
+      "step": 28217
+    },
+    {
+      "epoch": 0.6270666666666667,
+      "grad_norm": 1.5224628448486328,
+      "learning_rate": 7.460324516559237e-05,
+      "loss": 1.624,
+      "step": 28218
+    },
+    {
+      "epoch": 0.6270888888888889,
+      "grad_norm": 1.9173451662063599,
+      "learning_rate": 7.459879973327406e-05,
+      "loss": 2.2157,
+      "step": 28219
+    },
+    {
+      "epoch": 0.6271111111111111,
+      "grad_norm": 1.9144426584243774,
+      "learning_rate": 7.459435430095577e-05,
+      "loss": 2.6556,
+      "step": 28220
+    },
+    {
+      "epoch": 0.6271333333333333,
+      "grad_norm": 1.8819726705551147,
+      "learning_rate": 7.458990886863748e-05,
+      "loss": 2.0855,
+      "step": 28221
+    },
+    {
+      "epoch": 0.6271555555555556,
+      "grad_norm": 2.056820869445801,
+      "learning_rate": 7.458546343631918e-05,
+      "loss": 2.3333,
+      "step": 28222
+    },
+    {
+      "epoch": 0.6271777777777777,
+      "grad_norm": 1.8170225620269775,
+      "learning_rate": 7.458101800400089e-05,
+      "loss": 1.8746,
+      "step": 28223
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 1.4878606796264648,
+      "learning_rate": 7.45765725716826e-05,
+      "loss": 1.4699,
+      "step": 28224
+    },
+    {
+      "epoch": 0.6272222222222222,
+      "grad_norm": 2.008486747741699,
+      "learning_rate": 7.457212713936431e-05,
+      "loss": 2.4197,
+      "step": 28225
+    },
+    {
+      "epoch": 0.6272444444444445,
+      "grad_norm": 1.7091706991195679,
+      "learning_rate": 7.456768170704602e-05,
+      "loss": 2.0768,
+      "step": 28226
+    },
+    {
+      "epoch": 0.6272666666666666,
+      "grad_norm": 1.7279984951019287,
+      "learning_rate": 7.456323627472773e-05,
+      "loss": 1.8261,
+      "step": 28227
+    },
+    {
+      "epoch": 0.6272888888888889,
+      "grad_norm": 1.9838591814041138,
+      "learning_rate": 7.455879084240944e-05,
+      "loss": 1.7549,
+      "step": 28228
+    },
+    {
+      "epoch": 0.6273111111111112,
+      "grad_norm": 1.8756340742111206,
+      "learning_rate": 7.455434541009113e-05,
+      "loss": 1.8588,
+      "step": 28229
+    },
+    {
+      "epoch": 0.6273333333333333,
+      "grad_norm": 2.0535311698913574,
+      "learning_rate": 7.454989997777284e-05,
+      "loss": 2.3669,
+      "step": 28230
+    },
+    {
+      "epoch": 0.6273555555555556,
+      "grad_norm": 1.7964155673980713,
+      "learning_rate": 7.454545454545455e-05,
+      "loss": 1.7613,
+      "step": 28231
+    },
+    {
+      "epoch": 0.6273777777777778,
+      "grad_norm": 2.2725765705108643,
+      "learning_rate": 7.454100911313625e-05,
+      "loss": 2.5016,
+      "step": 28232
+    },
+    {
+      "epoch": 0.6274,
+      "grad_norm": 1.988587498664856,
+      "learning_rate": 7.453656368081796e-05,
+      "loss": 1.9788,
+      "step": 28233
+    },
+    {
+      "epoch": 0.6274222222222222,
+      "grad_norm": 1.7059104442596436,
+      "learning_rate": 7.453211824849967e-05,
+      "loss": 1.9183,
+      "step": 28234
+    },
+    {
+      "epoch": 0.6274444444444445,
+      "grad_norm": 1.7577725648880005,
+      "learning_rate": 7.452767281618138e-05,
+      "loss": 2.0483,
+      "step": 28235
+    },
+    {
+      "epoch": 0.6274666666666666,
+      "grad_norm": 1.6931707859039307,
+      "learning_rate": 7.452322738386309e-05,
+      "loss": 1.6994,
+      "step": 28236
+    },
+    {
+      "epoch": 0.6274888888888889,
+      "grad_norm": 1.6592637300491333,
+      "learning_rate": 7.45187819515448e-05,
+      "loss": 1.5483,
+      "step": 28237
+    },
+    {
+      "epoch": 0.6275111111111111,
+      "grad_norm": 1.3451963663101196,
+      "learning_rate": 7.45143365192265e-05,
+      "loss": 1.0713,
+      "step": 28238
+    },
+    {
+      "epoch": 0.6275333333333334,
+      "grad_norm": 1.2414436340332031,
+      "learning_rate": 7.45098910869082e-05,
+      "loss": 0.9416,
+      "step": 28239
+    },
+    {
+      "epoch": 0.6275555555555555,
+      "grad_norm": 1.9038292169570923,
+      "learning_rate": 7.450544565458991e-05,
+      "loss": 1.8877,
+      "step": 28240
+    },
+    {
+      "epoch": 0.6275777777777778,
+      "grad_norm": 1.881089210510254,
+      "learning_rate": 7.450100022227162e-05,
+      "loss": 1.6985,
+      "step": 28241
+    },
+    {
+      "epoch": 0.6276,
+      "grad_norm": 1.404329776763916,
+      "learning_rate": 7.449655478995332e-05,
+      "loss": 1.4275,
+      "step": 28242
+    },
+    {
+      "epoch": 0.6276222222222222,
+      "grad_norm": 1.655401349067688,
+      "learning_rate": 7.449210935763503e-05,
+      "loss": 2.0994,
+      "step": 28243
+    },
+    {
+      "epoch": 0.6276444444444444,
+      "grad_norm": 2.3409383296966553,
+      "learning_rate": 7.448766392531674e-05,
+      "loss": 2.3209,
+      "step": 28244
+    },
+    {
+      "epoch": 0.6276666666666667,
+      "grad_norm": 1.8269544839859009,
+      "learning_rate": 7.448321849299846e-05,
+      "loss": 1.9272,
+      "step": 28245
+    },
+    {
+      "epoch": 0.6276888888888889,
+      "grad_norm": 1.9554951190948486,
+      "learning_rate": 7.447877306068015e-05,
+      "loss": 2.1167,
+      "step": 28246
+    },
+    {
+      "epoch": 0.6277111111111111,
+      "grad_norm": 1.4238810539245605,
+      "learning_rate": 7.447432762836186e-05,
+      "loss": 1.0804,
+      "step": 28247
+    },
+    {
+      "epoch": 0.6277333333333334,
+      "grad_norm": 1.9184352159500122,
+      "learning_rate": 7.446988219604357e-05,
+      "loss": 1.945,
+      "step": 28248
+    },
+    {
+      "epoch": 0.6277555555555555,
+      "grad_norm": 1.8719063997268677,
+      "learning_rate": 7.446543676372527e-05,
+      "loss": 1.7791,
+      "step": 28249
+    },
+    {
+      "epoch": 0.6277777777777778,
+      "grad_norm": 1.7169616222381592,
+      "learning_rate": 7.446099133140698e-05,
+      "loss": 0.9318,
+      "step": 28250
+    },
+    {
+      "epoch": 0.6278,
+      "grad_norm": 1.0853959321975708,
+      "learning_rate": 7.445654589908869e-05,
+      "loss": 0.9651,
+      "step": 28251
+    },
+    {
+      "epoch": 0.6278222222222222,
+      "grad_norm": 1.5064071416854858,
+      "learning_rate": 7.44521004667704e-05,
+      "loss": 2.7501,
+      "step": 28252
+    },
+    {
+      "epoch": 0.6278444444444444,
+      "grad_norm": 1.6041384935379028,
+      "learning_rate": 7.44476550344521e-05,
+      "loss": 2.051,
+      "step": 28253
+    },
+    {
+      "epoch": 0.6278666666666667,
+      "grad_norm": 1.5042846202850342,
+      "learning_rate": 7.444320960213382e-05,
+      "loss": 2.0497,
+      "step": 28254
+    },
+    {
+      "epoch": 0.6278888888888889,
+      "grad_norm": 1.3406095504760742,
+      "learning_rate": 7.443876416981553e-05,
+      "loss": 2.2745,
+      "step": 28255
+    },
+    {
+      "epoch": 0.6279111111111111,
+      "grad_norm": 1.767425537109375,
+      "learning_rate": 7.443431873749722e-05,
+      "loss": 2.3426,
+      "step": 28256
+    },
+    {
+      "epoch": 0.6279333333333333,
+      "grad_norm": 1.8688346147537231,
+      "learning_rate": 7.442987330517893e-05,
+      "loss": 2.042,
+      "step": 28257
+    },
+    {
+      "epoch": 0.6279555555555556,
+      "grad_norm": 2.109346628189087,
+      "learning_rate": 7.442542787286064e-05,
+      "loss": 2.2495,
+      "step": 28258
+    },
+    {
+      "epoch": 0.6279777777777777,
+      "grad_norm": 2.2193105220794678,
+      "learning_rate": 7.442098244054234e-05,
+      "loss": 2.1782,
+      "step": 28259
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 1.6535050868988037,
+      "learning_rate": 7.441653700822405e-05,
+      "loss": 2.3659,
+      "step": 28260
+    },
+    {
+      "epoch": 0.6280222222222223,
+      "grad_norm": 0.8908278346061707,
+      "learning_rate": 7.441209157590576e-05,
+      "loss": 0.6832,
+      "step": 28261
+    },
+    {
+      "epoch": 0.6280444444444444,
+      "grad_norm": 1.612201452255249,
+      "learning_rate": 7.440764614358747e-05,
+      "loss": 2.3098,
+      "step": 28262
+    },
+    {
+      "epoch": 0.6280666666666667,
+      "grad_norm": 1.5728148221969604,
+      "learning_rate": 7.440320071126918e-05,
+      "loss": 1.993,
+      "step": 28263
+    },
+    {
+      "epoch": 0.6280888888888889,
+      "grad_norm": 1.5704433917999268,
+      "learning_rate": 7.439875527895089e-05,
+      "loss": 2.0002,
+      "step": 28264
+    },
+    {
+      "epoch": 0.6281111111111111,
+      "grad_norm": 1.6616250276565552,
+      "learning_rate": 7.43943098466326e-05,
+      "loss": 2.2513,
+      "step": 28265
+    },
+    {
+      "epoch": 0.6281333333333333,
+      "grad_norm": 1.6211087703704834,
+      "learning_rate": 7.438986441431429e-05,
+      "loss": 1.7105,
+      "step": 28266
+    },
+    {
+      "epoch": 0.6281555555555556,
+      "grad_norm": 1.5077217817306519,
+      "learning_rate": 7.4385418981996e-05,
+      "loss": 1.9318,
+      "step": 28267
+    },
+    {
+      "epoch": 0.6281777777777777,
+      "grad_norm": 1.5681517124176025,
+      "learning_rate": 7.438097354967771e-05,
+      "loss": 1.7493,
+      "step": 28268
+    },
+    {
+      "epoch": 0.6282,
+      "grad_norm": 1.7847591638565063,
+      "learning_rate": 7.437652811735941e-05,
+      "loss": 2.3265,
+      "step": 28269
+    },
+    {
+      "epoch": 0.6282222222222222,
+      "grad_norm": 1.3881011009216309,
+      "learning_rate": 7.437208268504112e-05,
+      "loss": 1.7574,
+      "step": 28270
+    },
+    {
+      "epoch": 0.6282444444444445,
+      "grad_norm": 1.2813900709152222,
+      "learning_rate": 7.436763725272283e-05,
+      "loss": 1.5656,
+      "step": 28271
+    },
+    {
+      "epoch": 0.6282666666666666,
+      "grad_norm": 1.3996785879135132,
+      "learning_rate": 7.436319182040454e-05,
+      "loss": 1.1872,
+      "step": 28272
+    },
+    {
+      "epoch": 0.6282888888888889,
+      "grad_norm": 1.6524691581726074,
+      "learning_rate": 7.435874638808625e-05,
+      "loss": 1.8171,
+      "step": 28273
+    },
+    {
+      "epoch": 0.6283111111111112,
+      "grad_norm": 1.781345009803772,
+      "learning_rate": 7.435430095576796e-05,
+      "loss": 2.0514,
+      "step": 28274
+    },
+    {
+      "epoch": 0.6283333333333333,
+      "grad_norm": 1.6943475008010864,
+      "learning_rate": 7.434985552344967e-05,
+      "loss": 2.1368,
+      "step": 28275
+    },
+    {
+      "epoch": 0.6283555555555556,
+      "grad_norm": 1.695285439491272,
+      "learning_rate": 7.434541009113136e-05,
+      "loss": 1.9909,
+      "step": 28276
+    },
+    {
+      "epoch": 0.6283777777777778,
+      "grad_norm": 1.768095850944519,
+      "learning_rate": 7.434096465881307e-05,
+      "loss": 2.0456,
+      "step": 28277
+    },
+    {
+      "epoch": 0.6284,
+      "grad_norm": 1.7602506875991821,
+      "learning_rate": 7.433651922649478e-05,
+      "loss": 2.0481,
+      "step": 28278
+    },
+    {
+      "epoch": 0.6284222222222222,
+      "grad_norm": 1.5751560926437378,
+      "learning_rate": 7.433207379417648e-05,
+      "loss": 1.7103,
+      "step": 28279
+    },
+    {
+      "epoch": 0.6284444444444445,
+      "grad_norm": 1.458869218826294,
+      "learning_rate": 7.432762836185819e-05,
+      "loss": 1.6771,
+      "step": 28280
+    },
+    {
+      "epoch": 0.6284666666666666,
+      "grad_norm": 1.6494125127792358,
+      "learning_rate": 7.43231829295399e-05,
+      "loss": 1.7331,
+      "step": 28281
+    },
+    {
+      "epoch": 0.6284888888888889,
+      "grad_norm": 1.9761335849761963,
+      "learning_rate": 7.431873749722162e-05,
+      "loss": 2.1751,
+      "step": 28282
+    },
+    {
+      "epoch": 0.6285111111111111,
+      "grad_norm": 1.5494686365127563,
+      "learning_rate": 7.431429206490332e-05,
+      "loss": 1.9114,
+      "step": 28283
+    },
+    {
+      "epoch": 0.6285333333333334,
+      "grad_norm": 1.946640133857727,
+      "learning_rate": 7.430984663258503e-05,
+      "loss": 1.9682,
+      "step": 28284
+    },
+    {
+      "epoch": 0.6285555555555555,
+      "grad_norm": 1.6203198432922363,
+      "learning_rate": 7.430540120026674e-05,
+      "loss": 1.9045,
+      "step": 28285
+    },
+    {
+      "epoch": 0.6285777777777778,
+      "grad_norm": 1.078140377998352,
+      "learning_rate": 7.430095576794843e-05,
+      "loss": 0.8918,
+      "step": 28286
+    },
+    {
+      "epoch": 0.6286,
+      "grad_norm": 1.5796608924865723,
+      "learning_rate": 7.429651033563014e-05,
+      "loss": 1.6461,
+      "step": 28287
+    },
+    {
+      "epoch": 0.6286222222222222,
+      "grad_norm": 2.0373597145080566,
+      "learning_rate": 7.429206490331185e-05,
+      "loss": 2.0898,
+      "step": 28288
+    },
+    {
+      "epoch": 0.6286444444444445,
+      "grad_norm": 1.6334196329116821,
+      "learning_rate": 7.428761947099355e-05,
+      "loss": 1.8604,
+      "step": 28289
+    },
+    {
+      "epoch": 0.6286666666666667,
+      "grad_norm": 1.8665622472763062,
+      "learning_rate": 7.428317403867526e-05,
+      "loss": 1.7337,
+      "step": 28290
+    },
+    {
+      "epoch": 0.6286888888888889,
+      "grad_norm": 1.641268014907837,
+      "learning_rate": 7.427872860635698e-05,
+      "loss": 2.0156,
+      "step": 28291
+    },
+    {
+      "epoch": 0.6287111111111111,
+      "grad_norm": 1.8151181936264038,
+      "learning_rate": 7.427428317403869e-05,
+      "loss": 2.0465,
+      "step": 28292
+    },
+    {
+      "epoch": 0.6287333333333334,
+      "grad_norm": 2.0497701168060303,
+      "learning_rate": 7.426983774172038e-05,
+      "loss": 2.0673,
+      "step": 28293
+    },
+    {
+      "epoch": 0.6287555555555555,
+      "grad_norm": 1.5897787809371948,
+      "learning_rate": 7.42653923094021e-05,
+      "loss": 1.8692,
+      "step": 28294
+    },
+    {
+      "epoch": 0.6287777777777778,
+      "grad_norm": 1.913669228553772,
+      "learning_rate": 7.42609468770838e-05,
+      "loss": 1.6572,
+      "step": 28295
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 1.6216816902160645,
+      "learning_rate": 7.42565014447655e-05,
+      "loss": 1.7076,
+      "step": 28296
+    },
+    {
+      "epoch": 0.6288222222222222,
+      "grad_norm": 1.8761065006256104,
+      "learning_rate": 7.425205601244721e-05,
+      "loss": 1.7828,
+      "step": 28297
+    },
+    {
+      "epoch": 0.6288444444444444,
+      "grad_norm": 1.8865710496902466,
+      "learning_rate": 7.424761058012892e-05,
+      "loss": 1.6179,
+      "step": 28298
+    },
+    {
+      "epoch": 0.6288666666666667,
+      "grad_norm": 2.1235756874084473,
+      "learning_rate": 7.424316514781063e-05,
+      "loss": 2.0475,
+      "step": 28299
+    },
+    {
+      "epoch": 0.6288888888888889,
+      "grad_norm": 2.2243402004241943,
+      "learning_rate": 7.423871971549234e-05,
+      "loss": 1.7504,
+      "step": 28300
+    },
+    {
+      "epoch": 0.6289111111111111,
+      "grad_norm": 1.769464373588562,
+      "learning_rate": 7.423427428317405e-05,
+      "loss": 2.6118,
+      "step": 28301
+    },
+    {
+      "epoch": 0.6289333333333333,
+      "grad_norm": 1.8621481657028198,
+      "learning_rate": 7.422982885085576e-05,
+      "loss": 1.1403,
+      "step": 28302
+    },
+    {
+      "epoch": 0.6289555555555556,
+      "grad_norm": 1.5293608903884888,
+      "learning_rate": 7.422538341853745e-05,
+      "loss": 2.2813,
+      "step": 28303
+    },
+    {
+      "epoch": 0.6289777777777777,
+      "grad_norm": 1.328939437866211,
+      "learning_rate": 7.422093798621916e-05,
+      "loss": 1.8391,
+      "step": 28304
+    },
+    {
+      "epoch": 0.629,
+      "grad_norm": 1.330453634262085,
+      "learning_rate": 7.421649255390087e-05,
+      "loss": 2.2479,
+      "step": 28305
+    },
+    {
+      "epoch": 0.6290222222222223,
+      "grad_norm": 1.6832678318023682,
+      "learning_rate": 7.421204712158257e-05,
+      "loss": 2.405,
+      "step": 28306
+    },
+    {
+      "epoch": 0.6290444444444444,
+      "grad_norm": 1.228036642074585,
+      "learning_rate": 7.420760168926428e-05,
+      "loss": 1.966,
+      "step": 28307
+    },
+    {
+      "epoch": 0.6290666666666667,
+      "grad_norm": 1.3511638641357422,
+      "learning_rate": 7.420315625694599e-05,
+      "loss": 2.014,
+      "step": 28308
+    },
+    {
+      "epoch": 0.6290888888888889,
+      "grad_norm": 1.5867738723754883,
+      "learning_rate": 7.41987108246277e-05,
+      "loss": 2.1594,
+      "step": 28309
+    },
+    {
+      "epoch": 0.6291111111111111,
+      "grad_norm": 1.6926277875900269,
+      "learning_rate": 7.419426539230941e-05,
+      "loss": 2.2151,
+      "step": 28310
+    },
+    {
+      "epoch": 0.6291333333333333,
+      "grad_norm": 2.021162271499634,
+      "learning_rate": 7.418981995999112e-05,
+      "loss": 2.1444,
+      "step": 28311
+    },
+    {
+      "epoch": 0.6291555555555556,
+      "grad_norm": 2.1428580284118652,
+      "learning_rate": 7.418537452767283e-05,
+      "loss": 2.575,
+      "step": 28312
+    },
+    {
+      "epoch": 0.6291777777777777,
+      "grad_norm": 1.5130378007888794,
+      "learning_rate": 7.418092909535452e-05,
+      "loss": 1.8039,
+      "step": 28313
+    },
+    {
+      "epoch": 0.6292,
+      "grad_norm": 1.5649083852767944,
+      "learning_rate": 7.417648366303623e-05,
+      "loss": 1.6873,
+      "step": 28314
+    },
+    {
+      "epoch": 0.6292222222222222,
+      "grad_norm": 1.5888735055923462,
+      "learning_rate": 7.417203823071794e-05,
+      "loss": 2.0904,
+      "step": 28315
+    },
+    {
+      "epoch": 0.6292444444444445,
+      "grad_norm": 1.8082695007324219,
+      "learning_rate": 7.416759279839964e-05,
+      "loss": 2.3681,
+      "step": 28316
+    },
+    {
+      "epoch": 0.6292666666666666,
+      "grad_norm": 1.7941712141036987,
+      "learning_rate": 7.416314736608135e-05,
+      "loss": 2.2749,
+      "step": 28317
+    },
+    {
+      "epoch": 0.6292888888888889,
+      "grad_norm": 1.4473190307617188,
+      "learning_rate": 7.415870193376306e-05,
+      "loss": 1.7183,
+      "step": 28318
+    },
+    {
+      "epoch": 0.6293111111111112,
+      "grad_norm": 1.321277141571045,
+      "learning_rate": 7.415425650144478e-05,
+      "loss": 1.3812,
+      "step": 28319
+    },
+    {
+      "epoch": 0.6293333333333333,
+      "grad_norm": 2.0893523693084717,
+      "learning_rate": 7.414981106912648e-05,
+      "loss": 2.7359,
+      "step": 28320
+    },
+    {
+      "epoch": 0.6293555555555556,
+      "grad_norm": 1.4881926774978638,
+      "learning_rate": 7.414536563680819e-05,
+      "loss": 1.6233,
+      "step": 28321
+    },
+    {
+      "epoch": 0.6293777777777778,
+      "grad_norm": 1.9817962646484375,
+      "learning_rate": 7.41409202044899e-05,
+      "loss": 2.1383,
+      "step": 28322
+    },
+    {
+      "epoch": 0.6294,
+      "grad_norm": 1.7774596214294434,
+      "learning_rate": 7.413647477217159e-05,
+      "loss": 2.2503,
+      "step": 28323
+    },
+    {
+      "epoch": 0.6294222222222222,
+      "grad_norm": 1.7046529054641724,
+      "learning_rate": 7.41320293398533e-05,
+      "loss": 2.0507,
+      "step": 28324
+    },
+    {
+      "epoch": 0.6294444444444445,
+      "grad_norm": 1.4740560054779053,
+      "learning_rate": 7.412758390753501e-05,
+      "loss": 1.6524,
+      "step": 28325
+    },
+    {
+      "epoch": 0.6294666666666666,
+      "grad_norm": 1.5283254384994507,
+      "learning_rate": 7.412313847521671e-05,
+      "loss": 2.2177,
+      "step": 28326
+    },
+    {
+      "epoch": 0.6294888888888889,
+      "grad_norm": 1.4951704740524292,
+      "learning_rate": 7.411869304289842e-05,
+      "loss": 1.9466,
+      "step": 28327
+    },
+    {
+      "epoch": 0.6295111111111111,
+      "grad_norm": 1.6033143997192383,
+      "learning_rate": 7.411424761058014e-05,
+      "loss": 1.9761,
+      "step": 28328
+    },
+    {
+      "epoch": 0.6295333333333333,
+      "grad_norm": 1.7867376804351807,
+      "learning_rate": 7.410980217826185e-05,
+      "loss": 1.8702,
+      "step": 28329
+    },
+    {
+      "epoch": 0.6295555555555555,
+      "grad_norm": 1.4749178886413574,
+      "learning_rate": 7.410535674594355e-05,
+      "loss": 1.5522,
+      "step": 28330
+    },
+    {
+      "epoch": 0.6295777777777778,
+      "grad_norm": 1.5684312582015991,
+      "learning_rate": 7.410091131362526e-05,
+      "loss": 1.2988,
+      "step": 28331
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 1.4363404512405396,
+      "learning_rate": 7.409646588130697e-05,
+      "loss": 1.7482,
+      "step": 28332
+    },
+    {
+      "epoch": 0.6296222222222222,
+      "grad_norm": 1.8172701597213745,
+      "learning_rate": 7.409202044898866e-05,
+      "loss": 1.7153,
+      "step": 28333
+    },
+    {
+      "epoch": 0.6296444444444445,
+      "grad_norm": 1.687948226928711,
+      "learning_rate": 7.408757501667037e-05,
+      "loss": 1.9432,
+      "step": 28334
+    },
+    {
+      "epoch": 0.6296666666666667,
+      "grad_norm": 1.816715955734253,
+      "learning_rate": 7.408312958435208e-05,
+      "loss": 1.9648,
+      "step": 28335
+    },
+    {
+      "epoch": 0.6296888888888889,
+      "grad_norm": 1.7293776273727417,
+      "learning_rate": 7.407868415203379e-05,
+      "loss": 1.9419,
+      "step": 28336
+    },
+    {
+      "epoch": 0.6297111111111111,
+      "grad_norm": 1.5770736932754517,
+      "learning_rate": 7.40742387197155e-05,
+      "loss": 1.8593,
+      "step": 28337
+    },
+    {
+      "epoch": 0.6297333333333334,
+      "grad_norm": 1.74238920211792,
+      "learning_rate": 7.406979328739721e-05,
+      "loss": 2.2409,
+      "step": 28338
+    },
+    {
+      "epoch": 0.6297555555555555,
+      "grad_norm": 1.8684200048446655,
+      "learning_rate": 7.406534785507892e-05,
+      "loss": 1.859,
+      "step": 28339
+    },
+    {
+      "epoch": 0.6297777777777778,
+      "grad_norm": 1.713669776916504,
+      "learning_rate": 7.406090242276062e-05,
+      "loss": 1.7262,
+      "step": 28340
+    },
+    {
+      "epoch": 0.6298,
+      "grad_norm": 1.7168631553649902,
+      "learning_rate": 7.405645699044232e-05,
+      "loss": 2.173,
+      "step": 28341
+    },
+    {
+      "epoch": 0.6298222222222222,
+      "grad_norm": 1.9766662120819092,
+      "learning_rate": 7.405201155812403e-05,
+      "loss": 1.9445,
+      "step": 28342
+    },
+    {
+      "epoch": 0.6298444444444444,
+      "grad_norm": 1.9945155382156372,
+      "learning_rate": 7.404756612580573e-05,
+      "loss": 1.865,
+      "step": 28343
+    },
+    {
+      "epoch": 0.6298666666666667,
+      "grad_norm": 1.841382622718811,
+      "learning_rate": 7.404312069348744e-05,
+      "loss": 1.8165,
+      "step": 28344
+    },
+    {
+      "epoch": 0.6298888888888889,
+      "grad_norm": 1.9434353113174438,
+      "learning_rate": 7.403867526116915e-05,
+      "loss": 1.7789,
+      "step": 28345
+    },
+    {
+      "epoch": 0.6299111111111111,
+      "grad_norm": 2.029454469680786,
+      "learning_rate": 7.403422982885086e-05,
+      "loss": 2.1717,
+      "step": 28346
+    },
+    {
+      "epoch": 0.6299333333333333,
+      "grad_norm": 2.521643877029419,
+      "learning_rate": 7.402978439653257e-05,
+      "loss": 2.3886,
+      "step": 28347
+    },
+    {
+      "epoch": 0.6299555555555556,
+      "grad_norm": 1.5658351182937622,
+      "learning_rate": 7.402533896421428e-05,
+      "loss": 1.7562,
+      "step": 28348
+    },
+    {
+      "epoch": 0.6299777777777777,
+      "grad_norm": 1.8885512351989746,
+      "learning_rate": 7.402089353189599e-05,
+      "loss": 1.4809,
+      "step": 28349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3135809898376465,
+      "learning_rate": 7.401644809957768e-05,
+      "loss": 0.8233,
+      "step": 28350
+    },
+    {
+      "epoch": 0.6300222222222223,
+      "grad_norm": 1.5412646532058716,
+      "learning_rate": 7.40120026672594e-05,
+      "loss": 2.08,
+      "step": 28351
+    },
+    {
+      "epoch": 0.6300444444444444,
+      "grad_norm": 0.986265242099762,
+      "learning_rate": 7.40075572349411e-05,
+      "loss": 0.8639,
+      "step": 28352
+    },
+    {
+      "epoch": 0.6300666666666667,
+      "grad_norm": 7.895050048828125,
+      "learning_rate": 7.40031118026228e-05,
+      "loss": 2.2206,
+      "step": 28353
+    },
+    {
+      "epoch": 0.6300888888888889,
+      "grad_norm": 1.8083707094192505,
+      "learning_rate": 7.399866637030451e-05,
+      "loss": 2.5821,
+      "step": 28354
+    },
+    {
+      "epoch": 0.6301111111111111,
+      "grad_norm": 1.620503544807434,
+      "learning_rate": 7.399422093798622e-05,
+      "loss": 2.2363,
+      "step": 28355
+    },
+    {
+      "epoch": 0.6301333333333333,
+      "grad_norm": 1.4891921281814575,
+      "learning_rate": 7.398977550566793e-05,
+      "loss": 1.24,
+      "step": 28356
+    },
+    {
+      "epoch": 0.6301555555555556,
+      "grad_norm": 1.437220573425293,
+      "learning_rate": 7.398533007334964e-05,
+      "loss": 2.2225,
+      "step": 28357
+    },
+    {
+      "epoch": 0.6301777777777777,
+      "grad_norm": 1.757455587387085,
+      "learning_rate": 7.398088464103135e-05,
+      "loss": 2.4739,
+      "step": 28358
+    },
+    {
+      "epoch": 0.6302,
+      "grad_norm": 1.5432251691818237,
+      "learning_rate": 7.397643920871306e-05,
+      "loss": 1.9358,
+      "step": 28359
+    },
+    {
+      "epoch": 0.6302222222222222,
+      "grad_norm": 1.6928318738937378,
+      "learning_rate": 7.397199377639475e-05,
+      "loss": 2.0875,
+      "step": 28360
+    },
+    {
+      "epoch": 0.6302444444444445,
+      "grad_norm": 1.3686691522598267,
+      "learning_rate": 7.396754834407646e-05,
+      "loss": 1.8747,
+      "step": 28361
+    },
+    {
+      "epoch": 0.6302666666666666,
+      "grad_norm": 1.6149375438690186,
+      "learning_rate": 7.396310291175817e-05,
+      "loss": 2.1837,
+      "step": 28362
+    },
+    {
+      "epoch": 0.6302888888888889,
+      "grad_norm": 1.6666003465652466,
+      "learning_rate": 7.395865747943987e-05,
+      "loss": 2.0275,
+      "step": 28363
+    },
+    {
+      "epoch": 0.6303111111111112,
+      "grad_norm": 1.603553295135498,
+      "learning_rate": 7.395421204712158e-05,
+      "loss": 2.3148,
+      "step": 28364
+    },
+    {
+      "epoch": 0.6303333333333333,
+      "grad_norm": 1.6051143407821655,
+      "learning_rate": 7.39497666148033e-05,
+      "loss": 2.2403,
+      "step": 28365
+    },
+    {
+      "epoch": 0.6303555555555556,
+      "grad_norm": 1.8546210527420044,
+      "learning_rate": 7.394532118248501e-05,
+      "loss": 2.2779,
+      "step": 28366
+    },
+    {
+      "epoch": 0.6303777777777778,
+      "grad_norm": 1.7641934156417847,
+      "learning_rate": 7.394087575016671e-05,
+      "loss": 2.2434,
+      "step": 28367
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 1.4324891567230225,
+      "learning_rate": 7.393643031784842e-05,
+      "loss": 1.8032,
+      "step": 28368
+    },
+    {
+      "epoch": 0.6304222222222222,
+      "grad_norm": 1.5796635150909424,
+      "learning_rate": 7.393198488553013e-05,
+      "loss": 1.8798,
+      "step": 28369
+    },
+    {
+      "epoch": 0.6304444444444445,
+      "grad_norm": 1.614767074584961,
+      "learning_rate": 7.392753945321182e-05,
+      "loss": 1.9635,
+      "step": 28370
+    },
+    {
+      "epoch": 0.6304666666666666,
+      "grad_norm": 1.954885482788086,
+      "learning_rate": 7.392309402089353e-05,
+      "loss": 2.3244,
+      "step": 28371
+    },
+    {
+      "epoch": 0.6304888888888889,
+      "grad_norm": 1.6378158330917358,
+      "learning_rate": 7.391864858857524e-05,
+      "loss": 2.1,
+      "step": 28372
+    },
+    {
+      "epoch": 0.6305111111111111,
+      "grad_norm": 1.7912436723709106,
+      "learning_rate": 7.391420315625695e-05,
+      "loss": 2.1831,
+      "step": 28373
+    },
+    {
+      "epoch": 0.6305333333333333,
+      "grad_norm": 1.5908149480819702,
+      "learning_rate": 7.390975772393866e-05,
+      "loss": 1.5954,
+      "step": 28374
+    },
+    {
+      "epoch": 0.6305555555555555,
+      "grad_norm": 1.681992769241333,
+      "learning_rate": 7.390531229162037e-05,
+      "loss": 2.1747,
+      "step": 28375
+    },
+    {
+      "epoch": 0.6305777777777778,
+      "grad_norm": 1.6367912292480469,
+      "learning_rate": 7.390086685930208e-05,
+      "loss": 2.3634,
+      "step": 28376
+    },
+    {
+      "epoch": 0.6306,
+      "grad_norm": 1.787254810333252,
+      "learning_rate": 7.389642142698378e-05,
+      "loss": 2.2269,
+      "step": 28377
+    },
+    {
+      "epoch": 0.6306222222222222,
+      "grad_norm": 1.8347268104553223,
+      "learning_rate": 7.389197599466549e-05,
+      "loss": 1.8225,
+      "step": 28378
+    },
+    {
+      "epoch": 0.6306444444444445,
+      "grad_norm": 1.7412904500961304,
+      "learning_rate": 7.38875305623472e-05,
+      "loss": 1.6935,
+      "step": 28379
+    },
+    {
+      "epoch": 0.6306666666666667,
+      "grad_norm": 1.8452202081680298,
+      "learning_rate": 7.388308513002889e-05,
+      "loss": 1.8675,
+      "step": 28380
+    },
+    {
+      "epoch": 0.6306888888888889,
+      "grad_norm": 1.8450442552566528,
+      "learning_rate": 7.38786396977106e-05,
+      "loss": 2.0222,
+      "step": 28381
+    },
+    {
+      "epoch": 0.6307111111111111,
+      "grad_norm": 1.9819624423980713,
+      "learning_rate": 7.387419426539231e-05,
+      "loss": 1.5903,
+      "step": 28382
+    },
+    {
+      "epoch": 0.6307333333333334,
+      "grad_norm": 2.1070683002471924,
+      "learning_rate": 7.386974883307402e-05,
+      "loss": 2.2376,
+      "step": 28383
+    },
+    {
+      "epoch": 0.6307555555555555,
+      "grad_norm": 2.4826745986938477,
+      "learning_rate": 7.386530340075573e-05,
+      "loss": 2.0857,
+      "step": 28384
+    },
+    {
+      "epoch": 0.6307777777777778,
+      "grad_norm": 1.4686522483825684,
+      "learning_rate": 7.386085796843744e-05,
+      "loss": 1.6862,
+      "step": 28385
+    },
+    {
+      "epoch": 0.6308,
+      "grad_norm": 1.6939491033554077,
+      "learning_rate": 7.385641253611915e-05,
+      "loss": 1.5444,
+      "step": 28386
+    },
+    {
+      "epoch": 0.6308222222222222,
+      "grad_norm": 1.892615556716919,
+      "learning_rate": 7.385196710380085e-05,
+      "loss": 1.6596,
+      "step": 28387
+    },
+    {
+      "epoch": 0.6308444444444444,
+      "grad_norm": 1.5864721536636353,
+      "learning_rate": 7.384752167148255e-05,
+      "loss": 1.8505,
+      "step": 28388
+    },
+    {
+      "epoch": 0.6308666666666667,
+      "grad_norm": 1.7852271795272827,
+      "learning_rate": 7.384307623916426e-05,
+      "loss": 2.1614,
+      "step": 28389
+    },
+    {
+      "epoch": 0.6308888888888889,
+      "grad_norm": 2.2662484645843506,
+      "learning_rate": 7.383863080684596e-05,
+      "loss": 1.9352,
+      "step": 28390
+    },
+    {
+      "epoch": 0.6309111111111111,
+      "grad_norm": 1.8422645330429077,
+      "learning_rate": 7.383418537452767e-05,
+      "loss": 1.5843,
+      "step": 28391
+    },
+    {
+      "epoch": 0.6309333333333333,
+      "grad_norm": 1.7014566659927368,
+      "learning_rate": 7.382973994220938e-05,
+      "loss": 1.9234,
+      "step": 28392
+    },
+    {
+      "epoch": 0.6309555555555556,
+      "grad_norm": 1.7305951118469238,
+      "learning_rate": 7.382529450989109e-05,
+      "loss": 2.058,
+      "step": 28393
+    },
+    {
+      "epoch": 0.6309777777777777,
+      "grad_norm": 1.942842721939087,
+      "learning_rate": 7.38208490775728e-05,
+      "loss": 2.0942,
+      "step": 28394
+    },
+    {
+      "epoch": 0.631,
+      "grad_norm": 1.5892136096954346,
+      "learning_rate": 7.381640364525451e-05,
+      "loss": 1.4916,
+      "step": 28395
+    },
+    {
+      "epoch": 0.6310222222222223,
+      "grad_norm": 1.6735446453094482,
+      "learning_rate": 7.381195821293622e-05,
+      "loss": 1.6573,
+      "step": 28396
+    },
+    {
+      "epoch": 0.6310444444444444,
+      "grad_norm": 1.4542067050933838,
+      "learning_rate": 7.380751278061791e-05,
+      "loss": 1.4651,
+      "step": 28397
+    },
+    {
+      "epoch": 0.6310666666666667,
+      "grad_norm": 1.794935941696167,
+      "learning_rate": 7.380306734829962e-05,
+      "loss": 1.508,
+      "step": 28398
+    },
+    {
+      "epoch": 0.6310888888888889,
+      "grad_norm": 1.9414783716201782,
+      "learning_rate": 7.379862191598133e-05,
+      "loss": 1.8128,
+      "step": 28399
+    },
+    {
+      "epoch": 0.6311111111111111,
+      "grad_norm": 1.4216324090957642,
+      "learning_rate": 7.379417648366303e-05,
+      "loss": 0.9064,
+      "step": 28400
+    },
+    {
+      "epoch": 0.6311333333333333,
+      "grad_norm": 1.4880794286727905,
+      "learning_rate": 7.378973105134474e-05,
+      "loss": 2.2743,
+      "step": 28401
+    },
+    {
+      "epoch": 0.6311555555555556,
+      "grad_norm": 1.5550297498703003,
+      "learning_rate": 7.378528561902646e-05,
+      "loss": 1.9959,
+      "step": 28402
+    },
+    {
+      "epoch": 0.6311777777777777,
+      "grad_norm": 1.4357762336730957,
+      "learning_rate": 7.378084018670816e-05,
+      "loss": 2.1209,
+      "step": 28403
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 1.5185269117355347,
+      "learning_rate": 7.377639475438987e-05,
+      "loss": 1.9705,
+      "step": 28404
+    },
+    {
+      "epoch": 0.6312222222222222,
+      "grad_norm": 1.398624300956726,
+      "learning_rate": 7.377194932207158e-05,
+      "loss": 1.5216,
+      "step": 28405
+    },
+    {
+      "epoch": 0.6312444444444445,
+      "grad_norm": 1.5025349855422974,
+      "learning_rate": 7.376750388975329e-05,
+      "loss": 2.089,
+      "step": 28406
+    },
+    {
+      "epoch": 0.6312666666666666,
+      "grad_norm": 1.5183087587356567,
+      "learning_rate": 7.376305845743498e-05,
+      "loss": 1.9401,
+      "step": 28407
+    },
+    {
+      "epoch": 0.6312888888888889,
+      "grad_norm": 1.4337327480316162,
+      "learning_rate": 7.375861302511669e-05,
+      "loss": 1.6506,
+      "step": 28408
+    },
+    {
+      "epoch": 0.6313111111111112,
+      "grad_norm": 1.5042972564697266,
+      "learning_rate": 7.37541675927984e-05,
+      "loss": 2.1496,
+      "step": 28409
+    },
+    {
+      "epoch": 0.6313333333333333,
+      "grad_norm": 2.031554698944092,
+      "learning_rate": 7.374972216048011e-05,
+      "loss": 2.2798,
+      "step": 28410
+    },
+    {
+      "epoch": 0.6313555555555556,
+      "grad_norm": 1.6585062742233276,
+      "learning_rate": 7.374527672816182e-05,
+      "loss": 2.2681,
+      "step": 28411
+    },
+    {
+      "epoch": 0.6313777777777778,
+      "grad_norm": 1.9968764781951904,
+      "learning_rate": 7.374083129584353e-05,
+      "loss": 2.1939,
+      "step": 28412
+    },
+    {
+      "epoch": 0.6314,
+      "grad_norm": 1.8754717111587524,
+      "learning_rate": 7.373638586352523e-05,
+      "loss": 2.2713,
+      "step": 28413
+    },
+    {
+      "epoch": 0.6314222222222222,
+      "grad_norm": 1.6223303079605103,
+      "learning_rate": 7.373194043120694e-05,
+      "loss": 2.0536,
+      "step": 28414
+    },
+    {
+      "epoch": 0.6314444444444445,
+      "grad_norm": 1.6929534673690796,
+      "learning_rate": 7.372749499888865e-05,
+      "loss": 1.9036,
+      "step": 28415
+    },
+    {
+      "epoch": 0.6314666666666666,
+      "grad_norm": 1.6671748161315918,
+      "learning_rate": 7.372304956657036e-05,
+      "loss": 1.9505,
+      "step": 28416
+    },
+    {
+      "epoch": 0.6314888888888889,
+      "grad_norm": 2.055662155151367,
+      "learning_rate": 7.371860413425205e-05,
+      "loss": 2.4632,
+      "step": 28417
+    },
+    {
+      "epoch": 0.6315111111111111,
+      "grad_norm": 1.5693776607513428,
+      "learning_rate": 7.371415870193376e-05,
+      "loss": 1.7384,
+      "step": 28418
+    },
+    {
+      "epoch": 0.6315333333333333,
+      "grad_norm": 1.3068820238113403,
+      "learning_rate": 7.370971326961547e-05,
+      "loss": 0.6118,
+      "step": 28419
+    },
+    {
+      "epoch": 0.6315555555555555,
+      "grad_norm": 1.467512845993042,
+      "learning_rate": 7.370526783729718e-05,
+      "loss": 2.0226,
+      "step": 28420
+    },
+    {
+      "epoch": 0.6315777777777778,
+      "grad_norm": 1.5475867986679077,
+      "learning_rate": 7.370082240497889e-05,
+      "loss": 1.931,
+      "step": 28421
+    },
+    {
+      "epoch": 0.6316,
+      "grad_norm": 1.5756876468658447,
+      "learning_rate": 7.36963769726606e-05,
+      "loss": 1.6972,
+      "step": 28422
+    },
+    {
+      "epoch": 0.6316222222222222,
+      "grad_norm": 1.9265605211257935,
+      "learning_rate": 7.369193154034231e-05,
+      "loss": 2.3031,
+      "step": 28423
+    },
+    {
+      "epoch": 0.6316444444444445,
+      "grad_norm": 1.5181357860565186,
+      "learning_rate": 7.3687486108024e-05,
+      "loss": 2.1034,
+      "step": 28424
+    },
+    {
+      "epoch": 0.6316666666666667,
+      "grad_norm": 1.7012661695480347,
+      "learning_rate": 7.368304067570572e-05,
+      "loss": 2.1504,
+      "step": 28425
+    },
+    {
+      "epoch": 0.6316888888888889,
+      "grad_norm": 1.4558168649673462,
+      "learning_rate": 7.367859524338743e-05,
+      "loss": 1.533,
+      "step": 28426
+    },
+    {
+      "epoch": 0.6317111111111111,
+      "grad_norm": 1.610884189605713,
+      "learning_rate": 7.367414981106912e-05,
+      "loss": 2.0492,
+      "step": 28427
+    },
+    {
+      "epoch": 0.6317333333333334,
+      "grad_norm": 1.6316362619400024,
+      "learning_rate": 7.366970437875083e-05,
+      "loss": 1.604,
+      "step": 28428
+    },
+    {
+      "epoch": 0.6317555555555555,
+      "grad_norm": 1.0601921081542969,
+      "learning_rate": 7.366525894643254e-05,
+      "loss": 0.8132,
+      "step": 28429
+    },
+    {
+      "epoch": 0.6317777777777778,
+      "grad_norm": 1.6722947359085083,
+      "learning_rate": 7.366081351411425e-05,
+      "loss": 1.6607,
+      "step": 28430
+    },
+    {
+      "epoch": 0.6318,
+      "grad_norm": 2.0034799575805664,
+      "learning_rate": 7.365636808179596e-05,
+      "loss": 1.9938,
+      "step": 28431
+    },
+    {
+      "epoch": 0.6318222222222222,
+      "grad_norm": 1.1785577535629272,
+      "learning_rate": 7.365192264947767e-05,
+      "loss": 0.8286,
+      "step": 28432
+    },
+    {
+      "epoch": 0.6318444444444444,
+      "grad_norm": 1.7210696935653687,
+      "learning_rate": 7.364747721715938e-05,
+      "loss": 1.7816,
+      "step": 28433
+    },
+    {
+      "epoch": 0.6318666666666667,
+      "grad_norm": 1.9599621295928955,
+      "learning_rate": 7.364303178484108e-05,
+      "loss": 2.3561,
+      "step": 28434
+    },
+    {
+      "epoch": 0.6318888888888889,
+      "grad_norm": 1.6815145015716553,
+      "learning_rate": 7.363858635252278e-05,
+      "loss": 1.9247,
+      "step": 28435
+    },
+    {
+      "epoch": 0.6319111111111111,
+      "grad_norm": 1.673786997795105,
+      "learning_rate": 7.36341409202045e-05,
+      "loss": 1.5815,
+      "step": 28436
+    },
+    {
+      "epoch": 0.6319333333333333,
+      "grad_norm": 1.7454754114151,
+      "learning_rate": 7.362969548788619e-05,
+      "loss": 1.6624,
+      "step": 28437
+    },
+    {
+      "epoch": 0.6319555555555556,
+      "grad_norm": 1.573633074760437,
+      "learning_rate": 7.36252500555679e-05,
+      "loss": 1.5709,
+      "step": 28438
+    },
+    {
+      "epoch": 0.6319777777777777,
+      "grad_norm": 1.6286156177520752,
+      "learning_rate": 7.362080462324962e-05,
+      "loss": 1.5022,
+      "step": 28439
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 1.5434051752090454,
+      "learning_rate": 7.361635919093132e-05,
+      "loss": 1.8551,
