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": 0.3152608573436737,
   "best_model_checkpoint": "./w2v-bert-2.0-igbo_naijavoices_100h/checkpoint-8000",
-  "epoch": 37.93103448275862,
+  "epoch": 41.37931034482759,
   "eval_steps": 1000,
-  "global_step": 11000,
+  "global_step": 12000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -77117,6 +77117,7016 @@
       "eval_steps_per_second": 0.169,
       "eval_wer": 0.2996894409937888,
       "step": 11000
+    },
+    {
+      "epoch": 37.93448275862069,
+      "grad_norm": 2.244279146194458,
+      "learning_rate": 2.8277701149425287e-05,
+      "loss": 0.1306,
+      "step": 11001
+    },
+    {
+      "epoch": 37.93793103448276,
+      "grad_norm": 0.8733891844749451,
+      "learning_rate": 2.8277241379310346e-05,
+      "loss": 0.1239,
+      "step": 11002
+    },
+    {
+      "epoch": 37.94137931034483,
+      "grad_norm": 1.3579665422439575,
+      "learning_rate": 2.8276781609195404e-05,
+      "loss": 0.1501,
+      "step": 11003
+    },
+    {
+      "epoch": 37.9448275862069,
+      "grad_norm": 0.8650179505348206,
+      "learning_rate": 2.827632183908046e-05,
+      "loss": 0.1442,
+      "step": 11004
+    },
+    {
+      "epoch": 37.94827586206897,
+      "grad_norm": 1.1297799348831177,
+      "learning_rate": 2.8275862068965518e-05,
+      "loss": 0.2019,
+      "step": 11005
+    },
+    {
+      "epoch": 37.95172413793104,
+      "grad_norm": 1.5747121572494507,
+      "learning_rate": 2.8275402298850577e-05,
+      "loss": 0.2113,
+      "step": 11006
+    },
+    {
+      "epoch": 37.9551724137931,
+      "grad_norm": 0.7033880949020386,
+      "learning_rate": 2.8274942528735632e-05,
+      "loss": 0.1966,
+      "step": 11007
+    },
+    {
+      "epoch": 37.95862068965517,
+      "grad_norm": 0.9622392654418945,
+      "learning_rate": 2.827448275862069e-05,
+      "loss": 0.203,
+      "step": 11008
+    },
+    {
+      "epoch": 37.96206896551724,
+      "grad_norm": 0.8751043081283569,
+      "learning_rate": 2.8274022988505746e-05,
+      "loss": 0.1979,
+      "step": 11009
+    },
+    {
+      "epoch": 37.96551724137931,
+      "grad_norm": 0.5318447947502136,
+      "learning_rate": 2.8273563218390805e-05,
+      "loss": 0.1627,
+      "step": 11010
+    },
+    {
+      "epoch": 37.96896551724138,
+      "grad_norm": 0.8166975975036621,
+      "learning_rate": 2.8273103448275864e-05,
+      "loss": 0.1566,
+      "step": 11011
+    },
+    {
+      "epoch": 37.97241379310345,
+      "grad_norm": 0.6261289119720459,
+      "learning_rate": 2.827264367816092e-05,
+      "loss": 0.1598,
+      "step": 11012
+    },
+    {
+      "epoch": 37.97586206896552,
+      "grad_norm": 0.5284278392791748,
+      "learning_rate": 2.8272183908045978e-05,
+      "loss": 0.1562,
+      "step": 11013
+    },
+    {
+      "epoch": 37.97931034482759,
+      "grad_norm": 0.6055872440338135,
+      "learning_rate": 2.8271724137931036e-05,
+      "loss": 0.1653,
+      "step": 11014
+    },
+    {
+      "epoch": 37.98275862068966,
+      "grad_norm": 1.0563055276870728,
+      "learning_rate": 2.827126436781609e-05,
+      "loss": 0.1318,
+      "step": 11015
+    },
+    {
+      "epoch": 37.98620689655172,
+      "grad_norm": 0.937923014163971,
+      "learning_rate": 2.827080459770115e-05,
+      "loss": 0.1388,
+      "step": 11016
+    },
+    {
+      "epoch": 37.98965517241379,
+      "grad_norm": 0.6607889533042908,
+      "learning_rate": 2.8270344827586206e-05,
+      "loss": 0.1404,
+      "step": 11017
+    },
+    {
+      "epoch": 37.99310344827586,
+      "grad_norm": 0.6529019474983215,
+      "learning_rate": 2.8269885057471268e-05,
+      "loss": 0.1232,
+      "step": 11018
+    },
+    {
+      "epoch": 37.99655172413793,
+      "grad_norm": 0.7910711765289307,
+      "learning_rate": 2.8269425287356323e-05,
+      "loss": 0.1587,
+      "step": 11019
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 1.2357171773910522,
+      "learning_rate": 2.8268965517241378e-05,
+      "loss": 0.1834,
+      "step": 11020
+    },
+    {
+      "epoch": 38.00344827586207,
+      "grad_norm": 1.288411259651184,
+      "learning_rate": 2.8268505747126437e-05,
+      "loss": 0.2112,
+      "step": 11021
+    },
+    {
+      "epoch": 38.00689655172414,
+      "grad_norm": 1.1342320442199707,
+      "learning_rate": 2.8268045977011496e-05,
+      "loss": 0.196,
+      "step": 11022
+    },
+    {
+      "epoch": 38.01034482758621,
+      "grad_norm": 0.6177983283996582,
+      "learning_rate": 2.8267586206896554e-05,
+      "loss": 0.1971,
+      "step": 11023
+    },
+    {
+      "epoch": 38.01379310344828,
+      "grad_norm": 1.423910140991211,
+      "learning_rate": 2.826712643678161e-05,
+      "loss": 0.1956,
+      "step": 11024
+    },
+    {
+      "epoch": 38.01724137931034,
+      "grad_norm": 0.557257354259491,
+      "learning_rate": 2.8266666666666665e-05,
+      "loss": 0.183,
+      "step": 11025
+    },
+    {
+      "epoch": 38.02068965517241,
+      "grad_norm": 0.6663224101066589,
+      "learning_rate": 2.8266206896551727e-05,
+      "loss": 0.1608,
+      "step": 11026
+    },
+    {
+      "epoch": 38.02413793103448,
+      "grad_norm": 0.7377750873565674,
+      "learning_rate": 2.8265747126436782e-05,
+      "loss": 0.1814,
+      "step": 11027
+    },
+    {
+      "epoch": 38.02758620689655,
+      "grad_norm": 1.8387467861175537,
+      "learning_rate": 2.826528735632184e-05,
+      "loss": 0.1764,
+      "step": 11028
+    },
+    {
+      "epoch": 38.03103448275862,
+      "grad_norm": 0.6119486093521118,
+      "learning_rate": 2.8264827586206896e-05,
+      "loss": 0.1495,
+      "step": 11029
+    },
+    {
+      "epoch": 38.03448275862069,
+      "grad_norm": 0.5738792419433594,
+      "learning_rate": 2.8264367816091955e-05,
+      "loss": 0.169,
+      "step": 11030
+    },
+    {
+      "epoch": 38.03793103448276,
+      "grad_norm": 1.0546351671218872,
+      "learning_rate": 2.8263908045977014e-05,
+      "loss": 0.167,
+      "step": 11031
+    },
+    {
+      "epoch": 38.04137931034483,
+      "grad_norm": 0.576999843120575,
+      "learning_rate": 2.826344827586207e-05,
+      "loss": 0.1506,
+      "step": 11032
+    },
+    {
+      "epoch": 38.0448275862069,
+      "grad_norm": 0.5732587575912476,
+      "learning_rate": 2.8262988505747128e-05,
+      "loss": 0.1443,
+      "step": 11033
+    },
+    {
+      "epoch": 38.04827586206896,
+      "grad_norm": 1.4059163331985474,
+      "learning_rate": 2.8262528735632186e-05,
+      "loss": 0.1423,
+      "step": 11034
+    },
+    {
+      "epoch": 38.05172413793103,
+      "grad_norm": 1.0474202632904053,
+      "learning_rate": 2.826206896551724e-05,
+      "loss": 0.1283,
+      "step": 11035
+    },
+    {
+      "epoch": 38.0551724137931,
+      "grad_norm": 0.6787098050117493,
+      "learning_rate": 2.82616091954023e-05,
+      "loss": 0.1193,
+      "step": 11036
+    },
+    {
+      "epoch": 38.05862068965517,
+      "grad_norm": 2.1983323097229004,
+      "learning_rate": 2.8261149425287355e-05,
+      "loss": 0.1413,
+      "step": 11037
+    },
+    {
+      "epoch": 38.06206896551724,
+      "grad_norm": 0.5369946360588074,
+      "learning_rate": 2.8260689655172418e-05,
+      "loss": 0.1137,
+      "step": 11038
+    },
+    {
+      "epoch": 38.06551724137931,
+      "grad_norm": 0.7499077916145325,
+      "learning_rate": 2.8260229885057473e-05,
+      "loss": 0.1353,
+      "step": 11039
+    },
+    {
+      "epoch": 38.06896551724138,
+      "grad_norm": 1.709822177886963,
+      "learning_rate": 2.8259770114942528e-05,
+      "loss": 0.13,
+      "step": 11040
+    },
+    {
+      "epoch": 38.07241379310345,
+      "grad_norm": 1.5337316989898682,
+      "learning_rate": 2.8259310344827587e-05,
+      "loss": 0.1139,
+      "step": 11041
+    },
+    {
+      "epoch": 38.07586206896552,
+      "grad_norm": 0.6773917078971863,
+      "learning_rate": 2.8258850574712645e-05,
+      "loss": 0.1149,
+      "step": 11042
+    },
+    {
+      "epoch": 38.07931034482758,
+      "grad_norm": 0.7857959270477295,
+      "learning_rate": 2.82583908045977e-05,
+      "loss": 0.1131,
+      "step": 11043
+    },
+    {
+      "epoch": 38.08275862068965,
+      "grad_norm": 1.449794054031372,
+      "learning_rate": 2.825793103448276e-05,
+      "loss": 0.133,
+      "step": 11044
+    },
+    {
+      "epoch": 38.08620689655172,
+      "grad_norm": 1.6826680898666382,
+      "learning_rate": 2.8257471264367815e-05,
+      "loss": 0.1824,
+      "step": 11045
+    },
+    {
+      "epoch": 38.08965517241379,
+      "grad_norm": 1.0662840604782104,
+      "learning_rate": 2.8257011494252877e-05,
+      "loss": 0.223,
+      "step": 11046
+    },
+    {
+      "epoch": 38.09310344827586,
+      "grad_norm": 0.9584540724754333,
+      "learning_rate": 2.8256551724137932e-05,
+      "loss": 0.1891,
+      "step": 11047
+    },
+    {
+      "epoch": 38.09655172413793,
+      "grad_norm": 0.53106689453125,
+      "learning_rate": 2.8256091954022987e-05,
+      "loss": 0.1755,
+      "step": 11048
+    },
+    {
+      "epoch": 38.1,
+      "grad_norm": 0.59113609790802,
+      "learning_rate": 2.8255632183908046e-05,
+      "loss": 0.1854,
+      "step": 11049
+    },
+    {
+      "epoch": 38.10344827586207,
+      "grad_norm": 1.7208354473114014,
+      "learning_rate": 2.8255172413793105e-05,
+      "loss": 0.1701,
+      "step": 11050
+    },
+    {
+      "epoch": 38.10689655172414,
+      "grad_norm": 0.7906379103660583,
+      "learning_rate": 2.8254712643678163e-05,
+      "loss": 0.1484,
+      "step": 11051
+    },
+    {
+      "epoch": 38.110344827586204,
+      "grad_norm": 0.6353622674942017,
+      "learning_rate": 2.825425287356322e-05,
+      "loss": 0.1662,
+      "step": 11052
+    },
+    {
+      "epoch": 38.11379310344827,
+      "grad_norm": 0.9228904247283936,
+      "learning_rate": 2.8253793103448274e-05,
+      "loss": 0.172,
+      "step": 11053
+    },
+    {
+      "epoch": 38.11724137931034,
+      "grad_norm": 1.6312378644943237,
+      "learning_rate": 2.8253333333333336e-05,
+      "loss": 0.1681,
+      "step": 11054
+    },
+    {
+      "epoch": 38.12068965517241,
+      "grad_norm": 0.7363501787185669,
+      "learning_rate": 2.825287356321839e-05,
+      "loss": 0.15,
+      "step": 11055
+    },
+    {
+      "epoch": 38.12413793103448,
+      "grad_norm": 1.122572660446167,
+      "learning_rate": 2.825241379310345e-05,
+      "loss": 0.1682,
+      "step": 11056
+    },
+    {
+      "epoch": 38.12758620689655,
+      "grad_norm": 2.118699073791504,
+      "learning_rate": 2.8251954022988505e-05,
+      "loss": 0.1745,
+      "step": 11057
+    },
+    {
+      "epoch": 38.13103448275862,
+      "grad_norm": 0.5879372358322144,
+      "learning_rate": 2.8251494252873564e-05,
+      "loss": 0.1575,
+      "step": 11058
+    },
+    {
+      "epoch": 38.13448275862069,
+      "grad_norm": 0.949296236038208,
+      "learning_rate": 2.8251034482758623e-05,
+      "loss": 0.1591,
+      "step": 11059
+    },
+    {
+      "epoch": 38.13793103448276,
+      "grad_norm": 0.6345550417900085,
+      "learning_rate": 2.8250574712643678e-05,
+      "loss": 0.1834,
+      "step": 11060
+    },
+    {
+      "epoch": 38.141379310344824,
+      "grad_norm": 0.9381221532821655,
+      "learning_rate": 2.8250114942528737e-05,
+      "loss": 0.1406,
+      "step": 11061
+    },
+    {
+      "epoch": 38.144827586206894,
+      "grad_norm": 0.6490294337272644,
+      "learning_rate": 2.8249655172413795e-05,
+      "loss": 0.1411,
+      "step": 11062
+    },
+    {
+      "epoch": 38.148275862068964,
+      "grad_norm": 0.621123194694519,
+      "learning_rate": 2.824919540229885e-05,
+      "loss": 0.1378,
+      "step": 11063
+    },
+    {
+      "epoch": 38.15172413793103,
+      "grad_norm": 0.6010726094245911,
+      "learning_rate": 2.824873563218391e-05,
+      "loss": 0.1251,
+      "step": 11064
+    },
+    {
+      "epoch": 38.1551724137931,
+      "grad_norm": 0.844870924949646,
+      "learning_rate": 2.8248275862068965e-05,
+      "loss": 0.1337,
+      "step": 11065
+    },
+    {
+      "epoch": 38.15862068965517,
+      "grad_norm": 0.9412791728973389,
+      "learning_rate": 2.8247816091954027e-05,
+      "loss": 0.1257,
+      "step": 11066
+    },
+    {
+      "epoch": 38.16206896551724,
+      "grad_norm": 3.892554521560669,
+      "learning_rate": 2.8247356321839082e-05,
+      "loss": 0.1335,
+      "step": 11067
+    },
+    {
+      "epoch": 38.16551724137931,
+      "grad_norm": 1.4252945184707642,
+      "learning_rate": 2.8246896551724137e-05,
+      "loss": 0.1256,
+      "step": 11068
+    },
+    {
+      "epoch": 38.16896551724138,
+      "grad_norm": 0.8877184391021729,
+      "learning_rate": 2.8246436781609196e-05,
+      "loss": 0.1303,
+      "step": 11069
+    },
+    {
+      "epoch": 38.172413793103445,
+      "grad_norm": 2.387632369995117,