+      "step": 28440
+    },
+    {
+      "epoch": 0.6320222222222223,
+      "grad_norm": 1.667515754699707,
+      "learning_rate": 7.361191375861303e-05,
+      "loss": 1.4745,
+      "step": 28441
+    },
+    {
+      "epoch": 0.6320444444444444,
+      "grad_norm": 1.844184160232544,
+      "learning_rate": 7.360746832629474e-05,
+      "loss": 1.8393,
+      "step": 28442
+    },
+    {
+      "epoch": 0.6320666666666667,
+      "grad_norm": 1.635435938835144,
+      "learning_rate": 7.360302289397645e-05,
+      "loss": 1.4518,
+      "step": 28443
+    },
+    {
+      "epoch": 0.6320888888888889,
+      "grad_norm": 2.1414828300476074,
+      "learning_rate": 7.359857746165814e-05,
+      "loss": 1.9476,
+      "step": 28444
+    },
+    {
+      "epoch": 0.6321111111111111,
+      "grad_norm": 1.8981890678405762,
+      "learning_rate": 7.359413202933985e-05,
+      "loss": 1.882,
+      "step": 28445
+    },
+    {
+      "epoch": 0.6321333333333333,
+      "grad_norm": 1.939850091934204,
+      "learning_rate": 7.358968659702156e-05,
+      "loss": 1.6742,
+      "step": 28446
+    },
+    {
+      "epoch": 0.6321555555555556,
+      "grad_norm": 1.9543533325195312,
+      "learning_rate": 7.358524116470327e-05,
+      "loss": 1.6316,
+      "step": 28447
+    },
+    {
+      "epoch": 0.6321777777777777,
+      "grad_norm": 1.5883665084838867,
+      "learning_rate": 7.358079573238498e-05,
+      "loss": 1.7852,
+      "step": 28448
+    },
+    {
+      "epoch": 0.6322,
+      "grad_norm": 2.1732687950134277,
+      "learning_rate": 7.357635030006669e-05,
+      "loss": 1.5089,
+      "step": 28449
+    },
+    {
+      "epoch": 0.6322222222222222,
+      "grad_norm": 2.041788101196289,
+      "learning_rate": 7.357190486774839e-05,
+      "loss": 1.3564,
+      "step": 28450
+    },
+    {
+      "epoch": 0.6322444444444445,
+      "grad_norm": 1.011060118675232,
+      "learning_rate": 7.35674594354301e-05,
+      "loss": 1.1324,
+      "step": 28451
+    },
+    {
+      "epoch": 0.6322666666666666,
+      "grad_norm": 2.093306064605713,
+      "learning_rate": 7.356301400311181e-05,
+      "loss": 2.4241,
+      "step": 28452
+    },
+    {
+      "epoch": 0.6322888888888889,
+      "grad_norm": 1.4873263835906982,
+      "learning_rate": 7.355856857079352e-05,
+      "loss": 2.338,
+      "step": 28453
+    },
+    {
+      "epoch": 0.6323111111111112,
+      "grad_norm": 2.758570432662964,
+      "learning_rate": 7.355412313847521e-05,
+      "loss": 1.6346,
+      "step": 28454
+    },
+    {
+      "epoch": 0.6323333333333333,
+      "grad_norm": 1.6452229022979736,
+      "learning_rate": 7.354967770615692e-05,
+      "loss": 1.3205,
+      "step": 28455
+    },
+    {
+      "epoch": 0.6323555555555556,
+      "grad_norm": 1.439057469367981,
+      "learning_rate": 7.354523227383863e-05,
+      "loss": 1.6358,
+      "step": 28456
+    },
+    {
+      "epoch": 0.6323777777777778,
+      "grad_norm": 1.928806185722351,
+      "learning_rate": 7.354078684152034e-05,
+      "loss": 2.5795,
+      "step": 28457
+    },
+    {
+      "epoch": 0.6324,
+      "grad_norm": 1.6794370412826538,
+      "learning_rate": 7.353634140920205e-05,
+      "loss": 2.255,
+      "step": 28458
+    },
+    {
+      "epoch": 0.6324222222222222,
+      "grad_norm": 1.8662517070770264,
+      "learning_rate": 7.353189597688376e-05,
+      "loss": 1.6204,
+      "step": 28459
+    },
+    {
+      "epoch": 0.6324444444444445,
+      "grad_norm": 1.5139602422714233,
+      "learning_rate": 7.352745054456546e-05,
+      "loss": 1.911,
+      "step": 28460
+    },
+    {
+      "epoch": 0.6324666666666666,
+      "grad_norm": 1.6585510969161987,
+      "learning_rate": 7.352300511224717e-05,
+      "loss": 2.3561,
+      "step": 28461
+    },
+    {
+      "epoch": 0.6324888888888889,
+      "grad_norm": 1.6817216873168945,
+      "learning_rate": 7.351855967992888e-05,
+      "loss": 2.1024,
+      "step": 28462
+    },
+    {
+      "epoch": 0.6325111111111111,
+      "grad_norm": 1.553934097290039,
+      "learning_rate": 7.351411424761059e-05,
+      "loss": 1.975,
+      "step": 28463
+    },
+    {
+      "epoch": 0.6325333333333333,
+      "grad_norm": 1.6149582862854004,
+      "learning_rate": 7.350966881529228e-05,
+      "loss": 2.3645,
+      "step": 28464
+    },
+    {
+      "epoch": 0.6325555555555555,
+      "grad_norm": 1.5581504106521606,
+      "learning_rate": 7.350522338297399e-05,
+      "loss": 2.1551,
+      "step": 28465
+    },
+    {
+      "epoch": 0.6325777777777778,
+      "grad_norm": 1.6197482347488403,
+      "learning_rate": 7.35007779506557e-05,
+      "loss": 2.0389,
+      "step": 28466
+    },
+    {
+      "epoch": 0.6326,
+      "grad_norm": 2.02691388130188,
+      "learning_rate": 7.349633251833741e-05,
+      "loss": 1.8135,
+      "step": 28467
+    },
+    {
+      "epoch": 0.6326222222222222,
+      "grad_norm": 1.5820540189743042,
+      "learning_rate": 7.349188708601912e-05,
+      "loss": 1.8064,
+      "step": 28468
+    },
+    {
+      "epoch": 0.6326444444444445,
+      "grad_norm": 1.6677908897399902,
+      "learning_rate": 7.348744165370083e-05,
+      "loss": 1.9711,
+      "step": 28469
+    },
+    {
+      "epoch": 0.6326666666666667,
+      "grad_norm": 1.846671223640442,
+      "learning_rate": 7.348299622138254e-05,
+      "loss": 1.7647,
+      "step": 28470
+    },
+    {
+      "epoch": 0.6326888888888889,
+      "grad_norm": 1.6714377403259277,
+      "learning_rate": 7.347855078906424e-05,
+      "loss": 1.6715,
+      "step": 28471
+    },
+    {
+      "epoch": 0.6327111111111111,
+      "grad_norm": 2.223407745361328,
+      "learning_rate": 7.347410535674595e-05,
+      "loss": 2.3233,
+      "step": 28472
+    },
+    {
+      "epoch": 0.6327333333333334,
+      "grad_norm": 1.6671288013458252,
+      "learning_rate": 7.346965992442766e-05,
+      "loss": 1.8601,
+      "step": 28473
+    },
+    {
+      "epoch": 0.6327555555555555,
+      "grad_norm": 1.7181869745254517,
+      "learning_rate": 7.346521449210935e-05,
+      "loss": 1.8613,
+      "step": 28474
+    },
+    {
+      "epoch": 0.6327777777777778,
+      "grad_norm": 1.8806204795837402,
+      "learning_rate": 7.346076905979106e-05,
+      "loss": 2.8201,
+      "step": 28475
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 1.6484366655349731,
+      "learning_rate": 7.345632362747278e-05,
+      "loss": 1.7206,
+      "step": 28476
+    },
+    {
+      "epoch": 0.6328222222222222,
+      "grad_norm": 1.443996548652649,
+      "learning_rate": 7.345187819515448e-05,
+      "loss": 1.6574,
+      "step": 28477
+    },
+    {
+      "epoch": 0.6328444444444444,
+      "grad_norm": 1.6608508825302124,
+      "learning_rate": 7.344743276283619e-05,
+      "loss": 1.6401,
+      "step": 28478
+    },
+    {
+      "epoch": 0.6328666666666667,
+      "grad_norm": 2.0114076137542725,
+      "learning_rate": 7.34429873305179e-05,
+      "loss": 2.4515,
+      "step": 28479
+    },
+    {
+      "epoch": 0.6328888888888888,
+      "grad_norm": 1.9120980501174927,
+      "learning_rate": 7.343854189819961e-05,
+      "loss": 2.0641,
+      "step": 28480
+    },
+    {
+      "epoch": 0.6329111111111111,
+      "grad_norm": 1.7209848165512085,
+      "learning_rate": 7.34340964658813e-05,
+      "loss": 1.9164,
+      "step": 28481
+    },
+    {
+      "epoch": 0.6329333333333333,
+      "grad_norm": 1.8291016817092896,
+      "learning_rate": 7.342965103356302e-05,
+      "loss": 2.1635,
+      "step": 28482
+    },
+    {
+      "epoch": 0.6329555555555556,
+      "grad_norm": 1.7169533967971802,
+      "learning_rate": 7.342520560124472e-05,
+      "loss": 1.9718,
+      "step": 28483
+    },
+    {
+      "epoch": 0.6329777777777778,
+      "grad_norm": 1.7712876796722412,
+      "learning_rate": 7.342076016892643e-05,
+      "loss": 1.8347,
+      "step": 28484
+    },
+    {
+      "epoch": 0.633,
+      "grad_norm": 1.8500925302505493,
+      "learning_rate": 7.341631473660814e-05,
+      "loss": 1.8675,
+      "step": 28485
+    },
+    {
+      "epoch": 0.6330222222222223,
+      "grad_norm": 1.7853578329086304,
+      "learning_rate": 7.341186930428985e-05,
+      "loss": 1.9601,
+      "step": 28486
+    },
+    {
+      "epoch": 0.6330444444444444,
+      "grad_norm": 1.840130090713501,
+      "learning_rate": 7.340742387197155e-05,
+      "loss": 2.0153,
+      "step": 28487
+    },
+    {
+      "epoch": 0.6330666666666667,
+      "grad_norm": 2.2379651069641113,
+      "learning_rate": 7.340297843965326e-05,
+      "loss": 2.0073,
+      "step": 28488
+    },
+    {
+      "epoch": 0.6330888888888889,
+      "grad_norm": 1.6185786724090576,
+      "learning_rate": 7.339853300733497e-05,
+      "loss": 1.7823,
+      "step": 28489
+    },
+    {
+      "epoch": 0.6331111111111111,
+      "grad_norm": 6.099532604217529,
+      "learning_rate": 7.339408757501668e-05,
+      "loss": 1.8272,
+      "step": 28490
+    },
+    {
+      "epoch": 0.6331333333333333,
+      "grad_norm": 1.8324321508407593,
+      "learning_rate": 7.338964214269837e-05,
+      "loss": 2.173,
+      "step": 28491
+    },
+    {
+      "epoch": 0.6331555555555556,
+      "grad_norm": 1.8781834840774536,
+      "learning_rate": 7.338519671038008e-05,
+      "loss": 1.8743,
+      "step": 28492
+    },
+    {
+      "epoch": 0.6331777777777777,
+      "grad_norm": 2.0203893184661865,
+      "learning_rate": 7.33807512780618e-05,
+      "loss": 2.1924,
+      "step": 28493
+    },
+    {
+      "epoch": 0.6332,
+      "grad_norm": 1.485815405845642,
+      "learning_rate": 7.33763058457435e-05,
+      "loss": 1.4762,
+      "step": 28494
+    },
+    {
+      "epoch": 0.6332222222222222,
+      "grad_norm": 2.0100176334381104,
+      "learning_rate": 7.337186041342521e-05,
+      "loss": 2.186,
+      "step": 28495
+    },
+    {
+      "epoch": 0.6332444444444445,
+      "grad_norm": 1.780142903327942,
+      "learning_rate": 7.336741498110692e-05,
+      "loss": 1.6422,
+      "step": 28496
+    },
+    {
+      "epoch": 0.6332666666666666,
+      "grad_norm": 2.057020425796509,
+      "learning_rate": 7.336296954878862e-05,
+      "loss": 1.7511,
+      "step": 28497
+    },
+    {
+      "epoch": 0.6332888888888889,
+      "grad_norm": 1.6470258235931396,
+      "learning_rate": 7.335852411647033e-05,
+      "loss": 1.7918,
+      "step": 28498
+    },
+    {
+      "epoch": 0.6333111111111112,
+      "grad_norm": 1.5285764932632446,
+      "learning_rate": 7.335407868415204e-05,
+      "loss": 1.5708,
+      "step": 28499
+    },
+    {
+      "epoch": 0.6333333333333333,
+      "grad_norm": 1.337584376335144,
+      "learning_rate": 7.334963325183375e-05,
+      "loss": 1.0091,
+      "step": 28500
+    },
+    {
+      "epoch": 0.6333555555555556,
+      "grad_norm": 1.5156707763671875,
+      "learning_rate": 7.334518781951544e-05,
+      "loss": 2.3721,
+      "step": 28501
+    },
+    {
+      "epoch": 0.6333777777777778,
+      "grad_norm": 1.9771760702133179,
+      "learning_rate": 7.334074238719715e-05,
+      "loss": 2.5071,
+      "step": 28502
+    },
+    {
+      "epoch": 0.6334,
+      "grad_norm": 1.4453988075256348,
+      "learning_rate": 7.333629695487886e-05,
+      "loss": 2.2172,
+      "step": 28503
+    },
+    {
+      "epoch": 0.6334222222222222,
+      "grad_norm": 1.5911939144134521,
+      "learning_rate": 7.333185152256057e-05,
+      "loss": 1.7264,
+      "step": 28504
+    },
+    {
+      "epoch": 0.6334444444444445,
+      "grad_norm": 1.421743631362915,
+      "learning_rate": 7.332740609024228e-05,
+      "loss": 1.9622,
+      "step": 28505
+    },
+    {
+      "epoch": 0.6334666666666666,
+      "grad_norm": 1.416795015335083,
+      "learning_rate": 7.332296065792399e-05,
+      "loss": 1.7702,
+      "step": 28506
+    },
+    {
+      "epoch": 0.6334888888888889,
+      "grad_norm": 1.7996914386749268,
+      "learning_rate": 7.331851522560569e-05,
+      "loss": 1.6658,
+      "step": 28507
+    },
+    {
+      "epoch": 0.6335111111111111,
+      "grad_norm": 1.7646291255950928,
+      "learning_rate": 7.33140697932874e-05,
+      "loss": 2.6733,
+      "step": 28508
+    },
+    {
+      "epoch": 0.6335333333333333,
+      "grad_norm": 1.6536164283752441,
+      "learning_rate": 7.330962436096911e-05,
+      "loss": 1.9984,
+      "step": 28509
+    },
+    {
+      "epoch": 0.6335555555555555,
+      "grad_norm": 1.6147255897521973,
+      "learning_rate": 7.330517892865082e-05,
+      "loss": 1.9761,
+      "step": 28510
+    },
+    {
+      "epoch": 0.6335777777777778,
+      "grad_norm": 1.5083590745925903,
+      "learning_rate": 7.330073349633251e-05,
+      "loss": 1.7461,
+      "step": 28511
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 1.7501157522201538,
+      "learning_rate": 7.329628806401422e-05,
+      "loss": 2.2495,
+      "step": 28512
+    },
+    {
+      "epoch": 0.6336222222222222,
+      "grad_norm": 1.9548697471618652,
+      "learning_rate": 7.329184263169595e-05,
+      "loss": 2.637,
+      "step": 28513
+    },
+    {
+      "epoch": 0.6336444444444445,
+      "grad_norm": 1.6512237787246704,
+      "learning_rate": 7.328739719937764e-05,
+      "loss": 2.0011,
+      "step": 28514
+    },
+    {
+      "epoch": 0.6336666666666667,
+      "grad_norm": 1.7872507572174072,
+      "learning_rate": 7.328295176705935e-05,
+      "loss": 2.345,
+      "step": 28515
+    },
+    {
+      "epoch": 0.6336888888888889,
+      "grad_norm": 1.7773979902267456,
+      "learning_rate": 7.327850633474106e-05,
+      "loss": 1.9037,
+      "step": 28516
+    },
+    {
+      "epoch": 0.6337111111111111,
+      "grad_norm": 2.1018433570861816,
+      "learning_rate": 7.327406090242276e-05,
+      "loss": 2.0568,
+      "step": 28517
+    },
+    {
+      "epoch": 0.6337333333333334,
+      "grad_norm": 1.9714939594268799,
+      "learning_rate": 7.326961547010447e-05,
+      "loss": 1.5208,
+      "step": 28518
+    },
+    {
+      "epoch": 0.6337555555555555,
+      "grad_norm": 1.7774789333343506,
+      "learning_rate": 7.326517003778618e-05,
+      "loss": 2.0657,
+      "step": 28519
+    },
+    {
+      "epoch": 0.6337777777777778,
+      "grad_norm": 1.48650324344635,
+      "learning_rate": 7.326072460546789e-05,
+      "loss": 2.0511,
+      "step": 28520
+    },
+    {
+      "epoch": 0.6338,
+      "grad_norm": 1.7070021629333496,
+      "learning_rate": 7.32562791731496e-05,
+      "loss": 2.0907,
+      "step": 28521
+    },
+    {
+      "epoch": 0.6338222222222222,
+      "grad_norm": 1.827212929725647,
+      "learning_rate": 7.32518337408313e-05,
+      "loss": 2.1822,
+      "step": 28522
+    },
+    {
+      "epoch": 0.6338444444444444,
+      "grad_norm": 1.3869909048080444,
+      "learning_rate": 7.324738830851301e-05,
+      "loss": 1.672,
+      "step": 28523
+    },
+    {
+      "epoch": 0.6338666666666667,
+      "grad_norm": 1.6951630115509033,
+      "learning_rate": 7.324294287619471e-05,
+      "loss": 1.6874,
+      "step": 28524
+    },
+    {
+      "epoch": 0.6338888888888888,
+      "grad_norm": 1.5163898468017578,
+      "learning_rate": 7.323849744387642e-05,
+      "loss": 2.0067,
+      "step": 28525
+    },
+    {
+      "epoch": 0.6339111111111111,
+      "grad_norm": 1.5956902503967285,
+      "learning_rate": 7.323405201155813e-05,
+      "loss": 1.7679,
+      "step": 28526
+    },
+    {
+      "epoch": 0.6339333333333333,
+      "grad_norm": 1.5486507415771484,
+      "learning_rate": 7.322960657923984e-05,
+      "loss": 1.7618,
+      "step": 28527
+    },
+    {
+      "epoch": 0.6339555555555556,
+      "grad_norm": 1.6986682415008545,
+      "learning_rate": 7.322516114692154e-05,
+      "loss": 2.2877,
+      "step": 28528
+    },
+    {
+      "epoch": 0.6339777777777778,
+      "grad_norm": 1.615833044052124,
+      "learning_rate": 7.322071571460325e-05,
+      "loss": 1.9716,
+      "step": 28529
+    },
+    {
+      "epoch": 0.634,
+      "grad_norm": 1.906920313835144,
+      "learning_rate": 7.321627028228495e-05,
+      "loss": 1.8578,
+      "step": 28530
+    },
+    {
+      "epoch": 0.6340222222222223,
+      "grad_norm": 1.7437498569488525,
+      "learning_rate": 7.321182484996666e-05,
+      "loss": 1.7576,
+      "step": 28531
+    },
+    {
+      "epoch": 0.6340444444444444,
+      "grad_norm": 1.6995993852615356,
+      "learning_rate": 7.320737941764837e-05,
+      "loss": 1.9975,
+      "step": 28532
+    },
+    {
+      "epoch": 0.6340666666666667,
+      "grad_norm": 1.756561279296875,
+      "learning_rate": 7.320293398533008e-05,
+      "loss": 0.8284,
+      "step": 28533
+    },
+    {
+      "epoch": 0.6340888888888889,
+      "grad_norm": 1.932459831237793,
+      "learning_rate": 7.319848855301178e-05,
+      "loss": 2.5953,
+      "step": 28534
+    },
+    {
+      "epoch": 0.6341111111111111,
+      "grad_norm": 1.888073444366455,
+      "learning_rate": 7.319404312069349e-05,
+      "loss": 2.0259,
+      "step": 28535
+    },
+    {
+      "epoch": 0.6341333333333333,
+      "grad_norm": 2.1089727878570557,
+      "learning_rate": 7.31895976883752e-05,
+      "loss": 2.2318,
+      "step": 28536
+    },
+    {
+      "epoch": 0.6341555555555556,
+      "grad_norm": 1.951337218284607,
+      "learning_rate": 7.318515225605691e-05,
+      "loss": 1.7462,
+      "step": 28537
+    },
+    {
+      "epoch": 0.6341777777777777,
+      "grad_norm": 1.523807168006897,
+      "learning_rate": 7.31807068237386e-05,
+      "loss": 1.8149,
+      "step": 28538
+    },
+    {
+      "epoch": 0.6342,
+      "grad_norm": 1.58031165599823,
+      "learning_rate": 7.317626139142031e-05,
+      "loss": 1.8405,
+      "step": 28539
+    },
+    {
+      "epoch": 0.6342222222222222,
+      "grad_norm": 1.4024109840393066,
+      "learning_rate": 7.317181595910202e-05,
+      "loss": 1.3146,
+      "step": 28540
+    },
+    {
+      "epoch": 0.6342444444444445,
+      "grad_norm": 2.466278076171875,
+      "learning_rate": 7.316737052678373e-05,
+      "loss": 1.946,
+      "step": 28541
+    },
+    {
+      "epoch": 0.6342666666666666,
+      "grad_norm": 1.586089849472046,
+      "learning_rate": 7.316292509446544e-05,
+      "loss": 1.9332,
+      "step": 28542
+    },
+    {
+      "epoch": 0.6342888888888889,
+      "grad_norm": 1.5611326694488525,
+      "learning_rate": 7.315847966214715e-05,
+      "loss": 1.6498,
+      "step": 28543
+    },
+    {
+      "epoch": 0.6343111111111112,
+      "grad_norm": 1.8813029527664185,
+      "learning_rate": 7.315403422982885e-05,
+      "loss": 1.7946,
+      "step": 28544
+    },
+    {
+      "epoch": 0.6343333333333333,
+      "grad_norm": 1.7111068964004517,
+      "learning_rate": 7.314958879751056e-05,
+      "loss": 1.7456,
+      "step": 28545
+    },
+    {
+      "epoch": 0.6343555555555556,
+      "grad_norm": 2.0285494327545166,
+      "learning_rate": 7.314514336519227e-05,
+      "loss": 1.8342,
+      "step": 28546
+    },
+    {
+      "epoch": 0.6343777777777778,
+      "grad_norm": 1.7374845743179321,
+      "learning_rate": 7.314069793287398e-05,
+      "loss": 1.7883,
+      "step": 28547
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 1.5826327800750732,
+      "learning_rate": 7.313625250055567e-05,
+      "loss": 1.3873,
+      "step": 28548
+    },
+    {
+      "epoch": 0.6344222222222222,
+      "grad_norm": 1.2411465644836426,
+      "learning_rate": 7.31318070682374e-05,
+      "loss": 0.7282,
+      "step": 28549
+    },
+    {
+      "epoch": 0.6344444444444445,
+      "grad_norm": 1.8891007900238037,
+      "learning_rate": 7.31273616359191e-05,
+      "loss": 1.24,
+      "step": 28550
+    },
+    {
+      "epoch": 0.6344666666666666,
+      "grad_norm": 1.4046944379806519,
+      "learning_rate": 7.31229162036008e-05,
+      "loss": 2.2538,
+      "step": 28551
+    },
+    {
+      "epoch": 0.6344888888888889,
+      "grad_norm": 1.4603004455566406,
+      "learning_rate": 7.311847077128251e-05,
+      "loss": 2.1524,
+      "step": 28552
+    },
+    {
+      "epoch": 0.6345111111111111,
+      "grad_norm": 1.4375442266464233,
+      "learning_rate": 7.311402533896422e-05,
+      "loss": 2.4616,
+      "step": 28553
+    },
+    {
+      "epoch": 0.6345333333333333,
+      "grad_norm": 1.6095811128616333,
+      "learning_rate": 7.310957990664592e-05,
+      "loss": 2.2056,
+      "step": 28554
+    },
+    {
+      "epoch": 0.6345555555555555,
+      "grad_norm": 1.7685246467590332,
+      "learning_rate": 7.310513447432763e-05,
+      "loss": 2.2868,
+      "step": 28555
+    },
+    {
+      "epoch": 0.6345777777777778,
+      "grad_norm": 0.9971567988395691,
+      "learning_rate": 7.310068904200934e-05,
+      "loss": 0.9219,
+      "step": 28556
+    },
+    {
+      "epoch": 0.6346,
+      "grad_norm": 1.743350625038147,
+      "learning_rate": 7.309624360969105e-05,
+      "loss": 2.5285,
+      "step": 28557
+    },
+    {
+      "epoch": 0.6346222222222222,
+      "grad_norm": 1.6786539554595947,
+      "learning_rate": 7.309179817737276e-05,
+      "loss": 1.6679,
+      "step": 28558
+    },
+    {
+      "epoch": 0.6346444444444445,
+      "grad_norm": 1.6530226469039917,
+      "learning_rate": 7.308735274505447e-05,
+      "loss": 1.8407,
+      "step": 28559
+    },
+    {
+      "epoch": 0.6346666666666667,
+      "grad_norm": 1.5131018161773682,
+      "learning_rate": 7.308290731273618e-05,
+      "loss": 2.0245,
+      "step": 28560
+    },
+    {
+      "epoch": 0.6346888888888889,
+      "grad_norm": 1.5645263195037842,
+      "learning_rate": 7.307846188041787e-05,
+      "loss": 1.9042,
+      "step": 28561
+    },
+    {
+      "epoch": 0.6347111111111111,
+      "grad_norm": 1.5649160146713257,
+      "learning_rate": 7.307401644809958e-05,
+      "loss": 2.1702,
+      "step": 28562
+    },
+    {
+      "epoch": 0.6347333333333334,
+      "grad_norm": 1.6197746992111206,
+      "learning_rate": 7.306957101578129e-05,
+      "loss": 2.2538,
+      "step": 28563
+    },
+    {
+      "epoch": 0.6347555555555555,
+      "grad_norm": 1.6465293169021606,
+      "learning_rate": 7.306512558346299e-05,
+      "loss": 2.389,
+      "step": 28564
+    },
+    {
+      "epoch": 0.6347777777777778,
+      "grad_norm": 1.5573149919509888,
+      "learning_rate": 7.30606801511447e-05,
+      "loss": 2.0219,
+      "step": 28565
+    },
+    {
+      "epoch": 0.6348,
+      "grad_norm": 1.5480624437332153,
+      "learning_rate": 7.30562347188264e-05,
+      "loss": 1.5438,
+      "step": 28566
+    },
+    {
+      "epoch": 0.6348222222222222,
+      "grad_norm": 1.59479558467865,
+      "learning_rate": 7.305178928650812e-05,
+      "loss": 1.1533,
+      "step": 28567
+    },
+    {
+      "epoch": 0.6348444444444444,
+      "grad_norm": 1.8960281610488892,
+      "learning_rate": 7.304734385418983e-05,
+      "loss": 2.0588,
+      "step": 28568
+    },
+    {
+      "epoch": 0.6348666666666667,
+      "grad_norm": 1.7194708585739136,
+      "learning_rate": 7.304289842187154e-05,
+      "loss": 2.1623,
+      "step": 28569
+    },
+    {
+      "epoch": 0.6348888888888888,
+      "grad_norm": 1.7426247596740723,
+      "learning_rate": 7.303845298955324e-05,
+      "loss": 2.4881,
+      "step": 28570
+    },
+    {
+      "epoch": 0.6349111111111111,
+      "grad_norm": 1.7681390047073364,
+      "learning_rate": 7.303400755723494e-05,
+      "loss": 1.7168,
+      "step": 28571
+    },
+    {
+      "epoch": 0.6349333333333333,
+      "grad_norm": 1.7570207118988037,
+      "learning_rate": 7.302956212491665e-05,
+      "loss": 1.8496,
+      "step": 28572
+    },
+    {
+      "epoch": 0.6349555555555556,
+      "grad_norm": 1.7781468629837036,
+      "learning_rate": 7.302511669259836e-05,
+      "loss": 1.9237,
+      "step": 28573
+    },
+    {
+      "epoch": 0.6349777777777778,
+      "grad_norm": 2.170135498046875,
+      "learning_rate": 7.302067126028006e-05,
+      "loss": 2.6366,
+      "step": 28574
+    },
+    {
+      "epoch": 0.635,
+      "grad_norm": 1.726895809173584,
+      "learning_rate": 7.301622582796177e-05,
+      "loss": 1.9751,
+      "step": 28575
+    },
+    {
+      "epoch": 0.6350222222222223,
+      "grad_norm": 1.7290587425231934,
+      "learning_rate": 7.301178039564348e-05,
+      "loss": 2.4883,
+      "step": 28576
+    },
+    {
+      "epoch": 0.6350444444444444,
+      "grad_norm": 1.654848337173462,
+      "learning_rate": 7.300733496332518e-05,
+      "loss": 2.0482,
+      "step": 28577
+    },
+    {
+      "epoch": 0.6350666666666667,
+      "grad_norm": 1.8695530891418457,
+      "learning_rate": 7.30028895310069e-05,
+      "loss": 1.987,
+      "step": 28578
+    },
+    {
+      "epoch": 0.6350888888888889,
+      "grad_norm": 1.7678459882736206,
+      "learning_rate": 7.29984440986886e-05,
+      "loss": 2.162,
+      "step": 28579
+    },
+    {
+      "epoch": 0.6351111111111111,
+      "grad_norm": 1.806392788887024,
+      "learning_rate": 7.299399866637031e-05,
+      "loss": 1.9873,
+      "step": 28580
+    },
+    {
+      "epoch": 0.6351333333333333,
+      "grad_norm": 1.878972053527832,
+      "learning_rate": 7.298955323405201e-05,
+      "loss": 2.217,
+      "step": 28581
+    },
+    {
+      "epoch": 0.6351555555555556,
+      "grad_norm": 1.5210174322128296,
+      "learning_rate": 7.298510780173372e-05,
+      "loss": 1.8135,
+      "step": 28582
+    },
+    {
+      "epoch": 0.6351777777777777,
+      "grad_norm": 1.3945029973983765,
+      "learning_rate": 7.298066236941543e-05,
+      "loss": 1.2624,
+      "step": 28583
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 1.7245527505874634,
+      "learning_rate": 7.297621693709714e-05,
+      "loss": 1.9487,
+      "step": 28584
+    },
+    {
+      "epoch": 0.6352222222222222,
+      "grad_norm": 1.6345226764678955,
+      "learning_rate": 7.297177150477883e-05,
+      "loss": 1.7475,
+      "step": 28585
+    },
+    {
+      "epoch": 0.6352444444444444,
+      "grad_norm": 1.9497441053390503,
+      "learning_rate": 7.296732607246056e-05,
+      "loss": 1.898,
+      "step": 28586
+    },
+    {
+      "epoch": 0.6352666666666666,
+      "grad_norm": 2.098942279815674,
+      "learning_rate": 7.296288064014227e-05,
+      "loss": 1.8164,
+      "step": 28587
+    },
+    {
+      "epoch": 0.6352888888888889,
+      "grad_norm": 2.027961492538452,
+      "learning_rate": 7.295843520782396e-05,
+      "loss": 2.0378,
+      "step": 28588
+    },
+    {
+      "epoch": 0.6353111111111112,
+      "grad_norm": 1.3239794969558716,
+      "learning_rate": 7.295398977550567e-05,
+      "loss": 1.2233,
+      "step": 28589
+    },
+    {
+      "epoch": 0.6353333333333333,
+      "grad_norm": 1.8406566381454468,
+      "learning_rate": 7.294954434318738e-05,
+      "loss": 1.9132,
+      "step": 28590
+    },
+    {
+      "epoch": 0.6353555555555556,
+      "grad_norm": 1.7158334255218506,
+      "learning_rate": 7.294509891086908e-05,
+      "loss": 1.5789,
+      "step": 28591
+    },
+    {
+      "epoch": 0.6353777777777778,
+      "grad_norm": 1.7892802953720093,
+      "learning_rate": 7.294065347855079e-05,
+      "loss": 1.9597,
+      "step": 28592
+    },
+    {
+      "epoch": 0.6354,
+      "grad_norm": 1.775794506072998,
+      "learning_rate": 7.29362080462325e-05,
+      "loss": 2.115,
+      "step": 28593
+    },
+    {
+      "epoch": 0.6354222222222222,
+      "grad_norm": 1.8277240991592407,
+      "learning_rate": 7.293176261391421e-05,
+      "loss": 1.7255,
+      "step": 28594
+    },
+    {
+      "epoch": 0.6354444444444445,
+      "grad_norm": 1.9447377920150757,
+      "learning_rate": 7.292731718159592e-05,
+      "loss": 1.9657,
+      "step": 28595
+    },
+    {
+      "epoch": 0.6354666666666666,
+      "grad_norm": 1.974876046180725,
+      "learning_rate": 7.292287174927763e-05,
+      "loss": 1.7476,
+      "step": 28596
+    },
+    {
+      "epoch": 0.6354888888888889,
+      "grad_norm": 2.3403728008270264,
+      "learning_rate": 7.291842631695934e-05,
+      "loss": 2.101,
+      "step": 28597
+    },
+    {
+      "epoch": 0.6355111111111111,
+      "grad_norm": 1.9675029516220093,
+      "learning_rate": 7.291398088464103e-05,
+      "loss": 2.0985,
+      "step": 28598
+    },
+    {
+      "epoch": 0.6355333333333333,
+      "grad_norm": 1.6708178520202637,
+      "learning_rate": 7.290953545232274e-05,
+      "loss": 1.4453,
+      "step": 28599
+    },
+    {
+      "epoch": 0.6355555555555555,
+      "grad_norm": 2.2771944999694824,
+      "learning_rate": 7.290509002000445e-05,
+      "loss": 1.4697,
+      "step": 28600
+    },
+    {
+      "epoch": 0.6355777777777778,
+      "grad_norm": 1.8394581079483032,
+      "learning_rate": 7.290064458768615e-05,
+      "loss": 2.1661,
+      "step": 28601
+    },
+    {
+      "epoch": 0.6356,
+      "grad_norm": 1.8911298513412476,
+      "learning_rate": 7.289619915536786e-05,
+      "loss": 2.4016,
+      "step": 28602
+    },
+    {
+      "epoch": 0.6356222222222222,
+      "grad_norm": 1.4113638401031494,
+      "learning_rate": 7.289175372304957e-05,
+      "loss": 2.1418,
+      "step": 28603
+    },
+    {
+      "epoch": 0.6356444444444445,
+      "grad_norm": 0.29859021306037903,
+      "learning_rate": 7.288730829073128e-05,
+      "loss": 0.0205,
+      "step": 28604
+    },
+    {
+      "epoch": 0.6356666666666667,
+      "grad_norm": 1.9816783666610718,
+      "learning_rate": 7.288286285841299e-05,
+      "loss": 1.8118,
+      "step": 28605
+    },
+    {
+      "epoch": 0.6356888888888889,
+      "grad_norm": 1.4286977052688599,
+      "learning_rate": 7.28784174260947e-05,
+      "loss": 1.9265,
+      "step": 28606
+    },
+    {
+      "epoch": 0.6357111111111111,
+      "grad_norm": 1.6861002445220947,
+      "learning_rate": 7.28739719937764e-05,
+      "loss": 2.3091,
+      "step": 28607
+    },
+    {
+      "epoch": 0.6357333333333334,
+      "grad_norm": 1.5298811197280884,
+      "learning_rate": 7.28695265614581e-05,
+      "loss": 2.1259,
+      "step": 28608
+    },
+    {
+      "epoch": 0.6357555555555555,
+      "grad_norm": 1.7088412046432495,
+      "learning_rate": 7.286508112913981e-05,
+      "loss": 1.7914,
+      "step": 28609
+    },
+    {
+      "epoch": 0.6357777777777778,
+      "grad_norm": 1.6431035995483398,
+      "learning_rate": 7.286063569682152e-05,
+      "loss": 2.349,
+      "step": 28610
+    },
+    {
+      "epoch": 0.6358,
+      "grad_norm": 1.3370503187179565,
+      "learning_rate": 7.285619026450322e-05,
+      "loss": 1.8768,
+      "step": 28611
+    },
+    {
+      "epoch": 0.6358222222222222,
+      "grad_norm": 1.6042437553405762,
+      "learning_rate": 7.285174483218493e-05,
+      "loss": 1.9377,
+      "step": 28612
+    },
+    {
+      "epoch": 0.6358444444444444,
+      "grad_norm": 0.9689003229141235,
+      "learning_rate": 7.284729939986664e-05,
+      "loss": 0.9605,
+      "step": 28613
+    },
+    {
+      "epoch": 0.6358666666666667,
+      "grad_norm": 1.4642431735992432,
+      "learning_rate": 7.284285396754835e-05,
+      "loss": 1.7675,
+      "step": 28614
+    },
+    {
+      "epoch": 0.6358888888888888,
+      "grad_norm": 1.6015061140060425,
+      "learning_rate": 7.283840853523006e-05,
+      "loss": 2.0033,
+      "step": 28615
+    },
+    {
+      "epoch": 0.6359111111111111,
+      "grad_norm": 1.7440736293792725,
+      "learning_rate": 7.283396310291177e-05,
+      "loss": 2.25,
+      "step": 28616
+    },
+    {
+      "epoch": 0.6359333333333334,
+      "grad_norm": 1.5764706134796143,
+      "learning_rate": 7.282951767059347e-05,
+      "loss": 1.6149,
+      "step": 28617
+    },
+    {
+      "epoch": 0.6359555555555556,
+      "grad_norm": 1.6555157899856567,
+      "learning_rate": 7.282507223827517e-05,
+      "loss": 2.1503,
+      "step": 28618
+    },
+    {
+      "epoch": 0.6359777777777778,
+      "grad_norm": 1.5757826566696167,
+      "learning_rate": 7.282062680595688e-05,
+      "loss": 1.9042,
+      "step": 28619
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 2.0371313095092773,
+      "learning_rate": 7.281618137363859e-05,
+      "loss": 2.2666,
+      "step": 28620
+    },
+    {
+      "epoch": 0.6360222222222223,
+      "grad_norm": 1.907381296157837,
+      "learning_rate": 7.281173594132029e-05,
+      "loss": 1.9729,
+      "step": 28621
+    },
+    {
+      "epoch": 0.6360444444444444,
+      "grad_norm": 1.9274381399154663,
+      "learning_rate": 7.2807290509002e-05,
+      "loss": 2.2086,
+      "step": 28622
+    },
+    {
+      "epoch": 0.6360666666666667,
+      "grad_norm": 1.6238957643508911,
+      "learning_rate": 7.280284507668372e-05,
+      "loss": 1.9729,
+      "step": 28623
+    },
+    {
+      "epoch": 0.6360888888888889,
+      "grad_norm": 1.978137731552124,
+      "learning_rate": 7.279839964436543e-05,
+      "loss": 2.1478,
+      "step": 28624
+    },
+    {
+      "epoch": 0.6361111111111111,
+      "grad_norm": 1.505639910697937,
+      "learning_rate": 7.279395421204712e-05,
+      "loss": 1.7419,
+      "step": 28625
+    },
+    {
+      "epoch": 0.6361333333333333,
+      "grad_norm": 1.651392936706543,
+      "learning_rate": 7.278950877972883e-05,
+      "loss": 1.9836,
+      "step": 28626
+    },
+    {
+      "epoch": 0.6361555555555556,
+      "grad_norm": 2.5465140342712402,
+      "learning_rate": 7.278506334741054e-05,
+      "loss": 2.0878,
+      "step": 28627
+    },
+    {
+      "epoch": 0.6361777777777777,
+      "grad_norm": 1.9380546808242798,
+      "learning_rate": 7.278061791509224e-05,
+      "loss": 2.1909,
+      "step": 28628
+    },
+    {
+      "epoch": 0.6362,
+      "grad_norm": 1.6133519411087036,
+      "learning_rate": 7.277617248277395e-05,
+      "loss": 1.9931,
+      "step": 28629
+    },
+    {
+      "epoch": 0.6362222222222222,
+      "grad_norm": 1.7483313083648682,
+      "learning_rate": 7.277172705045566e-05,
+      "loss": 1.6586,
+      "step": 28630
+    },
+    {
+      "epoch": 0.6362444444444444,
+      "grad_norm": 1.6443977355957031,
+      "learning_rate": 7.276728161813737e-05,
+      "loss": 1.9233,
+      "step": 28631
+    },
+    {
+      "epoch": 0.6362666666666666,
+      "grad_norm": 2.09979510307312,
+      "learning_rate": 7.276283618581908e-05,
+      "loss": 1.6895,
+      "step": 28632
+    },
+    {
+      "epoch": 0.6362888888888889,
+      "grad_norm": 1.843942642211914,
+      "learning_rate": 7.275839075350079e-05,
+      "loss": 1.7228,
+      "step": 28633
+    },
+    {
+      "epoch": 0.6363111111111112,
+      "grad_norm": 1.4837838411331177,
+      "learning_rate": 7.27539453211825e-05,
+      "loss": 1.6651,
+      "step": 28634
+    },
+    {
+      "epoch": 0.6363333333333333,
+      "grad_norm": 1.5741617679595947,
+      "learning_rate": 7.27494998888642e-05,
+      "loss": 1.6383,
+      "step": 28635
+    },
+    {
+      "epoch": 0.6363555555555556,
+      "grad_norm": 2.0141563415527344,
+      "learning_rate": 7.27450544565459e-05,
+      "loss": 2.3839,
+      "step": 28636
+    },
+    {
+      "epoch": 0.6363777777777778,
+      "grad_norm": 1.6650718450546265,
+      "learning_rate": 7.274060902422761e-05,
+      "loss": 1.8759,
+      "step": 28637
+    },
+    {
+      "epoch": 0.6364,
+      "grad_norm": 1.608781337738037,
+      "learning_rate": 7.273616359190931e-05,
+      "loss": 1.4356,
+      "step": 28638
+    },
+    {
+      "epoch": 0.6364222222222222,
+      "grad_norm": 1.8974065780639648,
+      "learning_rate": 7.273171815959102e-05,
+      "loss": 2.0662,
+      "step": 28639
+    },
+    {
+      "epoch": 0.6364444444444445,
+      "grad_norm": 1.67244291305542,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 1.7112,
+      "step": 28640
+    },
+    {
+      "epoch": 0.6364666666666666,
+      "grad_norm": 2.118295431137085,
+      "learning_rate": 7.272282729495444e-05,
+      "loss": 1.9977,
+      "step": 28641
+    },
+    {
+      "epoch": 0.6364888888888889,
+      "grad_norm": 2.3757729530334473,
+      "learning_rate": 7.271838186263615e-05,
+      "loss": 2.4056,
+      "step": 28642
+    },
+    {
+      "epoch": 0.6365111111111111,
+      "grad_norm": 1.6669895648956299,
+      "learning_rate": 7.271393643031786e-05,
+      "loss": 1.5257,
+      "step": 28643
+    },
+    {
+      "epoch": 0.6365333333333333,
+      "grad_norm": 1.4742716550827026,
+      "learning_rate": 7.270949099799957e-05,
+      "loss": 1.361,
+      "step": 28644
+    },
+    {
+      "epoch": 0.6365555555555555,
+      "grad_norm": 2.1252920627593994,
+      "learning_rate": 7.270504556568126e-05,
+      "loss": 1.7919,
+      "step": 28645
+    },
+    {
+      "epoch": 0.6365777777777778,
+      "grad_norm": 1.695860505104065,
+      "learning_rate": 7.270060013336297e-05,
+      "loss": 1.8634,
+      "step": 28646
+    },
+    {
+      "epoch": 0.6366,
+      "grad_norm": 2.3226397037506104,
+      "learning_rate": 7.269615470104468e-05,
+      "loss": 1.8964,
+      "step": 28647
+    },
+    {
+      "epoch": 0.6366222222222222,
+      "grad_norm": 1.852975606918335,
+      "learning_rate": 7.269170926872638e-05,
+      "loss": 1.5989,
+      "step": 28648
+    },
+    {
+      "epoch": 0.6366444444444445,
+      "grad_norm": 1.5640804767608643,
+      "learning_rate": 7.268726383640809e-05,
+      "loss": 1.5557,
+      "step": 28649
+    },
+    {
+      "epoch": 0.6366666666666667,
+      "grad_norm": 2.081127882003784,
+      "learning_rate": 7.26828184040898e-05,
+      "loss": 1.8513,
+      "step": 28650
+    },
+    {
+      "epoch": 0.6366888888888889,
+      "grad_norm": 1.040666103363037,
+      "learning_rate": 7.267837297177151e-05,
+      "loss": 0.0264,
+      "step": 28651
+    },
+    {
+      "epoch": 0.6367111111111111,
+      "grad_norm": 0.20972059667110443,
+      "learning_rate": 7.267392753945322e-05,
+      "loss": 0.0186,
+      "step": 28652
+    },
+    {
+      "epoch": 0.6367333333333334,
+      "grad_norm": 1.5281065702438354,
+      "learning_rate": 7.266948210713493e-05,
+      "loss": 1.9655,
+      "step": 28653
+    },
+    {
+      "epoch": 0.6367555555555555,
+      "grad_norm": 1.4604177474975586,
+      "learning_rate": 7.266503667481664e-05,
+      "loss": 1.9184,
+      "step": 28654
+    },
+    {
+      "epoch": 0.6367777777777778,
+      "grad_norm": 1.5252203941345215,
+      "learning_rate": 7.266059124249833e-05,
+      "loss": 1.9986,
+      "step": 28655
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 1.656562328338623,
+      "learning_rate": 7.265614581018004e-05,
+      "loss": 1.8098,
+      "step": 28656
+    },
+    {
+      "epoch": 0.6368222222222222,
+      "grad_norm": 1.582373857498169,
+      "learning_rate": 7.265170037786175e-05,
+      "loss": 1.4481,
+      "step": 28657
+    },
+    {
+      "epoch": 0.6368444444444444,
+      "grad_norm": 2.045361280441284,
+      "learning_rate": 7.264725494554345e-05,
+      "loss": 2.5536,
+      "step": 28658
+    },
+    {
+      "epoch": 0.6368666666666667,
+      "grad_norm": 1.7420368194580078,
+      "learning_rate": 7.264280951322516e-05,
+      "loss": 2.2029,
+      "step": 28659
+    },
+    {
+      "epoch": 0.6368888888888888,
+      "grad_norm": 1.4882012605667114,
+      "learning_rate": 7.263836408090688e-05,
+      "loss": 1.6128,
+      "step": 28660
+    },
+    {
+      "epoch": 0.6369111111111111,
+      "grad_norm": 1.2847424745559692,
+      "learning_rate": 7.263391864858859e-05,
+      "loss": 1.1251,
+      "step": 28661
+    },
+    {
+      "epoch": 0.6369333333333334,
+      "grad_norm": 1.6982362270355225,
+      "learning_rate": 7.262947321627029e-05,
+      "loss": 2.031,
+      "step": 28662
+    },
+    {
+      "epoch": 0.6369555555555556,
+      "grad_norm": 2.1967759132385254,
+      "learning_rate": 7.2625027783952e-05,
+      "loss": 2.1992,
+      "step": 28663