+      "learning_rate": 2.8245977011494255e-05,
+      "loss": 0.1634,
+      "step": 11070
+    },
+    {
+      "epoch": 38.175862068965515,
+      "grad_norm": 0.6913253664970398,
+      "learning_rate": 2.824551724137931e-05,
+      "loss": 0.2395,
+      "step": 11071
+    },
+    {
+      "epoch": 38.179310344827584,
+      "grad_norm": 0.519919753074646,
+      "learning_rate": 2.824505747126437e-05,
+      "loss": 0.2057,
+      "step": 11072
+    },
+    {
+      "epoch": 38.182758620689654,
+      "grad_norm": 0.8646356463432312,
+      "learning_rate": 2.8244597701149424e-05,
+      "loss": 0.1801,
+      "step": 11073
+    },
+    {
+      "epoch": 38.186206896551724,
+      "grad_norm": 0.6832991242408752,
+      "learning_rate": 2.8244137931034486e-05,
+      "loss": 0.173,
+      "step": 11074
+    },
+    {
+      "epoch": 38.189655172413794,
+      "grad_norm": 0.791799783706665,
+      "learning_rate": 2.824367816091954e-05,
+      "loss": 0.1578,
+      "step": 11075
+    },
+    {
+      "epoch": 38.19310344827586,
+      "grad_norm": 0.8161206841468811,
+      "learning_rate": 2.8243218390804597e-05,
+      "loss": 0.1594,
+      "step": 11076
+    },
+    {
+      "epoch": 38.19655172413793,
+      "grad_norm": 0.5462909936904907,
+      "learning_rate": 2.8242758620689655e-05,
+      "loss": 0.1881,
+      "step": 11077
+    },
+    {
+      "epoch": 38.2,
+      "grad_norm": 0.623665988445282,
+      "learning_rate": 2.8242298850574714e-05,
+      "loss": 0.174,
+      "step": 11078
+    },
+    {
+      "epoch": 38.203448275862065,
+      "grad_norm": 0.6887049078941345,
+      "learning_rate": 2.8241839080459773e-05,
+      "loss": 0.166,
+      "step": 11079
+    },
+    {
+      "epoch": 38.206896551724135,
+      "grad_norm": 0.7617496848106384,
+      "learning_rate": 2.8241379310344828e-05,
+      "loss": 0.1421,
+      "step": 11080
+    },
+    {
+      "epoch": 38.210344827586205,
+      "grad_norm": 1.2492121458053589,
+      "learning_rate": 2.8240919540229883e-05,
+      "loss": 0.1299,
+      "step": 11081
+    },
+    {
+      "epoch": 38.213793103448275,
+      "grad_norm": 0.5571796298027039,
+      "learning_rate": 2.8240459770114945e-05,
+      "loss": 0.1636,
+      "step": 11082
+    },
+    {
+      "epoch": 38.217241379310344,
+      "grad_norm": 0.548718273639679,
+      "learning_rate": 2.824e-05,
+      "loss": 0.1412,
+      "step": 11083
+    },
+    {
+      "epoch": 38.220689655172414,
+      "grad_norm": 0.5747441053390503,
+      "learning_rate": 2.823954022988506e-05,
+      "loss": 0.1289,
+      "step": 11084
+    },
+    {
+      "epoch": 38.224137931034484,
+      "grad_norm": 1.056795358657837,
+      "learning_rate": 2.8239080459770115e-05,
+      "loss": 0.1525,
+      "step": 11085
+    },
+    {
+      "epoch": 38.227586206896554,
+      "grad_norm": 0.6217753887176514,
+      "learning_rate": 2.8238620689655173e-05,
+      "loss": 0.1095,
+      "step": 11086
+    },
+    {
+      "epoch": 38.23103448275862,
+      "grad_norm": 0.5540239214897156,
+      "learning_rate": 2.8238160919540232e-05,
+      "loss": 0.1128,
+      "step": 11087
+    },
+    {
+      "epoch": 38.234482758620686,
+      "grad_norm": 0.7455006837844849,
+      "learning_rate": 2.8237701149425287e-05,
+      "loss": 0.1238,
+      "step": 11088
+    },
+    {
+      "epoch": 38.237931034482756,
+      "grad_norm": 0.701106071472168,
+      "learning_rate": 2.8237241379310346e-05,
+      "loss": 0.141,
+      "step": 11089
+    },
+    {
+      "epoch": 38.241379310344826,
+      "grad_norm": 0.6624748706817627,
+      "learning_rate": 2.8236781609195405e-05,
+      "loss": 0.1147,
+      "step": 11090
+    },
+    {
+      "epoch": 38.244827586206895,
+      "grad_norm": 1.6331150531768799,
+      "learning_rate": 2.823632183908046e-05,
+      "loss": 0.1235,
+      "step": 11091
+    },
+    {
+      "epoch": 38.248275862068965,
+      "grad_norm": 0.73601895570755,
+      "learning_rate": 2.823586206896552e-05,
+      "loss": 0.133,
+      "step": 11092
+    },
+    {
+      "epoch": 38.251724137931035,
+      "grad_norm": 0.6321110725402832,
+      "learning_rate": 2.8235402298850574e-05,
+      "loss": 0.1053,
+      "step": 11093
+    },
+    {
+      "epoch": 38.255172413793105,
+      "grad_norm": 0.8897535800933838,
+      "learning_rate": 2.8234942528735636e-05,
+      "loss": 0.1272,
+      "step": 11094
+    },
+    {
+      "epoch": 38.258620689655174,
+      "grad_norm": 1.3054646253585815,
+      "learning_rate": 2.823448275862069e-05,
+      "loss": 0.1575,
+      "step": 11095
+    },
+    {
+      "epoch": 38.262068965517244,
+      "grad_norm": 0.8358314037322998,
+      "learning_rate": 2.8234022988505747e-05,
+      "loss": 0.2264,
+      "step": 11096
+    },
+    {
+      "epoch": 38.265517241379314,
+      "grad_norm": 0.785050630569458,
+      "learning_rate": 2.8233563218390805e-05,
+      "loss": 0.1888,
+      "step": 11097
+    },
+    {
+      "epoch": 38.26896551724138,
+      "grad_norm": 0.5590908527374268,
+      "learning_rate": 2.8233103448275864e-05,
+      "loss": 0.1773,
+      "step": 11098
+    },
+    {
+      "epoch": 38.272413793103446,
+      "grad_norm": 0.7016497850418091,
+      "learning_rate": 2.823264367816092e-05,
+      "loss": 0.1844,
+      "step": 11099
+    },
+    {
+      "epoch": 38.275862068965516,
+      "grad_norm": 0.698447048664093,
+      "learning_rate": 2.8232183908045978e-05,
+      "loss": 0.174,
+      "step": 11100
+    },
+    {
+      "epoch": 38.279310344827586,
+      "grad_norm": 0.6120927929878235,
+      "learning_rate": 2.8231724137931033e-05,
+      "loss": 0.1828,
+      "step": 11101
+    },
+    {
+      "epoch": 38.282758620689656,
+      "grad_norm": 0.5784707069396973,
+      "learning_rate": 2.8231264367816095e-05,
+      "loss": 0.1687,
+      "step": 11102
+    },
+    {
+      "epoch": 38.286206896551725,
+      "grad_norm": 0.9990240931510925,
+      "learning_rate": 2.823080459770115e-05,
+      "loss": 0.1852,
+      "step": 11103
+    },
+    {
+      "epoch": 38.289655172413795,
+      "grad_norm": 0.6674383878707886,
+      "learning_rate": 2.8230344827586206e-05,
+      "loss": 0.1368,
+      "step": 11104
+    },
+    {
+      "epoch": 38.293103448275865,
+      "grad_norm": 0.5530248880386353,
+      "learning_rate": 2.8229885057471265e-05,
+      "loss": 0.1575,
+      "step": 11105
+    },
+    {
+      "epoch": 38.296551724137935,
+      "grad_norm": 0.9319071769714355,
+      "learning_rate": 2.8229425287356323e-05,
+      "loss": 0.1595,
+      "step": 11106
+    },
+    {
+      "epoch": 38.3,
+      "grad_norm": 1.5229263305664062,
+      "learning_rate": 2.8228965517241382e-05,
+      "loss": 0.1718,
+      "step": 11107
+    },
+    {
+      "epoch": 38.30344827586207,
+      "grad_norm": 0.8065203428268433,
+      "learning_rate": 2.8228505747126437e-05,
+      "loss": 0.1456,
+      "step": 11108
+    },
+    {
+      "epoch": 38.30689655172414,
+      "grad_norm": 0.522131621837616,
+      "learning_rate": 2.8228045977011493e-05,
+      "loss": 0.1281,
+      "step": 11109
+    },
+    {
+      "epoch": 38.310344827586206,
+      "grad_norm": 2.7977569103240967,
+      "learning_rate": 2.8227586206896555e-05,
+      "loss": 0.1345,
+      "step": 11110
+    },
+    {
+      "epoch": 38.313793103448276,
+      "grad_norm": 0.5845361351966858,
+      "learning_rate": 2.822712643678161e-05,
+      "loss": 0.1277,
+      "step": 11111
+    },
+    {
+      "epoch": 38.317241379310346,
+      "grad_norm": 1.2738466262817383,
+      "learning_rate": 2.822666666666667e-05,
+      "loss": 0.1467,
+      "step": 11112
+    },
+    {
+      "epoch": 38.320689655172416,
+      "grad_norm": 0.6461541056632996,
+      "learning_rate": 2.8226206896551724e-05,
+      "loss": 0.1256,
+      "step": 11113
+    },
+    {
+      "epoch": 38.324137931034485,
+      "grad_norm": 1.1052881479263306,
+      "learning_rate": 2.8225747126436783e-05,
+      "loss": 0.1244,
+      "step": 11114
+    },
+    {
+      "epoch": 38.327586206896555,
+      "grad_norm": 1.7344257831573486,
+      "learning_rate": 2.822528735632184e-05,
+      "loss": 0.1138,
+      "step": 11115
+    },
+    {
+      "epoch": 38.33103448275862,
+      "grad_norm": 1.1886646747589111,
+      "learning_rate": 2.8224827586206897e-05,
+      "loss": 0.1311,
+      "step": 11116
+    },
+    {
+      "epoch": 38.33448275862069,
+      "grad_norm": 1.0006709098815918,
+      "learning_rate": 2.8224367816091955e-05,
+      "loss": 0.1282,
+      "step": 11117
+    },
+    {
+      "epoch": 38.33793103448276,
+      "grad_norm": 1.5593358278274536,
+      "learning_rate": 2.8223908045977014e-05,
+      "loss": 0.1271,
+      "step": 11118
+    },
+    {
+      "epoch": 38.34137931034483,
+      "grad_norm": 5.1732940673828125,
+      "learning_rate": 2.822344827586207e-05,
+      "loss": 0.1236,
+      "step": 11119
+    },
+    {
+      "epoch": 38.3448275862069,
+      "grad_norm": 3.5342190265655518,
+      "learning_rate": 2.8222988505747128e-05,
+      "loss": 0.1547,
+      "step": 11120
+    },
+    {
+      "epoch": 38.34827586206897,
+      "grad_norm": 0.6172340512275696,
+      "learning_rate": 2.8222528735632183e-05,
+      "loss": 0.2441,
+      "step": 11121
+    },
+    {
+      "epoch": 38.351724137931036,
+      "grad_norm": 1.1009495258331299,
+      "learning_rate": 2.8222068965517245e-05,
+      "loss": 0.1905,
+      "step": 11122
+    },
+    {
+      "epoch": 38.355172413793106,
+      "grad_norm": 0.8726471662521362,
+      "learning_rate": 2.82216091954023e-05,
+      "loss": 0.1886,
+      "step": 11123
+    },
+    {
+      "epoch": 38.358620689655176,
+      "grad_norm": 0.8554416298866272,
+      "learning_rate": 2.8221149425287356e-05,
+      "loss": 0.2054,
+      "step": 11124
+    },
+    {
+      "epoch": 38.36206896551724,
+      "grad_norm": 1.0626298189163208,
+      "learning_rate": 2.8220689655172415e-05,
+      "loss": 0.1911,
+      "step": 11125
+    },
+    {
+      "epoch": 38.36551724137931,
+      "grad_norm": 0.6079199910163879,
+      "learning_rate": 2.8220229885057473e-05,
+      "loss": 0.1654,
+      "step": 11126
+    },
+    {
+      "epoch": 38.36896551724138,
+      "grad_norm": 0.6212504506111145,
+      "learning_rate": 2.8219770114942532e-05,
+      "loss": 0.1557,
+      "step": 11127
+    },
+    {
+      "epoch": 38.37241379310345,
+      "grad_norm": 0.496440052986145,
+      "learning_rate": 2.8219310344827587e-05,
+      "loss": 0.1584,
+      "step": 11128
+    },
+    {
+      "epoch": 38.37586206896552,
+      "grad_norm": 0.5732938647270203,
+      "learning_rate": 2.8218850574712642e-05,
+      "loss": 0.1365,
+      "step": 11129
+    },
+    {
+      "epoch": 38.37931034482759,
+      "grad_norm": 1.3416622877120972,
+      "learning_rate": 2.8218390804597705e-05,
+      "loss": 0.1582,
+      "step": 11130
+    },
+    {
+      "epoch": 38.38275862068966,
+      "grad_norm": 1.3807644844055176,
+      "learning_rate": 2.821793103448276e-05,
+      "loss": 0.1625,
+      "step": 11131
+    },
+    {
+      "epoch": 38.38620689655173,
+      "grad_norm": 0.9106972813606262,
+      "learning_rate": 2.8217471264367815e-05,
+      "loss": 0.147,
+      "step": 11132
+    },
+    {
+      "epoch": 38.389655172413796,
+      "grad_norm": 1.0389920473098755,
+      "learning_rate": 2.8217011494252874e-05,
+      "loss": 0.1461,
+      "step": 11133
+    },
+    {
+      "epoch": 38.39310344827586,
+      "grad_norm": 0.7202304005622864,
+      "learning_rate": 2.8216551724137933e-05,
+      "loss": 0.144,
+      "step": 11134
+    },
+    {
+      "epoch": 38.39655172413793,
+      "grad_norm": 0.555806577205658,
+      "learning_rate": 2.821609195402299e-05,
+      "loss": 0.1248,
+      "step": 11135
+    },
+    {
+      "epoch": 38.4,
+      "grad_norm": 0.7730570435523987,
+      "learning_rate": 2.8215632183908046e-05,
+      "loss": 0.1607,
+      "step": 11136
+    },
+    {
+      "epoch": 38.40344827586207,
+      "grad_norm": 0.7074033617973328,
+      "learning_rate": 2.8215172413793102e-05,
+      "loss": 0.127,
+      "step": 11137
+    },
+    {
+      "epoch": 38.40689655172414,
+      "grad_norm": 0.6953191161155701,
+      "learning_rate": 2.8214712643678164e-05,
+      "loss": 0.1298,
+      "step": 11138
+    },
+    {
+      "epoch": 38.41034482758621,
+      "grad_norm": 0.9239925742149353,
+      "learning_rate": 2.821425287356322e-05,
+      "loss": 0.1452,
+      "step": 11139
+    },
+    {
+      "epoch": 38.41379310344828,
+      "grad_norm": 1.2424966096878052,
+      "learning_rate": 2.8213793103448278e-05,
+      "loss": 0.1271,
+      "step": 11140
+    },
+    {
+      "epoch": 38.41724137931035,
+      "grad_norm": 2.8544068336486816,
+      "learning_rate": 2.8213333333333333e-05,
+      "loss": 0.1272,
+      "step": 11141
+    },
+    {
+      "epoch": 38.42068965517242,