+    },
+    {
+      "epoch": 0.6369777777777778,
+      "grad_norm": 1.2060320377349854,
+      "learning_rate": 7.26205823516337e-05,
+      "loss": 0.92,
+      "step": 28664
+    },
+    {
+      "epoch": 0.637,
+      "grad_norm": 1.366573691368103,
+      "learning_rate": 7.26161369193154e-05,
+      "loss": 1.611,
+      "step": 28665
+    },
+    {
+      "epoch": 0.6370222222222223,
+      "grad_norm": 1.792948842048645,
+      "learning_rate": 7.261169148699711e-05,
+      "loss": 1.9003,
+      "step": 28666
+    },
+    {
+      "epoch": 0.6370444444444444,
+      "grad_norm": 1.3751440048217773,
+      "learning_rate": 7.260724605467882e-05,
+      "loss": 1.7083,
+      "step": 28667
+    },
+    {
+      "epoch": 0.6370666666666667,
+      "grad_norm": 1.6102418899536133,
+      "learning_rate": 7.260280062236052e-05,
+      "loss": 1.9591,
+      "step": 28668
+    },
+    {
+      "epoch": 0.6370888888888889,
+      "grad_norm": 2.0655267238616943,
+      "learning_rate": 7.259835519004224e-05,
+      "loss": 2.3119,
+      "step": 28669
+    },
+    {
+      "epoch": 0.6371111111111111,
+      "grad_norm": 1.8483701944351196,
+      "learning_rate": 7.259390975772395e-05,
+      "loss": 2.1266,
+      "step": 28670
+    },
+    {
+      "epoch": 0.6371333333333333,
+      "grad_norm": 1.8502991199493408,
+      "learning_rate": 7.258946432540566e-05,
+      "loss": 2.212,
+      "step": 28671
+    },
+    {
+      "epoch": 0.6371555555555556,
+      "grad_norm": 1.7456656694412231,
+      "learning_rate": 7.258501889308735e-05,
+      "loss": 1.7485,
+      "step": 28672
+    },
+    {
+      "epoch": 0.6371777777777777,
+      "grad_norm": 1.6352704763412476,
+      "learning_rate": 7.258057346076906e-05,
+      "loss": 1.7125,
+      "step": 28673
+    },
+    {
+      "epoch": 0.6372,
+      "grad_norm": 1.7602741718292236,
+      "learning_rate": 7.257612802845077e-05,
+      "loss": 2.0005,
+      "step": 28674
+    },
+    {
+      "epoch": 0.6372222222222222,
+      "grad_norm": 1.2574185132980347,
+      "learning_rate": 7.257168259613247e-05,
+      "loss": 1.2635,
+      "step": 28675
+    },
+    {
+      "epoch": 0.6372444444444444,
+      "grad_norm": 1.5207939147949219,
+      "learning_rate": 7.256723716381418e-05,
+      "loss": 1.473,
+      "step": 28676
+    },
+    {
+      "epoch": 0.6372666666666666,
+      "grad_norm": 1.6920651197433472,
+      "learning_rate": 7.256279173149589e-05,
+      "loss": 1.9603,
+      "step": 28677
+    },
+    {
+      "epoch": 0.6372888888888889,
+      "grad_norm": 1.7033255100250244,
+      "learning_rate": 7.25583462991776e-05,
+      "loss": 1.6898,
+      "step": 28678
+    },
+    {
+      "epoch": 0.6373111111111112,
+      "grad_norm": 1.9950025081634521,
+      "learning_rate": 7.255390086685931e-05,
+      "loss": 2.1322,
+      "step": 28679
+    },
+    {
+      "epoch": 0.6373333333333333,
+      "grad_norm": 1.4635943174362183,
+      "learning_rate": 7.254945543454102e-05,
+      "loss": 1.3099,
+      "step": 28680
+    },
+    {
+      "epoch": 0.6373555555555556,
+      "grad_norm": 1.6194703578948975,
+      "learning_rate": 7.254501000222273e-05,
+      "loss": 1.7211,
+      "step": 28681
+    },
+    {
+      "epoch": 0.6373777777777778,
+      "grad_norm": 1.7633742094039917,
+      "learning_rate": 7.254056456990442e-05,
+      "loss": 1.8104,
+      "step": 28682
+    },
+    {
+      "epoch": 0.6374,
+      "grad_norm": 1.9420922994613647,
+      "learning_rate": 7.253611913758613e-05,
+      "loss": 1.694,
+      "step": 28683
+    },
+    {
+      "epoch": 0.6374222222222222,
+      "grad_norm": 1.8654829263687134,
+      "learning_rate": 7.253167370526784e-05,
+      "loss": 1.8214,
+      "step": 28684
+    },
+    {
+      "epoch": 0.6374444444444445,
+      "grad_norm": 2.2684643268585205,
+      "learning_rate": 7.252722827294954e-05,
+      "loss": 2.2231,
+      "step": 28685
+    },
+    {
+      "epoch": 0.6374666666666666,
+      "grad_norm": 2.0050923824310303,
+      "learning_rate": 7.252278284063125e-05,
+      "loss": 2.0718,
+      "step": 28686
+    },
+    {
+      "epoch": 0.6374888888888889,
+      "grad_norm": 1.8627880811691284,
+      "learning_rate": 7.251833740831296e-05,
+      "loss": 1.8782,
+      "step": 28687
+    },
+    {
+      "epoch": 0.6375111111111111,
+      "grad_norm": 1.5710184574127197,
+      "learning_rate": 7.251389197599467e-05,
+      "loss": 1.5789,
+      "step": 28688
+    },
+    {
+      "epoch": 0.6375333333333333,
+      "grad_norm": 1.7555488348007202,
+      "learning_rate": 7.250944654367638e-05,
+      "loss": 1.8122,
+      "step": 28689
+    },
+    {
+      "epoch": 0.6375555555555555,
+      "grad_norm": 1.7349294424057007,
+      "learning_rate": 7.250500111135809e-05,
+      "loss": 1.8575,
+      "step": 28690
+    },
+    {
+      "epoch": 0.6375777777777778,
+      "grad_norm": 1.8430469036102295,
+      "learning_rate": 7.25005556790398e-05,
+      "loss": 1.6942,
+      "step": 28691
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 1.8953179121017456,
+      "learning_rate": 7.249611024672149e-05,
+      "loss": 1.504,
+      "step": 28692
+    },
+    {
+      "epoch": 0.6376222222222222,
+      "grad_norm": 1.7643966674804688,
+      "learning_rate": 7.24916648144032e-05,
+      "loss": 1.7468,
+      "step": 28693
+    },
+    {
+      "epoch": 0.6376444444444445,
+      "grad_norm": 1.9069668054580688,
+      "learning_rate": 7.248721938208491e-05,
+      "loss": 2.0379,
+      "step": 28694
+    },
+    {
+      "epoch": 0.6376666666666667,
+      "grad_norm": 1.7193044424057007,
+      "learning_rate": 7.248277394976661e-05,
+      "loss": 1.5639,
+      "step": 28695
+    },
+    {
+      "epoch": 0.6376888888888889,
+      "grad_norm": 1.7010250091552734,
+      "learning_rate": 7.247832851744832e-05,
+      "loss": 1.7881,
+      "step": 28696
+    },
+    {
+      "epoch": 0.6377111111111111,
+      "grad_norm": 1.6988699436187744,
+      "learning_rate": 7.247388308513004e-05,
+      "loss": 1.9145,
+      "step": 28697
+    },
+    {
+      "epoch": 0.6377333333333334,
+      "grad_norm": 1.434458613395691,
+      "learning_rate": 7.246943765281175e-05,
+      "loss": 1.2502,
+      "step": 28698
+    },
+    {
+      "epoch": 0.6377555555555555,
+      "grad_norm": 1.992437481880188,
+      "learning_rate": 7.246499222049345e-05,
+      "loss": 1.8448,
+      "step": 28699
+    },
+    {
+      "epoch": 0.6377777777777778,
+      "grad_norm": 1.5702959299087524,
+      "learning_rate": 7.246054678817516e-05,
+      "loss": 1.3579,
+      "step": 28700
+    },
+    {
+      "epoch": 0.6378,
+      "grad_norm": 1.6802828311920166,
+      "learning_rate": 7.245610135585687e-05,
+      "loss": 2.341,
+      "step": 28701
+    },
+    {
+      "epoch": 0.6378222222222222,
+      "grad_norm": 1.627329707145691,
+      "learning_rate": 7.245165592353856e-05,
+      "loss": 2.8105,
+      "step": 28702
+    },
+    {
+      "epoch": 0.6378444444444444,
+      "grad_norm": 1.342093586921692,
+      "learning_rate": 7.244721049122027e-05,
+      "loss": 1.9832,
+      "step": 28703
+    },
+    {
+      "epoch": 0.6378666666666667,
+      "grad_norm": 1.5474640130996704,
+      "learning_rate": 7.244276505890198e-05,
+      "loss": 2.4607,
+      "step": 28704
+    },
+    {
+      "epoch": 0.6378888888888888,
+      "grad_norm": 1.531585454940796,
+      "learning_rate": 7.243831962658368e-05,
+      "loss": 2.0354,
+      "step": 28705
+    },
+    {
+      "epoch": 0.6379111111111111,
+      "grad_norm": 1.5042400360107422,
+      "learning_rate": 7.24338741942654e-05,
+      "loss": 2.1718,
+      "step": 28706
+    },
+    {
+      "epoch": 0.6379333333333334,
+      "grad_norm": 1.3284921646118164,
+      "learning_rate": 7.242942876194711e-05,
+      "loss": 1.6979,
+      "step": 28707
+    },
+    {
+      "epoch": 0.6379555555555556,
+      "grad_norm": 1.4425671100616455,
+      "learning_rate": 7.242498332962882e-05,
+      "loss": 1.9872,
+      "step": 28708
+    },
+    {
+      "epoch": 0.6379777777777778,
+      "grad_norm": 1.7384395599365234,
+      "learning_rate": 7.242053789731052e-05,
+      "loss": 2.2414,
+      "step": 28709
+    },
+    {
+      "epoch": 0.638,
+      "grad_norm": 1.554915189743042,
+      "learning_rate": 7.241609246499223e-05,
+      "loss": 2.3454,
+      "step": 28710
+    },
+    {
+      "epoch": 0.6380222222222223,
+      "grad_norm": 1.8486429452896118,
+      "learning_rate": 7.241164703267394e-05,
+      "loss": 2.168,
+      "step": 28711
+    },
+    {
+      "epoch": 0.6380444444444444,
+      "grad_norm": 1.7188366651535034,
+      "learning_rate": 7.240720160035563e-05,
+      "loss": 2.25,
+      "step": 28712
+    },
+    {
+      "epoch": 0.6380666666666667,
+      "grad_norm": 1.8007780313491821,
+      "learning_rate": 7.240275616803734e-05,
+      "loss": 2.4567,
+      "step": 28713
+    },
+    {
+      "epoch": 0.6380888888888889,
+      "grad_norm": 1.6362093687057495,
+      "learning_rate": 7.239831073571905e-05,
+      "loss": 2.0876,
+      "step": 28714
+    },
+    {
+      "epoch": 0.6381111111111111,
+      "grad_norm": 1.621285319328308,
+      "learning_rate": 7.239386530340076e-05,
+      "loss": 2.183,
+      "step": 28715
+    },
+    {
+      "epoch": 0.6381333333333333,
+      "grad_norm": 1.888231635093689,
+      "learning_rate": 7.238941987108247e-05,
+      "loss": 2.0216,
+      "step": 28716
+    },
+    {
+      "epoch": 0.6381555555555556,
+      "grad_norm": 1.4312587976455688,
+      "learning_rate": 7.238497443876418e-05,
+      "loss": 1.6035,
+      "step": 28717
+    },
+    {
+      "epoch": 0.6381777777777777,
+      "grad_norm": 1.1737911701202393,
+      "learning_rate": 7.238052900644589e-05,
+      "loss": 0.6206,
+      "step": 28718
+    },
+    {
+      "epoch": 0.6382,
+      "grad_norm": 1.8380234241485596,
+      "learning_rate": 7.237608357412758e-05,
+      "loss": 2.3758,
+      "step": 28719
+    },
+    {
+      "epoch": 0.6382222222222222,
+      "grad_norm": 1.6413174867630005,
+      "learning_rate": 7.23716381418093e-05,
+      "loss": 1.7652,
+      "step": 28720
+    },
+    {
+      "epoch": 0.6382444444444444,
+      "grad_norm": 1.5186172723770142,
+      "learning_rate": 7.2367192709491e-05,
+      "loss": 1.9663,
+      "step": 28721
+    },
+    {
+      "epoch": 0.6382666666666666,
+      "grad_norm": 1.8024673461914062,
+      "learning_rate": 7.23627472771727e-05,
+      "loss": 2.0255,
+      "step": 28722
+    },
+    {
+      "epoch": 0.6382888888888889,
+      "grad_norm": 1.5912338495254517,
+      "learning_rate": 7.235830184485441e-05,
+      "loss": 1.7869,
+      "step": 28723
+    },
+    {
+      "epoch": 0.6383111111111112,
+      "grad_norm": 1.888502836227417,
+      "learning_rate": 7.235385641253612e-05,
+      "loss": 2.0113,
+      "step": 28724
+    },
+    {
+      "epoch": 0.6383333333333333,
+      "grad_norm": 1.4829522371292114,
+      "learning_rate": 7.234941098021783e-05,
+      "loss": 1.7279,
+      "step": 28725
+    },
+    {
+      "epoch": 0.6383555555555556,
+      "grad_norm": 1.761410117149353,
+      "learning_rate": 7.234496554789954e-05,
+      "loss": 1.6857,
+      "step": 28726
+    },
+    {
+      "epoch": 0.6383777777777778,
+      "grad_norm": 1.7559994459152222,
+      "learning_rate": 7.234052011558125e-05,
+      "loss": 2.2451,
+      "step": 28727
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 1.6013765335083008,
+      "learning_rate": 7.233607468326296e-05,
+      "loss": 1.6825,
+      "step": 28728
+    },
+    {
+      "epoch": 0.6384222222222222,
+      "grad_norm": 1.9847111701965332,
+      "learning_rate": 7.233162925094465e-05,
+      "loss": 2.1672,
+      "step": 28729
+    },
+    {
+      "epoch": 0.6384444444444445,
+      "grad_norm": 1.9318690299987793,
+      "learning_rate": 7.232718381862636e-05,
+      "loss": 1.7432,
+      "step": 28730
+    },
+    {
+      "epoch": 0.6384666666666666,
+      "grad_norm": 1.307256817817688,
+      "learning_rate": 7.232273838630807e-05,
+      "loss": 0.9803,
+      "step": 28731
+    },
+    {
+      "epoch": 0.6384888888888889,
+      "grad_norm": 1.7495099306106567,
+      "learning_rate": 7.231829295398977e-05,
+      "loss": 1.606,
+      "step": 28732
+    },
+    {
+      "epoch": 0.6385111111111111,
+      "grad_norm": 1.1634020805358887,
+      "learning_rate": 7.231384752167148e-05,
+      "loss": 0.9347,
+      "step": 28733
+    },
+    {
+      "epoch": 0.6385333333333333,
+      "grad_norm": 2.0177419185638428,
+      "learning_rate": 7.23094020893532e-05,
+      "loss": 2.1003,
+      "step": 28734
+    },
+    {
+      "epoch": 0.6385555555555555,
+      "grad_norm": 1.7657959461212158,
+      "learning_rate": 7.23049566570349e-05,
+      "loss": 1.7731,
+      "step": 28735
+    },
+    {
+      "epoch": 0.6385777777777778,
+      "grad_norm": 1.8546388149261475,
+      "learning_rate": 7.230051122471661e-05,
+      "loss": 1.5486,
+      "step": 28736
+    },
+    {
+      "epoch": 0.6386,
+      "grad_norm": 1.8040077686309814,
+      "learning_rate": 7.229606579239832e-05,
+      "loss": 2.0225,
+      "step": 28737
+    },
+    {
+      "epoch": 0.6386222222222222,
+      "grad_norm": 1.7609293460845947,
+      "learning_rate": 7.229162036008003e-05,
+      "loss": 2.2548,
+      "step": 28738
+    },
+    {
+      "epoch": 0.6386444444444445,
+      "grad_norm": 2.3687591552734375,
+      "learning_rate": 7.228717492776172e-05,
+      "loss": 2.1545,
+      "step": 28739
+    },
+    {
+      "epoch": 0.6386666666666667,
+      "grad_norm": 3.026139259338379,
+      "learning_rate": 7.228272949544343e-05,
+      "loss": 2.0457,
+      "step": 28740
+    },
+    {
+      "epoch": 0.6386888888888889,
+      "grad_norm": 2.389371395111084,
+      "learning_rate": 7.227828406312514e-05,
+      "loss": 2.1006,
+      "step": 28741
+    },
+    {
+      "epoch": 0.6387111111111111,
+      "grad_norm": 1.8829636573791504,
+      "learning_rate": 7.227383863080684e-05,
+      "loss": 2.157,
+      "step": 28742
+    },
+    {
+      "epoch": 0.6387333333333334,
+      "grad_norm": 1.4691171646118164,
+      "learning_rate": 7.226939319848856e-05,
+      "loss": 1.1827,
+      "step": 28743
+    },
+    {
+      "epoch": 0.6387555555555555,
+      "grad_norm": 1.8725281953811646,
+      "learning_rate": 7.226494776617027e-05,
+      "loss": 2.0211,
+      "step": 28744
+    },
+    {
+      "epoch": 0.6387777777777778,
+      "grad_norm": 2.3019354343414307,
+      "learning_rate": 7.226050233385198e-05,
+      "loss": 1.9218,
+      "step": 28745
+    },
+    {
+      "epoch": 0.6388,
+      "grad_norm": 2.649057626724243,
+      "learning_rate": 7.225605690153368e-05,
+      "loss": 2.4973,
+      "step": 28746
+    },
+    {
+      "epoch": 0.6388222222222222,
+      "grad_norm": 2.0913870334625244,
+      "learning_rate": 7.225161146921539e-05,
+      "loss": 1.7488,
+      "step": 28747
+    },
+    {
+      "epoch": 0.6388444444444444,
+      "grad_norm": 1.7658172845840454,
+      "learning_rate": 7.22471660368971e-05,
+      "loss": 1.3801,
+      "step": 28748
+    },
+    {
+      "epoch": 0.6388666666666667,
+      "grad_norm": 1.7374156713485718,
+      "learning_rate": 7.224272060457879e-05,
+      "loss": 1.6473,
+      "step": 28749
+    },
+    {
+      "epoch": 0.6388888888888888,
+      "grad_norm": 1.4828306436538696,
+      "learning_rate": 7.22382751722605e-05,
+      "loss": 0.9041,
+      "step": 28750
+    },
+    {
+      "epoch": 0.6389111111111111,
+      "grad_norm": 1.3354045152664185,
+      "learning_rate": 7.223382973994221e-05,
+      "loss": 2.0015,
+      "step": 28751
+    },
+    {
+      "epoch": 0.6389333333333334,
+      "grad_norm": 2.6666393280029297,
+      "learning_rate": 7.222938430762392e-05,
+      "loss": 0.0448,
+      "step": 28752
+    },
+    {
+      "epoch": 0.6389555555555556,
+      "grad_norm": 1.3696006536483765,
+      "learning_rate": 7.222493887530563e-05,
+      "loss": 1.8281,
+      "step": 28753
+    },
+    {
+      "epoch": 0.6389777777777778,
+      "grad_norm": 1.4097464084625244,
+      "learning_rate": 7.222049344298734e-05,
+      "loss": 1.2002,
+      "step": 28754
+    },
+    {
+      "epoch": 0.639,
+      "grad_norm": 1.7855876684188843,
+      "learning_rate": 7.221604801066905e-05,
+      "loss": 2.0857,
+      "step": 28755
+    },
+    {
+      "epoch": 0.6390222222222223,
+      "grad_norm": 1.5594568252563477,
+      "learning_rate": 7.221160257835075e-05,
+      "loss": 1.6816,
+      "step": 28756
+    },
+    {
+      "epoch": 0.6390444444444444,
+      "grad_norm": 1.421626091003418,
+      "learning_rate": 7.220715714603246e-05,
+      "loss": 1.7824,
+      "step": 28757
+    },
+    {
+      "epoch": 0.6390666666666667,
+      "grad_norm": 2.0007307529449463,
+      "learning_rate": 7.220271171371417e-05,
+      "loss": 2.2619,
+      "step": 28758
+    },
+    {
+      "epoch": 0.6390888888888889,
+      "grad_norm": 1.9237414598464966,
+      "learning_rate": 7.219826628139586e-05,
+      "loss": 2.0441,
+      "step": 28759
+    },
+    {
+      "epoch": 0.6391111111111111,
+      "grad_norm": 1.8039405345916748,
+      "learning_rate": 7.219382084907757e-05,
+      "loss": 1.708,
+      "step": 28760
+    },
+    {
+      "epoch": 0.6391333333333333,
+      "grad_norm": 1.5210480690002441,
+      "learning_rate": 7.218937541675928e-05,
+      "loss": 2.385,
+      "step": 28761
+    },
+    {
+      "epoch": 0.6391555555555556,
+      "grad_norm": 1.5081499814987183,
+      "learning_rate": 7.218492998444099e-05,
+      "loss": 1.2788,
+      "step": 28762
+    },
+    {
+      "epoch": 0.6391777777777777,
+      "grad_norm": 1.9141021966934204,
+      "learning_rate": 7.21804845521227e-05,
+      "loss": 1.3567,
+      "step": 28763
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 3.3286590576171875,
+      "learning_rate": 7.217603911980441e-05,
+      "loss": 1.9065,
+      "step": 28764
+    },
+    {
+      "epoch": 0.6392222222222222,
+      "grad_norm": 0.9982495307922363,
+      "learning_rate": 7.217159368748612e-05,
+      "loss": 0.7088,
+      "step": 28765
+    },
+    {
+      "epoch": 0.6392444444444444,
+      "grad_norm": 1.485310673713684,
+      "learning_rate": 7.216714825516782e-05,
+      "loss": 1.1706,
+      "step": 28766
+    },
+    {
+      "epoch": 0.6392666666666666,
+      "grad_norm": 1.6460174322128296,
+      "learning_rate": 7.216270282284952e-05,
+      "loss": 1.7584,
+      "step": 28767
+    },
+    {
+      "epoch": 0.6392888888888889,
+      "grad_norm": 1.8000679016113281,
+      "learning_rate": 7.215825739053123e-05,
+      "loss": 1.9528,
+      "step": 28768
+    },
+    {
+      "epoch": 0.6393111111111112,
+      "grad_norm": 1.925246000289917,
+      "learning_rate": 7.215381195821293e-05,
+      "loss": 2.0873,
+      "step": 28769
+    },
+    {
+      "epoch": 0.6393333333333333,
+      "grad_norm": 1.6026747226715088,
+      "learning_rate": 7.214936652589464e-05,
+      "loss": 1.6128,
+      "step": 28770
+    },
+    {
+      "epoch": 0.6393555555555556,
+      "grad_norm": 2.411868095397949,
+      "learning_rate": 7.214492109357636e-05,
+      "loss": 2.344,
+      "step": 28771
+    },
+    {
+      "epoch": 0.6393777777777778,
+      "grad_norm": 1.7101866006851196,
+      "learning_rate": 7.214047566125806e-05,
+      "loss": 1.7117,
+      "step": 28772
+    },
+    {
+      "epoch": 0.6394,
+      "grad_norm": 1.8014975786209106,
+      "learning_rate": 7.213603022893977e-05,
+      "loss": 1.6977,
+      "step": 28773
+    },
+    {
+      "epoch": 0.6394222222222222,
+      "grad_norm": 1.1885337829589844,
+      "learning_rate": 7.213158479662148e-05,
+      "loss": 1.0145,
+      "step": 28774
+    },
+    {
+      "epoch": 0.6394444444444445,
+      "grad_norm": 1.8605204820632935,
+      "learning_rate": 7.212713936430319e-05,
+      "loss": 2.1966,
+      "step": 28775
+    },
+    {
+      "epoch": 0.6394666666666666,
+      "grad_norm": 1.8864420652389526,
+      "learning_rate": 7.212269393198488e-05,
+      "loss": 1.8072,
+      "step": 28776
+    },
+    {
+      "epoch": 0.6394888888888889,
+      "grad_norm": 1.7543187141418457,
+      "learning_rate": 7.21182484996666e-05,
+      "loss": 1.6595,
+      "step": 28777
+    },
+    {
+      "epoch": 0.6395111111111111,
+      "grad_norm": 1.7286779880523682,
+      "learning_rate": 7.21138030673483e-05,
+      "loss": 1.7274,
+      "step": 28778
+    },
+    {
+      "epoch": 0.6395333333333333,
+      "grad_norm": 1.0491361618041992,
+      "learning_rate": 7.210935763503e-05,
+      "loss": 0.7481,
+      "step": 28779
+    },
+    {
+      "epoch": 0.6395555555555555,
+      "grad_norm": 1.604767084121704,
+      "learning_rate": 7.210491220271172e-05,
+      "loss": 1.6491,
+      "step": 28780
+    },
+    {
+      "epoch": 0.6395777777777778,
+      "grad_norm": 1.9371610879898071,
+      "learning_rate": 7.210046677039343e-05,
+      "loss": 1.691,
+      "step": 28781
+    },
+    {
+      "epoch": 0.6396,
+      "grad_norm": 1.6699118614196777,
+      "learning_rate": 7.209602133807513e-05,
+      "loss": 1.7054,
+      "step": 28782
+    },
+    {
+      "epoch": 0.6396222222222222,
+      "grad_norm": 2.032633066177368,
+      "learning_rate": 7.209157590575684e-05,
+      "loss": 1.6071,
+      "step": 28783
+    },
+    {
+      "epoch": 0.6396444444444445,
+      "grad_norm": 2.003582239151001,
+      "learning_rate": 7.208713047343855e-05,
+      "loss": 2.2553,
+      "step": 28784
+    },
+    {
+      "epoch": 0.6396666666666667,
+      "grad_norm": 1.4603201150894165,
+      "learning_rate": 7.208268504112026e-05,
+      "loss": 1.6134,
+      "step": 28785
+    },
+    {
+      "epoch": 0.6396888888888889,
+      "grad_norm": 2.1268177032470703,
+      "learning_rate": 7.207823960880195e-05,
+      "loss": 1.8333,
+      "step": 28786
+    },
+    {
+      "epoch": 0.6397111111111111,
+      "grad_norm": 2.0532491207122803,
+      "learning_rate": 7.207379417648366e-05,
+      "loss": 1.9751,
+      "step": 28787
+    },
+    {
+      "epoch": 0.6397333333333334,
+      "grad_norm": 2.148063898086548,
+      "learning_rate": 7.206934874416537e-05,
+      "loss": 2.6461,
+      "step": 28788
+    },
+    {
+      "epoch": 0.6397555555555555,
+      "grad_norm": 2.0925416946411133,
+      "learning_rate": 7.206490331184708e-05,
+      "loss": 1.6654,
+      "step": 28789
+    },
+    {
+      "epoch": 0.6397777777777778,
+      "grad_norm": 1.6371979713439941,
+      "learning_rate": 7.206045787952879e-05,
+      "loss": 1.7088,
+      "step": 28790
+    },
+    {
+      "epoch": 0.6398,
+      "grad_norm": 2.2275896072387695,
+      "learning_rate": 7.20560124472105e-05,
+      "loss": 2.0222,
+      "step": 28791
+    },
+    {
+      "epoch": 0.6398222222222222,
+      "grad_norm": 1.7818533182144165,
+      "learning_rate": 7.20515670148922e-05,
+      "loss": 1.7862,
+      "step": 28792
+    },
+    {
+      "epoch": 0.6398444444444444,
+      "grad_norm": 1.8481240272521973,
+      "learning_rate": 7.204712158257391e-05,
+      "loss": 1.4648,
+      "step": 28793
+    },
+    {
+      "epoch": 0.6398666666666667,
+      "grad_norm": 1.8847888708114624,
+      "learning_rate": 7.204267615025562e-05,
+      "loss": 2.2493,
+      "step": 28794
+    },
+    {
+      "epoch": 0.6398888888888888,
+      "grad_norm": 2.177549362182617,
+      "learning_rate": 7.203823071793733e-05,
+      "loss": 2.3858,
+      "step": 28795
+    },
+    {
+      "epoch": 0.6399111111111111,
+      "grad_norm": 1.6145027875900269,
+      "learning_rate": 7.203378528561902e-05,
+      "loss": 1.475,
+      "step": 28796
+    },
+    {
+      "epoch": 0.6399333333333334,
+      "grad_norm": 1.8633002042770386,
+      "learning_rate": 7.202933985330073e-05,
+      "loss": 2.1072,
+      "step": 28797
+    },
+    {
+      "epoch": 0.6399555555555556,
+      "grad_norm": 1.697845220565796,
+      "learning_rate": 7.202489442098244e-05,
+      "loss": 1.5116,
+      "step": 28798
+    },
+    {
+      "epoch": 0.6399777777777778,
+      "grad_norm": 1.8304460048675537,
+      "learning_rate": 7.202044898866415e-05,
+      "loss": 2.1342,
+      "step": 28799
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6719493865966797,
+      "learning_rate": 7.201600355634586e-05,
+      "loss": 1.5713,
+      "step": 28800
+    },
+    {
+      "epoch": 0.6400222222222223,
+      "grad_norm": 1.6436892747879028,
+      "learning_rate": 7.201155812402757e-05,
+      "loss": 2.2734,
+      "step": 28801
+    },
+    {
+      "epoch": 0.6400444444444444,
+      "grad_norm": 0.18774180114269257,
+      "learning_rate": 7.200711269170928e-05,
+      "loss": 0.0162,
+      "step": 28802
+    },
+    {
+      "epoch": 0.6400666666666667,
+      "grad_norm": 1.8824647665023804,
+      "learning_rate": 7.200266725939098e-05,
+      "loss": 2.5967,
+      "step": 28803
+    },
+    {
+      "epoch": 0.6400888888888889,
+      "grad_norm": 1.5691263675689697,
+      "learning_rate": 7.199822182707269e-05,
+      "loss": 2.3886,
+      "step": 28804
+    },
+    {
+      "epoch": 0.6401111111111111,
+      "grad_norm": 1.7049403190612793,
+      "learning_rate": 7.19937763947544e-05,
+      "loss": 2.1433,
+      "step": 28805
+    },
+    {
+      "epoch": 0.6401333333333333,
+      "grad_norm": 1.461273193359375,
+      "learning_rate": 7.198933096243609e-05,
+      "loss": 1.9639,
+      "step": 28806
+    },
+    {
+      "epoch": 0.6401555555555556,
+      "grad_norm": 1.6055644750595093,
+      "learning_rate": 7.19848855301178e-05,
+      "loss": 2.2601,
+      "step": 28807
+    },
+    {
+      "epoch": 0.6401777777777777,
+      "grad_norm": 1.67502760887146,
+      "learning_rate": 7.198044009779952e-05,
+      "loss": 2.0551,
+      "step": 28808
+    },
+    {
+      "epoch": 0.6402,
+      "grad_norm": 1.8794890642166138,
+      "learning_rate": 7.197599466548122e-05,
+      "loss": 2.1515,
+      "step": 28809
+    },
+    {
+      "epoch": 0.6402222222222222,
+      "grad_norm": 1.5921684503555298,
+      "learning_rate": 7.197154923316293e-05,
+      "loss": 2.0577,
+      "step": 28810
+    },
+    {
+      "epoch": 0.6402444444444444,
+      "grad_norm": 1.6569201946258545,
+      "learning_rate": 7.196710380084464e-05,
+      "loss": 1.8993,
+      "step": 28811
+    },
+    {
+      "epoch": 0.6402666666666667,
+      "grad_norm": 1.6281111240386963,
+      "learning_rate": 7.196265836852635e-05,
+      "loss": 2.0419,
+      "step": 28812
+    },
+    {
+      "epoch": 0.6402888888888889,
+      "grad_norm": 1.9701122045516968,
+      "learning_rate": 7.195821293620805e-05,
+      "loss": 2.1372,
+      "step": 28813
+    },
+    {
+      "epoch": 0.6403111111111112,
+      "grad_norm": 1.0676236152648926,
+      "learning_rate": 7.195376750388975e-05,
+      "loss": 1.1958,
+      "step": 28814
+    },
+    {
+      "epoch": 0.6403333333333333,
+      "grad_norm": 1.5588091611862183,
+      "learning_rate": 7.194932207157146e-05,
+      "loss": 1.4684,
+      "step": 28815
+    },
+    {
+      "epoch": 0.6403555555555556,
+      "grad_norm": 1.7817189693450928,
+      "learning_rate": 7.194487663925316e-05,
+      "loss": 2.1338,
+      "step": 28816
+    },
+    {
+      "epoch": 0.6403777777777778,
+      "grad_norm": 1.8440243005752563,
+      "learning_rate": 7.194043120693488e-05,
+      "loss": 1.8603,
+      "step": 28817
+    },
+    {
+      "epoch": 0.6404,
+      "grad_norm": 1.5460695028305054,
+      "learning_rate": 7.19359857746166e-05,
+      "loss": 1.6515,
+      "step": 28818
+    },
+    {
+      "epoch": 0.6404222222222222,
+      "grad_norm": 1.8269490003585815,
+      "learning_rate": 7.193154034229829e-05,
+      "loss": 2.3875,
+      "step": 28819
+    },
+    {
+      "epoch": 0.6404444444444445,
+      "grad_norm": 1.5759791135787964,
+      "learning_rate": 7.192709490998e-05,
+      "loss": 1.9152,
+      "step": 28820
+    },
+    {
+      "epoch": 0.6404666666666666,
+      "grad_norm": 1.8602632284164429,
+      "learning_rate": 7.192264947766171e-05,
+      "loss": 2.1399,
+      "step": 28821
+    },
+    {
+      "epoch": 0.6404888888888889,
+      "grad_norm": 1.6041455268859863,
+      "learning_rate": 7.191820404534342e-05,
+      "loss": 1.9356,
+      "step": 28822
+    },
+    {
+      "epoch": 0.6405111111111111,
+      "grad_norm": 1.8999677896499634,
+      "learning_rate": 7.191375861302511e-05,
+      "loss": 2.1473,
+      "step": 28823
+    },
+    {
+      "epoch": 0.6405333333333333,
+      "grad_norm": 1.381639838218689,
+      "learning_rate": 7.190931318070682e-05,
+      "loss": 1.7347,
+      "step": 28824
+    },
+    {
+      "epoch": 0.6405555555555555,
+      "grad_norm": 2.126990795135498,
+      "learning_rate": 7.190486774838853e-05,
+      "loss": 2.4627,
+      "step": 28825
+    },
+    {
+      "epoch": 0.6405777777777778,
+      "grad_norm": 1.8442226648330688,
+      "learning_rate": 7.190042231607024e-05,
+      "loss": 1.6449,
+      "step": 28826
+    },
+    {
+      "epoch": 0.6406,
+      "grad_norm": 1.4026455879211426,
+      "learning_rate": 7.189597688375195e-05,
+      "loss": 1.5099,
+      "step": 28827
+    },
+    {
+      "epoch": 0.6406222222222222,
+      "grad_norm": 2.226283550262451,
+      "learning_rate": 7.189153145143366e-05,
+      "loss": 1.5443,
+      "step": 28828
+    },
+    {
+      "epoch": 0.6406444444444445,
+      "grad_norm": 1.6985044479370117,
+      "learning_rate": 7.188708601911536e-05,
+      "loss": 1.6088,
+      "step": 28829
+    },
+    {
+      "epoch": 0.6406666666666667,
+      "grad_norm": 1.2343090772628784,
+      "learning_rate": 7.188264058679707e-05,
+      "loss": 0.8093,
+      "step": 28830
+    },
+    {
+      "epoch": 0.6406888888888889,
+      "grad_norm": 0.5151288509368896,
+      "learning_rate": 7.187819515447878e-05,
+      "loss": 0.0383,
+      "step": 28831
+    },
+    {
+      "epoch": 0.6407111111111111,
+      "grad_norm": 1.7729060649871826,
+      "learning_rate": 7.187374972216049e-05,
+      "loss": 1.5273,
+      "step": 28832
+    },
+    {
+      "epoch": 0.6407333333333334,
+      "grad_norm": 2.00065541267395,
+      "learning_rate": 7.186930428984218e-05,
+      "loss": 2.1591,
+      "step": 28833
+    },
+    {
+      "epoch": 0.6407555555555555,
+      "grad_norm": 1.6418817043304443,
+      "learning_rate": 7.186485885752389e-05,
+      "loss": 2.2355,
+      "step": 28834
+    },
+    {
+      "epoch": 0.6407777777777778,
+      "grad_norm": 1.126987099647522,
+      "learning_rate": 7.18604134252056e-05,
+      "loss": 0.65,
+      "step": 28835
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 1.324263572692871,
+      "learning_rate": 7.185596799288731e-05,
+      "loss": 1.1635,
+      "step": 28836
+    },
+    {
+      "epoch": 0.6408222222222222,
+      "grad_norm": 1.718005895614624,
+      "learning_rate": 7.185152256056902e-05,
+      "loss": 1.6372,
+      "step": 28837
+    },
+    {
+      "epoch": 0.6408444444444444,
+      "grad_norm": 1.81002676486969,
+      "learning_rate": 7.184707712825073e-05,
+      "loss": 2.1625,
+      "step": 28838
+    },
+    {
+      "epoch": 0.6408666666666667,
+      "grad_norm": 1.6128833293914795,
+      "learning_rate": 7.184263169593243e-05,
+      "loss": 1.8423,
+      "step": 28839
+    },
+    {
+      "epoch": 0.6408888888888888,
+      "grad_norm": 1.8957856893539429,
+      "learning_rate": 7.183818626361414e-05,
+      "loss": 1.791,
+      "step": 28840
+    },
+    {
+      "epoch": 0.6409111111111111,
+      "grad_norm": 1.8511956930160522,
+      "learning_rate": 7.183374083129585e-05,
+      "loss": 1.7306,
+      "step": 28841
+    },
+    {
+      "epoch": 0.6409333333333334,
+      "grad_norm": 1.8138313293457031,
+      "learning_rate": 7.182929539897756e-05,
+      "loss": 1.9937,
+      "step": 28842
+    },
+    {
+      "epoch": 0.6409555555555555,
+      "grad_norm": 2.3374853134155273,
+      "learning_rate": 7.182484996665925e-05,
+      "loss": 2.0165,
+      "step": 28843
+    },
+    {
+      "epoch": 0.6409777777777778,
+      "grad_norm": 1.9973797798156738,
+      "learning_rate": 7.182040453434096e-05,
+      "loss": 2.0304,
+      "step": 28844
+    },
+    {
+      "epoch": 0.641,
+      "grad_norm": 1.9620558023452759,
+      "learning_rate": 7.181595910202269e-05,
+      "loss": 1.7683,
+      "step": 28845
+    },
+    {
+      "epoch": 0.6410222222222223,
+      "grad_norm": 1.861804723739624,
+      "learning_rate": 7.181151366970438e-05,
+      "loss": 1.8745,
+      "step": 28846
+    },
+    {
+      "epoch": 0.6410444444444444,
+      "grad_norm": 1.9343196153640747,
+      "learning_rate": 7.180706823738609e-05,
+      "loss": 2.1023,
+      "step": 28847
+    },
+    {
+      "epoch": 0.6410666666666667,
+      "grad_norm": 1.6749043464660645,
+      "learning_rate": 7.18026228050678e-05,
+      "loss": 1.4079,
+      "step": 28848
+    },
+    {
+      "epoch": 0.6410888888888889,
+      "grad_norm": 1.8276622295379639,
+      "learning_rate": 7.179817737274951e-05,
+      "loss": 1.7603,
+      "step": 28849
+    },
+    {
+      "epoch": 0.6411111111111111,
+      "grad_norm": 2.2006306648254395,
+      "learning_rate": 7.17937319404312e-05,
+      "loss": 1.4094,
+      "step": 28850
+    },
+    {
+      "epoch": 0.6411333333333333,
+      "grad_norm": 1.417955756187439,
+      "learning_rate": 7.178928650811292e-05,
+      "loss": 1.0786,
+      "step": 28851
+    },
+    {
+      "epoch": 0.6411555555555556,
+      "grad_norm": 1.5691580772399902,
+      "learning_rate": 7.178484107579463e-05,
+      "loss": 0.7504,
+      "step": 28852
+    },
+    {
+      "epoch": 0.6411777777777777,
+      "grad_norm": 1.7956329584121704,
+      "learning_rate": 7.178039564347632e-05,
+      "loss": 2.7077,
+      "step": 28853
+    },
+    {
+      "epoch": 0.6412,
+      "grad_norm": 0.820685088634491,
+      "learning_rate": 7.177595021115804e-05,
+      "loss": 0.9948,
+      "step": 28854
+    },
+    {
+      "epoch": 0.6412222222222222,
+      "grad_norm": 1.52669358253479,
+      "learning_rate": 7.177150477883975e-05,
+      "loss": 2.2596,
+      "step": 28855
+    },
+    {
+      "epoch": 0.6412444444444444,
+      "grad_norm": 1.7392785549163818,
+      "learning_rate": 7.176705934652145e-05,
+      "loss": 2.2223,
+      "step": 28856
+    },
+    {
+      "epoch": 0.6412666666666667,
+      "grad_norm": 1.745495319366455,
+      "learning_rate": 7.176261391420316e-05,
+      "loss": 0.0464,
+      "step": 28857
+    },
+    {
+      "epoch": 0.6412888888888889,
+      "grad_norm": 1.4235334396362305,
+      "learning_rate": 7.175816848188487e-05,
+      "loss": 2.0477,
+      "step": 28858
+    },
+    {
+      "epoch": 0.6413111111111112,
+      "grad_norm": 1.6636027097702026,
+      "learning_rate": 7.175372304956658e-05,
+      "loss": 2.2592,
+      "step": 28859
+    },
+    {
+      "epoch": 0.6413333333333333,
+      "grad_norm": 1.500736951828003,
+      "learning_rate": 7.174927761724828e-05,
+      "loss": 1.5237,
+      "step": 28860
+    },
+    {
+      "epoch": 0.6413555555555556,
+      "grad_norm": 1.7351739406585693,
+      "learning_rate": 7.174483218492998e-05,
+      "loss": 1.9768,
+      "step": 28861
+    },
+    {
+      "epoch": 0.6413777777777778,
+      "grad_norm": 1.5094364881515503,
+      "learning_rate": 7.17403867526117e-05,
+      "loss": 2.0292,
+      "step": 28862
+    },
+    {
+      "epoch": 0.6414,
+      "grad_norm": 1.594172716140747,
+      "learning_rate": 7.17359413202934e-05,
+      "loss": 2.2337,
+      "step": 28863
+    },
+    {
+      "epoch": 0.6414222222222222,
+      "grad_norm": 2.1875925064086914,
+      "learning_rate": 7.173149588797511e-05,
+      "loss": 2.3735,
+      "step": 28864
+    },
+    {
+      "epoch": 0.6414444444444445,
+      "grad_norm": 1.6282321214675903,
+      "learning_rate": 7.172705045565682e-05,
+      "loss": 2.1575,
+      "step": 28865
+    },
+    {
+      "epoch": 0.6414666666666666,
+      "grad_norm": 1.6158971786499023,
+      "learning_rate": 7.172260502333852e-05,
+      "loss": 1.9549,
+      "step": 28866
+    },
+    {
+      "epoch": 0.6414888888888889,
+      "grad_norm": 1.8376556634902954,
+      "learning_rate": 7.171815959102023e-05,
+      "loss": 1.9984,
+      "step": 28867
+    },
+    {
+      "epoch": 0.6415111111111111,
+      "grad_norm": 1.7575677633285522,
+      "learning_rate": 7.171371415870194e-05,
+      "loss": 2.0017,
+      "step": 28868
+    },
+    {
+      "epoch": 0.6415333333333333,
+      "grad_norm": 1.7002308368682861,
+      "learning_rate": 7.170926872638365e-05,
+      "loss": 1.9536,
+      "step": 28869
+    },
+    {
+      "epoch": 0.6415555555555555,
+      "grad_norm": 1.6438381671905518,
+      "learning_rate": 7.170482329406534e-05,
+      "loss": 1.4856,
+      "step": 28870
+    },
+    {
+      "epoch": 0.6415777777777778,
+      "grad_norm": 1.813004493713379,
+      "learning_rate": 7.170037786174705e-05,
+      "loss": 1.8227,
+      "step": 28871
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 1.6936862468719482,
+      "learning_rate": 7.169593242942876e-05,
+      "loss": 2.3135,
+      "step": 28872
+    },
+    {
+      "epoch": 0.6416222222222222,
+      "grad_norm": 1.5419673919677734,
+      "learning_rate": 7.169148699711047e-05,
+      "loss": 1.9524,
+      "step": 28873
+    },
+    {
+      "epoch": 0.6416444444444445,
+      "grad_norm": 1.786665916442871,
+      "learning_rate": 7.168704156479218e-05,
+      "loss": 2.0306,
+      "step": 28874
+    },
+    {
+      "epoch": 0.6416666666666667,
+      "grad_norm": 1.6579208374023438,
+      "learning_rate": 7.168259613247389e-05,
+      "loss": 1.9428,
+      "step": 28875
+    },
+    {
+      "epoch": 0.6416888888888889,
+      "grad_norm": 1.7095891237258911,
+      "learning_rate": 7.167815070015559e-05,
+      "loss": 1.9264,
+      "step": 28876
+    },
+    {
+      "epoch": 0.6417111111111111,
+      "grad_norm": 1.598313570022583,
+      "learning_rate": 7.16737052678373e-05,
+      "loss": 1.7624,
+      "step": 28877
+    },
+    {
+      "epoch": 0.6417333333333334,
+      "grad_norm": 1.8565610647201538,
+      "learning_rate": 7.166925983551901e-05,
+      "loss": 1.9579,
+      "step": 28878
+    },
+    {
+      "epoch": 0.6417555555555555,
+      "grad_norm": 1.7740898132324219,
+      "learning_rate": 7.166481440320072e-05,
+      "loss": 1.817,
+      "step": 28879
+    },
+    {
+      "epoch": 0.6417777777777778,
+      "grad_norm": 1.8127964735031128,
+      "learning_rate": 7.166036897088241e-05,
+      "loss": 2.2289,
+      "step": 28880
+    },
+    {
+      "epoch": 0.6418,
+      "grad_norm": 1.7820124626159668,
+      "learning_rate": 7.165592353856412e-05,
+      "loss": 1.7684,
+      "step": 28881
+    },
+    {
+      "epoch": 0.6418222222222222,
+      "grad_norm": 1.5686150789260864,
+      "learning_rate": 7.165147810624585e-05,
+      "loss": 2.0081,
+      "step": 28882
+    },
+    {
+      "epoch": 0.6418444444444444,
+      "grad_norm": 1.3537559509277344,
+      "learning_rate": 7.164703267392754e-05,
+      "loss": 1.4353,
+      "step": 28883
+    },
+    {
+      "epoch": 0.6418666666666667,
+      "grad_norm": 1.7080589532852173,
+      "learning_rate": 7.164258724160925e-05,
+      "loss": 1.9397,
+      "step": 28884
+    },
+    {
+      "epoch": 0.6418888888888888,
+      "grad_norm": 1.935981035232544,
+      "learning_rate": 7.163814180929096e-05,
+      "loss": 1.7046,
+      "step": 28885
+    },
+    {
+      "epoch": 0.6419111111111111,
+      "grad_norm": 1.734411597251892,
+      "learning_rate": 7.163369637697266e-05,
+      "loss": 2.0956,
+      "step": 28886
+    },
+    {
+      "epoch": 0.6419333333333334,
+      "grad_norm": 1.9430949687957764,
+      "learning_rate": 7.162925094465437e-05,
+      "loss": 1.619,
+      "step": 28887
+    },
+    {
+      "epoch": 0.6419555555555555,
+      "grad_norm": 1.137198805809021,
+      "learning_rate": 7.162480551233608e-05,
+      "loss": 0.6831,
+      "step": 28888
+    },
+    {
+      "epoch": 0.6419777777777778,
+      "grad_norm": 1.6652976274490356,
+      "learning_rate": 7.162036008001779e-05,
+      "loss": 1.8329,
+      "step": 28889
+    },
+    {
+      "epoch": 0.642,
+      "grad_norm": 1.7454249858856201,
+      "learning_rate": 7.161591464769948e-05,
+      "loss": 1.8221,
+      "step": 28890
+    },
+    {
+      "epoch": 0.6420222222222223,
+      "grad_norm": 1.548720359802246,
+      "learning_rate": 7.16114692153812e-05,
+      "loss": 1.8067,
+      "step": 28891
+    },
+    {
+      "epoch": 0.6420444444444444,
+      "grad_norm": 1.8045750856399536,
+      "learning_rate": 7.160702378306292e-05,
+      "loss": 1.3626,
+      "step": 28892
+    },
+    {
+      "epoch": 0.6420666666666667,
+      "grad_norm": 1.8985451459884644,
+      "learning_rate": 7.160257835074461e-05,
+      "loss": 1.9183,
+      "step": 28893
+    },
+    {
+      "epoch": 0.6420888888888889,
+      "grad_norm": 1.6870826482772827,
+      "learning_rate": 7.159813291842632e-05,
+      "loss": 1.8397,
+      "step": 28894
+    },
+    {
+      "epoch": 0.6421111111111111,
+      "grad_norm": 1.941114902496338,
+      "learning_rate": 7.159368748610803e-05,
+      "loss": 2.1521,
+      "step": 28895
+    },
+    {
+      "epoch": 0.6421333333333333,
+      "grad_norm": 1.5407544374465942,
+      "learning_rate": 7.158924205378973e-05,
+      "loss": 1.6867,
+      "step": 28896
+    },
+    {
+      "epoch": 0.6421555555555556,
+      "grad_norm": 1.2025772333145142,
+      "learning_rate": 7.158479662147144e-05,
+      "loss": 0.8019,
+      "step": 28897
+    },
+    {
+      "epoch": 0.6421777777777777,
+      "grad_norm": 1.5493876934051514,
+      "learning_rate": 7.158035118915315e-05,
+      "loss": 1.5128,
+      "step": 28898
+    },
+    {
+      "epoch": 0.6422,
+      "grad_norm": 1.7517184019088745,
+      "learning_rate": 7.157590575683486e-05,
+      "loss": 1.6096,
+      "step": 28899
+    },
+    {
+      "epoch": 0.6422222222222222,
+      "grad_norm": 1.934634804725647,
+      "learning_rate": 7.157146032451657e-05,
+      "loss": 1.3253,
+      "step": 28900
+    },
+    {
+      "epoch": 0.6422444444444444,
+      "grad_norm": 1.406434416770935,
+      "learning_rate": 7.156701489219827e-05,
+      "loss": 1.9834,
+      "step": 28901
+    },
+    {
+      "epoch": 0.6422666666666667,
+      "grad_norm": 1.0627119541168213,
+      "learning_rate": 7.156256945987998e-05,
+      "loss": 1.5022,
+      "step": 28902
+    },
+    {
+      "epoch": 0.6422888888888889,
+      "grad_norm": 1.4767934083938599,
+      "learning_rate": 7.155812402756168e-05,
+      "loss": 2.177,
+      "step": 28903
+    },
+    {
+      "epoch": 0.6423111111111112,
+      "grad_norm": 1.4371474981307983,
+      "learning_rate": 7.155367859524339e-05,
+      "loss": 2.339,
+      "step": 28904
+    },
+    {
+      "epoch": 0.6423333333333333,
+      "grad_norm": 1.7908233404159546,
+      "learning_rate": 7.15492331629251e-05,
+      "loss": 2.4955,
+      "step": 28905
+    },
+    {
+      "epoch": 0.6423555555555556,
+      "grad_norm": 1.7578809261322021,
+      "learning_rate": 7.154478773060681e-05,
+      "loss": 2.1955,
+      "step": 28906
+    },
+    {
+      "epoch": 0.6423777777777778,
+      "grad_norm": 1.7062569856643677,
+      "learning_rate": 7.15403422982885e-05,
+      "loss": 2.7585,
+      "step": 28907
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 1.4424821138381958,
+      "learning_rate": 7.153589686597021e-05,
+      "loss": 2.0638,
+      "step": 28908
+    },
+    {
+      "epoch": 0.6424222222222222,
+      "grad_norm": 1.797548770904541,
+      "learning_rate": 7.153145143365192e-05,
+      "loss": 2.3194,
+      "step": 28909
+    },
+    {
+      "epoch": 0.6424444444444445,
+      "grad_norm": 2.008291721343994,
+      "learning_rate": 7.152700600133363e-05,
+      "loss": 2.2287,
+      "step": 28910
+    },
+    {
+      "epoch": 0.6424666666666666,
+      "grad_norm": 1.5889003276824951,
+      "learning_rate": 7.152256056901534e-05,
+      "loss": 1.4463,
+      "step": 28911
+    },
+    {
+      "epoch": 0.6424888888888889,
+      "grad_norm": 1.6907141208648682,
+      "learning_rate": 7.151811513669705e-05,
+      "loss": 1.9515,
+      "step": 28912
+    },
+    {
+      "epoch": 0.6425111111111111,
+      "grad_norm": 1.3028802871704102,
+      "learning_rate": 7.151366970437875e-05,
+      "loss": 1.6399,
+      "step": 28913
+    },
+    {
+      "epoch": 0.6425333333333333,
+      "grad_norm": 1.5053068399429321,
+      "learning_rate": 7.150922427206046e-05,
+      "loss": 1.8337,
+      "step": 28914
+    },
+    {
+      "epoch": 0.6425555555555555,
+      "grad_norm": 1.4605801105499268,
+      "learning_rate": 7.150477883974217e-05,
+      "loss": 1.8094,
+      "step": 28915
+    },
+    {
+      "epoch": 0.6425777777777778,
+      "grad_norm": 1.552448034286499,
+      "learning_rate": 7.150033340742388e-05,
+      "loss": 1.659,
+      "step": 28916
+    },
+    {
+      "epoch": 0.6426,
+      "grad_norm": 1.7648966312408447,
+      "learning_rate": 7.149588797510557e-05,
+      "loss": 2.2957,
+      "step": 28917
+    },
+    {
+      "epoch": 0.6426222222222222,
+      "grad_norm": 1.7324978113174438,
+      "learning_rate": 7.149144254278728e-05,
+      "loss": 2.1141,
+      "step": 28918
+    },
+    {
+      "epoch": 0.6426444444444445,
+      "grad_norm": 2.1639018058776855,
+      "learning_rate": 7.148699711046901e-05,
+      "loss": 2.5954,
+      "step": 28919
+    },
+    {
+      "epoch": 0.6426666666666667,
+      "grad_norm": 1.7023972272872925,
+      "learning_rate": 7.14825516781507e-05,
+      "loss": 1.3876,
+      "step": 28920
+    },
+    {
+      "epoch": 0.6426888888888889,
+      "grad_norm": 2.214062213897705,
+      "learning_rate": 7.147810624583241e-05,
+      "loss": 2.5949,
+      "step": 28921
+    },
+    {
+      "epoch": 0.6427111111111111,
+      "grad_norm": 1.7753064632415771,
+      "learning_rate": 7.147366081351412e-05,
+      "loss": 2.0713,
+      "step": 28922
+    },
+    {
+      "epoch": 0.6427333333333334,
+      "grad_norm": 1.664215326309204,
+      "learning_rate": 7.146921538119582e-05,
+      "loss": 2.0425,
+      "step": 28923
+    },
+    {
+      "epoch": 0.6427555555555555,
+      "grad_norm": 1.6645792722702026,
+      "learning_rate": 7.146476994887753e-05,
+      "loss": 1.7387,
+      "step": 28924
+    },
+    {
+      "epoch": 0.6427777777777778,
+      "grad_norm": 1.8174889087677002,
+      "learning_rate": 7.146032451655924e-05,
+      "loss": 1.9186,
+      "step": 28925
+    },
+    {
+      "epoch": 0.6428,
+      "grad_norm": 1.781592607498169,
+      "learning_rate": 7.145587908424095e-05,
+      "loss": 2.0698,
+      "step": 28926
+    },
+    {
+      "epoch": 0.6428222222222222,
+      "grad_norm": 1.753316879272461,
+      "learning_rate": 7.145143365192264e-05,
+      "loss": 2.1797,
+      "step": 28927
+    },
+    {
+      "epoch": 0.6428444444444444,
+      "grad_norm": 1.7087000608444214,
+      "learning_rate": 7.144698821960437e-05,
+      "loss": 1.826,
+      "step": 28928
+    },
+    {
+      "epoch": 0.6428666666666667,
+      "grad_norm": 1.5608490705490112,
+      "learning_rate": 7.144254278728608e-05,
+      "loss": 1.7394,
+      "step": 28929
+    },
+    {
+      "epoch": 0.6428888888888888,
+      "grad_norm": 2.120506525039673,
+      "learning_rate": 7.143809735496777e-05,
+      "loss": 1.8856,
+      "step": 28930
+    },
+    {
+      "epoch": 0.6429111111111111,
+      "grad_norm": 2.612682342529297,
+      "learning_rate": 7.143365192264948e-05,
+      "loss": 2.0027,
+      "step": 28931
+    },
+    {
+      "epoch": 0.6429333333333334,
+      "grad_norm": 1.8839622735977173,
+      "learning_rate": 7.142920649033119e-05,
+      "loss": 1.8474,
+      "step": 28932
+    },
+    {
+      "epoch": 0.6429555555555555,
+      "grad_norm": 1.8573490381240845,
+      "learning_rate": 7.142476105801289e-05,
+      "loss": 2.2052,
+      "step": 28933
+    },
+    {
+      "epoch": 0.6429777777777778,
+      "grad_norm": 1.7652058601379395,
+      "learning_rate": 7.14203156256946e-05,
+      "loss": 1.7029,
+      "step": 28934
+    },
+    {
+      "epoch": 0.643,
+      "grad_norm": 1.7231435775756836,
+      "learning_rate": 7.141587019337631e-05,
+      "loss": 1.7502,
+      "step": 28935
+    },
+    {
+      "epoch": 0.6430222222222223,
+      "grad_norm": 1.7900536060333252,
+      "learning_rate": 7.141142476105802e-05,
+      "loss": 2.1998,
+      "step": 28936
+    },
+    {
+      "epoch": 0.6430444444444444,
+      "grad_norm": 1.664600133895874,
+      "learning_rate": 7.140697932873973e-05,
+      "loss": 1.7662,
+      "step": 28937
+    },
+    {
+      "epoch": 0.6430666666666667,
+      "grad_norm": 2.0292153358459473,
+      "learning_rate": 7.140253389642144e-05,
+      "loss": 2.5704,
+      "step": 28938
+    },
+    {
+      "epoch": 0.6430888888888889,
+      "grad_norm": 1.509979248046875,
+      "learning_rate": 7.139808846410315e-05,
+      "loss": 1.0218,
+      "step": 28939
+    },
+    {
+      "epoch": 0.6431111111111111,
+      "grad_norm": 0.9186019897460938,
+      "learning_rate": 7.139364303178484e-05,
+      "loss": 0.5662,
+      "step": 28940
+    },
+    {
+      "epoch": 0.6431333333333333,
+      "grad_norm": 1.5803574323654175,
+      "learning_rate": 7.138919759946655e-05,
+      "loss": 1.5752,
+      "step": 28941
+    },
+    {
+      "epoch": 0.6431555555555556,
+      "grad_norm": 1.9634132385253906,
+      "learning_rate": 7.138475216714826e-05,
+      "loss": 1.9959,
+      "step": 28942
+    },
+    {
+      "epoch": 0.6431777777777777,
+      "grad_norm": 1.919405221939087,
+      "learning_rate": 7.138030673482996e-05,
+      "loss": 2.1137,
+      "step": 28943
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 1.5899920463562012,
+      "learning_rate": 7.137586130251167e-05,
+      "loss": 1.9365,
+      "step": 28944
+    },
+    {
+      "epoch": 0.6432222222222223,
+      "grad_norm": 2.470597267150879,
+      "learning_rate": 7.137141587019338e-05,
+      "loss": 1.7956,
+      "step": 28945
+    },
+    {
+      "epoch": 0.6432444444444444,
+      "grad_norm": 1.7212203741073608,
+      "learning_rate": 7.136697043787509e-05,
+      "loss": 1.8821,
+      "step": 28946
+    },
+    {
+      "epoch": 0.6432666666666667,
+      "grad_norm": 1.7492668628692627,
+      "learning_rate": 7.13625250055568e-05,
+      "loss": 1.6948,
+      "step": 28947
+    },
+    {
+      "epoch": 0.6432888888888889,
+      "grad_norm": 1.3097518682479858,
+      "learning_rate": 7.13580795732385e-05,
+      "loss": 0.8045,
+      "step": 28948
+    },
+    {
+      "epoch": 0.6433111111111111,
+      "grad_norm": 1.538679838180542,
+      "learning_rate": 7.135363414092021e-05,
+      "loss": 1.4235,
+      "step": 28949
+    },
+    {
+      "epoch": 0.6433333333333333,
+      "grad_norm": 1.3475881814956665,
+      "learning_rate": 7.134918870860191e-05,
+      "loss": 0.836,
+      "step": 28950
+    },
+    {
+      "epoch": 0.6433555555555556,
+      "grad_norm": 1.8413110971450806,
+      "learning_rate": 7.134474327628362e-05,
+      "loss": 2.3781,
+      "step": 28951
+    },
+    {
+      "epoch": 0.6433777777777778,
+      "grad_norm": 1.575206995010376,
+      "learning_rate": 7.134029784396533e-05,
+      "loss": 2.3801,
+      "step": 28952
+    },
+    {
+      "epoch": 0.6434,
+      "grad_norm": 1.3493009805679321,
+      "learning_rate": 7.133585241164703e-05,
+      "loss": 1.9338,
+      "step": 28953
+    },
+    {
+      "epoch": 0.6434222222222222,
+      "grad_norm": 2.175967216491699,
+      "learning_rate": 7.133140697932874e-05,
+      "loss": 2.695,
+      "step": 28954
+    },
+    {
+      "epoch": 0.6434444444444445,
+      "grad_norm": 1.4450469017028809,
+      "learning_rate": 7.132696154701045e-05,
+      "loss": 1.7778,
+      "step": 28955
+    },
+    {
+      "epoch": 0.6434666666666666,
+      "grad_norm": 1.4620064496994019,
+      "learning_rate": 7.132251611469217e-05,
+      "loss": 1.841,
+      "step": 28956
+    },
+    {
+      "epoch": 0.6434888888888889,
+      "grad_norm": 1.7964931726455688,
+      "learning_rate": 7.131807068237386e-05,
+      "loss": 2.0961,
+      "step": 28957
+    },
+    {
+      "epoch": 0.6435111111111111,
+      "grad_norm": 1.6290562152862549,
+      "learning_rate": 7.131362525005557e-05,
+      "loss": 2.2417,
+      "step": 28958
+    },
+    {
+      "epoch": 0.6435333333333333,
+      "grad_norm": 1.3939601182937622,
+      "learning_rate": 7.130917981773728e-05,
+      "loss": 1.4903,
+      "step": 28959
+    },
+    {
+      "epoch": 0.6435555555555555,
+      "grad_norm": 1.6482905149459839,
+      "learning_rate": 7.130473438541898e-05,
+      "loss": 2.6648,
+      "step": 28960
+    },
+    {
+      "epoch": 0.6435777777777778,
+      "grad_norm": 2.1006858348846436,
+      "learning_rate": 7.130028895310069e-05,
+      "loss": 2.8309,
+      "step": 28961
+    },
+    {
+      "epoch": 0.6436,
+      "grad_norm": 1.5466973781585693,
+      "learning_rate": 7.12958435207824e-05,
+      "loss": 2.219,
+      "step": 28962
+    },
+    {
+      "epoch": 0.6436222222222222,
+      "grad_norm": 1.6188929080963135,
+      "learning_rate": 7.129139808846411e-05,
+      "loss": 1.6512,
+      "step": 28963
+    },
+    {
+      "epoch": 0.6436444444444445,
+      "grad_norm": 1.6394859552383423,
+      "learning_rate": 7.12869526561458e-05,
+      "loss": 1.9599,
+      "step": 28964
+    },
+    {
+      "epoch": 0.6436666666666667,
+      "grad_norm": 1.5032265186309814,
+      "learning_rate": 7.128250722382753e-05,
+      "loss": 2.1569,
+      "step": 28965
+    },
+    {
+      "epoch": 0.6436888888888889,
+      "grad_norm": 1.6356178522109985,
+      "learning_rate": 7.127806179150924e-05,
+      "loss": 2.1785,
+      "step": 28966
+    },
+    {
+      "epoch": 0.6437111111111111,
+      "grad_norm": 1.3364404439926147,
+      "learning_rate": 7.127361635919093e-05,
+      "loss": 1.3715,
+      "step": 28967
+    },
+    {
+      "epoch": 0.6437333333333334,
+      "grad_norm": 2.5799660682678223,
+      "learning_rate": 7.126917092687264e-05,
+      "loss": 1.8913,
+      "step": 28968
+    },
+    {
+      "epoch": 0.6437555555555555,
+      "grad_norm": 1.5566381216049194,
+      "learning_rate": 7.126472549455435e-05,
+      "loss": 1.7737,
+      "step": 28969
+    },
+    {
+      "epoch": 0.6437777777777778,
+      "grad_norm": 1.677599549293518,
+      "learning_rate": 7.126028006223605e-05,
+      "loss": 1.8859,
+      "step": 28970
+    },
+    {
+      "epoch": 0.6438,
+      "grad_norm": 1.5886155366897583,
+      "learning_rate": 7.125583462991776e-05,
+      "loss": 1.9092,
+      "step": 28971
+    },
+    {
+      "epoch": 0.6438222222222222,
+      "grad_norm": 1.4532115459442139,
+      "learning_rate": 7.125138919759947e-05,
+      "loss": 1.3206,
+      "step": 28972
+    },
+    {
+      "epoch": 0.6438444444444444,
+      "grad_norm": 1.7453795671463013,
+      "learning_rate": 7.124694376528118e-05,
+      "loss": 1.73,
+      "step": 28973
+    },
+    {
+      "epoch": 0.6438666666666667,
+      "grad_norm": 1.896207571029663,
+      "learning_rate": 7.124249833296289e-05,
+      "loss": 1.9166,
+      "step": 28974
+    },
+    {
+      "epoch": 0.6438888888888888,
+      "grad_norm": 1.6417932510375977,
+      "learning_rate": 7.12380529006446e-05,
+      "loss": 1.9912,
+      "step": 28975
+    },
+    {
+      "epoch": 0.6439111111111111,
+      "grad_norm": 2.062929630279541,
+      "learning_rate": 7.12336074683263e-05,
+      "loss": 2.1088,
+      "step": 28976
+    },
+    {
+      "epoch": 0.6439333333333334,
+      "grad_norm": 1.370573878288269,
+      "learning_rate": 7.1229162036008e-05,
+      "loss": 1.4623,
+      "step": 28977
+    },
+    {
+      "epoch": 0.6439555555555555,
+      "grad_norm": 1.727644443511963,
+      "learning_rate": 7.122471660368971e-05,
+      "loss": 1.547,
+      "step": 28978
+    },
+    {
+      "epoch": 0.6439777777777778,
+      "grad_norm": 1.6774473190307617,
+      "learning_rate": 7.122027117137142e-05,
+      "loss": 2.0309,
+      "step": 28979
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 1.7936469316482544,
+      "learning_rate": 7.121582573905312e-05,
+      "loss": 1.8048,
+      "step": 28980
+    },
+    {
+      "epoch": 0.6440222222222223,
+      "grad_norm": 1.6021851301193237,
+      "learning_rate": 7.121138030673483e-05,
+      "loss": 1.8476,
+      "step": 28981
+    },
+    {
+      "epoch": 0.6440444444444444,
+      "grad_norm": 1.6705433130264282,
+      "learning_rate": 7.120693487441654e-05,
+      "loss": 1.8819,
+      "step": 28982
+    },
+    {
+      "epoch": 0.6440666666666667,
+      "grad_norm": 2.086313009262085,
+      "learning_rate": 7.120248944209825e-05,
+      "loss": 2.1044,
+      "step": 28983
+    },
+    {
+      "epoch": 0.6440888888888889,
+      "grad_norm": 2.1750640869140625,
+      "learning_rate": 7.119804400977996e-05,
+      "loss": 1.7156,
+      "step": 28984
+    },
+    {
+      "epoch": 0.6441111111111111,
+      "grad_norm": 1.8231173753738403,
+      "learning_rate": 7.119359857746167e-05,
+      "loss": 1.8609,
+      "step": 28985
+    },
+    {
+      "epoch": 0.6441333333333333,
+      "grad_norm": 2.0555918216705322,
+      "learning_rate": 7.118915314514338e-05,
+      "loss": 2.4212,
+      "step": 28986
+    },
+    {
+      "epoch": 0.6441555555555556,
+      "grad_norm": 2.163342237472534,
+      "learning_rate": 7.118470771282507e-05,
+      "loss": 1.8877,
+      "step": 28987
+    },
+    {
+      "epoch": 0.6441777777777777,
+      "grad_norm": 1.6176615953445435,
+      "learning_rate": 7.118026228050678e-05,
+      "loss": 1.5502,
+      "step": 28988
+    },
+    {
+      "epoch": 0.6442,
+      "grad_norm": 1.878800630569458,
+      "learning_rate": 7.117581684818849e-05,
+      "loss": 1.7773,
+      "step": 28989
+    },
+    {
+      "epoch": 0.6442222222222223,
+      "grad_norm": 1.7244913578033447,
+      "learning_rate": 7.117137141587019e-05,
+      "loss": 1.9062,
+      "step": 28990
+    },
+    {
+      "epoch": 0.6442444444444444,
+      "grad_norm": 1.8530195951461792,
+      "learning_rate": 7.11669259835519e-05,
+      "loss": 1.9271,
+      "step": 28991
+    },
+    {
+      "epoch": 0.6442666666666667,
+      "grad_norm": 1.9446215629577637,
+      "learning_rate": 7.11624805512336e-05,
+      "loss": 2.044,
+      "step": 28992
+    },
+    {
+      "epoch": 0.6442888888888889,
+      "grad_norm": 1.711171269416809,
+      "learning_rate": 7.115803511891533e-05,
+      "loss": 1.6565,
+      "step": 28993
+    },
+    {
+      "epoch": 0.6443111111111111,
+      "grad_norm": 1.7592583894729614,
+      "learning_rate": 7.115358968659703e-05,
+      "loss": 1.8452,
+      "step": 28994
+    },
+    {
+      "epoch": 0.6443333333333333,
+      "grad_norm": 1.7510658502578735,
+      "learning_rate": 7.114914425427873e-05,
+      "loss": 1.9349,
+      "step": 28995
+    },
+    {
+      "epoch": 0.6443555555555556,
+      "grad_norm": 2.019502639770508,
+      "learning_rate": 7.114469882196044e-05,
+      "loss": 2.0444,
+      "step": 28996
+    },
+    {
+      "epoch": 0.6443777777777778,
+      "grad_norm": 1.7224868535995483,
+      "learning_rate": 7.114025338964214e-05,
+      "loss": 2.0977,
+      "step": 28997
+    },
+    {
+      "epoch": 0.6444,
+      "grad_norm": 0.34710970520973206,
+      "learning_rate": 7.113580795732385e-05,
+      "loss": 0.0366,
+      "step": 28998
+    },
+    {
+      "epoch": 0.6444222222222222,
+      "grad_norm": 2.390554428100586,
+      "learning_rate": 7.113136252500556e-05,
+      "loss": 2.1436,
+      "step": 28999
+    },
+    {
+      "epoch": 0.6444444444444445,
+      "grad_norm": 1.99862539768219,
+      "learning_rate": 7.112691709268726e-05,
+      "loss": 1.5966,
+      "step": 29000
+    },
+    {
+      "epoch": 0.6444666666666666,
+      "grad_norm": 1.686181902885437,
+      "learning_rate": 7.112247166036897e-05,
+      "loss": 2.3817,
+      "step": 29001
+    },
+    {
+      "epoch": 0.6444888888888889,
+      "grad_norm": 1.5259734392166138,
+      "learning_rate": 7.111802622805069e-05,
+      "loss": 1.845,
+      "step": 29002
+    },
+    {
+      "epoch": 0.6445111111111111,
+      "grad_norm": 1.4012967348098755,
+      "learning_rate": 7.11135807957324e-05,
+      "loss": 2.2198,
+      "step": 29003
+    },
+    {
+      "epoch": 0.6445333333333333,
+      "grad_norm": 1.4382514953613281,
+      "learning_rate": 7.11091353634141e-05,
+      "loss": 1.8372,
+      "step": 29004
+    },
+    {
+      "epoch": 0.6445555555555555,
+      "grad_norm": 1.8115453720092773,
+      "learning_rate": 7.11046899310958e-05,
+      "loss": 1.7997,
+      "step": 29005
+    },
+    {
+      "epoch": 0.6445777777777778,
+      "grad_norm": 1.711234211921692,
+      "learning_rate": 7.110024449877751e-05,
+      "loss": 2.125,
+      "step": 29006
+    },
+    {
+      "epoch": 0.6446,
+      "grad_norm": 1.3862836360931396,
+      "learning_rate": 7.109579906645921e-05,
+      "loss": 2.1216,
+      "step": 29007
+    },
+    {
+      "epoch": 0.6446222222222222,
+      "grad_norm": 1.8051152229309082,
+      "learning_rate": 7.109135363414092e-05,
+      "loss": 1.5909,
+      "step": 29008
+    },
+    {
+      "epoch": 0.6446444444444445,
+      "grad_norm": 1.8862489461898804,
+      "learning_rate": 7.108690820182263e-05,
+      "loss": 2.0823,
+      "step": 29009
+    },
+    {
+      "epoch": 0.6446666666666667,
+      "grad_norm": 1.46207594871521,
+      "learning_rate": 7.108246276950434e-05,
+      "loss": 2.1865,
+      "step": 29010
+    },
+    {
+      "epoch": 0.6446888888888889,
+      "grad_norm": 0.224213108420372,
+      "learning_rate": 7.107801733718605e-05,
+      "loss": 0.025,
+      "step": 29011
+    },
+    {
+      "epoch": 0.6447111111111111,
+      "grad_norm": 1.5012792348861694,
+      "learning_rate": 7.107357190486776e-05,
+      "loss": 2.1706,
+      "step": 29012
+    },
+    {
+      "epoch": 0.6447333333333334,
+      "grad_norm": 1.9338184595108032,
+      "learning_rate": 7.106912647254947e-05,
+      "loss": 2.2592,
+      "step": 29013
+    },
+    {
+      "epoch": 0.6447555555555555,
+      "grad_norm": 1.810837745666504,
+      "learning_rate": 7.106468104023116e-05,
+      "loss": 1.7871,
+      "step": 29014
+    },
+    {
+      "epoch": 0.6447777777777778,
+      "grad_norm": 1.4336885213851929,
+      "learning_rate": 7.106023560791287e-05,
+      "loss": 1.8581,
+      "step": 29015
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 1.5249687433242798,
+      "learning_rate": 7.105579017559458e-05,
+      "loss": 1.3569,
+      "step": 29016
+    },
+    {
+      "epoch": 0.6448222222222222,
+      "grad_norm": 2.3387036323547363,
+      "learning_rate": 7.105134474327628e-05,
+      "loss": 2.1901,
+      "step": 29017
+    },
+    {
+      "epoch": 0.6448444444444444,
+      "grad_norm": 1.683835744857788,
+      "learning_rate": 7.104689931095799e-05,
+      "loss": 2.242,
+      "step": 29018
+    },
+    {
+      "epoch": 0.6448666666666667,
+      "grad_norm": 1.5120434761047363,
+      "learning_rate": 7.10424538786397e-05,
+      "loss": 1.9611,
+      "step": 29019
+    },
+    {
+      "epoch": 0.6448888888888888,
+      "grad_norm": 1.8133294582366943,
+      "learning_rate": 7.103800844632141e-05,
+      "loss": 2.2278,
+      "step": 29020
+    },
+    {
+      "epoch": 0.6449111111111111,
+      "grad_norm": 1.528519630432129,
+      "learning_rate": 7.103356301400312e-05,
+      "loss": 1.7234,
+      "step": 29021
+    },
+    {
+      "epoch": 0.6449333333333334,
+      "grad_norm": 1.6087287664413452,
+      "learning_rate": 7.102911758168483e-05,
+      "loss": 1.8672,
+      "step": 29022
+    },
+    {
+      "epoch": 0.6449555555555555,
+      "grad_norm": 1.694104790687561,
+      "learning_rate": 7.102467214936654e-05,
+      "loss": 2.2416,
+      "step": 29023
+    },
+    {
+      "epoch": 0.6449777777777778,
+      "grad_norm": 2.1608152389526367,
+      "learning_rate": 7.102022671704823e-05,
+      "loss": 2.186,
+      "step": 29024
+    },
+    {
+      "epoch": 0.645,
+      "grad_norm": 1.8009564876556396,
+      "learning_rate": 7.101578128472994e-05,
+      "loss": 2.168,
+      "step": 29025
+    },
+    {
+      "epoch": 0.6450222222222223,
+      "grad_norm": 1.6181777715682983,
+      "learning_rate": 7.101133585241165e-05,
+      "loss": 2.2455,
+      "step": 29026
+    },
+    {
+      "epoch": 0.6450444444444444,
+      "grad_norm": 1.9097341299057007,
+      "learning_rate": 7.100689042009335e-05,
+      "loss": 2.0104,
+      "step": 29027
+    },
+    {
+      "epoch": 0.6450666666666667,
+      "grad_norm": 1.472928524017334,
+      "learning_rate": 7.100244498777506e-05,
+      "loss": 1.7127,
+      "step": 29028
+    },
+    {
+      "epoch": 0.6450888888888889,
+      "grad_norm": 1.8379104137420654,
+      "learning_rate": 7.099799955545677e-05,
+      "loss": 2.4129,
+      "step": 29029
+    },
+    {
+      "epoch": 0.6451111111111111,
+      "grad_norm": 1.4409955739974976,
+      "learning_rate": 7.099355412313849e-05,
+      "loss": 1.5263,
+      "step": 29030
+    },
+    {
+      "epoch": 0.6451333333333333,
+      "grad_norm": 1.8723595142364502,
+      "learning_rate": 7.098910869082019e-05,
+      "loss": 1.9664,
+      "step": 29031
+    },
+    {
+      "epoch": 0.6451555555555556,
+      "grad_norm": 0.896340012550354,
+      "learning_rate": 7.09846632585019e-05,
+      "loss": 0.6383,
+      "step": 29032
+    },
+    {
+      "epoch": 0.6451777777777777,
+      "grad_norm": 1.6818281412124634,
+      "learning_rate": 7.09802178261836e-05,
+      "loss": 2.082,
+      "step": 29033
+    },
+    {
+      "epoch": 0.6452,
+      "grad_norm": 1.591972827911377,
+      "learning_rate": 7.09757723938653e-05,
+      "loss": 1.7117,
+      "step": 29034
+    },
+    {
+      "epoch": 0.6452222222222223,
+      "grad_norm": 1.9729094505310059,
+      "learning_rate": 7.097132696154701e-05,
+      "loss": 2.056,
+      "step": 29035
+    },
+    {
+      "epoch": 0.6452444444444444,
+      "grad_norm": 1.9406650066375732,
+      "learning_rate": 7.096688152922872e-05,
+      "loss": 2.3653,
+      "step": 29036
+    },
+    {
+      "epoch": 0.6452666666666667,
+      "grad_norm": 1.635969877243042,
+      "learning_rate": 7.096243609691042e-05,
+      "loss": 1.4225,
+      "step": 29037
+    },
+    {
+      "epoch": 0.6452888888888889,
+      "grad_norm": 1.8061898946762085,
+      "learning_rate": 7.095799066459213e-05,
+      "loss": 2.1575,
+      "step": 29038
+    },
+    {
+      "epoch": 0.6453111111111111,
+      "grad_norm": 0.9610072374343872,
+      "learning_rate": 7.095354523227385e-05,
+      "loss": 0.7144,
+      "step": 29039
+    },
+    {
+      "epoch": 0.6453333333333333,
+      "grad_norm": 1.6564658880233765,
+      "learning_rate": 7.094909979995556e-05,
+      "loss": 1.6961,
+      "step": 29040
+    },
+    {
+      "epoch": 0.6453555555555556,
+      "grad_norm": 2.188620090484619,
+      "learning_rate": 7.094465436763726e-05,
+      "loss": 1.992,
+      "step": 29041
+    },
+    {
+      "epoch": 0.6453777777777778,
+      "grad_norm": 2.1346263885498047,
+      "learning_rate": 7.094020893531897e-05,
+      "loss": 1.7606,
+      "step": 29042
+    },
+    {
+      "epoch": 0.6454,
+      "grad_norm": 1.584122896194458,
+      "learning_rate": 7.093576350300067e-05,
+      "loss": 1.9503,
+      "step": 29043
+    },
+    {
+      "epoch": 0.6454222222222222,
+      "grad_norm": 2.2534492015838623,
+      "learning_rate": 7.093131807068237e-05,
+      "loss": 2.2768,
+      "step": 29044
+    },
+    {
+      "epoch": 0.6454444444444445,
+      "grad_norm": 1.9511295557022095,
+      "learning_rate": 7.092687263836408e-05,
+      "loss": 2.1601,
+      "step": 29045
+    },
+    {
+      "epoch": 0.6454666666666666,
+      "grad_norm": 1.7901535034179688,
+      "learning_rate": 7.092242720604579e-05,
+      "loss": 1.8901,
+      "step": 29046
+    },
+    {
+      "epoch": 0.6454888888888889,
+      "grad_norm": 1.7285199165344238,
+      "learning_rate": 7.09179817737275e-05,
+      "loss": 1.5072,
+      "step": 29047
+    },
+    {
+      "epoch": 0.6455111111111111,
+      "grad_norm": 1.8059173822402954,
+      "learning_rate": 7.091353634140921e-05,
+      "loss": 1.8646,
+      "step": 29048
+    },
+    {
+      "epoch": 0.6455333333333333,
+      "grad_norm": 1.863149881362915,
+      "learning_rate": 7.090909090909092e-05,
+      "loss": 1.971,
+      "step": 29049
+    },
+    {
+      "epoch": 0.6455555555555555,
+      "grad_norm": 1.723754644393921,
+      "learning_rate": 7.090464547677263e-05,
+      "loss": 1.5036,
+      "step": 29050
+    },
+    {
+      "epoch": 0.6455777777777778,
+      "grad_norm": 1.1470249891281128,
+      "learning_rate": 7.090020004445432e-05,
+      "loss": 1.0734,
+      "step": 29051
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 1.4927550554275513,
+      "learning_rate": 7.089575461213603e-05,
+      "loss": 2.4021,
+      "step": 29052
+    },
+    {
+      "epoch": 0.6456222222222222,
+      "grad_norm": 1.577982783317566,
+      "learning_rate": 7.089130917981774e-05,
+      "loss": 2.2874,
+      "step": 29053
+    },
+    {
+      "epoch": 0.6456444444444445,
+      "grad_norm": 1.6243376731872559,
+      "learning_rate": 7.088686374749944e-05,
+      "loss": 1.0384,
+      "step": 29054
+    },
+    {
+      "epoch": 0.6456666666666667,
+      "grad_norm": 1.6546640396118164,
+      "learning_rate": 7.088241831518115e-05,
+      "loss": 1.1537,
+      "step": 29055
+    },
+    {
+      "epoch": 0.6456888888888889,
+      "grad_norm": 1.5087653398513794,
+      "learning_rate": 7.087797288286286e-05,
+      "loss": 2.3756,
+      "step": 29056
+    },
+    {
+      "epoch": 0.6457111111111111,
+      "grad_norm": 1.4392410516738892,
+      "learning_rate": 7.087352745054457e-05,
+      "loss": 1.8451,
+      "step": 29057
+    },
+    {
+      "epoch": 0.6457333333333334,
+      "grad_norm": 1.604750156402588,
+      "learning_rate": 7.086908201822628e-05,
+      "loss": 2.3932,
+      "step": 29058
+    },
+    {
+      "epoch": 0.6457555555555555,
+      "grad_norm": 1.3697820901870728,
+      "learning_rate": 7.086463658590799e-05,
+      "loss": 1.99,
+      "step": 29059
+    },
+    {
+      "epoch": 0.6457777777777778,
+      "grad_norm": 1.5392061471939087,
+      "learning_rate": 7.08601911535897e-05,
+      "loss": 1.8891,
+      "step": 29060
+    },
+    {
+      "epoch": 0.6458,
+      "grad_norm": 1.8606305122375488,
+      "learning_rate": 7.08557457212714e-05,
+      "loss": 2.2007,
+      "step": 29061
+    },
+    {
+      "epoch": 0.6458222222222222,
+      "grad_norm": 1.821648359298706,
+      "learning_rate": 7.08513002889531e-05,
+      "loss": 2.2572,
+      "step": 29062
+    },
+    {
+      "epoch": 0.6458444444444444,
+      "grad_norm": 1.64143967628479,
+      "learning_rate": 7.084685485663481e-05,
+      "loss": 1.8566,
+      "step": 29063
+    },
+    {
+      "epoch": 0.6458666666666667,
+      "grad_norm": 1.5850093364715576,
+      "learning_rate": 7.084240942431651e-05,
+      "loss": 1.8496,
+      "step": 29064
+    },
+    {
+      "epoch": 0.6458888888888888,
+      "grad_norm": 1.6152373552322388,
+      "learning_rate": 7.083796399199822e-05,
+      "loss": 1.1328,
+      "step": 29065
+    },
+    {
+      "epoch": 0.6459111111111111,
+      "grad_norm": 1.1689858436584473,
+      "learning_rate": 7.083351855967993e-05,
+      "loss": 1.2043,
+      "step": 29066
+    },
+    {
+      "epoch": 0.6459333333333334,
+      "grad_norm": 1.7024846076965332,
+      "learning_rate": 7.082907312736165e-05,
+      "loss": 1.9357,
+      "step": 29067
+    },
+    {
+      "epoch": 0.6459555555555555,
+      "grad_norm": 1.3479031324386597,
+      "learning_rate": 7.082462769504335e-05,
+      "loss": 1.3946,
+      "step": 29068
+    },
+    {
+      "epoch": 0.6459777777777778,
+      "grad_norm": 1.5222748517990112,
+      "learning_rate": 7.082018226272506e-05,
+      "loss": 1.8374,
+      "step": 29069
+    },
+    {
+      "epoch": 0.646,
+      "grad_norm": 1.5448720455169678,
+      "learning_rate": 7.081573683040677e-05,
+      "loss": 1.7917,
+      "step": 29070
+    },
+    {
+      "epoch": 0.6460222222222223,
+      "grad_norm": 1.8289965391159058,
+      "learning_rate": 7.081129139808846e-05,
+      "loss": 1.3985,
+      "step": 29071
+    },
+    {
+      "epoch": 0.6460444444444444,
+      "grad_norm": 1.918680191040039,
+      "learning_rate": 7.080684596577017e-05,
+      "loss": 1.5956,
+      "step": 29072
+    },
+    {
+      "epoch": 0.6460666666666667,
+      "grad_norm": 1.5386064052581787,
+      "learning_rate": 7.080240053345188e-05,
+      "loss": 1.7289,
+      "step": 29073
+    },
+    {
+      "epoch": 0.6460888888888889,
+      "grad_norm": 1.6020234823226929,
+      "learning_rate": 7.079795510113358e-05,
+      "loss": 1.7265,
+      "step": 29074
+    },
+    {
+      "epoch": 0.6461111111111111,
+      "grad_norm": 1.6040480136871338,
+      "learning_rate": 7.079350966881529e-05,
+      "loss": 1.6565,
+      "step": 29075
+    },
+    {
+      "epoch": 0.6461333333333333,
+      "grad_norm": 1.5008918046951294,
+      "learning_rate": 7.078906423649701e-05,
+      "loss": 1.9662,
+      "step": 29076
+    },
+    {
+      "epoch": 0.6461555555555556,
+      "grad_norm": 1.6811549663543701,
+      "learning_rate": 7.078461880417872e-05,
+      "loss": 1.83,
+      "step": 29077
+    },
+    {
+      "epoch": 0.6461777777777777,
+      "grad_norm": 1.627158284187317,
+      "learning_rate": 7.078017337186042e-05,
+      "loss": 2.421,
+      "step": 29078
+    },
+    {
+      "epoch": 0.6462,
+      "grad_norm": 1.6065196990966797,
+      "learning_rate": 7.077572793954213e-05,
+      "loss": 1.8298,
+      "step": 29079
+    },
+    {
+      "epoch": 0.6462222222222223,
+      "grad_norm": 1.4701594114303589,
+      "learning_rate": 7.077128250722384e-05,
+      "loss": 1.3981,
+      "step": 29080
+    },
+    {
+      "epoch": 0.6462444444444444,
+      "grad_norm": 1.77341890335083,
+      "learning_rate": 7.076683707490553e-05,
+      "loss": 2.2195,
+      "step": 29081
+    },
+    {
+      "epoch": 0.6462666666666667,
+      "grad_norm": 1.2583370208740234,
+      "learning_rate": 7.076239164258724e-05,
+      "loss": 0.795,
+      "step": 29082
+    },
+    {
+      "epoch": 0.6462888888888889,
+      "grad_norm": 1.5209894180297852,
+      "learning_rate": 7.075794621026895e-05,
+      "loss": 1.8441,
+      "step": 29083
+    },
+    {
+      "epoch": 0.6463111111111111,
+      "grad_norm": 1.5885052680969238,
+      "learning_rate": 7.075350077795066e-05,
+      "loss": 2.0454,
+      "step": 29084
+    },
+    {
+      "epoch": 0.6463333333333333,
+      "grad_norm": 1.8185741901397705,
+      "learning_rate": 7.074905534563237e-05,
+      "loss": 1.8766,
+      "step": 29085
+    },
+    {
+      "epoch": 0.6463555555555556,
+      "grad_norm": 1.6103633642196655,
+      "learning_rate": 7.074460991331408e-05,
+      "loss": 1.5191,
+      "step": 29086
+    },
+    {
+      "epoch": 0.6463777777777778,
+      "grad_norm": 1.9912450313568115,
+      "learning_rate": 7.074016448099579e-05,
+      "loss": 2.1519,
+      "step": 29087
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 2.1047565937042236,
+      "learning_rate": 7.073571904867749e-05,
+      "loss": 2.1204,
+      "step": 29088
+    },
+    {
+      "epoch": 0.6464222222222222,
+      "grad_norm": 1.7408231496810913,
+      "learning_rate": 7.07312736163592e-05,
+      "loss": 1.687,
+      "step": 29089
+    },
+    {
+      "epoch": 0.6464444444444445,
+      "grad_norm": 2.2416300773620605,
+      "learning_rate": 7.07268281840409e-05,
+      "loss": 2.5354,
+      "step": 29090
+    },
+    {
+      "epoch": 0.6464666666666666,
+      "grad_norm": 2.1599998474121094,
+      "learning_rate": 7.07223827517226e-05,
+      "loss": 2.36,
+      "step": 29091
+    },
+    {
+      "epoch": 0.6464888888888889,
+      "grad_norm": 1.4476076364517212,
+      "learning_rate": 7.071793731940431e-05,
+      "loss": 1.1872,
+      "step": 29092
+    },
+    {
+      "epoch": 0.6465111111111111,
+      "grad_norm": 1.974564552307129,
+      "learning_rate": 7.071349188708602e-05,
+      "loss": 2.0381,
+      "step": 29093
+    },
+    {
+      "epoch": 0.6465333333333333,
+      "grad_norm": 1.7401434183120728,
+      "learning_rate": 7.070904645476773e-05,
+      "loss": 1.646,
+      "step": 29094
+    },
+    {
+      "epoch": 0.6465555555555556,
+      "grad_norm": 1.8770774602890015,
+      "learning_rate": 7.070460102244944e-05,
+      "loss": 1.6141,
+      "step": 29095
+    },
+    {
+      "epoch": 0.6465777777777778,
+      "grad_norm": 1.7361961603164673,
+      "learning_rate": 7.070015559013115e-05,
+      "loss": 1.3437,
+      "step": 29096
+    },
+    {
+      "epoch": 0.6466,
+      "grad_norm": 1.6656427383422852,
+      "learning_rate": 7.069571015781286e-05,
+      "loss": 1.5471,
+      "step": 29097
+    },
+    {
+      "epoch": 0.6466222222222222,
+      "grad_norm": 1.9020168781280518,
+      "learning_rate": 7.069126472549455e-05,
+      "loss": 1.9409,
+      "step": 29098
+    },
+    {
+      "epoch": 0.6466444444444445,
+      "grad_norm": 2.2219109535217285,
+      "learning_rate": 7.068681929317626e-05,
+      "loss": 2.1355,
+      "step": 29099
+    },
+    {
+      "epoch": 0.6466666666666666,
+      "grad_norm": 2.644800901412964,
+      "learning_rate": 7.068237386085797e-05,
+      "loss": 1.2337,
+      "step": 29100
+    },
+    {
+      "epoch": 0.6466888888888889,
+      "grad_norm": 1.3294848203659058,
+      "learning_rate": 7.067792842853967e-05,
+      "loss": 1.0227,
+      "step": 29101
+    },
+    {
+      "epoch": 0.6467111111111111,
+      "grad_norm": 1.480413556098938,
+      "learning_rate": 7.067348299622138e-05,
+      "loss": 2.1486,
+      "step": 29102
+    },
+    {
+      "epoch": 0.6467333333333334,
+      "grad_norm": 1.4161555767059326,
+      "learning_rate": 7.066903756390309e-05,
+      "loss": 1.8185,
+      "step": 29103
+    },
+    {
+      "epoch": 0.6467555555555555,
+      "grad_norm": 0.9094601273536682,
+      "learning_rate": 7.06645921315848e-05,
+      "loss": 1.1822,
+      "step": 29104
+    },
+    {
+      "epoch": 0.6467777777777778,
+      "grad_norm": 1.4612770080566406,
+      "learning_rate": 7.066014669926651e-05,
+      "loss": 2.2132,
+      "step": 29105
+    },
+    {
+      "epoch": 0.6468,
+      "grad_norm": 1.5787962675094604,
+      "learning_rate": 7.065570126694822e-05,
+      "loss": 2.2369,
+      "step": 29106
+    },
+    {
+      "epoch": 0.6468222222222222,
+      "grad_norm": 1.8740125894546509,
+      "learning_rate": 7.065125583462993e-05,
+      "loss": 2.5167,
+      "step": 29107
+    },
+    {
+      "epoch": 0.6468444444444444,
+      "grad_norm": 1.66010582447052,
+      "learning_rate": 7.064681040231162e-05,
+      "loss": 2.4951,
+      "step": 29108
+    },
+    {
+      "epoch": 0.6468666666666667,
+      "grad_norm": 1.5209728479385376,
+      "learning_rate": 7.064236496999333e-05,
+      "loss": 2.2145,
+      "step": 29109
+    },