+      "grad_norm": 0.7662988305091858,
+      "learning_rate": 2.8212873563218392e-05,
+      "loss": 0.1282,
+      "step": 11142
+    },
+    {
+      "epoch": 38.42413793103448,
+      "grad_norm": 0.9247673153877258,
+      "learning_rate": 2.821241379310345e-05,
+      "loss": 0.1239,
+      "step": 11143
+    },
+    {
+      "epoch": 38.42758620689655,
+      "grad_norm": 0.7490770816802979,
+      "learning_rate": 2.8211954022988506e-05,
+      "loss": 0.1217,
+      "step": 11144
+    },
+    {
+      "epoch": 38.43103448275862,
+      "grad_norm": 1.0376331806182861,
+      "learning_rate": 2.8211494252873564e-05,
+      "loss": 0.2088,
+      "step": 11145
+    },
+    {
+      "epoch": 38.43448275862069,
+      "grad_norm": 1.3446253538131714,
+      "learning_rate": 2.8211034482758623e-05,
+      "loss": 0.2585,
+      "step": 11146
+    },
+    {
+      "epoch": 38.43793103448276,
+      "grad_norm": 0.7463351488113403,
+      "learning_rate": 2.821057471264368e-05,
+      "loss": 0.1908,
+      "step": 11147
+    },
+    {
+      "epoch": 38.44137931034483,
+      "grad_norm": 0.9018734693527222,
+      "learning_rate": 2.8210114942528737e-05,
+      "loss": 0.2099,
+      "step": 11148
+    },
+    {
+      "epoch": 38.4448275862069,
+      "grad_norm": 0.5361537337303162,
+      "learning_rate": 2.8209655172413792e-05,
+      "loss": 0.1638,
+      "step": 11149
+    },
+    {
+      "epoch": 38.44827586206897,
+      "grad_norm": 0.693289577960968,
+      "learning_rate": 2.820919540229885e-05,
+      "loss": 0.1834,
+      "step": 11150
+    },
+    {
+      "epoch": 38.45172413793104,
+      "grad_norm": 0.7946658730506897,
+      "learning_rate": 2.820873563218391e-05,
+      "loss": 0.1651,
+      "step": 11151
+    },
+    {
+      "epoch": 38.4551724137931,
+      "grad_norm": 0.5924472808837891,
+      "learning_rate": 2.8208275862068965e-05,
+      "loss": 0.1621,
+      "step": 11152
+    },
+    {
+      "epoch": 38.45862068965517,
+      "grad_norm": 0.5958293676376343,
+      "learning_rate": 2.8207816091954024e-05,
+      "loss": 0.1666,
+      "step": 11153
+    },
+    {
+      "epoch": 38.46206896551724,
+      "grad_norm": 0.5247872471809387,
+      "learning_rate": 2.820735632183908e-05,
+      "loss": 0.157,
+      "step": 11154
+    },
+    {
+      "epoch": 38.46551724137931,
+      "grad_norm": 0.541880190372467,
+      "learning_rate": 2.820689655172414e-05,
+      "loss": 0.1612,
+      "step": 11155
+    },
+    {
+      "epoch": 38.46896551724138,
+      "grad_norm": 0.5855715870857239,
+      "learning_rate": 2.8206436781609196e-05,
+      "loss": 0.1473,
+      "step": 11156
+    },
+    {
+      "epoch": 38.47241379310345,
+      "grad_norm": 2.371551513671875,
+      "learning_rate": 2.8205977011494252e-05,
+      "loss": 0.1635,
+      "step": 11157
+    },
+    {
+      "epoch": 38.47586206896552,
+      "grad_norm": 0.6856994032859802,
+      "learning_rate": 2.820551724137931e-05,
+      "loss": 0.1512,
+      "step": 11158
+    },
+    {
+      "epoch": 38.47931034482759,
+      "grad_norm": 0.8492767810821533,
+      "learning_rate": 2.820505747126437e-05,
+      "loss": 0.148,
+      "step": 11159
+    },
+    {
+      "epoch": 38.48275862068966,
+      "grad_norm": 1.0750398635864258,
+      "learning_rate": 2.8204597701149424e-05,
+      "loss": 0.1501,
+      "step": 11160
+    },
+    {
+      "epoch": 38.48620689655172,
+      "grad_norm": 0.5936921238899231,
+      "learning_rate": 2.8204137931034483e-05,
+      "loss": 0.1534,
+      "step": 11161
+    },
+    {
+      "epoch": 38.48965517241379,
+      "grad_norm": 0.817501425743103,
+      "learning_rate": 2.820367816091954e-05,
+      "loss": 0.1196,
+      "step": 11162
+    },
+    {
+      "epoch": 38.49310344827586,
+      "grad_norm": 0.9979875087738037,
+      "learning_rate": 2.82032183908046e-05,
+      "loss": 0.1327,
+      "step": 11163
+    },
+    {
+      "epoch": 38.49655172413793,
+      "grad_norm": 0.7728829979896545,
+      "learning_rate": 2.8202758620689656e-05,
+      "loss": 0.1424,
+      "step": 11164
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 0.8059385418891907,
+      "learning_rate": 2.820229885057471e-05,
+      "loss": 0.1282,
+      "step": 11165
+    },
+    {
+      "epoch": 38.50344827586207,
+      "grad_norm": 0.7044061422348022,
+      "learning_rate": 2.820183908045977e-05,
+      "loss": 0.1546,
+      "step": 11166
+    },
+    {
+      "epoch": 38.50689655172414,
+      "grad_norm": 0.6805763840675354,
+      "learning_rate": 2.820137931034483e-05,
+      "loss": 0.1217,
+      "step": 11167
+    },
+    {
+      "epoch": 38.51034482758621,
+      "grad_norm": 1.2128427028656006,
+      "learning_rate": 2.8200919540229887e-05,
+      "loss": 0.1011,
+      "step": 11168
+    },
+    {
+      "epoch": 38.51379310344828,
+      "grad_norm": 0.8982587456703186,
+      "learning_rate": 2.8200459770114942e-05,
+      "loss": 0.1269,
+      "step": 11169
+    },
+    {
+      "epoch": 38.51724137931034,
+      "grad_norm": 1.0519921779632568,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.1741,
+      "step": 11170
+    },
+    {
+      "epoch": 38.52068965517241,
+      "grad_norm": 0.712051510810852,
+      "learning_rate": 2.819954022988506e-05,
+      "loss": 0.2228,
+      "step": 11171
+    },
+    {
+      "epoch": 38.52413793103448,
+      "grad_norm": 1.0443302392959595,
+      "learning_rate": 2.8199080459770115e-05,
+      "loss": 0.1965,
+      "step": 11172
+    },
+    {
+      "epoch": 38.52758620689655,
+      "grad_norm": 0.8447859883308411,
+      "learning_rate": 2.8198620689655174e-05,
+      "loss": 0.1917,
+      "step": 11173
+    },
+    {
+      "epoch": 38.53103448275862,
+      "grad_norm": 1.6274302005767822,
+      "learning_rate": 2.819816091954023e-05,
+      "loss": 0.1663,
+      "step": 11174
+    },
+    {
+      "epoch": 38.53448275862069,
+      "grad_norm": 0.6539332866668701,
+      "learning_rate": 2.8197701149425288e-05,
+      "loss": 0.184,
+      "step": 11175
+    },
+    {
+      "epoch": 38.53793103448276,
+      "grad_norm": 2.3592281341552734,
+      "learning_rate": 2.8197241379310346e-05,
+      "loss": 0.1568,
+      "step": 11176
+    },
+    {
+      "epoch": 38.54137931034483,
+      "grad_norm": 0.778899610042572,
+      "learning_rate": 2.81967816091954e-05,
+      "loss": 0.1901,
+      "step": 11177
+    },
+    {
+      "epoch": 38.5448275862069,