+    {
+      "epoch": 0.6468888888888888,
+      "grad_norm": 1.2550688982009888,
+      "learning_rate": 7.063791953767504e-05,
+      "loss": 1.5124,
+      "step": 29110
+    },
+    {
+      "epoch": 0.6469111111111111,
+      "grad_norm": 1.596299648284912,
+      "learning_rate": 7.063347410535674e-05,
+      "loss": 2.3205,
+      "step": 29111
+    },
+    {
+      "epoch": 0.6469333333333334,
+      "grad_norm": 1.7290644645690918,
+      "learning_rate": 7.062902867303845e-05,
+      "loss": 1.424,
+      "step": 29112
+    },
+    {
+      "epoch": 0.6469555555555555,
+      "grad_norm": 1.6497602462768555,
+      "learning_rate": 7.062458324072017e-05,
+      "loss": 1.7248,
+      "step": 29113
+    },
+    {
+      "epoch": 0.6469777777777778,
+      "grad_norm": 1.732588768005371,
+      "learning_rate": 7.062013780840187e-05,
+      "loss": 2.3204,
+      "step": 29114
+    },
+    {
+      "epoch": 0.647,
+      "grad_norm": 1.65004563331604,
+      "learning_rate": 7.061569237608358e-05,
+      "loss": 2.0739,
+      "step": 29115
+    },
+    {
+      "epoch": 0.6470222222222223,
+      "grad_norm": 1.4575328826904297,
+      "learning_rate": 7.061124694376529e-05,
+      "loss": 1.1476,
+      "step": 29116
+    },
+    {
+      "epoch": 0.6470444444444444,
+      "grad_norm": 1.8705321550369263,
+      "learning_rate": 7.0606801511447e-05,
+      "loss": 2.2397,
+      "step": 29117
+    },
+    {
+      "epoch": 0.6470666666666667,
+      "grad_norm": 1.4159882068634033,
+      "learning_rate": 7.060235607912869e-05,
+      "loss": 2.1124,
+      "step": 29118
+    },
+    {
+      "epoch": 0.6470888888888889,
+      "grad_norm": 1.5113205909729004,
+      "learning_rate": 7.05979106468104e-05,
+      "loss": 1.5342,
+      "step": 29119
+    },
+    {
+      "epoch": 0.6471111111111111,
+      "grad_norm": 1.9303370714187622,
+      "learning_rate": 7.059346521449211e-05,
+      "loss": 1.8784,
+      "step": 29120
+    },
+    {
+      "epoch": 0.6471333333333333,
+      "grad_norm": 1.6839933395385742,
+      "learning_rate": 7.058901978217382e-05,
+      "loss": 1.967,
+      "step": 29121
+    },
+    {
+      "epoch": 0.6471555555555556,
+      "grad_norm": 1.7867637872695923,
+      "learning_rate": 7.058457434985553e-05,
+      "loss": 2.0391,
+      "step": 29122
+    },
+    {
+      "epoch": 0.6471777777777777,
+      "grad_norm": 1.8341662883758545,
+      "learning_rate": 7.058012891753724e-05,
+      "loss": 1.9629,
+      "step": 29123
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 1.563461184501648,
+      "learning_rate": 7.057568348521895e-05,
+      "loss": 1.4533,
+      "step": 29124
+    },
+    {
+      "epoch": 0.6472222222222223,
+      "grad_norm": 1.56820547580719,
+      "learning_rate": 7.057123805290065e-05,
+      "loss": 2.1632,
+      "step": 29125
+    },
+    {
+      "epoch": 0.6472444444444444,
+      "grad_norm": 2.0252976417541504,
+      "learning_rate": 7.056679262058236e-05,
+      "loss": 2.6123,
+      "step": 29126
+    },
+    {
+      "epoch": 0.6472666666666667,
+      "grad_norm": 1.8063596487045288,
+      "learning_rate": 7.056234718826407e-05,
+      "loss": 1.9402,
+      "step": 29127
+    },
+    {
+      "epoch": 0.6472888888888889,
+      "grad_norm": 1.1451598405838013,
+      "learning_rate": 7.055790175594576e-05,
+      "loss": 1.0699,
+      "step": 29128
+    },
+    {
+      "epoch": 0.6473111111111111,
+      "grad_norm": 1.7466984987258911,
+      "learning_rate": 7.055345632362747e-05,
+      "loss": 1.846,
+      "step": 29129
+    },
+    {
+      "epoch": 0.6473333333333333,
+      "grad_norm": 1.801710605621338,
+      "learning_rate": 7.054901089130918e-05,
+      "loss": 1.5419,
+      "step": 29130
+    },
+    {
+      "epoch": 0.6473555555555556,
+      "grad_norm": 1.829551100730896,
+      "learning_rate": 7.054456545899089e-05,
+      "loss": 2.5472,
+      "step": 29131
+    },
+    {
+      "epoch": 0.6473777777777778,
+      "grad_norm": 2.013662099838257,
+      "learning_rate": 7.05401200266726e-05,
+      "loss": 2.0472,
+      "step": 29132
+    },
+    {
+      "epoch": 0.6474,
+      "grad_norm": 1.3499835729599,
+      "learning_rate": 7.053567459435431e-05,
+      "loss": 0.7647,
+      "step": 29133
+    },
+    {
+      "epoch": 0.6474222222222222,
+      "grad_norm": 2.016878843307495,
+      "learning_rate": 7.053122916203602e-05,
+      "loss": 2.2325,
+      "step": 29134
+    },
+    {
+      "epoch": 0.6474444444444445,
+      "grad_norm": 1.7505043745040894,
+      "learning_rate": 7.052678372971772e-05,
+      "loss": 2.1693,
+      "step": 29135
+    },
+    {
+      "epoch": 0.6474666666666666,
+      "grad_norm": 1.9273511171340942,
+      "learning_rate": 7.052233829739943e-05,
+      "loss": 1.8722,
+      "step": 29136
+    },
+    {
+      "epoch": 0.6474888888888889,
+      "grad_norm": 1.839815616607666,
+      "learning_rate": 7.051789286508113e-05,
+      "loss": 2.2612,
+      "step": 29137
+    },
+    {
+      "epoch": 0.6475111111111111,
+      "grad_norm": 1.7973805665969849,
+      "learning_rate": 7.051344743276283e-05,
+      "loss": 1.6756,
+      "step": 29138
+    },
+    {
+      "epoch": 0.6475333333333333,
+      "grad_norm": 1.2534786462783813,
+      "learning_rate": 7.050900200044454e-05,
+      "loss": 0.8786,
+      "step": 29139
+    },
+    {
+      "epoch": 0.6475555555555556,
+      "grad_norm": 1.5299891233444214,
+      "learning_rate": 7.050455656812625e-05,
+      "loss": 1.3872,
+      "step": 29140
+    },
+    {
+      "epoch": 0.6475777777777778,
+      "grad_norm": 1.5156304836273193,
+      "learning_rate": 7.050011113580796e-05,
+      "loss": 1.6405,
+      "step": 29141
+    },
+    {
+      "epoch": 0.6476,
+      "grad_norm": 1.7062276601791382,
+      "learning_rate": 7.049566570348967e-05,
+      "loss": 2.0486,
+      "step": 29142
+    },
+    {
+      "epoch": 0.6476222222222222,
+      "grad_norm": 1.7123117446899414,
+      "learning_rate": 7.049122027117138e-05,
+      "loss": 1.9122,
+      "step": 29143
+    },
+    {
+      "epoch": 0.6476444444444445,
+      "grad_norm": 1.8518922328948975,
+      "learning_rate": 7.048677483885309e-05,
+      "loss": 2.0751,
+      "step": 29144
+    },
+    {
+      "epoch": 0.6476666666666666,
+      "grad_norm": 1.8861069679260254,
+      "learning_rate": 7.048232940653478e-05,
+      "loss": 1.831,
+      "step": 29145
+    },
+    {
+      "epoch": 0.6476888888888889,
+      "grad_norm": 1.7695550918579102,
+      "learning_rate": 7.04778839742165e-05,
+      "loss": 1.9029,
+      "step": 29146
+    },
+    {
+      "epoch": 0.6477111111111111,
+      "grad_norm": 2.139415979385376,
+      "learning_rate": 7.04734385418982e-05,
+      "loss": 2.0564,
+      "step": 29147
+    },
+    {
+      "epoch": 0.6477333333333334,
+      "grad_norm": 1.5689301490783691,
+      "learning_rate": 7.04689931095799e-05,
+      "loss": 1.7203,
+      "step": 29148
+    },
+    {
+      "epoch": 0.6477555555555555,
+      "grad_norm": 1.9944018125534058,
+      "learning_rate": 7.046454767726161e-05,
+      "loss": 1.9389,
+      "step": 29149
+    },
+    {
+      "epoch": 0.6477777777777778,
+      "grad_norm": 2.172898054122925,
+      "learning_rate": 7.046010224494333e-05,
+      "loss": 1.5803,
+      "step": 29150
+    },
+    {
+      "epoch": 0.6478,
+      "grad_norm": 1.5711785554885864,
+      "learning_rate": 7.045565681262503e-05,
+      "loss": 2.575,
+      "step": 29151
+    },
+    {
+      "epoch": 0.6478222222222222,
+      "grad_norm": 1.8806873559951782,
+      "learning_rate": 7.045121138030674e-05,
+      "loss": 2.7309,
+      "step": 29152
+    },
+    {
+      "epoch": 0.6478444444444444,
+      "grad_norm": 1.4180324077606201,
+      "learning_rate": 7.044676594798845e-05,
+      "loss": 2.1319,
+      "step": 29153
+    },
+    {
+      "epoch": 0.6478666666666667,
+      "grad_norm": 1.5303846597671509,
+      "learning_rate": 7.044232051567016e-05,
+      "loss": 2.1468,
+      "step": 29154
+    },
+    {
+      "epoch": 0.6478888888888888,
+      "grad_norm": 1.6324728727340698,
+      "learning_rate": 7.043787508335185e-05,
+      "loss": 1.595,
+      "step": 29155
+    },
+    {
+      "epoch": 0.6479111111111111,
+      "grad_norm": 1.6591416597366333,
+      "learning_rate": 7.043342965103356e-05,
+      "loss": 2.4564,
+      "step": 29156
+    },
+    {
+      "epoch": 0.6479333333333334,
+      "grad_norm": 1.2871074676513672,
+      "learning_rate": 7.042898421871527e-05,
+      "loss": 1.7271,
+      "step": 29157
+    },
+    {
+      "epoch": 0.6479555555555555,
+      "grad_norm": 1.3471856117248535,
+      "learning_rate": 7.042453878639698e-05,
+      "loss": 2.1456,
+      "step": 29158
+    },
+    {
+      "epoch": 0.6479777777777778,
+      "grad_norm": 1.4971389770507812,
+      "learning_rate": 7.042009335407869e-05,
+      "loss": 1.8597,
+      "step": 29159
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 1.63566255569458,
+      "learning_rate": 7.04156479217604e-05,
+      "loss": 2.0335,
+      "step": 29160
+    },
+    {
+      "epoch": 0.6480222222222223,
+      "grad_norm": 1.4445345401763916,
+      "learning_rate": 7.04112024894421e-05,
+      "loss": 1.5174,
+      "step": 29161
+    },
+    {
+      "epoch": 0.6480444444444444,
+      "grad_norm": 1.633418083190918,
+      "learning_rate": 7.040675705712381e-05,
+      "loss": 1.92,
+      "step": 29162
+    },
+    {
+      "epoch": 0.6480666666666667,
+      "grad_norm": 0.33277857303619385,
+      "learning_rate": 7.040231162480552e-05,
+      "loss": 0.0243,
+      "step": 29163
+    },
+    {
+      "epoch": 0.6480888888888889,
+      "grad_norm": 1.222267746925354,
+      "learning_rate": 7.039786619248723e-05,
+      "loss": 1.1443,
+      "step": 29164
+    },
+    {
+      "epoch": 0.6481111111111111,
+      "grad_norm": 1.243935227394104,
+      "learning_rate": 7.039342076016892e-05,
+      "loss": 0.7909,
+      "step": 29165
+    },
+    {
+      "epoch": 0.6481333333333333,
+      "grad_norm": 1.5980591773986816,
+      "learning_rate": 7.038897532785063e-05,
+      "loss": 2.0629,
+      "step": 29166
+    },
+    {
+      "epoch": 0.6481555555555556,
+      "grad_norm": 1.7505806684494019,
+      "learning_rate": 7.038452989553234e-05,
+      "loss": 2.0045,
+      "step": 29167
+    },
+    {
+      "epoch": 0.6481777777777777,
+      "grad_norm": 1.5095384120941162,
+      "learning_rate": 7.038008446321405e-05,
+      "loss": 1.6692,
+      "step": 29168
+    },
+    {
+      "epoch": 0.6482,
+      "grad_norm": 1.7710570096969604,
+      "learning_rate": 7.037563903089576e-05,
+      "loss": 2.0661,
+      "step": 29169
+    },
+    {
+      "epoch": 0.6482222222222223,
+      "grad_norm": 1.5297256708145142,
+      "learning_rate": 7.037119359857747e-05,
+      "loss": 1.8635,
+      "step": 29170
+    },
+    {
+      "epoch": 0.6482444444444444,
+      "grad_norm": 1.8935678005218506,
+      "learning_rate": 7.036674816625917e-05,
+      "loss": 2.1259,
+      "step": 29171
+    },
+    {
+      "epoch": 0.6482666666666667,
+      "grad_norm": 1.6577333211898804,
+      "learning_rate": 7.036230273394088e-05,
+      "loss": 1.5838,
+      "step": 29172
+    },
+    {
+      "epoch": 0.6482888888888889,
+      "grad_norm": 1.8225229978561401,
+      "learning_rate": 7.035785730162259e-05,
+      "loss": 2.3188,
+      "step": 29173
+    },
+    {
+      "epoch": 0.6483111111111111,
+      "grad_norm": 1.8094227313995361,
+      "learning_rate": 7.03534118693043e-05,
+      "loss": 2.2508,
+      "step": 29174
+    },
+    {
+      "epoch": 0.6483333333333333,
+      "grad_norm": 1.651315689086914,
+      "learning_rate": 7.034896643698599e-05,
+      "loss": 1.8218,
+      "step": 29175
+    },
+    {
+      "epoch": 0.6483555555555556,
+      "grad_norm": 1.819054126739502,
+      "learning_rate": 7.03445210046677e-05,
+      "loss": 1.7227,
+      "step": 29176
+    },
+    {
+      "epoch": 0.6483777777777778,
+      "grad_norm": 1.9914813041687012,
+      "learning_rate": 7.034007557234941e-05,
+      "loss": 2.3753,
+      "step": 29177
+    },
+    {
+      "epoch": 0.6484,
+      "grad_norm": 2.263993501663208,
+      "learning_rate": 7.033563014003112e-05,
+      "loss": 1.438,
+      "step": 29178
+    },
+    {
+      "epoch": 0.6484222222222222,
+      "grad_norm": 1.7326619625091553,
+      "learning_rate": 7.033118470771283e-05,
+      "loss": 2.0003,
+      "step": 29179
+    },
+    {
+      "epoch": 0.6484444444444445,
+      "grad_norm": 2.027819871902466,
+      "learning_rate": 7.032673927539454e-05,
+      "loss": 2.2009,
+      "step": 29180
+    },
+    {
+      "epoch": 0.6484666666666666,
+      "grad_norm": 1.8078641891479492,
+      "learning_rate": 7.032229384307625e-05,
+      "loss": 1.7488,
+      "step": 29181
+    },
+    {
+      "epoch": 0.6484888888888889,
+      "grad_norm": 1.6092551946640015,
+      "learning_rate": 7.031784841075795e-05,
+      "loss": 1.5736,
+      "step": 29182
+    },
+    {
+      "epoch": 0.6485111111111111,
+      "grad_norm": 1.5341358184814453,
+      "learning_rate": 7.031340297843966e-05,
+      "loss": 1.9011,
+      "step": 29183
+    },
+    {
+      "epoch": 0.6485333333333333,
+      "grad_norm": 1.4618282318115234,
+      "learning_rate": 7.030895754612137e-05,
+      "loss": 1.5799,
+      "step": 29184
+    },
+    {
+      "epoch": 0.6485555555555556,
+      "grad_norm": 1.7872344255447388,
+      "learning_rate": 7.030451211380306e-05,
+      "loss": 1.6422,
+      "step": 29185
+    },
+    {
+      "epoch": 0.6485777777777778,
+      "grad_norm": 1.677580714225769,
+      "learning_rate": 7.030006668148477e-05,
+      "loss": 1.9763,
+      "step": 29186
+    },
+    {
+      "epoch": 0.6486,
+      "grad_norm": 1.6863222122192383,
+      "learning_rate": 7.02956212491665e-05,
+      "loss": 1.8678,
+      "step": 29187
+    },
+    {
+      "epoch": 0.6486222222222222,
+      "grad_norm": 1.24997079372406,
+      "learning_rate": 7.029117581684819e-05,
+      "loss": 1.3101,
+      "step": 29188
+    },
+    {
+      "epoch": 0.6486444444444445,
+      "grad_norm": 1.3846231698989868,
+      "learning_rate": 7.02867303845299e-05,
+      "loss": 1.7414,
+      "step": 29189
+    },
+    {
+      "epoch": 0.6486666666666666,
+      "grad_norm": 1.944069504737854,
+      "learning_rate": 7.028228495221161e-05,
+      "loss": 2.381,
+      "step": 29190
+    },
+    {
+      "epoch": 0.6486888888888889,
+      "grad_norm": 1.7412798404693604,
+      "learning_rate": 7.027783951989332e-05,
+      "loss": 1.7451,
+      "step": 29191
+    },
+    {
+      "epoch": 0.6487111111111111,
+      "grad_norm": 1.9458290338516235,
+      "learning_rate": 7.027339408757501e-05,
+      "loss": 2.1176,
+      "step": 29192
+    },
+    {
+      "epoch": 0.6487333333333334,
+      "grad_norm": 1.7362381219863892,
+      "learning_rate": 7.026894865525672e-05,
+      "loss": 1.4444,
+      "step": 29193
+    },
+    {
+      "epoch": 0.6487555555555555,
+      "grad_norm": 1.9106467962265015,
+      "learning_rate": 7.026450322293843e-05,
+      "loss": 1.8456,
+      "step": 29194
+    },
+    {
+      "epoch": 0.6487777777777778,
+      "grad_norm": 1.5858999490737915,
+      "learning_rate": 7.026005779062014e-05,
+      "loss": 1.6961,
+      "step": 29195
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 1.915849208831787,
+      "learning_rate": 7.025561235830185e-05,
+      "loss": 1.7903,
+      "step": 29196
+    },
+    {
+      "epoch": 0.6488222222222222,
+      "grad_norm": 1.544602394104004,
+      "learning_rate": 7.025116692598356e-05,
+      "loss": 1.1969,
+      "step": 29197
+    },
+    {
+      "epoch": 0.6488444444444444,
+      "grad_norm": 1.9226042032241821,
+      "learning_rate": 7.024672149366526e-05,
+      "loss": 1.826,
+      "step": 29198
+    },
+    {
+      "epoch": 0.6488666666666667,
+      "grad_norm": 1.8153972625732422,
+      "learning_rate": 7.024227606134697e-05,
+      "loss": 1.5014,
+      "step": 29199
+    },
+    {
+      "epoch": 0.6488888888888888,
+      "grad_norm": 1.5163156986236572,
+      "learning_rate": 7.023783062902868e-05,
+      "loss": 1.1973,
+      "step": 29200
+    },
+    {
+      "epoch": 0.6489111111111111,
+      "grad_norm": 1.4240121841430664,
+      "learning_rate": 7.023338519671039e-05,
+      "loss": 2.5114,
+      "step": 29201
+    },
+    {
+      "epoch": 0.6489333333333334,
+      "grad_norm": 1.366631031036377,
+      "learning_rate": 7.022893976439208e-05,
+      "loss": 2.4166,
+      "step": 29202
+    },
+    {
+      "epoch": 0.6489555555555555,
+      "grad_norm": 1.802046298980713,
+      "learning_rate": 7.02244943320738e-05,
+      "loss": 3.01,
+      "step": 29203
+    },
+    {
+      "epoch": 0.6489777777777778,
+      "grad_norm": 1.4731266498565674,
+      "learning_rate": 7.02200488997555e-05,
+      "loss": 2.2481,
+      "step": 29204
+    },
+    {
+      "epoch": 0.649,
+      "grad_norm": 1.8114069700241089,
+      "learning_rate": 7.021560346743721e-05,
+      "loss": 2.2668,
+      "step": 29205
+    },
+    {
+      "epoch": 0.6490222222222222,
+      "grad_norm": 1.574411153793335,
+      "learning_rate": 7.021115803511892e-05,
+      "loss": 2.3854,
+      "step": 29206
+    },
+    {
+      "epoch": 0.6490444444444444,
+      "grad_norm": 1.6780638694763184,
+      "learning_rate": 7.020671260280063e-05,
+      "loss": 2.4121,
+      "step": 29207
+    },
+    {
+      "epoch": 0.6490666666666667,
+      "grad_norm": 1.9986330270767212,
+      "learning_rate": 7.020226717048233e-05,
+      "loss": 2.5218,
+      "step": 29208
+    },
+    {
+      "epoch": 0.6490888888888889,
+      "grad_norm": 1.8758296966552734,
+      "learning_rate": 7.019782173816404e-05,
+      "loss": 1.8691,
+      "step": 29209
+    },
+    {
+      "epoch": 0.6491111111111111,
+      "grad_norm": 1.7497591972351074,
+      "learning_rate": 7.019337630584575e-05,
+      "loss": 2.0199,
+      "step": 29210
+    },
+    {
+      "epoch": 0.6491333333333333,
+      "grad_norm": 2.0398762226104736,
+      "learning_rate": 7.018893087352746e-05,
+      "loss": 2.1921,
+      "step": 29211
+    },
+    {
+      "epoch": 0.6491555555555556,
+      "grad_norm": 1.2331268787384033,
+      "learning_rate": 7.018448544120915e-05,
+      "loss": 0.9171,
+      "step": 29212
+    },
+    {
+      "epoch": 0.6491777777777777,
+      "grad_norm": 1.537521243095398,
+      "learning_rate": 7.018004000889086e-05,
+      "loss": 1.5539,
+      "step": 29213
+    },
+    {
+      "epoch": 0.6492,
+      "grad_norm": 1.4166390895843506,
+      "learning_rate": 7.017559457657257e-05,
+      "loss": 1.7561,
+      "step": 29214
+    },
+    {
+      "epoch": 0.6492222222222223,
+      "grad_norm": 1.5925884246826172,
+      "learning_rate": 7.017114914425428e-05,
+      "loss": 1.7955,
+      "step": 29215
+    },
+    {
+      "epoch": 0.6492444444444444,
+      "grad_norm": 1.5275793075561523,
+      "learning_rate": 7.016670371193599e-05,
+      "loss": 1.911,
+      "step": 29216
+    },
+    {
+      "epoch": 0.6492666666666667,
+      "grad_norm": 2.0152485370635986,
+      "learning_rate": 7.01622582796177e-05,
+      "loss": 1.8886,
+      "step": 29217
+    },
+    {
+      "epoch": 0.6492888888888889,
+      "grad_norm": 1.747724175453186,
+      "learning_rate": 7.01578128472994e-05,
+      "loss": 2.0026,
+      "step": 29218
+    },
+    {
+      "epoch": 0.6493111111111111,
+      "grad_norm": 1.6655484437942505,
+      "learning_rate": 7.015336741498111e-05,
+      "loss": 2.2749,
+      "step": 29219
+    },
+    {
+      "epoch": 0.6493333333333333,
+      "grad_norm": 1.7254546880722046,
+      "learning_rate": 7.014892198266282e-05,
+      "loss": 1.9314,
+      "step": 29220
+    },
+    {
+      "epoch": 0.6493555555555556,
+      "grad_norm": 1.608613133430481,
+      "learning_rate": 7.014447655034453e-05,
+      "loss": 1.9615,
+      "step": 29221
+    },
+    {
+      "epoch": 0.6493777777777778,
+      "grad_norm": 1.9358397722244263,
+      "learning_rate": 7.014003111802622e-05,
+      "loss": 2.3336,
+      "step": 29222
+    },
+    {
+      "epoch": 0.6494,
+      "grad_norm": 1.6523768901824951,
+      "learning_rate": 7.013558568570793e-05,
+      "loss": 2.1089,
+      "step": 29223
+    },
+    {
+      "epoch": 0.6494222222222222,
+      "grad_norm": 2.0118861198425293,
+      "learning_rate": 7.013114025338965e-05,
+      "loss": 2.4732,
+      "step": 29224
+    },
+    {
+      "epoch": 0.6494444444444445,
+      "grad_norm": 1.8435945510864258,
+      "learning_rate": 7.012669482107135e-05,
+      "loss": 1.9642,
+      "step": 29225
+    },
+    {
+      "epoch": 0.6494666666666666,
+      "grad_norm": 1.514154076576233,
+      "learning_rate": 7.012224938875306e-05,
+      "loss": 1.6705,
+      "step": 29226
+    },
+    {
+      "epoch": 0.6494888888888889,
+      "grad_norm": 1.780166745185852,
+      "learning_rate": 7.011780395643477e-05,
+      "loss": 1.512,
+      "step": 29227
+    },
+    {
+      "epoch": 0.6495111111111112,
+      "grad_norm": 1.433593511581421,
+      "learning_rate": 7.011335852411648e-05,
+      "loss": 1.7877,
+      "step": 29228
+    },
+    {
+      "epoch": 0.6495333333333333,
+      "grad_norm": 1.7081892490386963,
+      "learning_rate": 7.010891309179818e-05,
+      "loss": 1.9008,
+      "step": 29229
+    },
+    {
+      "epoch": 0.6495555555555556,
+      "grad_norm": 1.8547171354293823,
+      "learning_rate": 7.010446765947989e-05,
+      "loss": 2.4596,
+      "step": 29230
+    },
+    {
+      "epoch": 0.6495777777777778,
+      "grad_norm": 1.7276898622512817,
+      "learning_rate": 7.01000222271616e-05,
+      "loss": 1.9145,
+      "step": 29231
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 2.084520101547241,
+      "learning_rate": 7.00955767948433e-05,
+      "loss": 2.3716,
+      "step": 29232
+    },
+    {
+      "epoch": 0.6496222222222222,
+      "grad_norm": 1.6999021768569946,
+      "learning_rate": 7.009113136252501e-05,
+      "loss": 1.7156,
+      "step": 29233
+    },
+    {
+      "epoch": 0.6496444444444445,
+      "grad_norm": 1.4727013111114502,
+      "learning_rate": 7.008668593020672e-05,
+      "loss": 1.7079,
+      "step": 29234
+    },
+    {
+      "epoch": 0.6496666666666666,
+      "grad_norm": 1.6207228899002075,
+      "learning_rate": 7.008224049788842e-05,
+      "loss": 1.4759,
+      "step": 29235
+    },
+    {
+      "epoch": 0.6496888888888889,
+      "grad_norm": 1.706359624862671,
+      "learning_rate": 7.007779506557013e-05,
+      "loss": 1.8351,
+      "step": 29236
+    },
+    {
+      "epoch": 0.6497111111111111,
+      "grad_norm": 1.4816555976867676,
+      "learning_rate": 7.007334963325184e-05,
+      "loss": 1.5791,
+      "step": 29237
+    },
+    {
+      "epoch": 0.6497333333333334,
+      "grad_norm": 1.713639259338379,
+      "learning_rate": 7.006890420093355e-05,
+      "loss": 1.7698,
+      "step": 29238
+    },
+    {
+      "epoch": 0.6497555555555555,
+      "grad_norm": 1.8787704706192017,
+      "learning_rate": 7.006445876861525e-05,
+      "loss": 1.9489,
+      "step": 29239
+    },
+    {
+      "epoch": 0.6497777777777778,
+      "grad_norm": 1.6118589639663696,
+      "learning_rate": 7.006001333629695e-05,
+      "loss": 1.2888,
+      "step": 29240
+    },
+    {
+      "epoch": 0.6498,
+      "grad_norm": 1.7838908433914185,
+      "learning_rate": 7.005556790397866e-05,
+      "loss": 2.0457,
+      "step": 29241
+    },
+    {
+      "epoch": 0.6498222222222222,
+      "grad_norm": 2.315001964569092,
+      "learning_rate": 7.005112247166037e-05,
+      "loss": 2.044,
+      "step": 29242
+    },
+    {
+      "epoch": 0.6498444444444444,
+      "grad_norm": 1.7782366275787354,
+      "learning_rate": 7.004667703934208e-05,
+      "loss": 1.7675,
+      "step": 29243
+    },
+    {
+      "epoch": 0.6498666666666667,
+      "grad_norm": 1.7710777521133423,
+      "learning_rate": 7.004223160702379e-05,
+      "loss": 1.7482,
+      "step": 29244
+    },
+    {
+      "epoch": 0.6498888888888888,
+      "grad_norm": 2.2236485481262207,
+      "learning_rate": 7.003778617470549e-05,
+      "loss": 1.7651,
+      "step": 29245
+    },
+    {
+      "epoch": 0.6499111111111111,
+      "grad_norm": 1.7950549125671387,
+      "learning_rate": 7.00333407423872e-05,
+      "loss": 1.6507,
+      "step": 29246
+    },
+    {
+      "epoch": 0.6499333333333334,
+      "grad_norm": 1.7493401765823364,
+      "learning_rate": 7.002889531006891e-05,
+      "loss": 1.5964,
+      "step": 29247
+    },
+    {
+      "epoch": 0.6499555555555555,
+      "grad_norm": 2.2070393562316895,
+      "learning_rate": 7.002444987775062e-05,
+      "loss": 1.8369,
+      "step": 29248
+    },
+    {
+      "epoch": 0.6499777777777778,
+      "grad_norm": 2.0941531658172607,
+      "learning_rate": 7.002000444543231e-05,
+      "loss": 2.0079,
+      "step": 29249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0086233615875244,
+      "learning_rate": 7.001555901311402e-05,
+      "loss": 1.8565,
+      "step": 29250
+    },
+    {
+      "epoch": 0.6500222222222222,
+      "grad_norm": 1.5135737657546997,
+      "learning_rate": 7.001111358079573e-05,
+      "loss": 2.3019,
+      "step": 29251
+    },
+    {
+      "epoch": 0.6500444444444444,
+      "grad_norm": 1.2567074298858643,
+      "learning_rate": 7.000666814847744e-05,
+      "loss": 1.8777,
+      "step": 29252
+    },
+    {
+      "epoch": 0.6500666666666667,
+      "grad_norm": 0.17498348653316498,
+      "learning_rate": 7.000222271615915e-05,
+      "loss": 0.018,
+      "step": 29253
+    },
+    {
+      "epoch": 0.6500888888888889,
+      "grad_norm": 1.7214148044586182,
+      "learning_rate": 6.999777728384086e-05,
+      "loss": 2.4307,
+      "step": 29254
+    },
+    {
+      "epoch": 0.6501111111111111,
+      "grad_norm": 1.6908961534500122,
+      "learning_rate": 6.999333185152256e-05,
+      "loss": 2.3565,
+      "step": 29255
+    },
+    {
+      "epoch": 0.6501333333333333,
+      "grad_norm": 2.346676826477051,
+      "learning_rate": 6.998888641920427e-05,
+      "loss": 2.1449,
+      "step": 29256
+    },
+    {
+      "epoch": 0.6501555555555556,
+      "grad_norm": 1.7402980327606201,
+      "learning_rate": 6.998444098688598e-05,
+      "loss": 2.0284,
+      "step": 29257
+    },
+    {
+      "epoch": 0.6501777777777777,
+      "grad_norm": 1.891755223274231,
+      "learning_rate": 6.997999555456769e-05,
+      "loss": 2.0583,
+      "step": 29258
+    },
+    {
+      "epoch": 0.6502,
+      "grad_norm": 1.889899730682373,
+      "learning_rate": 6.997555012224938e-05,
+      "loss": 1.9627,
+      "step": 29259
+    },
+    {
+      "epoch": 0.6502222222222223,
+      "grad_norm": 1.6953115463256836,
+      "learning_rate": 6.997110468993109e-05,
+      "loss": 2.3591,
+      "step": 29260
+    },
+    {
+      "epoch": 0.6502444444444444,
+      "grad_norm": 1.7562839984893799,
+      "learning_rate": 6.996665925761282e-05,
+      "loss": 2.2369,
+      "step": 29261
+    },
+    {
+      "epoch": 0.6502666666666667,
+      "grad_norm": 1.4049495458602905,
+      "learning_rate": 6.996221382529451e-05,
+      "loss": 2.1355,
+      "step": 29262
+    },
+    {
+      "epoch": 0.6502888888888889,
+      "grad_norm": 1.5074312686920166,
+      "learning_rate": 6.995776839297622e-05,
+      "loss": 2.13,
+      "step": 29263
+    },
+    {
+      "epoch": 0.6503111111111111,
+      "grad_norm": 1.7136636972427368,
+      "learning_rate": 6.995332296065793e-05,
+      "loss": 1.8946,
+      "step": 29264
+    },
+    {
+      "epoch": 0.6503333333333333,
+      "grad_norm": 1.636090636253357,
+      "learning_rate": 6.994887752833963e-05,
+      "loss": 2.1277,
+      "step": 29265
+    },
+    {
+      "epoch": 0.6503555555555556,
+      "grad_norm": 2.028170347213745,
+      "learning_rate": 6.994443209602134e-05,
+      "loss": 2.0288,
+      "step": 29266
+    },
+    {
+      "epoch": 0.6503777777777778,
+      "grad_norm": 1.5597490072250366,
+      "learning_rate": 6.993998666370305e-05,
+      "loss": 1.7009,
+      "step": 29267
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 1.6211811304092407,
+      "learning_rate": 6.993554123138476e-05,
+      "loss": 2.1396,
+      "step": 29268
+    },
+    {
+      "epoch": 0.6504222222222222,
+      "grad_norm": 1.767333745956421,
+      "learning_rate": 6.993109579906647e-05,
+      "loss": 2.4782,
+      "step": 29269
+    },
+    {
+      "epoch": 0.6504444444444445,
+      "grad_norm": 1.4722760915756226,
+      "learning_rate": 6.992665036674818e-05,
+      "loss": 1.9787,
+      "step": 29270
+    },
+    {
+      "epoch": 0.6504666666666666,
+      "grad_norm": 1.608223557472229,
+      "learning_rate": 6.992220493442989e-05,
+      "loss": 2.326,
+      "step": 29271
+    },
+    {
+      "epoch": 0.6504888888888889,
+      "grad_norm": 1.988062858581543,
+      "learning_rate": 6.991775950211158e-05,
+      "loss": 2.6228,
+      "step": 29272
+    },
+    {
+      "epoch": 0.6505111111111112,
+      "grad_norm": 1.5838956832885742,
+      "learning_rate": 6.991331406979329e-05,
+      "loss": 1.75,
+      "step": 29273
+    },
+    {
+      "epoch": 0.6505333333333333,
+      "grad_norm": 1.4592198133468628,
+      "learning_rate": 6.9908868637475e-05,
+      "loss": 1.7813,
+      "step": 29274
+    },
+    {
+      "epoch": 0.6505555555555556,
+      "grad_norm": 1.73031485080719,
+      "learning_rate": 6.99044232051567e-05,
+      "loss": 2.2714,
+      "step": 29275
+    },
+    {
+      "epoch": 0.6505777777777778,
+      "grad_norm": 1.753232717514038,
+      "learning_rate": 6.98999777728384e-05,
+      "loss": 1.9807,
+      "step": 29276
+    },
+    {
+      "epoch": 0.6506,
+      "grad_norm": 1.6164661645889282,
+      "learning_rate": 6.989553234052012e-05,
+      "loss": 1.4149,
+      "step": 29277
+    },
+    {
+      "epoch": 0.6506222222222222,
+      "grad_norm": 1.8052868843078613,
+      "learning_rate": 6.989108690820183e-05,
+      "loss": 1.9208,
+      "step": 29278
+    },
+    {
+      "epoch": 0.6506444444444445,
+      "grad_norm": 1.4451102018356323,
+      "learning_rate": 6.988664147588353e-05,
+      "loss": 2.1409,
+      "step": 29279
+    },
+    {
+      "epoch": 0.6506666666666666,
+      "grad_norm": 1.5738019943237305,
+      "learning_rate": 6.988219604356524e-05,
+      "loss": 1.9391,
+      "step": 29280
+    },
+    {
+      "epoch": 0.6506888888888889,
+      "grad_norm": 1.637847661972046,
+      "learning_rate": 6.987775061124695e-05,
+      "loss": 1.9187,
+      "step": 29281
+    },
+    {
+      "epoch": 0.6507111111111111,
+      "grad_norm": 1.3089234828948975,
+      "learning_rate": 6.987330517892865e-05,
+      "loss": 0.885,
+      "step": 29282
+    },
+    {
+      "epoch": 0.6507333333333334,
+      "grad_norm": 1.7746233940124512,
+      "learning_rate": 6.986885974661036e-05,
+      "loss": 1.9604,
+      "step": 29283
+    },
+    {
+      "epoch": 0.6507555555555555,
+      "grad_norm": 1.536644458770752,
+      "learning_rate": 6.986441431429207e-05,
+      "loss": 1.9093,
+      "step": 29284
+    },
+    {
+      "epoch": 0.6507777777777778,
+      "grad_norm": 1.560268759727478,
+      "learning_rate": 6.985996888197378e-05,
+      "loss": 1.367,
+      "step": 29285
+    },
+    {
+      "epoch": 0.6508,
+      "grad_norm": 1.840164065361023,
+      "learning_rate": 6.985552344965548e-05,
+      "loss": 1.9394,
+      "step": 29286
+    },
+    {
+      "epoch": 0.6508222222222222,
+      "grad_norm": 1.6413569450378418,
+      "learning_rate": 6.985107801733718e-05,
+      "loss": 1.4838,
+      "step": 29287
+    },
+    {
+      "epoch": 0.6508444444444444,
+      "grad_norm": 1.576214075088501,
+      "learning_rate": 6.98466325850189e-05,
+      "loss": 1.8984,
+      "step": 29288
+    },
+    {
+      "epoch": 0.6508666666666667,
+      "grad_norm": 1.3859788179397583,
+      "learning_rate": 6.98421871527006e-05,
+      "loss": 1.4899,
+      "step": 29289
+    },
+    {
+      "epoch": 0.6508888888888889,
+      "grad_norm": 1.5814870595932007,
+      "learning_rate": 6.983774172038231e-05,
+      "loss": 1.4805,
+      "step": 29290
+    },
+    {
+      "epoch": 0.6509111111111111,
+      "grad_norm": 1.8000617027282715,
+      "learning_rate": 6.983329628806402e-05,
+      "loss": 1.8311,
+      "step": 29291
+    },
+    {
+      "epoch": 0.6509333333333334,
+      "grad_norm": 1.9427064657211304,
+      "learning_rate": 6.982885085574572e-05,
+      "loss": 1.9634,
+      "step": 29292
+    },
+    {
+      "epoch": 0.6509555555555555,
+      "grad_norm": 1.5224941968917847,
+      "learning_rate": 6.982440542342743e-05,
+      "loss": 1.4516,
+      "step": 29293
+    },
+    {
+      "epoch": 0.6509777777777778,
+      "grad_norm": 1.658179521560669,
+      "learning_rate": 6.981995999110914e-05,
+      "loss": 1.682,
+      "step": 29294
+    },
+    {
+      "epoch": 0.651,
+      "grad_norm": 1.6062688827514648,
+      "learning_rate": 6.981551455879085e-05,
+      "loss": 1.7161,
+      "step": 29295
+    },
+    {
+      "epoch": 0.6510222222222222,
+      "grad_norm": 2.0045063495635986,
+      "learning_rate": 6.981106912647254e-05,
+      "loss": 2.073,
+      "step": 29296
+    },
+    {
+      "epoch": 0.6510444444444444,
+      "grad_norm": 1.722097396850586,
+      "learning_rate": 6.980662369415425e-05,
+      "loss": 1.7127,
+      "step": 29297
+    },
+    {
+      "epoch": 0.6510666666666667,
+      "grad_norm": 1.8101304769515991,
+      "learning_rate": 6.980217826183598e-05,
+      "loss": 1.6494,
+      "step": 29298
+    },
+    {
+      "epoch": 0.6510888888888889,
+      "grad_norm": 1.9141604900360107,
+      "learning_rate": 6.979773282951767e-05,
+      "loss": 1.4045,
+      "step": 29299
+    },
+    {
+      "epoch": 0.6511111111111111,
+      "grad_norm": 2.0712220668792725,
+      "learning_rate": 6.979328739719938e-05,
+      "loss": 1.5308,
+      "step": 29300
+    },
+    {
+      "epoch": 0.6511333333333333,
+      "grad_norm": 1.2344119548797607,
+      "learning_rate": 6.978884196488109e-05,
+      "loss": 1.1758,
+      "step": 29301
+    },
+    {
+      "epoch": 0.6511555555555556,
+      "grad_norm": 1.4297155141830444,
+      "learning_rate": 6.978439653256279e-05,
+      "loss": 2.1887,
+      "step": 29302
+    },
+    {
+      "epoch": 0.6511777777777777,
+      "grad_norm": 0.8754487633705139,
+      "learning_rate": 6.97799511002445e-05,
+      "loss": 0.8814,
+      "step": 29303
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 1.4717179536819458,
+      "learning_rate": 6.977550566792621e-05,
+      "loss": 2.1349,
+      "step": 29304
+    },
+    {
+      "epoch": 0.6512222222222223,
+      "grad_norm": 1.4106658697128296,
+      "learning_rate": 6.977106023560792e-05,
+      "loss": 2.1573,
+      "step": 29305
+    },
+    {
+      "epoch": 0.6512444444444444,
+      "grad_norm": 1.9039552211761475,
+      "learning_rate": 6.976661480328963e-05,
+      "loss": 1.8249,
+      "step": 29306
+    },
+    {
+      "epoch": 0.6512666666666667,
+      "grad_norm": 1.3200098276138306,
+      "learning_rate": 6.976216937097134e-05,
+      "loss": 1.7248,
+      "step": 29307
+    },
+    {
+      "epoch": 0.6512888888888889,
+      "grad_norm": 1.861236810684204,
+      "learning_rate": 6.975772393865305e-05,
+      "loss": 2.4081,
+      "step": 29308
+    },
+    {
+      "epoch": 0.6513111111111111,
+      "grad_norm": 1.6973249912261963,
+      "learning_rate": 6.975327850633474e-05,
+      "loss": 1.9996,
+      "step": 29309
+    },
+    {
+      "epoch": 0.6513333333333333,
+      "grad_norm": 1.8404150009155273,
+      "learning_rate": 6.974883307401645e-05,
+      "loss": 2.4652,
+      "step": 29310
+    },
+    {
+      "epoch": 0.6513555555555556,
+      "grad_norm": 1.7387663125991821,
+      "learning_rate": 6.974438764169816e-05,
+      "loss": 1.4416,
+      "step": 29311
+    },
+    {
+      "epoch": 0.6513777777777778,
+      "grad_norm": 1.6357557773590088,
+      "learning_rate": 6.973994220937986e-05,
+      "loss": 1.6109,
+      "step": 29312
+    },
+    {
+      "epoch": 0.6514,
+      "grad_norm": 1.6437475681304932,
+      "learning_rate": 6.973549677706157e-05,
+      "loss": 1.9761,
+      "step": 29313
+    },
+    {
+      "epoch": 0.6514222222222222,
+      "grad_norm": 1.6953421831130981,
+      "learning_rate": 6.973105134474328e-05,
+      "loss": 2.1356,
+      "step": 29314
+    },
+    {
+      "epoch": 0.6514444444444445,
+      "grad_norm": 1.6046780347824097,
+      "learning_rate": 6.972660591242499e-05,
+      "loss": 2.1672,
+      "step": 29315
+    },
+    {
+      "epoch": 0.6514666666666666,
+      "grad_norm": 1.550767421722412,
+      "learning_rate": 6.97221604801067e-05,
+      "loss": 1.653,
+      "step": 29316
+    },
+    {
+      "epoch": 0.6514888888888889,
+      "grad_norm": 1.5348842144012451,
+      "learning_rate": 6.97177150477884e-05,
+      "loss": 1.8861,
+      "step": 29317
+    },
+    {
+      "epoch": 0.6515111111111112,
+      "grad_norm": 1.7914471626281738,
+      "learning_rate": 6.971326961547012e-05,
+      "loss": 2.2415,
+      "step": 29318
+    },
+    {
+      "epoch": 0.6515333333333333,
+      "grad_norm": 1.428574800491333,
+      "learning_rate": 6.970882418315181e-05,
+      "loss": 1.4008,
+      "step": 29319
+    },
+    {
+      "epoch": 0.6515555555555556,
+      "grad_norm": 1.382114052772522,
+      "learning_rate": 6.970437875083352e-05,
+      "loss": 1.7083,
+      "step": 29320
+    },
+    {
+      "epoch": 0.6515777777777778,
+      "grad_norm": 2.0291502475738525,
+      "learning_rate": 6.969993331851523e-05,
+      "loss": 1.5898,
+      "step": 29321
+    },
+    {
+      "epoch": 0.6516,
+      "grad_norm": 1.5214177370071411,
+      "learning_rate": 6.969548788619693e-05,
+      "loss": 1.9985,
+      "step": 29322
+    },
+    {
+      "epoch": 0.6516222222222222,
+      "grad_norm": 1.516836166381836,
+      "learning_rate": 6.969104245387864e-05,
+      "loss": 1.6018,
+      "step": 29323
+    },
+    {
+      "epoch": 0.6516444444444445,
+      "grad_norm": 1.7093638181686401,
+      "learning_rate": 6.968659702156035e-05,
+      "loss": 2.0104,
+      "step": 29324
+    },
+    {
+      "epoch": 0.6516666666666666,
+      "grad_norm": 1.858313798904419,
+      "learning_rate": 6.968215158924206e-05,
+      "loss": 2.0254,
+      "step": 29325
+    },
+    {
+      "epoch": 0.6516888888888889,
+      "grad_norm": 2.083031177520752,
+      "learning_rate": 6.967770615692377e-05,
+      "loss": 2.307,
+      "step": 29326
+    },
+    {
+      "epoch": 0.6517111111111111,
+      "grad_norm": 1.7572811841964722,
+      "learning_rate": 6.967326072460547e-05,
+      "loss": 1.9557,
+      "step": 29327
+    },
+    {
+      "epoch": 0.6517333333333334,
+      "grad_norm": 1.7568235397338867,
+      "learning_rate": 6.966881529228718e-05,
+      "loss": 2.1862,
+      "step": 29328
+    },
+    {
+      "epoch": 0.6517555555555555,
+      "grad_norm": 1.9669950008392334,
+      "learning_rate": 6.966436985996888e-05,
+      "loss": 1.9378,
+      "step": 29329
+    },
+    {
+      "epoch": 0.6517777777777778,
+      "grad_norm": 1.1136808395385742,
+      "learning_rate": 6.965992442765059e-05,
+      "loss": 0.8185,
+      "step": 29330
+    },
+    {
+      "epoch": 0.6518,
+      "grad_norm": 1.4626590013504028,
+      "learning_rate": 6.96554789953323e-05,
+      "loss": 1.6177,
+      "step": 29331
+    },
+    {
+      "epoch": 0.6518222222222222,
+      "grad_norm": 1.4172871112823486,
+      "learning_rate": 6.9651033563014e-05,
+      "loss": 1.1709,
+      "step": 29332
+    },
+    {
+      "epoch": 0.6518444444444444,
+      "grad_norm": 1.4712958335876465,
+      "learning_rate": 6.96465881306957e-05,
+      "loss": 1.5426,
+      "step": 29333
+    },
+    {
+      "epoch": 0.6518666666666667,
+      "grad_norm": 1.5287601947784424,
+      "learning_rate": 6.964214269837741e-05,
+      "loss": 1.3882,
+      "step": 29334
+    },
+    {
+      "epoch": 0.6518888888888889,
+      "grad_norm": 1.9561853408813477,
+      "learning_rate": 6.963769726605914e-05,
+      "loss": 1.9832,
+      "step": 29335
+    },
+    {
+      "epoch": 0.6519111111111111,
+      "grad_norm": 1.5903472900390625,
+      "learning_rate": 6.963325183374083e-05,
+      "loss": 1.5891,
+      "step": 29336
+    },
+    {
+      "epoch": 0.6519333333333334,
+      "grad_norm": 1.7223854064941406,
+      "learning_rate": 6.962880640142254e-05,
+      "loss": 2.1605,
+      "step": 29337
+    },
+    {
+      "epoch": 0.6519555555555555,
+      "grad_norm": 1.8184937238693237,
+      "learning_rate": 6.962436096910425e-05,
+      "loss": 1.7764,
+      "step": 29338
+    },
+    {
+      "epoch": 0.6519777777777778,
+      "grad_norm": 2.029592752456665,
+      "learning_rate": 6.961991553678595e-05,
+      "loss": 2.023,
+      "step": 29339
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 1.8846449851989746,
+      "learning_rate": 6.961547010446766e-05,
+      "loss": 1.8893,
+      "step": 29340
+    },
+    {
+      "epoch": 0.6520222222222222,
+      "grad_norm": 1.637250542640686,
+      "learning_rate": 6.961102467214937e-05,
+      "loss": 1.5786,
+      "step": 29341
+    },
+    {
+      "epoch": 0.6520444444444444,
+      "grad_norm": 1.8435598611831665,
+      "learning_rate": 6.960657923983108e-05,
+      "loss": 1.6511,
+      "step": 29342
+    },
+    {
+      "epoch": 0.6520666666666667,
+      "grad_norm": 1.94805109500885,
+      "learning_rate": 6.960213380751279e-05,
+      "loss": 1.8842,
+      "step": 29343
+    },
+    {
+      "epoch": 0.6520888888888889,
+      "grad_norm": 1.7755887508392334,
+      "learning_rate": 6.95976883751945e-05,
+      "loss": 2.0506,
+      "step": 29344
+    },
+    {
+      "epoch": 0.6521111111111111,
+      "grad_norm": 2.023674964904785,
+      "learning_rate": 6.959324294287621e-05,
+      "loss": 2.1951,
+      "step": 29345
+    },
+    {
+      "epoch": 0.6521333333333333,
+      "grad_norm": 1.8801541328430176,
+      "learning_rate": 6.95887975105579e-05,
+      "loss": 2.0137,
+      "step": 29346
+    },
+    {
+      "epoch": 0.6521555555555556,
+      "grad_norm": 1.7253317832946777,
+      "learning_rate": 6.958435207823961e-05,
+      "loss": 1.9039,
+      "step": 29347
+    },
+    {
+      "epoch": 0.6521777777777777,
+      "grad_norm": 1.8239405155181885,
+      "learning_rate": 6.957990664592132e-05,
+      "loss": 1.5129,
+      "step": 29348
+    },
+    {
+      "epoch": 0.6522,
+      "grad_norm": 1.4606218338012695,
+      "learning_rate": 6.957546121360302e-05,
+      "loss": 0.9085,
+      "step": 29349
+    },
+    {
+      "epoch": 0.6522222222222223,
+      "grad_norm": 2.2330949306488037,
+      "learning_rate": 6.957101578128473e-05,
+      "loss": 1.3404,
+      "step": 29350
+    },
+    {
+      "epoch": 0.6522444444444444,
+      "grad_norm": 1.858919382095337,
+      "learning_rate": 6.956657034896644e-05,
+      "loss": 2.718,
+      "step": 29351
+    },
+    {
+      "epoch": 0.6522666666666667,
+      "grad_norm": 0.23877856135368347,
+      "learning_rate": 6.956212491664815e-05,
+      "loss": 0.0166,
+      "step": 29352
+    },
+    {
+      "epoch": 0.6522888888888889,
+      "grad_norm": 0.24272207915782928,
+      "learning_rate": 6.955767948432986e-05,
+      "loss": 0.0162,
+      "step": 29353
+    },
+    {
+      "epoch": 0.6523111111111111,
+      "grad_norm": 1.4344298839569092,
+      "learning_rate": 6.955323405201157e-05,
+      "loss": 2.3705,
+      "step": 29354
+    },
+    {
+      "epoch": 0.6523333333333333,
+      "grad_norm": 1.3044419288635254,
+      "learning_rate": 6.954878861969328e-05,
+      "loss": 2.1032,
+      "step": 29355
+    },
+    {
+      "epoch": 0.6523555555555556,
+      "grad_norm": 1.8512741327285767,
+      "learning_rate": 6.954434318737497e-05,
+      "loss": 2.3973,
+      "step": 29356
+    },
+    {
+      "epoch": 0.6523777777777777,
+      "grad_norm": 1.987452745437622,
+      "learning_rate": 6.953989775505668e-05,
+      "loss": 1.2002,
+      "step": 29357
+    },
+    {
+      "epoch": 0.6524,
+      "grad_norm": 1.310753583908081,
+      "learning_rate": 6.953545232273839e-05,
+      "loss": 1.9942,
+      "step": 29358
+    },
+    {
+      "epoch": 0.6524222222222222,
+      "grad_norm": 1.4221051931381226,
+      "learning_rate": 6.953100689042009e-05,
+      "loss": 1.8441,
+      "step": 29359
+    },
+    {
+      "epoch": 0.6524444444444445,
+      "grad_norm": 1.6217492818832397,
+      "learning_rate": 6.95265614581018e-05,
+      "loss": 2.3557,
+      "step": 29360
+    },
+    {
+      "epoch": 0.6524666666666666,
+      "grad_norm": 1.9682235717773438,
+      "learning_rate": 6.952211602578351e-05,
+      "loss": 1.7465,
+      "step": 29361
+    },
+    {
+      "epoch": 0.6524888888888889,
+      "grad_norm": 1.491863489151001,
+      "learning_rate": 6.951767059346522e-05,
+      "loss": 1.6969,
+      "step": 29362
+    },
+    {
+      "epoch": 0.6525111111111112,
+      "grad_norm": 1.5127167701721191,
+      "learning_rate": 6.951322516114693e-05,
+      "loss": 2.0036,
+      "step": 29363
+    },
+    {
+      "epoch": 0.6525333333333333,
+      "grad_norm": 1.4923393726348877,
+      "learning_rate": 6.950877972882864e-05,
+      "loss": 1.715,
+      "step": 29364
+    },
+    {
+      "epoch": 0.6525555555555556,
+      "grad_norm": 1.6749690771102905,
+      "learning_rate": 6.950433429651035e-05,
+      "loss": 2.1604,
+      "step": 29365
+    },
+    {
+      "epoch": 0.6525777777777778,
+      "grad_norm": 1.8960293531417847,
+      "learning_rate": 6.949988886419204e-05,
+      "loss": 1.997,
+      "step": 29366
+    },
+    {
+      "epoch": 0.6526,
+      "grad_norm": 1.4093700647354126,
+      "learning_rate": 6.949544343187375e-05,
+      "loss": 2.076,
+      "step": 29367
+    },
+    {
+      "epoch": 0.6526222222222222,
+      "grad_norm": 1.5399001836776733,
+      "learning_rate": 6.949099799955546e-05,
+      "loss": 1.9729,
+      "step": 29368
+    },
+    {
+      "epoch": 0.6526444444444445,
+      "grad_norm": 1.6583781242370605,
+      "learning_rate": 6.948655256723716e-05,
+      "loss": 1.9131,
+      "step": 29369
+    },
+    {
+      "epoch": 0.6526666666666666,
+      "grad_norm": 1.5324383974075317,
+      "learning_rate": 6.948210713491887e-05,
+      "loss": 1.7619,
+      "step": 29370
+    },
+    {
+      "epoch": 0.6526888888888889,
+      "grad_norm": 1.7012816667556763,
+      "learning_rate": 6.947766170260059e-05,
+      "loss": 2.1841,
+      "step": 29371
+    },
+    {
+      "epoch": 0.6527111111111111,
+      "grad_norm": 1.626502275466919,
+      "learning_rate": 6.94732162702823e-05,
+      "loss": 1.9042,
+      "step": 29372
+    },
+    {
+      "epoch": 0.6527333333333334,
+      "grad_norm": 1.9852179288864136,
+      "learning_rate": 6.9468770837964e-05,
+      "loss": 1.9659,
+      "step": 29373
+    },
+    {
+      "epoch": 0.6527555555555555,
+      "grad_norm": 1.7495781183242798,
+      "learning_rate": 6.94643254056457e-05,
+      "loss": 2.356,
+      "step": 29374
+    },
+    {
+      "epoch": 0.6527777777777778,
+      "grad_norm": 1.5500662326812744,
+      "learning_rate": 6.945987997332741e-05,
+      "loss": 1.9053,
+      "step": 29375
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 1.7778003215789795,
+      "learning_rate": 6.945543454100911e-05,
+      "loss": 1.7993,
+      "step": 29376
+    },
+    {
+      "epoch": 0.6528222222222222,
+      "grad_norm": 1.6801308393478394,
+      "learning_rate": 6.945098910869082e-05,
+      "loss": 1.7258,
+      "step": 29377
+    },
+    {
+      "epoch": 0.6528444444444444,
+      "grad_norm": 1.6005624532699585,
+      "learning_rate": 6.944654367637253e-05,
+      "loss": 1.1401,
+      "step": 29378
+    },
+    {
+      "epoch": 0.6528666666666667,
+      "grad_norm": 2.02657151222229,
+      "learning_rate": 6.944209824405423e-05,
+      "loss": 1.6297,
+      "step": 29379
+    },
+    {
+      "epoch": 0.6528888888888889,
+      "grad_norm": 1.8680578470230103,
+      "learning_rate": 6.943765281173595e-05,
+      "loss": 2.0188,
+      "step": 29380
+    },
+    {
+      "epoch": 0.6529111111111111,
+      "grad_norm": 1.6274778842926025,
+      "learning_rate": 6.943320737941766e-05,
+      "loss": 1.6062,
+      "step": 29381
+    },
+    {
+      "epoch": 0.6529333333333334,
+      "grad_norm": 1.5393065214157104,
+      "learning_rate": 6.942876194709937e-05,
+      "loss": 0.8731,
+      "step": 29382
+    },
+    {
+      "epoch": 0.6529555555555555,
+      "grad_norm": 1.7478028535842896,
+      "learning_rate": 6.942431651478106e-05,
+      "loss": 2.1741,
+      "step": 29383
+    },
+    {
+      "epoch": 0.6529777777777778,
+      "grad_norm": 1.7424736022949219,
+      "learning_rate": 6.941987108246277e-05,
+      "loss": 2.0114,
+      "step": 29384
+    },
+    {
+      "epoch": 0.653,
+      "grad_norm": 1.6595288515090942,
+      "learning_rate": 6.941542565014448e-05,
+      "loss": 1.7124,
+      "step": 29385
+    },
+    {
+      "epoch": 0.6530222222222222,
+      "grad_norm": 1.862471103668213,
+      "learning_rate": 6.941098021782618e-05,
+      "loss": 1.3923,
+      "step": 29386
+    },
+    {
+      "epoch": 0.6530444444444444,
+      "grad_norm": 1.7211300134658813,
+      "learning_rate": 6.940653478550789e-05,
+      "loss": 2.1222,
+      "step": 29387
+    },
+    {
+      "epoch": 0.6530666666666667,
+      "grad_norm": 1.7551764249801636,
+      "learning_rate": 6.94020893531896e-05,
+      "loss": 2.3051,
+      "step": 29388
+    },
+    {
+      "epoch": 0.6530888888888889,
+      "grad_norm": 2.34627628326416,
+      "learning_rate": 6.939764392087131e-05,
+      "loss": 2.3686,
+      "step": 29389
+    },
+    {
+      "epoch": 0.6531111111111111,
+      "grad_norm": 1.6594637632369995,
+      "learning_rate": 6.939319848855302e-05,
+      "loss": 2.0798,
+      "step": 29390
+    },
+    {
+      "epoch": 0.6531333333333333,
+      "grad_norm": 1.7297558784484863,
+      "learning_rate": 6.938875305623473e-05,
+      "loss": 1.7862,
+      "step": 29391
+    },
+    {
+      "epoch": 0.6531555555555556,
+      "grad_norm": 1.5960259437561035,
+      "learning_rate": 6.938430762391644e-05,
+      "loss": 1.7059,
+      "step": 29392
+    },
+    {
+      "epoch": 0.6531777777777777,
+      "grad_norm": 1.6867831945419312,
+      "learning_rate": 6.937986219159813e-05,
+      "loss": 1.8962,
+      "step": 29393
+    },
+    {
+      "epoch": 0.6532,
+      "grad_norm": 1.8363149166107178,
+      "learning_rate": 6.937541675927984e-05,
+      "loss": 1.6911,
+      "step": 29394
+    },
+    {
+      "epoch": 0.6532222222222223,
+      "grad_norm": 1.8267245292663574,
+      "learning_rate": 6.937097132696155e-05,
+      "loss": 1.9225,
+      "step": 29395
+    },
+    {
+      "epoch": 0.6532444444444444,
+      "grad_norm": 1.8893781900405884,
+      "learning_rate": 6.936652589464325e-05,
+      "loss": 1.8525,
+      "step": 29396
+    },
+    {
+      "epoch": 0.6532666666666667,
+      "grad_norm": 1.3311126232147217,
+      "learning_rate": 6.936208046232496e-05,
+      "loss": 1.0012,
+      "step": 29397
+    },
+    {
+      "epoch": 0.6532888888888889,
+      "grad_norm": 1.7276990413665771,
+      "learning_rate": 6.935763503000667e-05,
+      "loss": 1.8642,
+      "step": 29398
+    },
+    {
+      "epoch": 0.6533111111111111,
+      "grad_norm": 1.8954962491989136,
+      "learning_rate": 6.935318959768838e-05,
+      "loss": 1.9947,
+      "step": 29399
+    },
+    {
+      "epoch": 0.6533333333333333,
+      "grad_norm": 2.5243430137634277,
+      "learning_rate": 6.934874416537009e-05,
+      "loss": 1.7075,
+      "step": 29400
+    },
+    {
+      "epoch": 0.6533555555555556,
+      "grad_norm": 1.5330208539962769,
+      "learning_rate": 6.93442987330518e-05,
+      "loss": 2.5989,
+      "step": 29401
+    },
+    {
+      "epoch": 0.6533777777777777,
+      "grad_norm": 1.3771998882293701,
+      "learning_rate": 6.93398533007335e-05,
+      "loss": 2.0617,
+      "step": 29402
+    },
+    {
+      "epoch": 0.6534,
+      "grad_norm": 1.2721847295761108,
+      "learning_rate": 6.93354078684152e-05,
+      "loss": 1.001,
+      "step": 29403
+    },
+    {
+      "epoch": 0.6534222222222222,
+      "grad_norm": 1.885785698890686,
+      "learning_rate": 6.933096243609691e-05,
+      "loss": 2.2274,
+      "step": 29404
+    },
+    {
+      "epoch": 0.6534444444444445,
+      "grad_norm": 1.6090855598449707,
+      "learning_rate": 6.932651700377862e-05,
+      "loss": 2.017,
+      "step": 29405
+    },
+    {
+      "epoch": 0.6534666666666666,
+      "grad_norm": 1.3860564231872559,
+      "learning_rate": 6.932207157146032e-05,
+      "loss": 2.1833,
+      "step": 29406
+    },
+    {
+      "epoch": 0.6534888888888889,
+      "grad_norm": 1.5746746063232422,
+      "learning_rate": 6.931762613914203e-05,
+      "loss": 1.9449,
+      "step": 29407
+    },
+    {
+      "epoch": 0.6535111111111112,
+      "grad_norm": 1.605556607246399,
+      "learning_rate": 6.931318070682375e-05,
+      "loss": 2.3436,
+      "step": 29408
+    },
+    {
+      "epoch": 0.6535333333333333,
+      "grad_norm": 1.6281412839889526,
+      "learning_rate": 6.930873527450546e-05,
+      "loss": 1.9975,
+      "step": 29409
+    },
+    {
+      "epoch": 0.6535555555555556,
+      "grad_norm": 1.5633022785186768,
+      "learning_rate": 6.930428984218716e-05,
+      "loss": 1.875,
+      "step": 29410
+    },
+    {
+      "epoch": 0.6535777777777778,
+      "grad_norm": 1.6272820234298706,
+      "learning_rate": 6.929984440986887e-05,
+      "loss": 2.2241,
+      "step": 29411
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 1.4241611957550049,
+      "learning_rate": 6.929539897755058e-05,
+      "loss": 1.9127,
+      "step": 29412
+    },
+    {
+      "epoch": 0.6536222222222222,
+      "grad_norm": 1.8110153675079346,
+      "learning_rate": 6.929095354523227e-05,
+      "loss": 2.2737,
+      "step": 29413
+    },
+    {
+      "epoch": 0.6536444444444445,
+      "grad_norm": 1.5204417705535889,
+      "learning_rate": 6.928650811291398e-05,
+      "loss": 1.8794,
+      "step": 29414
+    },
+    {
+      "epoch": 0.6536666666666666,
+      "grad_norm": 1.456246256828308,
+      "learning_rate": 6.928206268059569e-05,
+      "loss": 2.1589,
+      "step": 29415
+    },
+    {
+      "epoch": 0.6536888888888889,
+      "grad_norm": 1.581737756729126,
+      "learning_rate": 6.927761724827739e-05,
+      "loss": 2.0204,
+      "step": 29416
+    },
+    {
+      "epoch": 0.6537111111111111,
+      "grad_norm": 1.9131207466125488,
+      "learning_rate": 6.927317181595911e-05,
+      "loss": 2.2468,
+      "step": 29417
+    },
+    {
+      "epoch": 0.6537333333333334,
+      "grad_norm": 1.531806468963623,
+      "learning_rate": 6.926872638364082e-05,
+      "loss": 1.9102,
+      "step": 29418
+    },
+    {
+      "epoch": 0.6537555555555555,
+      "grad_norm": 1.7774204015731812,
+      "learning_rate": 6.926428095132253e-05,
+      "loss": 2.2159,
+      "step": 29419
+    },
+    {
+      "epoch": 0.6537777777777778,
+      "grad_norm": 1.632561206817627,
+      "learning_rate": 6.925983551900423e-05,
+      "loss": 1.7789,
+      "step": 29420
+    },
+    {
+      "epoch": 0.6538,
+      "grad_norm": 1.4562997817993164,
+      "learning_rate": 6.925539008668593e-05,
+      "loss": 1.7284,
+      "step": 29421
+    },
+    {
+      "epoch": 0.6538222222222222,
+      "grad_norm": 1.8392057418823242,
+      "learning_rate": 6.925094465436764e-05,
+      "loss": 1.7259,
+      "step": 29422
+    },
+    {
+      "epoch": 0.6538444444444445,
+      "grad_norm": 1.8033900260925293,
+      "learning_rate": 6.924649922204934e-05,
+      "loss": 2.0409,
+      "step": 29423
+    },
+    {
+      "epoch": 0.6538666666666667,
+      "grad_norm": 1.638187289237976,
+      "learning_rate": 6.924205378973105e-05,
+      "loss": 1.9508,
+      "step": 29424
+    },
+    {
+      "epoch": 0.6538888888888889,
+      "grad_norm": 2.0364911556243896,
+      "learning_rate": 6.923760835741276e-05,
+      "loss": 2.0946,
+      "step": 29425
+    },
+    {
+      "epoch": 0.6539111111111111,
+      "grad_norm": 1.5568419694900513,
+      "learning_rate": 6.923316292509447e-05,
+      "loss": 1.9651,
+      "step": 29426
+    },
+    {
+      "epoch": 0.6539333333333334,
+      "grad_norm": 1.6838810443878174,
+      "learning_rate": 6.922871749277618e-05,
+      "loss": 2.1729,
+      "step": 29427
+    },
+    {
+      "epoch": 0.6539555555555555,
+      "grad_norm": 1.4847619533538818,
+      "learning_rate": 6.922427206045789e-05,
+      "loss": 1.8529,
+      "step": 29428
+    },
+    {
+      "epoch": 0.6539777777777778,
+      "grad_norm": 1.2320556640625,
+      "learning_rate": 6.92198266281396e-05,
+      "loss": 0.7937,
+      "step": 29429
+    },
+    {
+      "epoch": 0.654,
+      "grad_norm": 1.8546630144119263,
+      "learning_rate": 6.92153811958213e-05,
+      "loss": 1.6273,
+      "step": 29430
+    },
+    {
+      "epoch": 0.6540222222222222,
+      "grad_norm": 1.7034964561462402,
+      "learning_rate": 6.9210935763503e-05,
+      "loss": 1.9266,
+      "step": 29431
+    },
+    {
+      "epoch": 0.6540444444444444,
+      "grad_norm": 1.6947805881500244,
+      "learning_rate": 6.920649033118471e-05,
+      "loss": 1.9812,
+      "step": 29432
+    },
+    {
+      "epoch": 0.6540666666666667,
+      "grad_norm": 1.985138177871704,
+      "learning_rate": 6.920204489886641e-05,
+      "loss": 2.1235,
+      "step": 29433
+    },
+    {
+      "epoch": 0.6540888888888889,
+      "grad_norm": 1.9874866008758545,
+      "learning_rate": 6.919759946654812e-05,
+      "loss": 1.7606,
+      "step": 29434
+    },
+    {
+      "epoch": 0.6541111111111111,
+      "grad_norm": 1.6188753843307495,
+      "learning_rate": 6.919315403422983e-05,
+      "loss": 1.6674,
+      "step": 29435
+    },
+    {
+      "epoch": 0.6541333333333333,
+      "grad_norm": 1.4949100017547607,
+      "learning_rate": 6.918870860191154e-05,
+      "loss": 1.8744,
+      "step": 29436
+    },
+    {
+      "epoch": 0.6541555555555556,
+      "grad_norm": 1.6058498620986938,
+      "learning_rate": 6.918426316959325e-05,
+      "loss": 1.6328,
+      "step": 29437
+    },
+    {
+      "epoch": 0.6541777777777777,
+      "grad_norm": 1.6887714862823486,
+      "learning_rate": 6.917981773727496e-05,
+      "loss": 1.6641,
+      "step": 29438
+    },
+    {
+      "epoch": 0.6542,
+      "grad_norm": 1.9345225095748901,
+      "learning_rate": 6.917537230495667e-05,
+      "loss": 1.7995,
+      "step": 29439
+    },
+    {
+      "epoch": 0.6542222222222223,
+      "grad_norm": 1.9309325218200684,
+      "learning_rate": 6.917092687263836e-05,
+      "loss": 1.731,
+      "step": 29440
+    },
+    {
+      "epoch": 0.6542444444444444,
+      "grad_norm": 1.5484614372253418,
+      "learning_rate": 6.916648144032007e-05,
+      "loss": 1.8296,
+      "step": 29441
+    },
+    {
+      "epoch": 0.6542666666666667,
+      "grad_norm": 1.6554497480392456,
+      "learning_rate": 6.916203600800178e-05,
+      "loss": 1.7379,
+      "step": 29442
+    },
+    {
+      "epoch": 0.6542888888888889,
+      "grad_norm": 2.0249950885772705,
+      "learning_rate": 6.915759057568348e-05,
+      "loss": 1.827,
+      "step": 29443
+    },
+    {
+      "epoch": 0.6543111111111111,
+      "grad_norm": 1.835463523864746,
+      "learning_rate": 6.915314514336519e-05,
+      "loss": 1.9925,
+      "step": 29444
+    },
+    {
+      "epoch": 0.6543333333333333,
+      "grad_norm": 2.0246810913085938,
+      "learning_rate": 6.914869971104691e-05,
+      "loss": 2.0876,
+      "step": 29445
+    },
+    {
+      "epoch": 0.6543555555555556,
+      "grad_norm": 2.1907918453216553,
+      "learning_rate": 6.914425427872862e-05,
+      "loss": 1.8498,
+      "step": 29446
+    },
+    {
+      "epoch": 0.6543777777777777,
+      "grad_norm": 2.0203280448913574,
+      "learning_rate": 6.913980884641032e-05,
+      "loss": 1.9762,
+      "step": 29447
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 2.1965038776397705,
+      "learning_rate": 6.913536341409203e-05,
+      "loss": 2.0273,
+      "step": 29448
+    },
+    {
+      "epoch": 0.6544222222222222,
+      "grad_norm": 2.1303956508636475,
+      "learning_rate": 6.913091798177374e-05,
+      "loss": 1.9217,
+      "step": 29449
+    },
+    {
+      "epoch": 0.6544444444444445,
+      "grad_norm": 1.9999853372573853,
+      "learning_rate": 6.912647254945543e-05,
+      "loss": 1.485,
+      "step": 29450
+    },
+    {
+      "epoch": 0.6544666666666666,
+      "grad_norm": 1.4747973680496216,
+      "learning_rate": 6.912202711713714e-05,
+      "loss": 2.2189,
+      "step": 29451
+    },
+    {
+      "epoch": 0.6544888888888889,
+      "grad_norm": 1.6959315538406372,
+      "learning_rate": 6.911758168481885e-05,
+      "loss": 2.6387,
+      "step": 29452
+    },
+    {
+      "epoch": 0.6545111111111112,
+      "grad_norm": 1.709522008895874,
+      "learning_rate": 6.911313625250055e-05,
+      "loss": 2.4668,
+      "step": 29453
+    },
+    {
+      "epoch": 0.6545333333333333,
+      "grad_norm": 1.0732325315475464,
+      "learning_rate": 6.910869082018227e-05,
+      "loss": 1.189,
+      "step": 29454
+    },
+    {
+      "epoch": 0.6545555555555556,
+      "grad_norm": 1.5143368244171143,
+      "learning_rate": 6.910424538786398e-05,
+      "loss": 2.1954,
+      "step": 29455
+    },
+    {
+      "epoch": 0.6545777777777778,
+      "grad_norm": 1.6703866720199585,
+      "learning_rate": 6.909979995554569e-05,
+      "loss": 2.5903,
+      "step": 29456
+    },
+    {
+      "epoch": 0.6546,
+      "grad_norm": 1.806408166885376,
+      "learning_rate": 6.909535452322739e-05,
+      "loss": 2.0729,
+      "step": 29457
+    },
+    {
+      "epoch": 0.6546222222222222,
+      "grad_norm": 1.9158971309661865,
+      "learning_rate": 6.90909090909091e-05,
+      "loss": 2.732,
+      "step": 29458
+    },
+    {
+      "epoch": 0.6546444444444445,
+      "grad_norm": 1.7235842943191528,
+      "learning_rate": 6.90864636585908e-05,
+      "loss": 1.8045,
+      "step": 29459
+    },
+    {
+      "epoch": 0.6546666666666666,
+      "grad_norm": 2.2138214111328125,
+      "learning_rate": 6.90820182262725e-05,
+      "loss": 2.4499,
+      "step": 29460
+    },
+    {
+      "epoch": 0.6546888888888889,
+      "grad_norm": 1.5769482851028442,
+      "learning_rate": 6.907757279395421e-05,
+      "loss": 1.8384,
+      "step": 29461
+    },
+    {
+      "epoch": 0.6547111111111111,
+      "grad_norm": 1.7914637327194214,
+      "learning_rate": 6.907312736163592e-05,
+      "loss": 2.0621,
+      "step": 29462
+    },
+    {
+      "epoch": 0.6547333333333333,
+      "grad_norm": 1.412074327468872,
+      "learning_rate": 6.906868192931763e-05,
+      "loss": 1.6855,
+      "step": 29463
+    },
+    {
+      "epoch": 0.6547555555555555,
+      "grad_norm": 1.8703691959381104,
+      "learning_rate": 6.906423649699934e-05,
+      "loss": 2.2053,
+      "step": 29464
+    },
+    {
+      "epoch": 0.6547777777777778,
+      "grad_norm": 1.5958142280578613,
+      "learning_rate": 6.905979106468105e-05,
+      "loss": 2.143,
+      "step": 29465
+    },
+    {
+      "epoch": 0.6548,
+      "grad_norm": 1.4695812463760376,
+      "learning_rate": 6.905534563236276e-05,
+      "loss": 1.7765,
+      "step": 29466
+    },
+    {
+      "epoch": 0.6548222222222222,
+      "grad_norm": 1.472147822380066,
+      "learning_rate": 6.905090020004446e-05,
+      "loss": 1.1977,
+      "step": 29467
+    },
+    {
+      "epoch": 0.6548444444444445,
+      "grad_norm": 1.6705714464187622,
+      "learning_rate": 6.904645476772617e-05,
+      "loss": 2.2475,
+      "step": 29468
+    },
+    {
+      "epoch": 0.6548666666666667,
+      "grad_norm": 1.4583940505981445,
+      "learning_rate": 6.904200933540787e-05,
+      "loss": 1.7501,
+      "step": 29469
+    },
+    {
+      "epoch": 0.6548888888888889,
+      "grad_norm": 1.5646191835403442,
+      "learning_rate": 6.903756390308957e-05,
+      "loss": 1.6958,
+      "step": 29470
+    },
+    {
+      "epoch": 0.6549111111111111,
+      "grad_norm": 1.6197688579559326,
+      "learning_rate": 6.903311847077128e-05,
+      "loss": 2.0638,
+      "step": 29471
+    },
+    {
+      "epoch": 0.6549333333333334,
+      "grad_norm": 1.7141706943511963,
+      "learning_rate": 6.902867303845299e-05,
+      "loss": 1.8158,
+      "step": 29472
+    },
+    {
+      "epoch": 0.6549555555555555,
+      "grad_norm": 1.6812819242477417,
+      "learning_rate": 6.90242276061347e-05,
+      "loss": 2.1096,
+      "step": 29473
+    },
+    {
+      "epoch": 0.6549777777777778,
+      "grad_norm": 1.5177154541015625,
+      "learning_rate": 6.901978217381641e-05,
+      "loss": 1.5467,
+      "step": 29474
+    },
+    {
+      "epoch": 0.655,
+      "grad_norm": 1.914524793624878,
+      "learning_rate": 6.901533674149812e-05,
+      "loss": 2.4122,
+      "step": 29475
+    },
+    {
+      "epoch": 0.6550222222222222,
+      "grad_norm": 1.8134791851043701,
+      "learning_rate": 6.901089130917983e-05,
+      "loss": 1.9971,
+      "step": 29476
+    },
+    {
+      "epoch": 0.6550444444444444,
+      "grad_norm": 2.0190398693084717,
+      "learning_rate": 6.900644587686152e-05,
+      "loss": 2.315,
+      "step": 29477
+    },
+    {
+      "epoch": 0.6550666666666667,
+      "grad_norm": 1.5984684228897095,
+      "learning_rate": 6.900200044454323e-05,
+      "loss": 2.2506,
+      "step": 29478
+    },
+    {
+      "epoch": 0.6550888888888889,
+      "grad_norm": 1.6222925186157227,
+      "learning_rate": 6.899755501222494e-05,
+      "loss": 2.2784,
+      "step": 29479
+    },
+    {
+      "epoch": 0.6551111111111111,
+      "grad_norm": 1.7684059143066406,
+      "learning_rate": 6.899310957990664e-05,
+      "loss": 1.4624,
+      "step": 29480
+    },
+    {
+      "epoch": 0.6551333333333333,
+      "grad_norm": 1.797141671180725,
+      "learning_rate": 6.898866414758835e-05,
+      "loss": 1.712,
+      "step": 29481
+    },
+    {
+      "epoch": 0.6551555555555556,
+      "grad_norm": 1.7618950605392456,
+      "learning_rate": 6.898421871527007e-05,
+      "loss": 1.8547,
+      "step": 29482
+    },
+    {
+      "epoch": 0.6551777777777777,
+      "grad_norm": 1.7545932531356812,
+      "learning_rate": 6.897977328295177e-05,
+      "loss": 2.2939,
+      "step": 29483
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 2.0590357780456543,
+      "learning_rate": 6.897532785063348e-05,
+      "loss": 2.5174,
+      "step": 29484
+    },
+    {
+      "epoch": 0.6552222222222223,
+      "grad_norm": 1.5144672393798828,
+      "learning_rate": 6.897088241831519e-05,
+      "loss": 1.5754,
+      "step": 29485
+    },
+    {
+      "epoch": 0.6552444444444444,
+      "grad_norm": 0.1713659167289734,
+      "learning_rate": 6.89664369859969e-05,
+      "loss": 0.0272,
+      "step": 29486
+    },
+    {
+      "epoch": 0.6552666666666667,
+      "grad_norm": 1.5620533227920532,
+      "learning_rate": 6.89619915536786e-05,
+      "loss": 1.5878,
+      "step": 29487
+    },
+    {
+      "epoch": 0.6552888888888889,
+      "grad_norm": 1.8646880388259888,
+      "learning_rate": 6.89575461213603e-05,
+      "loss": 1.9633,
+      "step": 29488
+    },
+    {
+      "epoch": 0.6553111111111111,
+      "grad_norm": 1.9776960611343384,
+      "learning_rate": 6.895310068904201e-05,
+      "loss": 2.3859,
+      "step": 29489
+    },
+    {
+      "epoch": 0.6553333333333333,
+      "grad_norm": 2.0050442218780518,
+      "learning_rate": 6.894865525672371e-05,
+      "loss": 1.9738,
+      "step": 29490
+    },
+    {
+      "epoch": 0.6553555555555556,
+      "grad_norm": 1.5091774463653564,
+      "learning_rate": 6.894420982440543e-05,
+      "loss": 1.4317,
+      "step": 29491
+    },
+    {
+      "epoch": 0.6553777777777777,
+      "grad_norm": 1.7000620365142822,
+      "learning_rate": 6.893976439208714e-05,
+      "loss": 2.1876,
+      "step": 29492
+    },
+    {
+      "epoch": 0.6554,
+      "grad_norm": 1.7755290269851685,
+      "learning_rate": 6.893531895976884e-05,
+      "loss": 1.7824,
+      "step": 29493
+    },
+    {
+      "epoch": 0.6554222222222222,
+      "grad_norm": 1.6783421039581299,
+      "learning_rate": 6.893087352745055e-05,
+      "loss": 1.6337,
+      "step": 29494
+    },
+    {
+      "epoch": 0.6554444444444445,
+      "grad_norm": 2.275012254714966,
+      "learning_rate": 6.892642809513226e-05,
+      "loss": 1.7613,
+      "step": 29495
+    },
+    {
+      "epoch": 0.6554666666666666,
+      "grad_norm": 1.8266583681106567,
+      "learning_rate": 6.892198266281397e-05,
+      "loss": 2.0478,
+      "step": 29496
+    },
+    {
+      "epoch": 0.6554888888888889,
+      "grad_norm": 1.857939600944519,
+      "learning_rate": 6.891753723049566e-05,
+      "loss": 1.874,
+      "step": 29497
+    },
+    {
+      "epoch": 0.6555111111111112,
+      "grad_norm": 2.2057530879974365,
+      "learning_rate": 6.891309179817737e-05,
+      "loss": 2.0207,
+      "step": 29498
+    },
+    {
+      "epoch": 0.6555333333333333,
+      "grad_norm": 2.1313059329986572,
+      "learning_rate": 6.890864636585908e-05,
+      "loss": 2.0267,
+      "step": 29499
+    },
+    {
+      "epoch": 0.6555555555555556,
+      "grad_norm": 1.5375425815582275,
+      "learning_rate": 6.890420093354079e-05,
+      "loss": 1.0636,
+      "step": 29500
+    },
+    {
+      "epoch": 0.6555777777777778,
+      "grad_norm": 1.5120632648468018,
+      "learning_rate": 6.88997555012225e-05,
+      "loss": 1.5956,
+      "step": 29501
+    },
+    {
+      "epoch": 0.6556,
+      "grad_norm": 1.6631743907928467,
+      "learning_rate": 6.889531006890421e-05,
+      "loss": 2.4788,
+      "step": 29502
+    },
+    {
+      "epoch": 0.6556222222222222,
+      "grad_norm": 1.5857105255126953,
+      "learning_rate": 6.889086463658592e-05,
+      "loss": 2.9431,
+      "step": 29503
+    },
+    {
+      "epoch": 0.6556444444444445,
+      "grad_norm": 1.843294382095337,
+      "learning_rate": 6.888641920426762e-05,
+      "loss": 1.8507,
+      "step": 29504
+    },
+    {
+      "epoch": 0.6556666666666666,
+      "grad_norm": 2.2526984214782715,
+      "learning_rate": 6.888197377194933e-05,
+      "loss": 2.0335,
+      "step": 29505
+    },
+    {
+      "epoch": 0.6556888888888889,
+      "grad_norm": 1.6078662872314453,
+      "learning_rate": 6.887752833963104e-05,
+      "loss": 2.5839,
+      "step": 29506
+    },
+    {
+      "epoch": 0.6557111111111111,
+      "grad_norm": 1.6662286520004272,
+      "learning_rate": 6.887308290731273e-05,
+      "loss": 2.2819,
+      "step": 29507
+    },
+    {
+      "epoch": 0.6557333333333333,
+      "grad_norm": 1.4887229204177856,
+      "learning_rate": 6.886863747499444e-05,
+      "loss": 2.1366,
+      "step": 29508
+    },
+    {
+      "epoch": 0.6557555555555555,
+      "grad_norm": 1.8119449615478516,
+      "learning_rate": 6.886419204267615e-05,
+      "loss": 2.6567,
+      "step": 29509
+    },
+    {
+      "epoch": 0.6557777777777778,
+      "grad_norm": 1.9445807933807373,
+      "learning_rate": 6.885974661035786e-05,
+      "loss": 2.3424,
+      "step": 29510
+    },
+    {
+      "epoch": 0.6558,
+      "grad_norm": 1.461766004562378,
+      "learning_rate": 6.885530117803957e-05,
+      "loss": 1.5104,
+      "step": 29511
+    },
+    {
+      "epoch": 0.6558222222222222,
+      "grad_norm": 1.448891520500183,
+      "learning_rate": 6.885085574572128e-05,
+      "loss": 1.9026,
+      "step": 29512
+    },
+    {
+      "epoch": 0.6558444444444445,
+      "grad_norm": 1.5717157125473022,
+      "learning_rate": 6.884641031340299e-05,
+      "loss": 2.3704,
+      "step": 29513
+    },
+    {
+      "epoch": 0.6558666666666667,
+      "grad_norm": 1.8957597017288208,
+      "learning_rate": 6.884196488108469e-05,
+      "loss": 2.3059,
+      "step": 29514
+    },
+    {
+      "epoch": 0.6558888888888889,
+      "grad_norm": 1.5091511011123657,
+      "learning_rate": 6.88375194487664e-05,
+      "loss": 2.1742,
+      "step": 29515
+    },
+    {
+      "epoch": 0.6559111111111111,
+      "grad_norm": 1.6314390897750854,
+      "learning_rate": 6.88330740164481e-05,
+      "loss": 1.9497,
+      "step": 29516
+    },
+    {
+      "epoch": 0.6559333333333334,
+      "grad_norm": 1.682511568069458,
+      "learning_rate": 6.88286285841298e-05,
+      "loss": 2.5928,
+      "step": 29517
+    },
+    {
+      "epoch": 0.6559555555555555,
+      "grad_norm": 1.6695607900619507,
+      "learning_rate": 6.882418315181151e-05,
+      "loss": 2.4622,
+      "step": 29518
+    },
+    {
+      "epoch": 0.6559777777777778,
+      "grad_norm": 2.0376710891723633,
+      "learning_rate": 6.881973771949323e-05,
+      "loss": 2.1434,
+      "step": 29519
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 1.7357299327850342,
+      "learning_rate": 6.881529228717493e-05,
+      "loss": 1.02,
+      "step": 29520
+    },
+    {
+      "epoch": 0.6560222222222222,
+      "grad_norm": 1.7393115758895874,
+      "learning_rate": 6.881084685485664e-05,
+      "loss": 2.2984,
+      "step": 29521
+    },
+    {
+      "epoch": 0.6560444444444444,
+      "grad_norm": 1.6580684185028076,
+      "learning_rate": 6.880640142253835e-05,
+      "loss": 2.1213,
+      "step": 29522
+    },
+    {
+      "epoch": 0.6560666666666667,
+      "grad_norm": 1.7607481479644775,
+      "learning_rate": 6.880195599022006e-05,
+      "loss": 1.9981,
+      "step": 29523
+    },
+    {
+      "epoch": 0.6560888888888889,
+      "grad_norm": 1.508754014968872,
+      "learning_rate": 6.879751055790175e-05,
+      "loss": 1.4985,
+      "step": 29524
+    },
+    {
+      "epoch": 0.6561111111111111,
+      "grad_norm": 1.6106895208358765,
+      "learning_rate": 6.879306512558346e-05,
+      "loss": 2.0425,
+      "step": 29525
+    },
+    {
+      "epoch": 0.6561333333333333,
+      "grad_norm": 1.9137887954711914,
+      "learning_rate": 6.878861969326517e-05,
+      "loss": 1.9829,
+      "step": 29526
+    },
+    {
+      "epoch": 0.6561555555555556,
+      "grad_norm": 1.5618802309036255,
+      "learning_rate": 6.878417426094687e-05,
+      "loss": 1.7582,
+      "step": 29527
+    },
+    {
+      "epoch": 0.6561777777777777,
+      "grad_norm": 1.4684094190597534,
+      "learning_rate": 6.877972882862859e-05,
+      "loss": 1.7052,
+      "step": 29528
+    },
+    {
+      "epoch": 0.6562,
+      "grad_norm": 1.937695026397705,
+      "learning_rate": 6.87752833963103e-05,
+      "loss": 2.0951,
+      "step": 29529
+    },
+    {
+      "epoch": 0.6562222222222223,
+      "grad_norm": 1.691579818725586,
+      "learning_rate": 6.8770837963992e-05,
+      "loss": 1.9227,
+      "step": 29530
+    },
+    {
+      "epoch": 0.6562444444444444,
+      "grad_norm": 1.5953030586242676,
+      "learning_rate": 6.876639253167371e-05,
+      "loss": 1.5954,
+      "step": 29531
+    },
+    {
+      "epoch": 0.6562666666666667,
+      "grad_norm": 1.897382378578186,
+      "learning_rate": 6.876194709935542e-05,
+      "loss": 2.3019,
+      "step": 29532
+    },
+    {
+      "epoch": 0.6562888888888889,
+      "grad_norm": 1.8537498712539673,
+      "learning_rate": 6.875750166703713e-05,
+      "loss": 2.1086,
+      "step": 29533
+    },
+    {
+      "epoch": 0.6563111111111111,
+      "grad_norm": 1.8957946300506592,
+      "learning_rate": 6.875305623471882e-05,
+      "loss": 1.8338,
+      "step": 29534
+    },
+    {
+      "epoch": 0.6563333333333333,
+      "grad_norm": 1.5951693058013916,
+      "learning_rate": 6.874861080240053e-05,
+      "loss": 1.7943,
+      "step": 29535
+    },
+    {
+      "epoch": 0.6563555555555556,
+      "grad_norm": 1.6607433557510376,
+      "learning_rate": 6.874416537008224e-05,
+      "loss": 1.9024,
+      "step": 29536
+    },
+    {
+      "epoch": 0.6563777777777777,
+      "grad_norm": 1.5129278898239136,
+      "learning_rate": 6.873971993776395e-05,
+      "loss": 1.636,
+      "step": 29537
+    },
+    {
+      "epoch": 0.6564,
+      "grad_norm": 1.4948451519012451,
+      "learning_rate": 6.873527450544566e-05,
+      "loss": 1.9594,
+      "step": 29538
+    },
+    {
+      "epoch": 0.6564222222222222,
+      "grad_norm": 2.0458195209503174,
+      "learning_rate": 6.873082907312737e-05,
+      "loss": 2.0937,
+      "step": 29539
+    },
+    {
+      "epoch": 0.6564444444444445,
+      "grad_norm": 1.569808840751648,
+      "learning_rate": 6.872638364080907e-05,
+      "loss": 1.3614,
+      "step": 29540
+    },
+    {
+      "epoch": 0.6564666666666666,
+      "grad_norm": 1.8471095561981201,
+      "learning_rate": 6.872193820849078e-05,
+      "loss": 1.6355,
+      "step": 29541
+    },
+    {
+      "epoch": 0.6564888888888889,
+      "grad_norm": 1.5188415050506592,
+      "learning_rate": 6.871749277617249e-05,
+      "loss": 1.2968,
+      "step": 29542
+    },
+    {
+      "epoch": 0.6565111111111112,
+      "grad_norm": 1.6712753772735596,
+      "learning_rate": 6.87130473438542e-05,
+      "loss": 1.5632,
+      "step": 29543
+    },
+    {
+      "epoch": 0.6565333333333333,
+      "grad_norm": 1.6549158096313477,
+      "learning_rate": 6.870860191153589e-05,
+      "loss": 2.0788,
+      "step": 29544
+    },
+    {
+      "epoch": 0.6565555555555556,
+      "grad_norm": 1.7481123208999634,
+      "learning_rate": 6.87041564792176e-05,
+      "loss": 1.4649,
+      "step": 29545
+    },
+    {
+      "epoch": 0.6565777777777778,
+      "grad_norm": 1.8728467226028442,
+      "learning_rate": 6.869971104689931e-05,
+      "loss": 1.9749,
+      "step": 29546
+    },
+    {
+      "epoch": 0.6566,
+      "grad_norm": 1.7548919916152954,
+      "learning_rate": 6.869526561458102e-05,
+      "loss": 2.0778,
+      "step": 29547
+    },
+    {
+      "epoch": 0.6566222222222222,
+      "grad_norm": 1.713470220565796,
+      "learning_rate": 6.869082018226273e-05,
+      "loss": 1.6491,
+      "step": 29548
+    },
+    {
+      "epoch": 0.6566444444444445,
+      "grad_norm": 1.8220151662826538,
+      "learning_rate": 6.868637474994444e-05,
+      "loss": 2.0425,
+      "step": 29549
+    },
+    {
+      "epoch": 0.6566666666666666,
+      "grad_norm": 1.7327014207839966,
+      "learning_rate": 6.868192931762614e-05,
+      "loss": 1.7185,
+      "step": 29550
+    },
+    {
+      "epoch": 0.6566888888888889,
+      "grad_norm": 0.9733756184577942,
+      "learning_rate": 6.867748388530785e-05,
+      "loss": 1.1545,
+      "step": 29551
+    },
+    {
+      "epoch": 0.6567111111111111,
+      "grad_norm": 1.092980146408081,
+      "learning_rate": 6.867303845298956e-05,
+      "loss": 1.2458,
+      "step": 29552
+    },
+    {
+      "epoch": 0.6567333333333333,
+      "grad_norm": 1.8089306354522705,
+      "learning_rate": 6.866859302067127e-05,
+      "loss": 1.9041,
+      "step": 29553
+    },
+    {
+      "epoch": 0.6567555555555555,
+      "grad_norm": 1.3253194093704224,
+      "learning_rate": 6.866414758835296e-05,
+      "loss": 1.9735,
+      "step": 29554
+    },
+    {
+      "epoch": 0.6567777777777778,
+      "grad_norm": 2.6289620399475098,
+      "learning_rate": 6.865970215603467e-05,
+      "loss": 1.8887,
+      "step": 29555