+      "grad_norm": 0.66343092918396,
+      "learning_rate": 2.819632183908046e-05,
+      "loss": 0.1642,
+      "step": 11178
+    },
+    {
+      "epoch": 38.54827586206896,
+      "grad_norm": 1.4064241647720337,
+      "learning_rate": 2.819586206896552e-05,
+      "loss": 0.1544,
+      "step": 11179
+    },
+    {
+      "epoch": 38.55172413793103,
+      "grad_norm": 0.7369663715362549,
+      "learning_rate": 2.8195402298850574e-05,
+      "loss": 0.138,
+      "step": 11180
+    },
+    {
+      "epoch": 38.5551724137931,
+      "grad_norm": 0.7397128343582153,
+      "learning_rate": 2.8194942528735633e-05,
+      "loss": 0.1493,
+      "step": 11181
+    },
+    {
+      "epoch": 38.55862068965517,
+      "grad_norm": 0.9013453722000122,
+      "learning_rate": 2.819448275862069e-05,
+      "loss": 0.1754,
+      "step": 11182
+    },
+    {
+      "epoch": 38.56206896551724,
+      "grad_norm": 0.7493063807487488,
+      "learning_rate": 2.819402298850575e-05,
+      "loss": 0.1474,
+      "step": 11183
+    },
+    {
+      "epoch": 38.56551724137931,
+      "grad_norm": 0.9977419972419739,
+      "learning_rate": 2.8193563218390806e-05,
+      "loss": 0.1724,
+      "step": 11184
+    },
+    {
+      "epoch": 38.56896551724138,
+      "grad_norm": 1.036911964416504,
+      "learning_rate": 2.819310344827586e-05,
+      "loss": 0.1644,
+      "step": 11185
+    },
+    {
+      "epoch": 38.57241379310345,
+      "grad_norm": 1.0633352994918823,
+      "learning_rate": 2.819264367816092e-05,
+      "loss": 0.1588,
+      "step": 11186
+    },
+    {
+      "epoch": 38.57586206896552,
+      "grad_norm": 1.0141501426696777,
+      "learning_rate": 2.819218390804598e-05,
+      "loss": 0.1336,
+      "step": 11187
+    },
+    {
+      "epoch": 38.57931034482758,
+      "grad_norm": 0.8761879205703735,
+      "learning_rate": 2.8191724137931034e-05,
+      "loss": 0.1414,
+      "step": 11188
+    },
+    {
+      "epoch": 38.58275862068965,
+      "grad_norm": 0.6292900443077087,
+      "learning_rate": 2.8191264367816092e-05,
+      "loss": 0.1258,
+      "step": 11189
+    },
+    {
+      "epoch": 38.58620689655172,
+      "grad_norm": 1.840221881866455,
+      "learning_rate": 2.8190804597701148e-05,
+      "loss": 0.1381,
+      "step": 11190
+    },
+    {
+      "epoch": 38.58965517241379,
+      "grad_norm": 0.7856280207633972,
+      "learning_rate": 2.819034482758621e-05,
+      "loss": 0.1404,
+      "step": 11191
+    },
+    {
+      "epoch": 38.59310344827586,
+      "grad_norm": 0.9368109107017517,
+      "learning_rate": 2.8189885057471265e-05,
+      "loss": 0.1211,
+      "step": 11192
+    },
+    {
+      "epoch": 38.59655172413793,
+      "grad_norm": 0.7232136130332947,
+      "learning_rate": 2.818942528735632e-05,
+      "loss": 0.1263,
+      "step": 11193
+    },
+    {
+      "epoch": 38.6,
+      "grad_norm": 0.7170366048812866,
+      "learning_rate": 2.818896551724138e-05,
+      "loss": 0.1318,
+      "step": 11194
+    },
+    {
+      "epoch": 38.60344827586207,
+      "grad_norm": 1.0634135007858276,
+      "learning_rate": 2.8188505747126438e-05,
+      "loss": 0.1605,
+      "step": 11195
+    },
+    {
+      "epoch": 38.60689655172414,
+      "grad_norm": 0.5701091289520264,
+      "learning_rate": 2.8188045977011496e-05,
+      "loss": 0.235,
+      "step": 11196
+    },
+    {
+      "epoch": 38.610344827586204,
+      "grad_norm": 0.7595037221908569,
+      "learning_rate": 2.818758620689655e-05,
+      "loss": 0.2031,
+      "step": 11197
+    },
+    {
+      "epoch": 38.61379310344827,
+      "grad_norm": 0.7609796524047852,
+      "learning_rate": 2.8187126436781607e-05,
+      "loss": 0.1746,
+      "step": 11198
+    },
+    {
+      "epoch": 38.61724137931034,
+      "grad_norm": 0.741205632686615,
+      "learning_rate": 2.818666666666667e-05,
+      "loss": 0.1755,
+      "step": 11199
+    },
+    {
+      "epoch": 38.62068965517241,
+      "grad_norm": 0.5766764283180237,
+      "learning_rate": 2.8186206896551724e-05,
+      "loss": 0.169,
+      "step": 11200
+    },
+    {
+      "epoch": 38.62413793103448,
+      "grad_norm": 0.8556082844734192,
+      "learning_rate": 2.8185747126436783e-05,
+      "loss": 0.1857,
+      "step": 11201
+    },
+    {
+      "epoch": 38.62758620689655,
+      "grad_norm": 0.8908397555351257,
+      "learning_rate": 2.8185287356321838e-05,
+      "loss": 0.1601,
+      "step": 11202
+    },
+    {
+      "epoch": 38.63103448275862,
+      "grad_norm": 1.4185616970062256,
+      "learning_rate": 2.8184827586206897e-05,
+      "loss": 0.1557,
+      "step": 11203
+    },
+    {
+      "epoch": 38.63448275862069,
+      "grad_norm": 0.6099755167961121,
+      "learning_rate": 2.8184367816091956e-05,
+      "loss": 0.1468,
+      "step": 11204
+    },
+    {
+      "epoch": 38.63793103448276,
+      "grad_norm": 0.7966406345367432,
+      "learning_rate": 2.818390804597701e-05,
+      "loss": 0.1514,
+      "step": 11205
+    },
+    {
+      "epoch": 38.641379310344824,
+      "grad_norm": 1.1357074975967407,
+      "learning_rate": 2.818344827586207e-05,
+      "loss": 0.1456,
+      "step": 11206
+    },
+    {
+      "epoch": 38.644827586206894,
+      "grad_norm": 1.0193994045257568,
+      "learning_rate": 2.8182988505747128e-05,
+      "loss": 0.16,
+      "step": 11207
+    },
+    {
+      "epoch": 38.648275862068964,
+      "grad_norm": 1.0271649360656738,
+      "learning_rate": 2.8182528735632184e-05,
+      "loss": 0.1416,
+      "step": 11208
+    },
+    {
+      "epoch": 38.65172413793103,
+      "grad_norm": 0.6540797352790833,
+      "learning_rate": 2.8182068965517242e-05,
+      "loss": 0.1504,
+      "step": 11209
+    },
+    {
+      "epoch": 38.6551724137931,
+      "grad_norm": 0.669283926486969,
+      "learning_rate": 2.8181609195402298e-05,
+      "loss": 0.1587,
+      "step": 11210
+    },
+    {
+      "epoch": 38.65862068965517,
+      "grad_norm": 3.3099746704101562,
+      "learning_rate": 2.818114942528736e-05,
+      "loss": 0.138,
+      "step": 11211
+    },
+    {
+      "epoch": 38.66206896551724,
+      "grad_norm": 0.7048425078392029,
+      "learning_rate": 2.8180689655172415e-05,
+      "loss": 0.1377,
+      "step": 11212
+    },
+    {
+      "epoch": 38.66551724137931,
+      "grad_norm": 3.8509955406188965,
+      "learning_rate": 2.818022988505747e-05,
+      "loss": 0.1572,
+      "step": 11213
+    },
+    {
+      "epoch": 38.66896551724138,
+      "grad_norm": 0.6190562844276428,
+      "learning_rate": 2.817977011494253e-05,
+      "loss": 0.117,
+      "step": 11214
+    },
+    {
+      "epoch": 38.672413793103445,
+      "grad_norm": 0.7845790386199951,
+      "learning_rate": 2.8179310344827588e-05,
+      "loss": 0.1339,
+      "step": 11215
+    },
+    {
+      "epoch": 38.675862068965515,
+      "grad_norm": 1.3034923076629639,
+      "learning_rate": 2.8178850574712646e-05,
+      "loss": 0.1236,
+      "step": 11216
+    },
+    {
+      "epoch": 38.679310344827584,
+      "grad_norm": 0.6990228295326233,
+      "learning_rate": 2.81783908045977e-05,
+      "loss": 0.1069,
+      "step": 11217
+    },
+    {
+      "epoch": 38.682758620689654,
+      "grad_norm": 0.9575676321983337,
+      "learning_rate": 2.8177931034482757e-05,
+      "loss": 0.1258,
+      "step": 11218
+    },
+    {
+      "epoch": 38.686206896551724,
+      "grad_norm": 0.7768901586532593,
+      "learning_rate": 2.817747126436782e-05,
+      "loss": 0.1392,
+      "step": 11219
+    },
+    {
+      "epoch": 38.689655172413794,
+      "grad_norm": 1.1386855840682983,
+      "learning_rate": 2.8177011494252874e-05,
+      "loss": 0.1685,
+      "step": 11220
+    },
+    {
+      "epoch": 38.69310344827586,
+      "grad_norm": 0.674602746963501,
+      "learning_rate": 2.817655172413793e-05,
+      "loss": 0.2232,
+      "step": 11221
+    },
+    {
+      "epoch": 38.69655172413793,
+      "grad_norm": 0.7140346765518188,
+      "learning_rate": 2.8176091954022988e-05,
+      "loss": 0.2039,
+      "step": 11222
+    },
+    {
+      "epoch": 38.7,
+      "grad_norm": 0.6936749219894409,
+      "learning_rate": 2.8175632183908047e-05,
+      "loss": 0.1843,
+      "step": 11223
+    },
+    {
+      "epoch": 38.703448275862065,
+      "grad_norm": 0.7373558282852173,
+      "learning_rate": 2.8175172413793106e-05,
+      "loss": 0.1581,
+      "step": 11224
+    },
+    {
+      "epoch": 38.706896551724135,
+      "grad_norm": 1.226414442062378,
+      "learning_rate": 2.817471264367816e-05,
+      "loss": 0.1753,
+      "step": 11225
+    },
+    {
+      "epoch": 38.710344827586205,
+      "grad_norm": 1.0694701671600342,
+      "learning_rate": 2.8174252873563216e-05,
+      "loss": 0.1643,
+      "step": 11226
+    },
+    {
+      "epoch": 38.713793103448275,
+      "grad_norm": 0.8881990909576416,
+      "learning_rate": 2.8173793103448278e-05,
+      "loss": 0.1766,
+      "step": 11227
+    },
+    {
+      "epoch": 38.717241379310344,
+      "grad_norm": 0.8632447719573975,
+      "learning_rate": 2.8173333333333334e-05,
+      "loss": 0.1886,
+      "step": 11228
+    },
+    {
+      "epoch": 38.720689655172414,
+      "grad_norm": 0.9795783758163452,
+      "learning_rate": 2.8172873563218392e-05,
+      "loss": 0.1837,
+      "step": 11229
+    },
+    {
+      "epoch": 38.724137931034484,
+      "grad_norm": 1.17487633228302,
+      "learning_rate": 2.8172413793103447e-05,
+      "loss": 0.145,
+      "step": 11230
+    },
+    {
+      "epoch": 38.727586206896554,
+      "grad_norm": 0.6265134811401367,
+      "learning_rate": 2.8171954022988506e-05,
+      "loss": 0.1517,
+      "step": 11231
+    },
+    {
+      "epoch": 38.73103448275862,
+      "grad_norm": 1.9006118774414062,
+      "learning_rate": 2.8171494252873565e-05,
+      "loss": 0.1642,
+      "step": 11232
+    },
+    {
+      "epoch": 38.734482758620686,
+      "grad_norm": 0.6617832779884338,
+      "learning_rate": 2.817103448275862e-05,
+      "loss": 0.1569,
+      "step": 11233
+    },
+    {
+      "epoch": 38.737931034482756,
+      "grad_norm": 0.9698827862739563,
+      "learning_rate": 2.817057471264368e-05,
+      "loss": 0.1408,
+      "step": 11234
+    },
+    {
+      "epoch": 38.741379310344826,
+      "grad_norm": 0.6907277703285217,
+      "learning_rate": 2.8170114942528738e-05,
+      "loss": 0.1459,
+      "step": 11235
+    },
+    {
+      "epoch": 38.744827586206895,
+      "grad_norm": 0.6659308075904846,
+      "learning_rate": 2.8169655172413793e-05,
+      "loss": 0.1455,
+      "step": 11236
+    },
+    {
+      "epoch": 38.748275862068965,
+      "grad_norm": 0.9962143898010254,
+      "learning_rate": 2.816919540229885e-05,
+      "loss": 0.1424,
+      "step": 11237
+    },
+    {
+      "epoch": 38.751724137931035,
+      "grad_norm": 0.5742040872573853,
+      "learning_rate": 2.8168735632183907e-05,
+      "loss": 0.1398,
+      "step": 11238
+    },
+    {
+      "epoch": 38.755172413793105,
+      "grad_norm": 0.9889978170394897,
+      "learning_rate": 2.816827586206897e-05,
+      "loss": 0.1347,
+      "step": 11239
+    },
+    {
+      "epoch": 38.758620689655174,
+      "grad_norm": 0.7384815812110901,
+      "learning_rate": 2.8167816091954024e-05,
+      "loss": 0.1177,
+      "step": 11240
+    },
+    {
+      "epoch": 38.762068965517244,
+      "grad_norm": 0.8404116034507751,
+      "learning_rate": 2.816735632183908e-05,
+      "loss": 0.129,
+      "step": 11241
+    },
+    {
+      "epoch": 38.765517241379314,
+      "grad_norm": 0.6753665804862976,
+      "learning_rate": 2.8166896551724138e-05,
+      "loss": 0.1138,
+      "step": 11242
+    },
+    {
+      "epoch": 38.76896551724138,
+      "grad_norm": 0.9615057706832886,
+      "learning_rate": 2.8166436781609197e-05,
+      "loss": 0.1261,
+      "step": 11243
+    },
+    {
+      "epoch": 38.772413793103446,
+      "grad_norm": 1.154229998588562,
+      "learning_rate": 2.8165977011494255e-05,
+      "loss": 0.1195,
+      "step": 11244
+    },
+    {
+      "epoch": 38.775862068965516,
+      "grad_norm": 1.661726474761963,
+      "learning_rate": 2.816551724137931e-05,
+      "loss": 0.1417,
+      "step": 11245
+    },
+    {
+      "epoch": 38.779310344827586,
+      "grad_norm": 2.558492660522461,
+      "learning_rate": 2.8165057471264366e-05,
+      "loss": 0.2247,
+      "step": 11246
+    },
+    {
+      "epoch": 38.782758620689656,
+      "grad_norm": 0.5422631502151489,
+      "learning_rate": 2.8164597701149428e-05,
+      "loss": 0.1866,
+      "step": 11247
+    },
+    {
+      "epoch": 38.786206896551725,
+      "grad_norm": 0.5166971683502197,
+      "learning_rate": 2.8164137931034483e-05,
+      "loss": 0.201,
+      "step": 11248
+    },
+    {
+      "epoch": 38.789655172413795,
+      "grad_norm": 0.8265708088874817,
+      "learning_rate": 2.816367816091954e-05,
+      "loss": 0.1776,
+      "step": 11249
+    },
+    {