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 1.636061191558838,
+      "learning_rate": 6.86552567237164e-05,
+      "loss": 2.3801,
+      "step": 29556
+    },
+    {
+      "epoch": 0.6568222222222222,
+      "grad_norm": 1.466566801071167,
+      "learning_rate": 6.865081129139809e-05,
+      "loss": 1.0834,
+      "step": 29557
+    },
+    {
+      "epoch": 0.6568444444444445,
+      "grad_norm": 1.4488520622253418,
+      "learning_rate": 6.86463658590798e-05,
+      "loss": 1.9759,
+      "step": 29558
+    },
+    {
+      "epoch": 0.6568666666666667,
+      "grad_norm": 1.8230533599853516,
+      "learning_rate": 6.864192042676151e-05,
+      "loss": 2.4713,
+      "step": 29559
+    },
+    {
+      "epoch": 0.6568888888888889,
+      "grad_norm": 1.695978045463562,
+      "learning_rate": 6.863747499444322e-05,
+      "loss": 2.4114,
+      "step": 29560
+    },
+    {
+      "epoch": 0.6569111111111111,
+      "grad_norm": 1.4497877359390259,
+      "learning_rate": 6.863302956212492e-05,
+      "loss": 2.0624,
+      "step": 29561
+    },
+    {
+      "epoch": 0.6569333333333334,
+      "grad_norm": 1.780235767364502,
+      "learning_rate": 6.862858412980663e-05,
+      "loss": 1.9743,
+      "step": 29562
+    },
+    {
+      "epoch": 0.6569555555555555,
+      "grad_norm": 1.5658392906188965,
+      "learning_rate": 6.862413869748833e-05,
+      "loss": 1.8545,
+      "step": 29563
+    },
+    {
+      "epoch": 0.6569777777777778,
+      "grad_norm": 1.272002100944519,
+      "learning_rate": 6.861969326517003e-05,
+      "loss": 1.5825,
+      "step": 29564
+    },
+    {
+      "epoch": 0.657,
+      "grad_norm": 1.6310688257217407,
+      "learning_rate": 6.861524783285175e-05,
+      "loss": 2.2191,
+      "step": 29565
+    },
+    {
+      "epoch": 0.6570222222222222,
+      "grad_norm": 1.7843564748764038,
+      "learning_rate": 6.861080240053346e-05,
+      "loss": 2.2956,
+      "step": 29566
+    },
+    {
+      "epoch": 0.6570444444444444,
+      "grad_norm": 1.9026020765304565,
+      "learning_rate": 6.860635696821516e-05,
+      "loss": 2.1435,
+      "step": 29567
+    },
+    {
+      "epoch": 0.6570666666666667,
+      "grad_norm": 1.5643749237060547,
+      "learning_rate": 6.860191153589687e-05,
+      "loss": 2.0786,
+      "step": 29568
+    },
+    {
+      "epoch": 0.6570888888888888,
+      "grad_norm": 1.9233726263046265,
+      "learning_rate": 6.859746610357858e-05,
+      "loss": 1.6316,
+      "step": 29569
+    },
+    {
+      "epoch": 0.6571111111111111,
+      "grad_norm": 1.754144549369812,
+      "learning_rate": 6.859302067126029e-05,
+      "loss": 2.0501,
+      "step": 29570
+    },
+    {
+      "epoch": 0.6571333333333333,
+      "grad_norm": 1.6095532178878784,
+      "learning_rate": 6.858857523894198e-05,
+      "loss": 1.973,
+      "step": 29571
+    },
+    {
+      "epoch": 0.6571555555555556,
+      "grad_norm": 1.6992303133010864,
+      "learning_rate": 6.85841298066237e-05,
+      "loss": 1.8782,
+      "step": 29572
+    },
+    {
+      "epoch": 0.6571777777777777,
+      "grad_norm": 2.066359043121338,
+      "learning_rate": 6.85796843743054e-05,
+      "loss": 2.3603,
+      "step": 29573
+    },
+    {
+      "epoch": 0.6572,
+      "grad_norm": 1.9374815225601196,
+      "learning_rate": 6.857523894198711e-05,
+      "loss": 1.8272,
+      "step": 29574
+    },
+    {
+      "epoch": 0.6572222222222223,
+      "grad_norm": 1.6850558519363403,
+      "learning_rate": 6.857079350966882e-05,
+      "loss": 2.108,
+      "step": 29575
+    },
+    {
+      "epoch": 0.6572444444444444,
+      "grad_norm": 1.5528672933578491,
+      "learning_rate": 6.856634807735053e-05,
+      "loss": 1.6883,
+      "step": 29576
+    },
+    {
+      "epoch": 0.6572666666666667,
+      "grad_norm": 1.6508656740188599,
+      "learning_rate": 6.856190264503223e-05,
+      "loss": 1.9709,
+      "step": 29577
+    },
+    {
+      "epoch": 0.6572888888888889,
+      "grad_norm": 1.7889775037765503,
+      "learning_rate": 6.855745721271394e-05,
+      "loss": 1.8791,
+      "step": 29578
+    },
+    {
+      "epoch": 0.6573111111111111,
+      "grad_norm": 2.086968421936035,
+      "learning_rate": 6.855301178039565e-05,
+      "loss": 2.4607,
+      "step": 29579
+    },
+    {
+      "epoch": 0.6573333333333333,
+      "grad_norm": 1.2391284704208374,
+      "learning_rate": 6.854856634807736e-05,
+      "loss": 0.9328,
+      "step": 29580
+    },
+    {
+      "epoch": 0.6573555555555556,
+      "grad_norm": 1.818240761756897,
+      "learning_rate": 6.854412091575905e-05,
+      "loss": 1.5874,
+      "step": 29581
+    },
+    {
+      "epoch": 0.6573777777777777,
+      "grad_norm": 1.5828113555908203,
+      "learning_rate": 6.853967548344076e-05,
+      "loss": 1.6431,
+      "step": 29582
+    },
+    {
+      "epoch": 0.6574,
+      "grad_norm": 1.8962316513061523,
+      "learning_rate": 6.853523005112247e-05,
+      "loss": 2.2709,
+      "step": 29583
+    },
+    {
+      "epoch": 0.6574222222222222,
+      "grad_norm": 1.5729304552078247,
+      "learning_rate": 6.853078461880418e-05,
+      "loss": 1.5839,
+      "step": 29584
+    },
+    {
+      "epoch": 0.6574444444444445,
+      "grad_norm": 2.0226142406463623,
+      "learning_rate": 6.852633918648589e-05,
+      "loss": 2.0012,
+      "step": 29585
+    },
+    {
+      "epoch": 0.6574666666666666,
+      "grad_norm": 1.720937728881836,
+      "learning_rate": 6.85218937541676e-05,
+      "loss": 1.717,
+      "step": 29586
+    },
+    {
+      "epoch": 0.6574888888888889,
+      "grad_norm": 1.8267759084701538,
+      "learning_rate": 6.85174483218493e-05,
+      "loss": 1.6517,
+      "step": 29587
+    },
+    {
+      "epoch": 0.6575111111111112,
+      "grad_norm": 1.7748316526412964,
+      "learning_rate": 6.851300288953101e-05,
+      "loss": 1.7497,
+      "step": 29588
+    },
+    {
+      "epoch": 0.6575333333333333,
+      "grad_norm": 1.5970873832702637,
+      "learning_rate": 6.850855745721272e-05,
+      "loss": 1.9194,
+      "step": 29589
+    },
+    {
+      "epoch": 0.6575555555555556,
+      "grad_norm": 1.6664336919784546,
+      "learning_rate": 6.850411202489443e-05,
+      "loss": 1.8788,
+      "step": 29590
+    },
+    {
+      "epoch": 0.6575777777777778,
+      "grad_norm": 1.6667951345443726,
+      "learning_rate": 6.849966659257612e-05,
+      "loss": 1.8426,
+      "step": 29591
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 1.5941898822784424,
+      "learning_rate": 6.849522116025783e-05,
+      "loss": 1.5277,
+      "step": 29592
+    },
+    {
+      "epoch": 0.6576222222222222,
+      "grad_norm": 1.630222201347351,
+      "learning_rate": 6.849077572793956e-05,
+      "loss": 1.6931,
+      "step": 29593
+    },
+    {
+      "epoch": 0.6576444444444445,
+      "grad_norm": 1.4186203479766846,
+      "learning_rate": 6.848633029562125e-05,
+      "loss": 1.4202,
+      "step": 29594
+    },
+    {
+      "epoch": 0.6576666666666666,
+      "grad_norm": 1.4970699548721313,
+      "learning_rate": 6.848188486330296e-05,
+      "loss": 1.3966,
+      "step": 29595
+    },
+    {
+      "epoch": 0.6576888888888889,
+      "grad_norm": 1.7796052694320679,
+      "learning_rate": 6.847743943098467e-05,
+      "loss": 1.799,
+      "step": 29596
+    },
+    {
+      "epoch": 0.6577111111111111,
+      "grad_norm": 1.919205904006958,
+      "learning_rate": 6.847299399866637e-05,
+      "loss": 1.7903,
+      "step": 29597
+    },
+    {
+      "epoch": 0.6577333333333333,
+      "grad_norm": 1.979422926902771,
+      "learning_rate": 6.846854856634808e-05,
+      "loss": 1.9165,
+      "step": 29598
+    },
+    {
+      "epoch": 0.6577555555555555,
+      "grad_norm": 1.5138273239135742,
+      "learning_rate": 6.846410313402979e-05,
+      "loss": 1.5262,
+      "step": 29599
+    },
+    {
+      "epoch": 0.6577777777777778,
+      "grad_norm": 1.6930948495864868,
+      "learning_rate": 6.84596577017115e-05,
+      "loss": 1.0738,
+      "step": 29600
+    },
+    {
+      "epoch": 0.6578,
+      "grad_norm": 1.5009393692016602,
+      "learning_rate": 6.845521226939319e-05,
+      "loss": 1.9558,
+      "step": 29601
+    },
+    {
+      "epoch": 0.6578222222222222,
+      "grad_norm": 1.225827693939209,
+      "learning_rate": 6.845076683707492e-05,
+      "loss": 1.5175,
+      "step": 29602
+    },
+    {
+      "epoch": 0.6578444444444445,
+      "grad_norm": 1.4594206809997559,
+      "learning_rate": 6.844632140475662e-05,
+      "loss": 2.0969,
+      "step": 29603
+    },
+    {
+      "epoch": 0.6578666666666667,
+      "grad_norm": 1.7906701564788818,
+      "learning_rate": 6.844187597243832e-05,
+      "loss": 2.1246,
+      "step": 29604
+    },
+    {
+      "epoch": 0.6578888888888889,
+      "grad_norm": 1.4477447271347046,
+      "learning_rate": 6.843743054012003e-05,
+      "loss": 1.8149,
+      "step": 29605
+    },
+    {
+      "epoch": 0.6579111111111111,
+      "grad_norm": 1.7752348184585571,
+      "learning_rate": 6.843298510780174e-05,
+      "loss": 2.6142,
+      "step": 29606
+    },
+    {
+      "epoch": 0.6579333333333334,
+      "grad_norm": 1.4473702907562256,
+      "learning_rate": 6.842853967548345e-05,
+      "loss": 1.5544,
+      "step": 29607
+    },
+    {
+      "epoch": 0.6579555555555555,
+      "grad_norm": 1.456236720085144,
+      "learning_rate": 6.842409424316515e-05,
+      "loss": 2.0661,
+      "step": 29608
+    },
+    {
+      "epoch": 0.6579777777777778,
+      "grad_norm": 1.6258362531661987,
+      "learning_rate": 6.841964881084686e-05,
+      "loss": 1.7835,
+      "step": 29609
+    },
+    {
+      "epoch": 0.658,
+      "grad_norm": 1.791027307510376,
+      "learning_rate": 6.841520337852857e-05,
+      "loss": 2.0022,
+      "step": 29610
+    },
+    {
+      "epoch": 0.6580222222222222,
+      "grad_norm": 2.289262056350708,
+      "learning_rate": 6.841075794621027e-05,
+      "loss": 1.7229,
+      "step": 29611
+    },
+    {
+      "epoch": 0.6580444444444444,
+      "grad_norm": 1.7048383951187134,
+      "learning_rate": 6.840631251389198e-05,
+      "loss": 2.1676,
+      "step": 29612
+    },
+    {
+      "epoch": 0.6580666666666667,
+      "grad_norm": 1.8120087385177612,
+      "learning_rate": 6.84018670815737e-05,
+      "loss": 2.9178,
+      "step": 29613
+    },
+    {
+      "epoch": 0.6580888888888888,
+      "grad_norm": 1.4685407876968384,
+      "learning_rate": 6.839742164925539e-05,
+      "loss": 1.1795,
+      "step": 29614
+    },
+    {
+      "epoch": 0.6581111111111111,
+      "grad_norm": 1.9184573888778687,
+      "learning_rate": 6.83929762169371e-05,
+      "loss": 2.4737,
+      "step": 29615
+    },
+    {
+      "epoch": 0.6581333333333333,
+      "grad_norm": 1.8086332082748413,
+      "learning_rate": 6.838853078461881e-05,
+      "loss": 1.844,
+      "step": 29616
+    },
+    {
+      "epoch": 0.6581555555555556,
+      "grad_norm": 1.885086178779602,
+      "learning_rate": 6.838408535230052e-05,
+      "loss": 1.7616,
+      "step": 29617
+    },
+    {
+      "epoch": 0.6581777777777778,
+      "grad_norm": 1.7152103185653687,
+      "learning_rate": 6.837963991998221e-05,
+      "loss": 2.3195,
+      "step": 29618
+    },
+    {
+      "epoch": 0.6582,
+      "grad_norm": 2.053940773010254,
+      "learning_rate": 6.837519448766392e-05,
+      "loss": 2.235,
+      "step": 29619
+    },
+    {
+      "epoch": 0.6582222222222223,
+      "grad_norm": 1.9013762474060059,
+      "learning_rate": 6.837074905534563e-05,
+      "loss": 2.3344,
+      "step": 29620
+    },
+    {
+      "epoch": 0.6582444444444444,
+      "grad_norm": 1.1086158752441406,
+      "learning_rate": 6.836630362302734e-05,
+      "loss": 0.9228,
+      "step": 29621
+    },
+    {
+      "epoch": 0.6582666666666667,
+      "grad_norm": 1.6535439491271973,
+      "learning_rate": 6.836185819070905e-05,
+      "loss": 1.9664,
+      "step": 29622
+    },
+    {
+      "epoch": 0.6582888888888889,
+      "grad_norm": 1.2858459949493408,
+      "learning_rate": 6.835741275839076e-05,
+      "loss": 0.9314,
+      "step": 29623
+    },
+    {
+      "epoch": 0.6583111111111111,
+      "grad_norm": 1.4932442903518677,
+      "learning_rate": 6.835296732607246e-05,
+      "loss": 1.8218,
+      "step": 29624
+    },
+    {
+      "epoch": 0.6583333333333333,
+      "grad_norm": 1.3918441534042358,
+      "learning_rate": 6.834852189375417e-05,
+      "loss": 1.7047,
+      "step": 29625
+    },
+    {
+      "epoch": 0.6583555555555556,
+      "grad_norm": 2.063030242919922,
+      "learning_rate": 6.834407646143588e-05,
+      "loss": 2.3652,
+      "step": 29626
+    },
+    {
+      "epoch": 0.6583777777777777,
+      "grad_norm": 2.071507453918457,
+      "learning_rate": 6.833963102911759e-05,
+      "loss": 1.9228,
+      "step": 29627
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 1.5261411666870117,
+      "learning_rate": 6.833518559679928e-05,
+      "loss": 1.667,
+      "step": 29628
+    },
+    {
+      "epoch": 0.6584222222222222,
+      "grad_norm": 1.715639591217041,
+      "learning_rate": 6.8330740164481e-05,
+      "loss": 1.8281,
+      "step": 29629
+    },
+    {
+      "epoch": 0.6584444444444445,
+      "grad_norm": 1.7601677179336548,
+      "learning_rate": 6.832629473216272e-05,
+      "loss": 2.1907,
+      "step": 29630
+    },
+    {
+      "epoch": 0.6584666666666666,
+      "grad_norm": 1.5039981603622437,
+      "learning_rate": 6.832184929984441e-05,
+      "loss": 1.7468,
+      "step": 29631
+    },
+    {
+      "epoch": 0.6584888888888889,
+      "grad_norm": 1.8185672760009766,
+      "learning_rate": 6.831740386752612e-05,
+      "loss": 1.5048,
+      "step": 29632
+    },
+    {
+      "epoch": 0.6585111111111112,
+      "grad_norm": 1.8447353839874268,
+      "learning_rate": 6.831295843520783e-05,
+      "loss": 1.9108,
+      "step": 29633
+    },
+    {
+      "epoch": 0.6585333333333333,
+      "grad_norm": 2.0437545776367188,
+      "learning_rate": 6.830851300288953e-05,
+      "loss": 2.3156,
+      "step": 29634
+    },
+    {
+      "epoch": 0.6585555555555556,
+      "grad_norm": 1.719464659690857,
+      "learning_rate": 6.830406757057124e-05,
+      "loss": 1.751,
+      "step": 29635
+    },
+    {
+      "epoch": 0.6585777777777778,
+      "grad_norm": 1.796617865562439,
+      "learning_rate": 6.829962213825295e-05,
+      "loss": 1.6782,
+      "step": 29636
+    },
+    {
+      "epoch": 0.6586,
+      "grad_norm": 1.913540244102478,
+      "learning_rate": 6.829517670593466e-05,
+      "loss": 2.1384,
+      "step": 29637
+    },
+    {
+      "epoch": 0.6586222222222222,
+      "grad_norm": 1.398923397064209,
+      "learning_rate": 6.829073127361635e-05,
+      "loss": 1.5806,
+      "step": 29638
+    },
+    {
+      "epoch": 0.6586444444444445,
+      "grad_norm": 1.6090314388275146,
+      "learning_rate": 6.828628584129808e-05,
+      "loss": 1.4752,
+      "step": 29639
+    },
+    {
+      "epoch": 0.6586666666666666,
+      "grad_norm": 2.2967145442962646,
+      "learning_rate": 6.828184040897979e-05,
+      "loss": 2.758,
+      "step": 29640
+    },
+    {
+      "epoch": 0.6586888888888889,
+      "grad_norm": 1.6961264610290527,
+      "learning_rate": 6.827739497666148e-05,
+      "loss": 1.7788,
+      "step": 29641
+    },
+    {
+      "epoch": 0.6587111111111111,
+      "grad_norm": 1.4948949813842773,
+      "learning_rate": 6.827294954434319e-05,
+      "loss": 1.3806,
+      "step": 29642
+    },
+    {
+      "epoch": 0.6587333333333333,
+      "grad_norm": 1.6879496574401855,
+      "learning_rate": 6.82685041120249e-05,
+      "loss": 2.105,
+      "step": 29643
+    },
+    {
+      "epoch": 0.6587555555555555,
+      "grad_norm": 1.7877131700515747,
+      "learning_rate": 6.82640586797066e-05,
+      "loss": 1.9023,
+      "step": 29644
+    },
+    {
+      "epoch": 0.6587777777777778,
+      "grad_norm": 1.640920639038086,
+      "learning_rate": 6.825961324738831e-05,
+      "loss": 1.5046,
+      "step": 29645
+    },
+    {
+      "epoch": 0.6588,
+      "grad_norm": 1.6340906620025635,
+      "learning_rate": 6.825516781507002e-05,
+      "loss": 1.6889,
+      "step": 29646
+    },
+    {
+      "epoch": 0.6588222222222222,
+      "grad_norm": 1.9517959356307983,
+      "learning_rate": 6.825072238275173e-05,
+      "loss": 1.8836,
+      "step": 29647
+    },
+    {
+      "epoch": 0.6588444444444445,
+      "grad_norm": 2.099510669708252,
+      "learning_rate": 6.824627695043344e-05,
+      "loss": 1.684,
+      "step": 29648
+    },
+    {
+      "epoch": 0.6588666666666667,
+      "grad_norm": 1.5277636051177979,
+      "learning_rate": 6.824183151811515e-05,
+      "loss": 1.368,
+      "step": 29649
+    },
+    {
+      "epoch": 0.6588888888888889,
+      "grad_norm": 2.4442501068115234,
+      "learning_rate": 6.823738608579685e-05,
+      "loss": 2.1152,
+      "step": 29650
+    },
+    {
+      "epoch": 0.6589111111111111,
+      "grad_norm": 0.9391299486160278,
+      "learning_rate": 6.823294065347855e-05,
+      "loss": 1.2782,
+      "step": 29651
+    },
+    {
+      "epoch": 0.6589333333333334,
+      "grad_norm": 1.7166022062301636,
+      "learning_rate": 6.822849522116026e-05,
+      "loss": 2.52,
+      "step": 29652
+    },
+    {
+      "epoch": 0.6589555555555555,
+      "grad_norm": 0.9608427286148071,
+      "learning_rate": 6.822404978884197e-05,
+      "loss": 0.9397,
+      "step": 29653
+    },
+    {
+      "epoch": 0.6589777777777778,
+      "grad_norm": 0.9675276279449463,
+      "learning_rate": 6.821960435652367e-05,
+      "loss": 0.9316,
+      "step": 29654
+    },
+    {
+      "epoch": 0.659,
+      "grad_norm": 1.6645857095718384,
+      "learning_rate": 6.821515892420538e-05,
+      "loss": 2.3069,
+      "step": 29655
+    },
+    {
+      "epoch": 0.6590222222222222,
+      "grad_norm": 1.3648736476898193,
+      "learning_rate": 6.821071349188709e-05,
+      "loss": 1.6865,
+      "step": 29656
+    },
+    {
+      "epoch": 0.6590444444444444,
+      "grad_norm": 1.2589092254638672,
+      "learning_rate": 6.82062680595688e-05,
+      "loss": 2.0334,
+      "step": 29657
+    },
+    {
+      "epoch": 0.6590666666666667,
+      "grad_norm": 1.7342948913574219,
+      "learning_rate": 6.82018226272505e-05,
+      "loss": 2.4378,
+      "step": 29658
+    },
+    {
+      "epoch": 0.6590888888888888,
+      "grad_norm": 1.6458474397659302,
+      "learning_rate": 6.819737719493221e-05,
+      "loss": 1.6645,
+      "step": 29659
+    },
+    {
+      "epoch": 0.6591111111111111,
+      "grad_norm": 1.6723769903182983,
+      "learning_rate": 6.819293176261392e-05,
+      "loss": 1.7515,
+      "step": 29660
+    },
+    {
+      "epoch": 0.6591333333333333,
+      "grad_norm": 1.7363861799240112,
+      "learning_rate": 6.818848633029562e-05,
+      "loss": 2.6189,
+      "step": 29661
+    },
+    {
+      "epoch": 0.6591555555555556,
+      "grad_norm": 1.4819042682647705,
+      "learning_rate": 6.818404089797733e-05,
+      "loss": 1.5243,
+      "step": 29662
+    },
+    {
+      "epoch": 0.6591777777777778,
+      "grad_norm": 1.7086176872253418,
+      "learning_rate": 6.817959546565904e-05,
+      "loss": 2.3058,
+      "step": 29663
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 1.4702590703964233,
+      "learning_rate": 6.817515003334075e-05,
+      "loss": 1.3932,
+      "step": 29664
+    },
+    {
+      "epoch": 0.6592222222222223,
+      "grad_norm": 1.6508628129959106,
+      "learning_rate": 6.817070460102244e-05,
+      "loss": 2.3402,
+      "step": 29665
+    },
+    {
+      "epoch": 0.6592444444444444,
+      "grad_norm": 1.8322089910507202,
+      "learning_rate": 6.816625916870415e-05,
+      "loss": 2.0941,
+      "step": 29666
+    },
+    {
+      "epoch": 0.6592666666666667,
+      "grad_norm": 1.5012660026550293,
+      "learning_rate": 6.816181373638588e-05,
+      "loss": 1.861,
+      "step": 29667
+    },
+    {
+      "epoch": 0.6592888888888889,
+      "grad_norm": 1.5552964210510254,
+      "learning_rate": 6.815736830406757e-05,
+      "loss": 1.7842,
+      "step": 29668
+    },
+    {
+      "epoch": 0.6593111111111111,
+      "grad_norm": 1.3963440656661987,
+      "learning_rate": 6.815292287174928e-05,
+      "loss": 1.2243,
+      "step": 29669
+    },
+    {
+      "epoch": 0.6593333333333333,
+      "grad_norm": 1.7548903226852417,
+      "learning_rate": 6.814847743943099e-05,
+      "loss": 1.9696,
+      "step": 29670
+    },
+    {
+      "epoch": 0.6593555555555556,
+      "grad_norm": 1.929069995880127,
+      "learning_rate": 6.814403200711269e-05,
+      "loss": 2.0156,
+      "step": 29671
+    },
+    {
+      "epoch": 0.6593777777777777,
+      "grad_norm": 1.6079157590866089,
+      "learning_rate": 6.81395865747944e-05,
+      "loss": 2.0887,
+      "step": 29672
+    },
+    {
+      "epoch": 0.6594,
+      "grad_norm": 2.9583001136779785,
+      "learning_rate": 6.813514114247611e-05,
+      "loss": 1.8755,
+      "step": 29673
+    },
+    {
+      "epoch": 0.6594222222222222,
+      "grad_norm": 1.3937418460845947,
+      "learning_rate": 6.813069571015782e-05,
+      "loss": 1.1554,
+      "step": 29674
+    },
+    {
+      "epoch": 0.6594444444444445,
+      "grad_norm": 1.1485793590545654,
+      "learning_rate": 6.812625027783951e-05,
+      "loss": 1.0502,
+      "step": 29675
+    },
+    {
+      "epoch": 0.6594666666666666,
+      "grad_norm": 1.6714928150177002,
+      "learning_rate": 6.812180484552124e-05,
+      "loss": 1.9522,
+      "step": 29676
+    },
+    {
+      "epoch": 0.6594888888888889,
+      "grad_norm": 1.5369539260864258,
+      "learning_rate": 6.811735941320295e-05,
+      "loss": 1.8008,
+      "step": 29677
+    },
+    {
+      "epoch": 0.6595111111111112,
+      "grad_norm": 1.638718843460083,
+      "learning_rate": 6.811291398088464e-05,
+      "loss": 1.6649,
+      "step": 29678
+    },
+    {
+      "epoch": 0.6595333333333333,
+      "grad_norm": 2.0578885078430176,
+      "learning_rate": 6.810846854856635e-05,
+      "loss": 2.3217,
+      "step": 29679
+    },
+    {
+      "epoch": 0.6595555555555556,
+      "grad_norm": 1.6226433515548706,
+      "learning_rate": 6.810402311624806e-05,
+      "loss": 1.8039,
+      "step": 29680
+    },
+    {
+      "epoch": 0.6595777777777778,
+      "grad_norm": 1.656620740890503,
+      "learning_rate": 6.809957768392976e-05,
+      "loss": 1.5798,
+      "step": 29681
+    },
+    {
+      "epoch": 0.6596,
+      "grad_norm": 1.6753404140472412,
+      "learning_rate": 6.809513225161147e-05,
+      "loss": 1.861,
+      "step": 29682
+    },
+    {
+      "epoch": 0.6596222222222222,
+      "grad_norm": 1.7894854545593262,
+      "learning_rate": 6.809068681929318e-05,
+      "loss": 1.6227,
+      "step": 29683
+    },
+    {
+      "epoch": 0.6596444444444445,
+      "grad_norm": 1.937540888786316,
+      "learning_rate": 6.808624138697489e-05,
+      "loss": 2.182,
+      "step": 29684
+    },
+    {
+      "epoch": 0.6596666666666666,
+      "grad_norm": 1.7234638929367065,
+      "learning_rate": 6.80817959546566e-05,
+      "loss": 2.4864,
+      "step": 29685
+    },
+    {
+      "epoch": 0.6596888888888889,
+      "grad_norm": 1.9724160432815552,
+      "learning_rate": 6.80773505223383e-05,
+      "loss": 2.3914,
+      "step": 29686
+    },
+    {
+      "epoch": 0.6597111111111111,
+      "grad_norm": 1.872678518295288,
+      "learning_rate": 6.807290509002002e-05,
+      "loss": 1.9283,
+      "step": 29687
+    },
+    {
+      "epoch": 0.6597333333333333,
+      "grad_norm": 1.7543476819992065,
+      "learning_rate": 6.806845965770171e-05,
+      "loss": 1.6584,
+      "step": 29688
+    },
+    {
+      "epoch": 0.6597555555555555,
+      "grad_norm": 1.8167245388031006,
+      "learning_rate": 6.806401422538342e-05,
+      "loss": 1.7004,
+      "step": 29689
+    },
+    {
+      "epoch": 0.6597777777777778,
+      "grad_norm": 1.886895775794983,
+      "learning_rate": 6.805956879306513e-05,
+      "loss": 1.8245,
+      "step": 29690
+    },
+    {
+      "epoch": 0.6598,
+      "grad_norm": 1.7559186220169067,
+      "learning_rate": 6.805512336074683e-05,
+      "loss": 1.9585,
+      "step": 29691
+    },
+    {
+      "epoch": 0.6598222222222222,
+      "grad_norm": 1.783811330795288,
+      "learning_rate": 6.805067792842854e-05,
+      "loss": 1.9343,
+      "step": 29692
+    },
+    {
+      "epoch": 0.6598444444444445,
+      "grad_norm": 1.630228877067566,
+      "learning_rate": 6.804623249611025e-05,
+      "loss": 1.4084,
+      "step": 29693
+    },
+    {
+      "epoch": 0.6598666666666667,
+      "grad_norm": 1.8122942447662354,
+      "learning_rate": 6.804178706379196e-05,
+      "loss": 1.8979,
+      "step": 29694
+    },
+    {
+      "epoch": 0.6598888888888889,
+      "grad_norm": 2.2165892124176025,
+      "learning_rate": 6.803734163147367e-05,
+      "loss": 2.2585,
+      "step": 29695
+    },
+    {
+      "epoch": 0.6599111111111111,
+      "grad_norm": 1.2415522336959839,
+      "learning_rate": 6.803289619915538e-05,
+      "loss": 1.1165,
+      "step": 29696
+    },
+    {
+      "epoch": 0.6599333333333334,
+      "grad_norm": 1.784896969795227,
+      "learning_rate": 6.802845076683709e-05,
+      "loss": 1.7723,
+      "step": 29697
+    },
+    {
+      "epoch": 0.6599555555555555,
+      "grad_norm": 2.093536615371704,
+      "learning_rate": 6.802400533451878e-05,
+      "loss": 1.4105,
+      "step": 29698
+    },
+    {
+      "epoch": 0.6599777777777778,
+      "grad_norm": 1.7735121250152588,
+      "learning_rate": 6.801955990220049e-05,
+      "loss": 1.7523,
+      "step": 29699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0457617044448853,
+      "learning_rate": 6.80151144698822e-05,
+      "loss": 0.6506,
+      "step": 29700
+    },
+    {
+      "epoch": 0.6600222222222222,
+      "grad_norm": 1.3916157484054565,
+      "learning_rate": 6.80106690375639e-05,
+      "loss": 2.4461,
+      "step": 29701
+    },
+    {
+      "epoch": 0.6600444444444444,
+      "grad_norm": 1.639994502067566,
+      "learning_rate": 6.80062236052456e-05,
+      "loss": 2.5601,
+      "step": 29702
+    },
+    {
+      "epoch": 0.6600666666666667,
+      "grad_norm": 1.5185401439666748,
+      "learning_rate": 6.800177817292732e-05,
+      "loss": 2.6104,
+      "step": 29703
+    },
+    {
+      "epoch": 0.6600888888888888,
+      "grad_norm": 1.7068262100219727,
+      "learning_rate": 6.799733274060904e-05,
+      "loss": 1.9843,
+      "step": 29704
+    },
+    {
+      "epoch": 0.6601111111111111,
+      "grad_norm": 1.891868233680725,
+      "learning_rate": 6.799288730829073e-05,
+      "loss": 3.0391,
+      "step": 29705
+    },
+    {
+      "epoch": 0.6601333333333333,
+      "grad_norm": 1.5163244009017944,
+      "learning_rate": 6.798844187597244e-05,
+      "loss": 2.0175,
+      "step": 29706
+    },
+    {
+      "epoch": 0.6601555555555556,
+      "grad_norm": 1.5959887504577637,
+      "learning_rate": 6.798399644365415e-05,
+      "loss": 2.1754,
+      "step": 29707
+    },
+    {
+      "epoch": 0.6601777777777778,
+      "grad_norm": 1.3765621185302734,
+      "learning_rate": 6.797955101133585e-05,
+      "loss": 2.1678,
+      "step": 29708
+    },
+    {
+      "epoch": 0.6602,
+      "grad_norm": 1.5355892181396484,
+      "learning_rate": 6.797510557901756e-05,
+      "loss": 1.8159,
+      "step": 29709
+    },
+    {
+      "epoch": 0.6602222222222223,
+      "grad_norm": 1.7196125984191895,
+      "learning_rate": 6.797066014669927e-05,
+      "loss": 1.6837,
+      "step": 29710
+    },
+    {
+      "epoch": 0.6602444444444444,
+      "grad_norm": 1.511362910270691,
+      "learning_rate": 6.796621471438097e-05,
+      "loss": 2.4021,
+      "step": 29711
+    },
+    {
+      "epoch": 0.6602666666666667,
+      "grad_norm": 1.6308833360671997,
+      "learning_rate": 6.796176928206268e-05,
+      "loss": 1.6508,
+      "step": 29712
+    },
+    {
+      "epoch": 0.6602888888888889,
+      "grad_norm": 1.5413122177124023,
+      "learning_rate": 6.79573238497444e-05,
+      "loss": 2.0749,
+      "step": 29713
+    },
+    {
+      "epoch": 0.6603111111111111,
+      "grad_norm": 1.774201512336731,
+      "learning_rate": 6.795287841742611e-05,
+      "loss": 1.9161,
+      "step": 29714
+    },
+    {
+      "epoch": 0.6603333333333333,
+      "grad_norm": 1.5780080556869507,
+      "learning_rate": 6.79484329851078e-05,
+      "loss": 2.2094,
+      "step": 29715
+    },
+    {
+      "epoch": 0.6603555555555556,
+      "grad_norm": 1.2944010496139526,
+      "learning_rate": 6.794398755278951e-05,
+      "loss": 1.785,
+      "step": 29716
+    },
+    {
+      "epoch": 0.6603777777777777,
+      "grad_norm": 1.6052595376968384,
+      "learning_rate": 6.793954212047122e-05,
+      "loss": 2.0328,
+      "step": 29717
+    },
+    {
+      "epoch": 0.6604,
+      "grad_norm": 1.6314421892166138,
+      "learning_rate": 6.793509668815292e-05,
+      "loss": 1.9605,
+      "step": 29718
+    },
+    {
+      "epoch": 0.6604222222222222,
+      "grad_norm": 1.7301579713821411,
+      "learning_rate": 6.793065125583463e-05,
+      "loss": 1.9727,
+      "step": 29719
+    },
+    {
+      "epoch": 0.6604444444444444,
+      "grad_norm": 1.8894845247268677,
+      "learning_rate": 6.792620582351634e-05,
+      "loss": 2.1291,
+      "step": 29720
+    },
+    {
+      "epoch": 0.6604666666666666,
+      "grad_norm": 1.5691381692886353,
+      "learning_rate": 6.792176039119805e-05,
+      "loss": 1.5028,
+      "step": 29721
+    },
+    {
+      "epoch": 0.6604888888888889,
+      "grad_norm": 1.8337043523788452,
+      "learning_rate": 6.791731495887976e-05,
+      "loss": 1.5781,
+      "step": 29722
+    },
+    {
+      "epoch": 0.6605111111111112,
+      "grad_norm": 1.8304129838943481,
+      "learning_rate": 6.791286952656147e-05,
+      "loss": 2.2604,
+      "step": 29723
+    },
+    {
+      "epoch": 0.6605333333333333,
+      "grad_norm": 1.7375537157058716,
+      "learning_rate": 6.790842409424318e-05,
+      "loss": 2.2333,
+      "step": 29724
+    },
+    {
+      "epoch": 0.6605555555555556,
+      "grad_norm": 1.5077272653579712,
+      "learning_rate": 6.790397866192487e-05,
+      "loss": 1.47,
+      "step": 29725
+    },
+    {
+      "epoch": 0.6605777777777778,
+      "grad_norm": 1.7780345678329468,
+      "learning_rate": 6.789953322960658e-05,
+      "loss": 1.7045,
+      "step": 29726
+    },
+    {
+      "epoch": 0.6606,
+      "grad_norm": 1.7432992458343506,
+      "learning_rate": 6.789508779728829e-05,
+      "loss": 1.9861,
+      "step": 29727
+    },
+    {
+      "epoch": 0.6606222222222222,
+      "grad_norm": 1.5569027662277222,
+      "learning_rate": 6.789064236496999e-05,
+      "loss": 1.6343,
+      "step": 29728
+    },
+    {
+      "epoch": 0.6606444444444445,
+      "grad_norm": 1.9147087335586548,
+      "learning_rate": 6.78861969326517e-05,
+      "loss": 1.8007,
+      "step": 29729
+    },
+    {
+      "epoch": 0.6606666666666666,
+      "grad_norm": 1.8529517650604248,
+      "learning_rate": 6.788175150033341e-05,
+      "loss": 2.0208,
+      "step": 29730
+    },
+    {
+      "epoch": 0.6606888888888889,
+      "grad_norm": 1.5898747444152832,
+      "learning_rate": 6.787730606801512e-05,
+      "loss": 1.9657,
+      "step": 29731
+    },
+    {
+      "epoch": 0.6607111111111111,
+      "grad_norm": 2.055760622024536,
+      "learning_rate": 6.787286063569683e-05,
+      "loss": 2.0554,
+      "step": 29732
+    },
+    {
+      "epoch": 0.6607333333333333,
+      "grad_norm": 2.072956085205078,
+      "learning_rate": 6.786841520337854e-05,
+      "loss": 2.523,
+      "step": 29733
+    },
+    {
+      "epoch": 0.6607555555555555,
+      "grad_norm": 1.6523401737213135,
+      "learning_rate": 6.786396977106025e-05,
+      "loss": 1.6765,
+      "step": 29734
+    },
+    {
+      "epoch": 0.6607777777777778,
+      "grad_norm": 1.8183640241622925,
+      "learning_rate": 6.785952433874194e-05,
+      "loss": 2.185,
+      "step": 29735
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 1.4948359727859497,
+      "learning_rate": 6.785507890642365e-05,
+      "loss": 1.467,
+      "step": 29736
+    },
+    {
+      "epoch": 0.6608222222222222,
+      "grad_norm": 1.782305121421814,
+      "learning_rate": 6.785063347410536e-05,
+      "loss": 2.0008,
+      "step": 29737
+    },
+    {
+      "epoch": 0.6608444444444445,
+      "grad_norm": 1.9820513725280762,
+      "learning_rate": 6.784618804178706e-05,
+      "loss": 1.8032,
+      "step": 29738
+    },
+    {
+      "epoch": 0.6608666666666667,
+      "grad_norm": 1.8400990962982178,
+      "learning_rate": 6.784174260946877e-05,
+      "loss": 2.0825,
+      "step": 29739
+    },
+    {
+      "epoch": 0.6608888888888889,
+      "grad_norm": 1.880311131477356,
+      "learning_rate": 6.783729717715048e-05,
+      "loss": 2.0094,
+      "step": 29740
+    },
+    {
+      "epoch": 0.6609111111111111,
+      "grad_norm": 1.5467296838760376,
+      "learning_rate": 6.78328517448322e-05,
+      "loss": 1.53,
+      "step": 29741
+    },
+    {
+      "epoch": 0.6609333333333334,
+      "grad_norm": 1.9151309728622437,
+      "learning_rate": 6.78284063125139e-05,
+      "loss": 2.0012,
+      "step": 29742
+    },
+    {
+      "epoch": 0.6609555555555555,
+      "grad_norm": 1.48209547996521,
+      "learning_rate": 6.78239608801956e-05,
+      "loss": 1.5927,
+      "step": 29743
+    },
+    {
+      "epoch": 0.6609777777777778,
+      "grad_norm": 2.134293556213379,
+      "learning_rate": 6.781951544787732e-05,
+      "loss": 1.918,
+      "step": 29744
+    },
+    {
+      "epoch": 0.661,
+      "grad_norm": 1.6198315620422363,
+      "learning_rate": 6.781507001555901e-05,
+      "loss": 1.6545,
+      "step": 29745
+    },
+    {
+      "epoch": 0.6610222222222222,
+      "grad_norm": 1.608734130859375,
+      "learning_rate": 6.781062458324072e-05,
+      "loss": 1.5489,
+      "step": 29746
+    },
+    {
+      "epoch": 0.6610444444444444,
+      "grad_norm": 1.6038824319839478,
+      "learning_rate": 6.780617915092243e-05,
+      "loss": 1.4575,
+      "step": 29747
+    },
+    {
+      "epoch": 0.6610666666666667,
+      "grad_norm": 1.6160492897033691,
+      "learning_rate": 6.780173371860413e-05,
+      "loss": 1.6108,
+      "step": 29748
+    },
+    {
+      "epoch": 0.6610888888888888,
+      "grad_norm": 1.9526495933532715,
+      "learning_rate": 6.779728828628584e-05,
+      "loss": 2.0151,
+      "step": 29749
+    },
+    {
+      "epoch": 0.6611111111111111,
+      "grad_norm": 1.8424237966537476,
+      "learning_rate": 6.779284285396756e-05,
+      "loss": 1.6796,
+      "step": 29750
+    },
+    {
+      "epoch": 0.6611333333333334,
+      "grad_norm": 1.365674376487732,
+      "learning_rate": 6.778839742164927e-05,
+      "loss": 2.4502,
+      "step": 29751
+    },
+    {
+      "epoch": 0.6611555555555556,
+      "grad_norm": 1.7497609853744507,
+      "learning_rate": 6.778395198933096e-05,
+      "loss": 2.7399,
+      "step": 29752
+    },
+    {
+      "epoch": 0.6611777777777778,
+      "grad_norm": 1.6796815395355225,
+      "learning_rate": 6.777950655701267e-05,
+      "loss": 2.0876,
+      "step": 29753
+    },
+    {
+      "epoch": 0.6612,
+      "grad_norm": 1.3631795644760132,
+      "learning_rate": 6.777506112469438e-05,
+      "loss": 1.9842,
+      "step": 29754
+    },
+    {
+      "epoch": 0.6612222222222223,
+      "grad_norm": 1.490350365638733,
+      "learning_rate": 6.777061569237608e-05,
+      "loss": 1.9246,
+      "step": 29755
+    },
+    {
+      "epoch": 0.6612444444444444,
+      "grad_norm": 1.5525401830673218,
+      "learning_rate": 6.776617026005779e-05,
+      "loss": 1.9768,
+      "step": 29756
+    },
+    {
+      "epoch": 0.6612666666666667,
+      "grad_norm": 2.369767904281616,
+      "learning_rate": 6.77617248277395e-05,
+      "loss": 1.8494,
+      "step": 29757
+    },
+    {
+      "epoch": 0.6612888888888889,
+      "grad_norm": 1.6807938814163208,
+      "learning_rate": 6.775727939542121e-05,
+      "loss": 2.0463,
+      "step": 29758
+    },
+    {
+      "epoch": 0.6613111111111111,
+      "grad_norm": 1.5899170637130737,
+      "learning_rate": 6.775283396310292e-05,
+      "loss": 2.4831,
+      "step": 29759
+    },
+    {
+      "epoch": 0.6613333333333333,
+      "grad_norm": 1.462839961051941,
+      "learning_rate": 6.774838853078463e-05,
+      "loss": 2.0605,
+      "step": 29760
+    },
+    {
+      "epoch": 0.6613555555555556,
+      "grad_norm": 1.7161911725997925,
+      "learning_rate": 6.774394309846634e-05,
+      "loss": 2.0418,
+      "step": 29761
+    },
+    {
+      "epoch": 0.6613777777777777,
+      "grad_norm": 1.541351318359375,
+      "learning_rate": 6.773949766614803e-05,
+      "loss": 1.7762,
+      "step": 29762
+    },
+    {
+      "epoch": 0.6614,
+      "grad_norm": 1.732049822807312,
+      "learning_rate": 6.773505223382974e-05,
+      "loss": 2.1298,
+      "step": 29763
+    },
+    {
+      "epoch": 0.6614222222222222,
+      "grad_norm": 1.5366971492767334,
+      "learning_rate": 6.773060680151145e-05,
+      "loss": 1.7534,
+      "step": 29764
+    },
+    {
+      "epoch": 0.6614444444444444,
+      "grad_norm": 1.4801750183105469,
+      "learning_rate": 6.772616136919315e-05,
+      "loss": 1.811,
+      "step": 29765
+    },
+    {
+      "epoch": 0.6614666666666666,
+      "grad_norm": 1.6979097127914429,
+      "learning_rate": 6.772171593687486e-05,
+      "loss": 2.5855,
+      "step": 29766
+    },
+    {
+      "epoch": 0.6614888888888889,
+      "grad_norm": 1.709199070930481,
+      "learning_rate": 6.771727050455657e-05,
+      "loss": 2.179,
+      "step": 29767
+    },
+    {
+      "epoch": 0.6615111111111112,
+      "grad_norm": 1.6825566291809082,
+      "learning_rate": 6.771282507223828e-05,
+      "loss": 2.3518,
+      "step": 29768
+    },
+    {
+      "epoch": 0.6615333333333333,
+      "grad_norm": 1.5685398578643799,
+      "learning_rate": 6.770837963991999e-05,
+      "loss": 1.6449,
+      "step": 29769
+    },
+    {
+      "epoch": 0.6615555555555556,
+      "grad_norm": 2.124915838241577,
+      "learning_rate": 6.77039342076017e-05,
+      "loss": 0.7097,
+      "step": 29770
+    },
+    {
+      "epoch": 0.6615777777777778,
+      "grad_norm": 1.4958488941192627,
+      "learning_rate": 6.769948877528341e-05,
+      "loss": 1.2729,
+      "step": 29771
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 2.002549409866333,
+      "learning_rate": 6.76950433429651e-05,
+      "loss": 2.3574,
+      "step": 29772
+    },
+    {
+      "epoch": 0.6616222222222222,
+      "grad_norm": 1.654229760169983,
+      "learning_rate": 6.769059791064681e-05,
+      "loss": 2.0806,
+      "step": 29773
+    },
+    {
+      "epoch": 0.6616444444444445,
+      "grad_norm": 1.7626105546951294,
+      "learning_rate": 6.768615247832852e-05,
+      "loss": 1.9968,
+      "step": 29774
+    },
+    {
+      "epoch": 0.6616666666666666,
+      "grad_norm": 1.9881726503372192,
+      "learning_rate": 6.768170704601022e-05,
+      "loss": 2.1822,
+      "step": 29775
+    },
+    {
+      "epoch": 0.6616888888888889,
+      "grad_norm": 1.5868173837661743,
+      "learning_rate": 6.767726161369193e-05,
+      "loss": 1.8892,
+      "step": 29776
+    },
+    {
+      "epoch": 0.6617111111111111,
+      "grad_norm": 1.8402587175369263,
+      "learning_rate": 6.767281618137364e-05,
+      "loss": 1.5504,
+      "step": 29777
+    },
+    {
+      "epoch": 0.6617333333333333,
+      "grad_norm": 1.462702751159668,
+      "learning_rate": 6.766837074905536e-05,
+      "loss": 2.1335,
+      "step": 29778
+    },
+    {
+      "epoch": 0.6617555555555555,
+      "grad_norm": 1.7339056730270386,
+      "learning_rate": 6.766392531673706e-05,
+      "loss": 2.0455,
+      "step": 29779
+    },
+    {
+      "epoch": 0.6617777777777778,
+      "grad_norm": 1.7997807264328003,
+      "learning_rate": 6.765947988441877e-05,
+      "loss": 1.7998,
+      "step": 29780
+    },
+    {
+      "epoch": 0.6618,
+      "grad_norm": 1.674707055091858,
+      "learning_rate": 6.765503445210048e-05,
+      "loss": 1.8069,
+      "step": 29781
+    },
+    {
+      "epoch": 0.6618222222222222,
+      "grad_norm": 1.5384355783462524,
+      "learning_rate": 6.765058901978217e-05,
+      "loss": 1.8479,
+      "step": 29782
+    },
+    {
+      "epoch": 0.6618444444444445,
+      "grad_norm": 2.0576751232147217,
+      "learning_rate": 6.764614358746388e-05,
+      "loss": 1.7632,
+      "step": 29783
+    },
+    {
+      "epoch": 0.6618666666666667,
+      "grad_norm": 1.397933840751648,
+      "learning_rate": 6.764169815514559e-05,
+      "loss": 0.9408,
+      "step": 29784
+    },
+    {
+      "epoch": 0.6618888888888889,
+      "grad_norm": 1.7931183576583862,
+      "learning_rate": 6.763725272282729e-05,
+      "loss": 1.9449,
+      "step": 29785
+    },
+    {
+      "epoch": 0.6619111111111111,
+      "grad_norm": 1.3988561630249023,
+      "learning_rate": 6.7632807290509e-05,
+      "loss": 1.2232,
+      "step": 29786
+    },
+    {
+      "epoch": 0.6619333333333334,
+      "grad_norm": 1.7735909223556519,
+      "learning_rate": 6.762836185819072e-05,
+      "loss": 1.8449,
+      "step": 29787
+    },
+    {
+      "epoch": 0.6619555555555555,
+      "grad_norm": 1.4180495738983154,
+      "learning_rate": 6.762391642587243e-05,
+      "loss": 1.471,
+      "step": 29788
+    },
+    {
+      "epoch": 0.6619777777777778,
+      "grad_norm": 1.846642255783081,
+      "learning_rate": 6.761947099355413e-05,
+      "loss": 2.0046,
+      "step": 29789
+    },
+    {
+      "epoch": 0.662,
+      "grad_norm": 1.1563552618026733,
+      "learning_rate": 6.761502556123584e-05,
+      "loss": 0.9774,
+      "step": 29790
+    },
+    {
+      "epoch": 0.6620222222222222,
+      "grad_norm": 1.5811171531677246,
+      "learning_rate": 6.761058012891755e-05,
+      "loss": 1.7614,
+      "step": 29791
+    },
+    {
+      "epoch": 0.6620444444444444,
+      "grad_norm": 1.9264756441116333,
+      "learning_rate": 6.760613469659924e-05,
+      "loss": 1.8206,
+      "step": 29792
+    },
+    {
+      "epoch": 0.6620666666666667,
+      "grad_norm": 1.555372953414917,
+      "learning_rate": 6.760168926428095e-05,
+      "loss": 1.3343,
+      "step": 29793
+    },
+    {
+      "epoch": 0.6620888888888888,
+      "grad_norm": 2.0197927951812744,
+      "learning_rate": 6.759724383196266e-05,
+      "loss": 1.5174,
+      "step": 29794
+    },
+    {
+      "epoch": 0.6621111111111111,
+      "grad_norm": 1.9780981540679932,
+      "learning_rate": 6.759279839964437e-05,
+      "loss": 1.6895,
+      "step": 29795
+    },
+    {
+      "epoch": 0.6621333333333334,
+      "grad_norm": 1.7140216827392578,
+      "learning_rate": 6.758835296732608e-05,
+      "loss": 1.5978,
+      "step": 29796
+    },
+    {
+      "epoch": 0.6621555555555556,
+      "grad_norm": 2.1798079013824463,
+      "learning_rate": 6.758390753500779e-05,
+      "loss": 1.1335,
+      "step": 29797
+    },
+    {
+      "epoch": 0.6621777777777778,
+      "grad_norm": 1.043080449104309,
+      "learning_rate": 6.75794621026895e-05,
+      "loss": 0.719,
+      "step": 29798
+    },
+    {
+      "epoch": 0.6622,
+      "grad_norm": 2.2078845500946045,
+      "learning_rate": 6.75750166703712e-05,
+      "loss": 2.1393,
+      "step": 29799
+    },
+    {
+      "epoch": 0.6622222222222223,
+      "grad_norm": 2.1851742267608643,
+      "learning_rate": 6.75705712380529e-05,
+      "loss": 1.9053,
+      "step": 29800
+    },
+    {
+      "epoch": 0.6622444444444444,
+      "grad_norm": 2.015225648880005,
+      "learning_rate": 6.756612580573461e-05,
+      "loss": 3.1311,
+      "step": 29801
+    },
+    {
+      "epoch": 0.6622666666666667,
+      "grad_norm": 1.6868116855621338,
+      "learning_rate": 6.756168037341631e-05,
+      "loss": 2.4089,
+      "step": 29802
+    },
+    {
+      "epoch": 0.6622888888888889,
+      "grad_norm": 1.5408720970153809,
+      "learning_rate": 6.755723494109802e-05,
+      "loss": 1.1948,
+      "step": 29803
+    },
+    {
+      "epoch": 0.6623111111111111,
+      "grad_norm": 1.2142361402511597,
+      "learning_rate": 6.755278950877973e-05,
+      "loss": 1.1249,
+      "step": 29804
+    },
+    {
+      "epoch": 0.6623333333333333,
+      "grad_norm": 1.530510425567627,
+      "learning_rate": 6.754834407646144e-05,
+      "loss": 2.4555,
+      "step": 29805
+    },
+    {
+      "epoch": 0.6623555555555556,
+      "grad_norm": 1.4871829748153687,
+      "learning_rate": 6.754389864414315e-05,
+      "loss": 2.1614,
+      "step": 29806
+    },
+    {
+      "epoch": 0.6623777777777777,
+      "grad_norm": 1.8419668674468994,
+      "learning_rate": 6.753945321182486e-05,
+      "loss": 2.5486,
+      "step": 29807
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 1.5700019598007202,
+      "learning_rate": 6.753500777950657e-05,
+      "loss": 1.9956,
+      "step": 29808
+    },
+    {
+      "epoch": 0.6624222222222222,
+      "grad_norm": 1.4468845129013062,
+      "learning_rate": 6.753056234718826e-05,
+      "loss": 1.6379,
+      "step": 29809
+    },
+    {
+      "epoch": 0.6624444444444444,
+      "grad_norm": 1.574005365371704,
+      "learning_rate": 6.752611691486997e-05,
+      "loss": 2.0273,
+      "step": 29810
+    },
+    {
+      "epoch": 0.6624666666666666,
+      "grad_norm": 1.4133375883102417,
+      "learning_rate": 6.752167148255168e-05,
+      "loss": 1.6354,
+      "step": 29811
+    },
+    {
+      "epoch": 0.6624888888888889,
+      "grad_norm": 1.696982502937317,
+      "learning_rate": 6.751722605023338e-05,
+      "loss": 2.2724,
+      "step": 29812
+    },
+    {
+      "epoch": 0.6625111111111112,
+      "grad_norm": 1.4631052017211914,
+      "learning_rate": 6.751278061791509e-05,
+      "loss": 2.0235,
+      "step": 29813
+    },
+    {
+      "epoch": 0.6625333333333333,
+      "grad_norm": 1.6357436180114746,
+      "learning_rate": 6.75083351855968e-05,
+      "loss": 1.7853,
+      "step": 29814
+    },
+    {
+      "epoch": 0.6625555555555556,
+      "grad_norm": 1.5390557050704956,
+      "learning_rate": 6.750388975327851e-05,
+      "loss": 1.7458,
+      "step": 29815
+    },
+    {
+      "epoch": 0.6625777777777778,
+      "grad_norm": 1.714276909828186,
+      "learning_rate": 6.749944432096022e-05,
+      "loss": 1.8244,
+      "step": 29816
+    },
+    {
+      "epoch": 0.6626,
+      "grad_norm": 1.4015839099884033,
+      "learning_rate": 6.749499888864193e-05,
+      "loss": 1.6646,
+      "step": 29817
+    },
+    {
+      "epoch": 0.6626222222222222,
+      "grad_norm": 1.5322781801223755,
+      "learning_rate": 6.749055345632364e-05,
+      "loss": 1.5663,
+      "step": 29818
+    },
+    {
+      "epoch": 0.6626444444444445,
+      "grad_norm": 1.8195284605026245,
+      "learning_rate": 6.748610802400533e-05,
+      "loss": 1.7811,
+      "step": 29819
+    },
+    {
+      "epoch": 0.6626666666666666,
+      "grad_norm": 1.9035009145736694,
+      "learning_rate": 6.748166259168704e-05,
+      "loss": 1.8438,
+      "step": 29820
+    },
+    {
+      "epoch": 0.6626888888888889,
+      "grad_norm": 1.8950119018554688,
+      "learning_rate": 6.747721715936875e-05,
+      "loss": 1.9304,
+      "step": 29821
+    },
+    {
+      "epoch": 0.6627111111111111,
+      "grad_norm": 1.8354183435440063,
+      "learning_rate": 6.747277172705045e-05,
+      "loss": 1.8252,
+      "step": 29822
+    },
+    {
+      "epoch": 0.6627333333333333,
+      "grad_norm": 1.8345448970794678,
+      "learning_rate": 6.746832629473216e-05,
+      "loss": 1.8646,
+      "step": 29823
+    },
+    {
+      "epoch": 0.6627555555555555,
+      "grad_norm": 1.8485467433929443,
+      "learning_rate": 6.746388086241388e-05,
+      "loss": 2.2144,
+      "step": 29824
+    },
+    {
+      "epoch": 0.6627777777777778,
+      "grad_norm": 1.7133246660232544,
+      "learning_rate": 6.745943543009559e-05,
+      "loss": 1.832,
+      "step": 29825
+    },
+    {
+      "epoch": 0.6628,
+      "grad_norm": 1.6095284223556519,
+      "learning_rate": 6.745498999777729e-05,
+      "loss": 1.5356,
+      "step": 29826
+    },
+    {
+      "epoch": 0.6628222222222222,
+      "grad_norm": 1.4461475610733032,
+      "learning_rate": 6.7450544565459e-05,
+      "loss": 1.6122,
+      "step": 29827
+    },
+    {
+      "epoch": 0.6628444444444445,
+      "grad_norm": 1.8822609186172485,
+      "learning_rate": 6.74460991331407e-05,
+      "loss": 1.7766,
+      "step": 29828
+    },
+    {
+      "epoch": 0.6628666666666667,
+      "grad_norm": 1.944409966468811,
+      "learning_rate": 6.74416537008224e-05,
+      "loss": 1.8874,
+      "step": 29829
+    },
+    {
+      "epoch": 0.6628888888888889,
+      "grad_norm": 1.6179362535476685,
+      "learning_rate": 6.743720826850411e-05,
+      "loss": 1.8056,
+      "step": 29830
+    },
+    {
+      "epoch": 0.6629111111111111,
+      "grad_norm": 2.0456881523132324,
+      "learning_rate": 6.743276283618582e-05,
+      "loss": 1.6853,
+      "step": 29831
+    },
+    {
+      "epoch": 0.6629333333333334,
+      "grad_norm": 1.715853214263916,
+      "learning_rate": 6.742831740386753e-05,
+      "loss": 1.886,
+      "step": 29832
+    },
+    {
+      "epoch": 0.6629555555555555,
+      "grad_norm": 1.9937889575958252,
+      "learning_rate": 6.742387197154924e-05,
+      "loss": 1.998,
+      "step": 29833
+    },
+    {
+      "epoch": 0.6629777777777778,
+      "grad_norm": 1.687606692314148,
+      "learning_rate": 6.741942653923095e-05,
+      "loss": 1.9692,
+      "step": 29834
+    },
+    {
+      "epoch": 0.663,
+      "grad_norm": 1.661903977394104,
+      "learning_rate": 6.741498110691266e-05,
+      "loss": 1.675,
+      "step": 29835
+    },
+    {
+      "epoch": 0.6630222222222222,
+      "grad_norm": 2.170902729034424,
+      "learning_rate": 6.741053567459436e-05,
+      "loss": 2.1342,
+      "step": 29836
+    },
+    {
+      "epoch": 0.6630444444444444,
+      "grad_norm": 0.8865088820457458,
+      "learning_rate": 6.740609024227607e-05,
+      "loss": 0.7062,
+      "step": 29837
+    },
+    {
+      "epoch": 0.6630666666666667,
+      "grad_norm": 2.08332896232605,
+      "learning_rate": 6.740164480995778e-05,
+      "loss": 2.2277,
+      "step": 29838
+    },
+    {
+      "epoch": 0.6630888888888888,
+      "grad_norm": 1.8963425159454346,
+      "learning_rate": 6.739719937763947e-05,
+      "loss": 2.13,
+      "step": 29839
+    },
+    {
+      "epoch": 0.6631111111111111,
+      "grad_norm": 1.5987082719802856,
+      "learning_rate": 6.739275394532118e-05,
+      "loss": 1.5478,
+      "step": 29840
+    },
+    {
+      "epoch": 0.6631333333333334,
+      "grad_norm": 1.8892338275909424,
+      "learning_rate": 6.738830851300289e-05,
+      "loss": 2.3368,
+      "step": 29841
+    },
+    {
+      "epoch": 0.6631555555555556,
+      "grad_norm": 1.368084192276001,
+      "learning_rate": 6.73838630806846e-05,
+      "loss": 1.2946,
+      "step": 29842
+    },
+    {
+      "epoch": 0.6631777777777778,
+      "grad_norm": 2.3991076946258545,
+      "learning_rate": 6.737941764836631e-05,
+      "loss": 2.455,
+      "step": 29843
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 1.8040879964828491,
+      "learning_rate": 6.737497221604802e-05,
+      "loss": 1.7914,
+      "step": 29844
+    },
+    {
+      "epoch": 0.6632222222222223,
+      "grad_norm": 1.668006181716919,
+      "learning_rate": 6.737052678372973e-05,
+      "loss": 1.7811,
+      "step": 29845
+    },
+    {
+      "epoch": 0.6632444444444444,
+      "grad_norm": 1.5654646158218384,
+      "learning_rate": 6.736608135141143e-05,
+      "loss": 1.384,
+      "step": 29846
+    },
+    {
+      "epoch": 0.6632666666666667,
+      "grad_norm": 2.4598138332366943,
+      "learning_rate": 6.736163591909313e-05,
+      "loss": 1.9209,
+      "step": 29847
+    },
+    {
+      "epoch": 0.6632888888888889,
+      "grad_norm": 1.0733261108398438,
+      "learning_rate": 6.735719048677484e-05,
+      "loss": 0.789,
+      "step": 29848
+    },
+    {
+      "epoch": 0.6633111111111111,
+      "grad_norm": 1.8876014947891235,
+      "learning_rate": 6.735274505445654e-05,
+      "loss": 1.9376,
+      "step": 29849
+    },
+    {
+      "epoch": 0.6633333333333333,
+      "grad_norm": 2.126352548599243,
+      "learning_rate": 6.734829962213825e-05,
+      "loss": 1.7392,
+      "step": 29850
+    },
+    {
+      "epoch": 0.6633555555555556,
+      "grad_norm": 0.9994049668312073,
+      "learning_rate": 6.734385418981996e-05,
+      "loss": 1.4126,
+      "step": 29851
+    },
+    {
+      "epoch": 0.6633777777777777,
+      "grad_norm": 1.020012378692627,
+      "learning_rate": 6.733940875750167e-05,
+      "loss": 1.054,
+      "step": 29852
+    },
+    {
+      "epoch": 0.6634,
+      "grad_norm": 1.5305405855178833,
+      "learning_rate": 6.733496332518338e-05,
+      "loss": 2.3882,
+      "step": 29853
+    },
+    {
+      "epoch": 0.6634222222222222,
+      "grad_norm": 1.1586368083953857,
+      "learning_rate": 6.733051789286509e-05,
+      "loss": 1.6271,
+      "step": 29854
+    },
+    {
+      "epoch": 0.6634444444444444,
+      "grad_norm": 1.5261671543121338,
+      "learning_rate": 6.73260724605468e-05,
+      "loss": 1.771,
+      "step": 29855
+    },
+    {
+      "epoch": 0.6634666666666666,
+      "grad_norm": 1.3771371841430664,
+      "learning_rate": 6.73216270282285e-05,
+      "loss": 1.4175,
+      "step": 29856
+    },
+    {
+      "epoch": 0.6634888888888889,
+      "grad_norm": 1.535698413848877,
+      "learning_rate": 6.73171815959102e-05,
+      "loss": 1.0387,
+      "step": 29857
+    },
+    {
+      "epoch": 0.6635111111111112,
+      "grad_norm": 1.5857796669006348,
+      "learning_rate": 6.731273616359191e-05,
+      "loss": 1.9065,
+      "step": 29858
+    },
+    {
+      "epoch": 0.6635333333333333,
+      "grad_norm": 1.5167961120605469,
+      "learning_rate": 6.730829073127361e-05,
+      "loss": 1.8251,
+      "step": 29859
+    },
+    {
+      "epoch": 0.6635555555555556,
+      "grad_norm": 1.5984530448913574,
+      "learning_rate": 6.730384529895532e-05,
+      "loss": 1.9663,
+      "step": 29860
+    },
+    {
+      "epoch": 0.6635777777777778,
+      "grad_norm": 1.4273908138275146,
+      "learning_rate": 6.729939986663704e-05,
+      "loss": 1.9222,
+      "step": 29861
+    },
+    {
+      "epoch": 0.6636,
+      "grad_norm": 1.4247217178344727,
+      "learning_rate": 6.729495443431874e-05,
+      "loss": 1.7142,
+      "step": 29862
+    },
+    {
+      "epoch": 0.6636222222222222,
+      "grad_norm": 1.9986554384231567,
+      "learning_rate": 6.729050900200045e-05,
+      "loss": 2.0371,
+      "step": 29863
+    },
+    {
+      "epoch": 0.6636444444444445,
+      "grad_norm": 1.9894723892211914,
+      "learning_rate": 6.728606356968216e-05,
+      "loss": 2.0388,
+      "step": 29864
+    },
+    {
+      "epoch": 0.6636666666666666,
+      "grad_norm": 2.0761454105377197,
+      "learning_rate": 6.728161813736387e-05,
+      "loss": 1.1724,
+      "step": 29865
+    },
+    {
+      "epoch": 0.6636888888888889,
+      "grad_norm": 1.590334415435791,
+      "learning_rate": 6.727717270504556e-05,
+      "loss": 1.8133,
+      "step": 29866
+    },
+    {
+      "epoch": 0.6637111111111111,
+      "grad_norm": 1.7695194482803345,
+      "learning_rate": 6.727272727272727e-05,
+      "loss": 1.9001,
+      "step": 29867
+    },
+    {
+      "epoch": 0.6637333333333333,
+      "grad_norm": 1.504016399383545,
+      "learning_rate": 6.726828184040898e-05,
+      "loss": 2.0085,
+      "step": 29868
+    },
+    {
+      "epoch": 0.6637555555555555,
+      "grad_norm": 1.9227676391601562,
+      "learning_rate": 6.726383640809069e-05,
+      "loss": 1.9244,
+      "step": 29869
+    },
+    {
+      "epoch": 0.6637777777777778,
+      "grad_norm": 1.641811728477478,
+      "learning_rate": 6.72593909757724e-05,
+      "loss": 1.7412,
+      "step": 29870
+    },
+    {
+      "epoch": 0.6638,
+      "grad_norm": 2.3565618991851807,
+      "learning_rate": 6.725494554345411e-05,
+      "loss": 1.8609,
+      "step": 29871
+    },
+    {
+      "epoch": 0.6638222222222222,
+      "grad_norm": 2.0382018089294434,
+      "learning_rate": 6.725050011113581e-05,
+      "loss": 2.0848,
+      "step": 29872
+    },
+    {
+      "epoch": 0.6638444444444445,
+      "grad_norm": 2.107382297515869,
+      "learning_rate": 6.724605467881752e-05,
+      "loss": 1.7436,
+      "step": 29873
+    },
+    {
+      "epoch": 0.6638666666666667,
+      "grad_norm": 1.684561848640442,
+      "learning_rate": 6.724160924649923e-05,
+      "loss": 1.8383,
+      "step": 29874
+    },
+    {
+      "epoch": 0.6638888888888889,
+      "grad_norm": 1.0685566663742065,
+      "learning_rate": 6.723716381418094e-05,
+      "loss": 0.9968,
+      "step": 29875
+    },
+    {
+      "epoch": 0.6639111111111111,
+      "grad_norm": 2.0837154388427734,
+      "learning_rate": 6.723271838186263e-05,
+      "loss": 2.1355,
+      "step": 29876
+    },
+    {
+      "epoch": 0.6639333333333334,
+      "grad_norm": 1.7028639316558838,
+      "learning_rate": 6.722827294954434e-05,
+      "loss": 2.1359,
+      "step": 29877
+    },
+    {
+      "epoch": 0.6639555555555555,
+      "grad_norm": 1.3876352310180664,
+      "learning_rate": 6.722382751722605e-05,
+      "loss": 1.3667,
+      "step": 29878
+    },
+    {
+      "epoch": 0.6639777777777778,
+      "grad_norm": 1.7416679859161377,
+      "learning_rate": 6.721938208490776e-05,
+      "loss": 1.8754,
+      "step": 29879
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 1.6241660118103027,
+      "learning_rate": 6.721493665258947e-05,
+      "loss": 1.6207,
+      "step": 29880
+    },
+    {
+      "epoch": 0.6640222222222222,
+      "grad_norm": 1.7644826173782349,
+      "learning_rate": 6.721049122027118e-05,
+      "loss": 1.7229,
+      "step": 29881
+    },
+    {
+      "epoch": 0.6640444444444444,
+      "grad_norm": 1.7775423526763916,
+      "learning_rate": 6.720604578795289e-05,
+      "loss": 1.9053,
+      "step": 29882
+    },
+    {
+      "epoch": 0.6640666666666667,
+      "grad_norm": 1.7751245498657227,
+      "learning_rate": 6.720160035563459e-05,
+      "loss": 1.8768,
+      "step": 29883
+    },
+    {
+      "epoch": 0.6640888888888888,
+      "grad_norm": 2.0301311016082764,
+      "learning_rate": 6.71971549233163e-05,
+      "loss": 2.4465,
+      "step": 29884
+    },
+    {
+      "epoch": 0.6641111111111111,
+      "grad_norm": 1.7487518787384033,
+      "learning_rate": 6.7192709490998e-05,
+      "loss": 1.6696,
+      "step": 29885
+    },
+    {
+      "epoch": 0.6641333333333334,
+      "grad_norm": 2.3853297233581543,
+      "learning_rate": 6.71882640586797e-05,
+      "loss": 2.2918,
+      "step": 29886
+    },
+    {
+      "epoch": 0.6641555555555556,
+      "grad_norm": 2.0184779167175293,
+      "learning_rate": 6.718381862636141e-05,
+      "loss": 1.7051,
+      "step": 29887
+    },
+    {
+      "epoch": 0.6641777777777778,
+      "grad_norm": 1.8643485307693481,
+      "learning_rate": 6.717937319404312e-05,
+      "loss": 1.9038,
+      "step": 29888
+    },
+    {
+      "epoch": 0.6642,
+      "grad_norm": 2.1959586143493652,
+      "learning_rate": 6.717492776172483e-05,
+      "loss": 1.5555,
+      "step": 29889
+    },
+    {
+      "epoch": 0.6642222222222223,
+      "grad_norm": 1.7016730308532715,
+      "learning_rate": 6.717048232940654e-05,
+      "loss": 1.5396,
+      "step": 29890
+    },
+    {
+      "epoch": 0.6642444444444444,
+      "grad_norm": 2.1555588245391846,
+      "learning_rate": 6.716603689708825e-05,
+      "loss": 2.1698,
+      "step": 29891
+    },
+    {
+      "epoch": 0.6642666666666667,
+      "grad_norm": 1.6451393365859985,
+      "learning_rate": 6.716159146476996e-05,
+      "loss": 1.3959,
+      "step": 29892
+    },
+    {
+      "epoch": 0.6642888888888889,
+      "grad_norm": 1.5219990015029907,
+      "learning_rate": 6.715714603245166e-05,
+      "loss": 1.5174,
+      "step": 29893
+    },
+    {
+      "epoch": 0.6643111111111111,
+      "grad_norm": 1.9876725673675537,
+      "learning_rate": 6.715270060013336e-05,
+      "loss": 2.0229,
+      "step": 29894
+    },
+    {
+      "epoch": 0.6643333333333333,
+      "grad_norm": 1.7877013683319092,
+      "learning_rate": 6.714825516781507e-05,
+      "loss": 1.8763,
+      "step": 29895
+    },
+    {
+      "epoch": 0.6643555555555556,
+      "grad_norm": 2.0413553714752197,
+      "learning_rate": 6.714380973549677e-05,
+      "loss": 1.671,
+      "step": 29896
+    },
+    {
+      "epoch": 0.6643777777777777,
+      "grad_norm": 2.1173155307769775,
+      "learning_rate": 6.713936430317848e-05,
+      "loss": 2.178,
+      "step": 29897
+    },
+    {
+      "epoch": 0.6644,
+      "grad_norm": 1.778635859489441,
+      "learning_rate": 6.71349188708602e-05,
+      "loss": 1.5821,
+      "step": 29898
+    },
+    {
+      "epoch": 0.6644222222222222,
+      "grad_norm": 1.8207401037216187,
+      "learning_rate": 6.71304734385419e-05,
+      "loss": 1.6871,
+      "step": 29899
+    },
+    {
+      "epoch": 0.6644444444444444,
+      "grad_norm": 1.787549376487732,
+      "learning_rate": 6.712602800622361e-05,
+      "loss": 1.468,
+      "step": 29900
+    },
+    {
+      "epoch": 0.6644666666666666,
+      "grad_norm": 1.3397949934005737,
+      "learning_rate": 6.712158257390532e-05,
+      "loss": 2.3533,
+      "step": 29901
+    },
+    {
+      "epoch": 0.6644888888888889,
+      "grad_norm": 1.6535183191299438,
+      "learning_rate": 6.711713714158703e-05,
+      "loss": 2.1798,
+      "step": 29902
+    },
+    {
+      "epoch": 0.6645111111111112,
+      "grad_norm": 1.3034520149230957,
+      "learning_rate": 6.711269170926872e-05,
+      "loss": 2.2776,
+      "step": 29903
+    },
+    {
+      "epoch": 0.6645333333333333,
+      "grad_norm": 1.4036054611206055,
+      "learning_rate": 6.710824627695043e-05,
+      "loss": 2.2437,
+      "step": 29904
+    },
+    {
+      "epoch": 0.6645555555555556,
+      "grad_norm": 1.8223963975906372,
+      "learning_rate": 6.710380084463214e-05,
+      "loss": 2.2021,
+      "step": 29905
+    },
+    {
+      "epoch": 0.6645777777777778,
+      "grad_norm": 2.307182788848877,
+      "learning_rate": 6.709935541231385e-05,
+      "loss": 1.7815,
+      "step": 29906
+    },
+    {
+      "epoch": 0.6646,
+      "grad_norm": 1.9300739765167236,
+      "learning_rate": 6.709490997999556e-05,
+      "loss": 2.2553,
+      "step": 29907
+    },
+    {
+      "epoch": 0.6646222222222222,
+      "grad_norm": 1.8276855945587158,
+      "learning_rate": 6.709046454767727e-05,
+      "loss": 2.4208,
+      "step": 29908
+    },
+    {
+      "epoch": 0.6646444444444445,
+      "grad_norm": 2.037367820739746,
+      "learning_rate": 6.708601911535897e-05,
+      "loss": 2.1848,
+      "step": 29909
+    },
+    {
+      "epoch": 0.6646666666666666,
+      "grad_norm": 1.8194398880004883,
+      "learning_rate": 6.708157368304068e-05,
+      "loss": 2.4673,
+      "step": 29910
+    },
+    {
+      "epoch": 0.6646888888888889,
+      "grad_norm": 1.8202264308929443,
+      "learning_rate": 6.707712825072239e-05,
+      "loss": 2.5849,
+      "step": 29911
+    },
+    {
+      "epoch": 0.6647111111111111,
+      "grad_norm": 1.4894503355026245,
+      "learning_rate": 6.70726828184041e-05,
+      "loss": 1.5413,
+      "step": 29912
+    },
+    {
+      "epoch": 0.6647333333333333,
+      "grad_norm": 1.4747028350830078,
+      "learning_rate": 6.70682373860858e-05,
+      "loss": 1.8897,
+      "step": 29913
+    },
+    {
+      "epoch": 0.6647555555555555,
+      "grad_norm": 1.6377867460250854,
+      "learning_rate": 6.70637919537675e-05,
+      "loss": 2.2015,
+      "step": 29914
+    },
+    {
+      "epoch": 0.6647777777777778,
+      "grad_norm": 1.5551769733428955,
+      "learning_rate": 6.705934652144921e-05,
+      "loss": 2.0039,
+      "step": 29915
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 1.4956742525100708,
+      "learning_rate": 6.705490108913092e-05,
+      "loss": 1.9786,
+      "step": 29916
+    },
+    {
+      "epoch": 0.6648222222222222,
+      "grad_norm": 1.1861640214920044,
+      "learning_rate": 6.705045565681263e-05,
+      "loss": 1.133,
+      "step": 29917
+    },
+    {
+      "epoch": 0.6648444444444445,
+      "grad_norm": 1.2967723608016968,
+      "learning_rate": 6.704601022449434e-05,
+      "loss": 1.218,
+      "step": 29918
+    },
+    {
+      "epoch": 0.6648666666666667,
+      "grad_norm": 1.7160927057266235,
+      "learning_rate": 6.704156479217604e-05,
+      "loss": 2.0236,
+      "step": 29919
+    },
+    {
+      "epoch": 0.6648888888888889,
+      "grad_norm": 2.133582353591919,
+      "learning_rate": 6.703711935985775e-05,
+      "loss": 1.696,
+      "step": 29920
+    },
+    {
+      "epoch": 0.6649111111111111,
+      "grad_norm": 1.6214841604232788,
+      "learning_rate": 6.703267392753946e-05,
+      "loss": 2.2105,
+      "step": 29921
+    },
+    {
+      "epoch": 0.6649333333333334,
+      "grad_norm": 1.700264811515808,
+      "learning_rate": 6.702822849522117e-05,
+      "loss": 2.034,
+      "step": 29922
+    },
+    {
+      "epoch": 0.6649555555555555,
+      "grad_norm": 1.8656233549118042,
+      "learning_rate": 6.702378306290286e-05,
+      "loss": 1.8581,
+      "step": 29923
+    },
+    {
+      "epoch": 0.6649777777777778,
+      "grad_norm": 1.5610407590866089,
+      "learning_rate": 6.701933763058457e-05,
+      "loss": 1.808,
+      "step": 29924
+    },
+    {
+      "epoch": 0.665,
+      "grad_norm": 1.8312547206878662,
+      "learning_rate": 6.701489219826628e-05,
+      "loss": 2.0996,
+      "step": 29925
+    },
+    {
+      "epoch": 0.6650222222222222,
+      "grad_norm": 1.595229983329773,
+      "learning_rate": 6.701044676594799e-05,
+      "loss": 1.597,
+      "step": 29926
+    },
+    {
+      "epoch": 0.6650444444444444,
+      "grad_norm": 1.6952110528945923,
+      "learning_rate": 6.70060013336297e-05,
+      "loss": 1.5797,
+      "step": 29927
+    },
+    {
+      "epoch": 0.6650666666666667,
+      "grad_norm": 1.8080729246139526,
+      "learning_rate": 6.700155590131141e-05,
+      "loss": 2.0955,
+      "step": 29928
+    },
+    {
+      "epoch": 0.6650888888888888,
+      "grad_norm": 1.844958782196045,
+      "learning_rate": 6.69971104689931e-05,
+      "loss": 2.0051,
+      "step": 29929
+    },
+    {
+      "epoch": 0.6651111111111111,
+      "grad_norm": 1.6383330821990967,
+      "learning_rate": 6.699266503667482e-05,
+      "loss": 1.7088,
+      "step": 29930
+    },
+    {
+      "epoch": 0.6651333333333334,
+      "grad_norm": 1.9203208684921265,
+      "learning_rate": 6.698821960435653e-05,
+      "loss": 2.2203,
+      "step": 29931
+    },
+    {
+      "epoch": 0.6651555555555556,
+      "grad_norm": 1.5951393842697144,
+      "learning_rate": 6.698377417203824e-05,
+      "loss": 1.7076,
+      "step": 29932
+    },
+    {
+      "epoch": 0.6651777777777778,
+      "grad_norm": 1.5344406366348267,
+      "learning_rate": 6.697932873971993e-05,
+      "loss": 1.6437,
+      "step": 29933
+    },
+    {
+      "epoch": 0.6652,
+      "grad_norm": 1.807194709777832,
+      "learning_rate": 6.697488330740164e-05,
+      "loss": 2.2013,
+      "step": 29934
+    },
+    {
+      "epoch": 0.6652222222222223,
+      "grad_norm": 1.5656250715255737,
+      "learning_rate": 6.697043787508336e-05,
+      "loss": 1.6515,
+      "step": 29935
+    },
+    {
+      "epoch": 0.6652444444444444,
+      "grad_norm": 2.082768678665161,
+      "learning_rate": 6.696599244276506e-05,
+      "loss": 2.3978,
+      "step": 29936
+    },
+    {
+      "epoch": 0.6652666666666667,
+      "grad_norm": 1.8907321691513062,
+      "learning_rate": 6.696154701044677e-05,
+      "loss": 1.5345,
+      "step": 29937
+    },
+    {
+      "epoch": 0.6652888888888889,
+      "grad_norm": 2.222169876098633,
+      "learning_rate": 6.695710157812848e-05,
+      "loss": 1.7997,
+      "step": 29938
+    },
+    {
+      "epoch": 0.6653111111111111,
+      "grad_norm": 1.9868658781051636,
+      "learning_rate": 6.695265614581019e-05,
+      "loss": 2.0057,
+      "step": 29939
+    },
+    {
+      "epoch": 0.6653333333333333,
+      "grad_norm": 1.4449933767318726,
+      "learning_rate": 6.694821071349189e-05,
+      "loss": 1.4912,
+      "step": 29940
+    },
+    {
+      "epoch": 0.6653555555555556,
+      "grad_norm": 1.9717739820480347,
+      "learning_rate": 6.69437652811736e-05,
+      "loss": 2.2573,
+      "step": 29941
+    },
+    {
+      "epoch": 0.6653777777777777,
+      "grad_norm": 2.050900936126709,
+      "learning_rate": 6.69393198488553e-05,
+      "loss": 1.6612,
+      "step": 29942
+    },
+    {
+      "epoch": 0.6654,
+      "grad_norm": 1.7138173580169678,
+      "learning_rate": 6.693487441653701e-05,
+      "loss": 1.7574,
+      "step": 29943
+    },
+    {
+      "epoch": 0.6654222222222222,
+      "grad_norm": 1.945564866065979,
+      "learning_rate": 6.693042898421872e-05,
+      "loss": 2.1009,
+      "step": 29944
+    },
+    {
+      "epoch": 0.6654444444444444,
+      "grad_norm": 1.7058165073394775,
+      "learning_rate": 6.692598355190043e-05,
+      "loss": 1.7286,
+      "step": 29945
+    },
+    {
+      "epoch": 0.6654666666666667,
+      "grad_norm": 1.0451241731643677,
+      "learning_rate": 6.692153811958213e-05,
+      "loss": 0.7271,
+      "step": 29946
+    },
+    {
+      "epoch": 0.6654888888888889,
+      "grad_norm": 1.9874522686004639,
+      "learning_rate": 6.691709268726384e-05,
+      "loss": 2.2219,
+      "step": 29947
+    },
+    {
+      "epoch": 0.6655111111111112,
+      "grad_norm": 1.9588372707366943,
+      "learning_rate": 6.691264725494555e-05,
+      "loss": 1.9927,
+      "step": 29948
+    },
+    {
+      "epoch": 0.6655333333333333,
+      "grad_norm": 2.0496973991394043,
+      "learning_rate": 6.690820182262726e-05,
+      "loss": 2.1667,
+      "step": 29949
+    },
+    {
+      "epoch": 0.6655555555555556,
+      "grad_norm": 1.554182767868042,
+      "learning_rate": 6.690375639030895e-05,
+      "loss": 0.7582,
+      "step": 29950
+    },
+    {
+      "epoch": 0.6655777777777778,
+      "grad_norm": 0.9341526031494141,
+      "learning_rate": 6.689931095799066e-05,
+      "loss": 1.1776,
+      "step": 29951
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 1.5718581676483154,
+      "learning_rate": 6.689486552567237e-05,
+      "loss": 2.8773,
+      "step": 29952
+    },
+    {
+      "epoch": 0.6656222222222222,
+      "grad_norm": 1.5068167448043823,
+      "learning_rate": 6.689042009335408e-05,
+      "loss": 2.6586,
+      "step": 29953
+    },
+    {
+      "epoch": 0.6656444444444445,
+      "grad_norm": 1.533138394355774,
+      "learning_rate": 6.688597466103579e-05,
+      "loss": 2.4295,
+      "step": 29954
+    },
+    {
+      "epoch": 0.6656666666666666,
+      "grad_norm": 1.6191515922546387,
+      "learning_rate": 6.68815292287175e-05,
+      "loss": 2.2264,
+      "step": 29955
+    },
+    {
+      "epoch": 0.6656888888888889,
+      "grad_norm": 1.8643866777420044,
+      "learning_rate": 6.68770837963992e-05,
+      "loss": 2.3549,
+      "step": 29956
+    },
+    {
+      "epoch": 0.6657111111111111,
+      "grad_norm": 2.198772430419922,
+      "learning_rate": 6.687263836408091e-05,
+      "loss": 1.9581,
+      "step": 29957
+    },
+    {
+      "epoch": 0.6657333333333333,
+      "grad_norm": 1.8021020889282227,
+      "learning_rate": 6.686819293176262e-05,
+      "loss": 2.2015,
+      "step": 29958
+    },
+    {
+      "epoch": 0.6657555555555555,
+      "grad_norm": 1.3431627750396729,
+      "learning_rate": 6.686374749944433e-05,
+      "loss": 1.7711,
+      "step": 29959
+    },
+    {
+      "epoch": 0.6657777777777778,
+      "grad_norm": 1.6012687683105469,
+      "learning_rate": 6.685930206712602e-05,
+      "loss": 1.8353,
+      "step": 29960
+    },
+    {
+      "epoch": 0.6658,
+      "grad_norm": 1.1746835708618164,
+      "learning_rate": 6.685485663480773e-05,
+      "loss": 0.8971,
+      "step": 29961
+    },
+    {
+      "epoch": 0.6658222222222222,
+      "grad_norm": 1.4977556467056274,
+      "learning_rate": 6.685041120248944e-05,
+      "loss": 1.6268,
+      "step": 29962
+    },
+    {
+      "epoch": 0.6658444444444445,
+      "grad_norm": 1.93625009059906,
+      "learning_rate": 6.684596577017115e-05,
+      "loss": 1.6674,
+      "step": 29963
+    },
+    {
+      "epoch": 0.6658666666666667,
+      "grad_norm": 1.4827030897140503,
+      "learning_rate": 6.684152033785286e-05,
+      "loss": 1.3579,
+      "step": 29964
+    },
+    {
+      "epoch": 0.6658888888888889,
+      "grad_norm": 1.4736237525939941,
+      "learning_rate": 6.683707490553457e-05,
+      "loss": 1.5683,
+      "step": 29965
+    },
+    {
+      "epoch": 0.6659111111111111,
+      "grad_norm": 1.9679960012435913,
+      "learning_rate": 6.683262947321627e-05,
+      "loss": 1.6845,
+      "step": 29966
+    },
+    {
+      "epoch": 0.6659333333333334,
+      "grad_norm": 1.2724868059158325,
+      "learning_rate": 6.682818404089798e-05,
+      "loss": 1.2223,
+      "step": 29967
+    },
+    {
+      "epoch": 0.6659555555555555,
+      "grad_norm": 1.369015097618103,
+      "learning_rate": 6.682373860857969e-05,
+      "loss": 1.5704,
+      "step": 29968
+    },
+    {
+      "epoch": 0.6659777777777778,
+      "grad_norm": 1.5374850034713745,
+      "learning_rate": 6.68192931762614e-05,
+      "loss": 2.1142,
+      "step": 29969
+    },
+    {
+      "epoch": 0.666,
+      "grad_norm": 0.48563823103904724,
+      "learning_rate": 6.681484774394309e-05,
+      "loss": 0.0228,
+      "step": 29970
+    },
+    {
+      "epoch": 0.6660222222222222,
+      "grad_norm": 1.7181105613708496,
+      "learning_rate": 6.68104023116248e-05,
+      "loss": 1.9205,
+      "step": 29971
+    },
+    {
+      "epoch": 0.6660444444444444,
+      "grad_norm": 1.7128758430480957,
+      "learning_rate": 6.680595687930653e-05,
+      "loss": 2.1458,
+      "step": 29972
+    },
+    {
+      "epoch": 0.6660666666666667,
+      "grad_norm": 1.8142510652542114,
+      "learning_rate": 6.680151144698822e-05,
+      "loss": 1.9488,
+      "step": 29973
+    },
+    {
+      "epoch": 0.6660888888888888,
+      "grad_norm": 1.813297986984253,
+      "learning_rate": 6.679706601466993e-05,
+      "loss": 2.392,
+      "step": 29974
+    },
+    {
+      "epoch": 0.6661111111111111,
+      "grad_norm": 1.5008163452148438,
+      "learning_rate": 6.679262058235164e-05,
+      "loss": 1.7734,
+      "step": 29975
+    },
+    {
+      "epoch": 0.6661333333333334,
+      "grad_norm": 2.0730113983154297,
+      "learning_rate": 6.678817515003334e-05,
+      "loss": 1.8511,
+      "step": 29976
+    },
+    {
+      "epoch": 0.6661555555555555,
+      "grad_norm": 2.2350423336029053,
+      "learning_rate": 6.678372971771505e-05,
+      "loss": 1.9735,
+      "step": 29977
+    },
+    {
+      "epoch": 0.6661777777777778,
+      "grad_norm": 1.4494450092315674,
+      "learning_rate": 6.677928428539676e-05,
+      "loss": 1.6645,
+      "step": 29978
+    },
+    {
+      "epoch": 0.6662,
+      "grad_norm": 1.9463860988616943,
+      "learning_rate": 6.677483885307847e-05,
+      "loss": 1.9717,
+      "step": 29979
+    },
+    {
+      "epoch": 0.6662222222222223,
+      "grad_norm": 1.710606336593628,
+      "learning_rate": 6.677039342076018e-05,
+      "loss": 1.9876,
+      "step": 29980
+    },
+    {
+      "epoch": 0.6662444444444444,
+      "grad_norm": 1.8882859945297241,
+      "learning_rate": 6.676594798844188e-05,
+      "loss": 1.5495,
+      "step": 29981
+    },
+    {
+      "epoch": 0.6662666666666667,
+      "grad_norm": 1.8677129745483398,
+      "learning_rate": 6.67615025561236e-05,
+      "loss": 2.0978,
+      "step": 29982
+    },
+    {
+      "epoch": 0.6662888888888889,
+      "grad_norm": 1.759299397468567,
+      "learning_rate": 6.675705712380529e-05,
+      "loss": 1.9628,
+      "step": 29983
+    },
+    {
+      "epoch": 0.6663111111111111,
+      "grad_norm": 2.7281928062438965,
+      "learning_rate": 6.6752611691487e-05,
+      "loss": 2.2807,
+      "step": 29984
+    },
+    {
+      "epoch": 0.6663333333333333,
+      "grad_norm": 1.6504846811294556,
+      "learning_rate": 6.674816625916871e-05,
+      "loss": 1.9276,
+      "step": 29985
+    },
+    {
+      "epoch": 0.6663555555555556,
+      "grad_norm": 1.9323383569717407,
+      "learning_rate": 6.674372082685042e-05,
+      "loss": 2.0836,
+      "step": 29986
+    },
+    {
+      "epoch": 0.6663777777777777,
+      "grad_norm": 1.8040101528167725,
+      "learning_rate": 6.673927539453212e-05,
+      "loss": 2.2852,
+      "step": 29987
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 1.518248200416565,
+      "learning_rate": 6.673482996221383e-05,
+      "loss": 1.6423,
+      "step": 29988
+    },
+    {
+      "epoch": 0.6664222222222222,
+      "grad_norm": 1.5416886806488037,
+      "learning_rate": 6.673038452989553e-05,
+      "loss": 1.4105,
+      "step": 29989
+    },
+    {
+      "epoch": 0.6664444444444444,
+      "grad_norm": 1.7516173124313354,
+      "learning_rate": 6.672593909757724e-05,
+      "loss": 1.8827,
+      "step": 29990
+    },
+    {
+      "epoch": 0.6664666666666667,
+      "grad_norm": 1.6546638011932373,
+      "learning_rate": 6.672149366525895e-05,
+      "loss": 1.7423,
+      "step": 29991
+    },
+    {
+      "epoch": 0.6664888888888889,
+      "grad_norm": 1.5564340353012085,
+      "learning_rate": 6.671704823294066e-05,
+      "loss": 1.6448,
+      "step": 29992
+    },
+    {
+      "epoch": 0.6665111111111112,
+      "grad_norm": 1.8928364515304565,
+      "learning_rate": 6.671260280062236e-05,
+      "loss": 1.7723,
+      "step": 29993
+    },
+    {
+      "epoch": 0.6665333333333333,
+      "grad_norm": 1.846963882446289,
+      "learning_rate": 6.670815736830407e-05,
+      "loss": 2.0836,
+      "step": 29994
+    },
+    {
+      "epoch": 0.6665555555555556,
+      "grad_norm": 1.6592938899993896,
+      "learning_rate": 6.670371193598578e-05,
+      "loss": 1.7189,
+      "step": 29995
+    },
+    {
+      "epoch": 0.6665777777777778,
+      "grad_norm": 1.7135471105575562,
+      "learning_rate": 6.669926650366749e-05,
+      "loss": 1.8557,
+      "step": 29996
+    },
+    {
+      "epoch": 0.6666,
+      "grad_norm": 1.8523329496383667,
+      "learning_rate": 6.669482107134918e-05,
+      "loss": 1.9346,
+      "step": 29997
+    },
+    {
+      "epoch": 0.6666222222222222,
+      "grad_norm": 1.719085693359375,
+      "learning_rate": 6.66903756390309e-05,
+      "loss": 1.7526,
+      "step": 29998
+    },
+    {
+      "epoch": 0.6666444444444445,
+      "grad_norm": 2.335409164428711,
+      "learning_rate": 6.66859302067126e-05,
+      "loss": 1.9978,
+      "step": 29999
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 1.673316240310669,
+      "learning_rate": 6.668148477439431e-05,
+      "loss": 1.5546,
+      "step": 30000
     }
   ],
   "logging_steps": 1,
@@ -175042,7 +210050,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.470864784640553e+17,
+  "total_flos": 2.96467995936897e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null