+      "epoch": 38.793103448275865,
+      "grad_norm": 1.0693845748901367,
+      "learning_rate": 2.8163218390804597e-05,
+      "loss": 0.1796,
+      "step": 11250
+    },
+    {
+      "epoch": 38.796551724137935,
+      "grad_norm": 0.6405310034751892,
+      "learning_rate": 2.8162758620689656e-05,
+      "loss": 0.1773,
+      "step": 11251
+    },
+    {
+      "epoch": 38.8,
+      "grad_norm": 0.7177334427833557,
+      "learning_rate": 2.8162298850574715e-05,
+      "loss": 0.1751,
+      "step": 11252
+    },
+    {
+      "epoch": 38.80344827586207,
+      "grad_norm": 0.7087022662162781,
+      "learning_rate": 2.816183908045977e-05,
+      "loss": 0.1711,
+      "step": 11253
+    },
+    {
+      "epoch": 38.80689655172414,
+      "grad_norm": 0.6847899556159973,
+      "learning_rate": 2.8161379310344825e-05,
+      "loss": 0.1366,
+      "step": 11254
+    },
+    {
+      "epoch": 38.810344827586206,
+      "grad_norm": 0.6740031838417053,
+      "learning_rate": 2.8160919540229887e-05,
+      "loss": 0.1679,
+      "step": 11255
+    },
+    {
+      "epoch": 38.813793103448276,
+      "grad_norm": 0.8846389651298523,
+      "learning_rate": 2.8160459770114943e-05,
+      "loss": 0.1506,
+      "step": 11256
+    },
+    {
+      "epoch": 38.817241379310346,
+      "grad_norm": 1.050904393196106,
+      "learning_rate": 2.816e-05,
+      "loss": 0.1412,
+      "step": 11257
+    },
+    {
+      "epoch": 38.820689655172416,
+      "grad_norm": 0.5958207845687866,
+      "learning_rate": 2.8159540229885057e-05,
+      "loss": 0.146,
+      "step": 11258
+    },
+    {
+      "epoch": 38.824137931034485,
+      "grad_norm": 1.9387481212615967,
+      "learning_rate": 2.8159080459770115e-05,
+      "loss": 0.1524,
+      "step": 11259
+    },
+    {
+      "epoch": 38.827586206896555,
+      "grad_norm": 0.675316572189331,
+      "learning_rate": 2.8158620689655174e-05,
+      "loss": 0.1672,
+      "step": 11260
+    },
+    {
+      "epoch": 38.83103448275862,
+      "grad_norm": 0.8230834007263184,
+      "learning_rate": 2.815816091954023e-05,
+      "loss": 0.1504,
+      "step": 11261
+    },
+    {
+      "epoch": 38.83448275862069,
+      "grad_norm": 0.6209582090377808,
+      "learning_rate": 2.8157701149425288e-05,
+      "loss": 0.1349,
+      "step": 11262
+    },
+    {
+      "epoch": 38.83793103448276,
+      "grad_norm": 1.9919519424438477,
+      "learning_rate": 2.8157241379310347e-05,
+      "loss": 0.1511,
+      "step": 11263
+    },
+    {
+      "epoch": 38.84137931034483,
+      "grad_norm": 0.7193061113357544,
+      "learning_rate": 2.8156781609195402e-05,
+      "loss": 0.1433,
+      "step": 11264
+    },
+    {
+      "epoch": 38.8448275862069,
+      "grad_norm": 1.1764130592346191,
+      "learning_rate": 2.815632183908046e-05,
+      "loss": 0.1382,
+      "step": 11265
+    },
+    {
+      "epoch": 38.84827586206897,
+      "grad_norm": 2.513911485671997,
+      "learning_rate": 2.8155862068965516e-05,
+      "loss": 0.1382,
+      "step": 11266
+    },
+    {
+      "epoch": 38.851724137931036,
+      "grad_norm": 0.906965970993042,
+      "learning_rate": 2.8155402298850578e-05,
+      "loss": 0.1118,
+      "step": 11267
+    },
+    {
+      "epoch": 38.855172413793106,
+      "grad_norm": 0.846393883228302,
+      "learning_rate": 2.8154942528735633e-05,
+      "loss": 0.121,
+      "step": 11268
+    },
+    {
+      "epoch": 38.858620689655176,
+      "grad_norm": 1.06280517578125,
+      "learning_rate": 2.815448275862069e-05,
+      "loss": 0.1765,
+      "step": 11269
+    },
+    {
+      "epoch": 38.86206896551724,
+      "grad_norm": 1.3685685396194458,
+      "learning_rate": 2.8154022988505747e-05,
+      "loss": 0.1882,
+      "step": 11270
+    },
+    {
+      "epoch": 38.86551724137931,
+      "grad_norm": 0.7835462689399719,
+      "learning_rate": 2.8153563218390806e-05,
+      "loss": 0.2189,
+      "step": 11271
+    },
+    {
+      "epoch": 38.86896551724138,
+      "grad_norm": 0.5669861435890198,
+      "learning_rate": 2.8153103448275865e-05,
+      "loss": 0.1614,
+      "step": 11272
+    },
+    {
+      "epoch": 38.87241379310345,
+      "grad_norm": 0.8797566294670105,
+      "learning_rate": 2.815264367816092e-05,
+      "loss": 0.1926,
+      "step": 11273
+    },
+    {
+      "epoch": 38.87586206896552,
+      "grad_norm": 0.6453063488006592,
+      "learning_rate": 2.8152183908045975e-05,
+      "loss": 0.1725,
+      "step": 11274
+    },
+    {
+      "epoch": 38.87931034482759,
+      "grad_norm": 0.6178017258644104,
+      "learning_rate": 2.8151724137931037e-05,
+      "loss": 0.1767,
+      "step": 11275
+    },
+    {
+      "epoch": 38.88275862068966,
+      "grad_norm": 0.8502981662750244,
+      "learning_rate": 2.8151264367816093e-05,
+      "loss": 0.1924,
+      "step": 11276
+    },
+    {
+      "epoch": 38.88620689655173,
+      "grad_norm": 0.5046601295471191,
+      "learning_rate": 2.8150804597701148e-05,
+      "loss": 0.1677,
+      "step": 11277
+    },
+    {
+      "epoch": 38.889655172413796,
+      "grad_norm": 0.8824790716171265,
+      "learning_rate": 2.8150344827586207e-05,
+      "loss": 0.1649,
+      "step": 11278
+    },
+    {
+      "epoch": 38.89310344827586,
+      "grad_norm": 0.547552764415741,
+      "learning_rate": 2.8149885057471265e-05,
+      "loss": 0.169,
+      "step": 11279
+    },
+    {
+      "epoch": 38.89655172413793,
+      "grad_norm": 0.5297574400901794,
+      "learning_rate": 2.8149425287356324e-05,
+      "loss": 0.1518,
+      "step": 11280
+    },
+    {
+      "epoch": 38.9,
+      "grad_norm": 0.6053862571716309,
+      "learning_rate": 2.814896551724138e-05,
+      "loss": 0.1612,
+      "step": 11281
+    },
+    {
+      "epoch": 38.90344827586207,
+      "grad_norm": 0.7442946434020996,
+      "learning_rate": 2.8148505747126435e-05,
+      "loss": 0.1471,
+      "step": 11282
+    },
+    {
+      "epoch": 38.90689655172414,
+      "grad_norm": 0.6418216228485107,
+      "learning_rate": 2.8148045977011497e-05,
+      "loss": 0.1492,
+      "step": 11283
+    },
+    {
+      "epoch": 38.91034482758621,
+      "grad_norm": 0.5186421275138855,
+      "learning_rate": 2.8147586206896552e-05,
+      "loss": 0.1468,
+      "step": 11284
+    },
+    {
+      "epoch": 38.91379310344828,
+      "grad_norm": 0.6584075093269348,
+      "learning_rate": 2.814712643678161e-05,
+      "loss": 0.1599,
+      "step": 11285
+    },
+    {
+      "epoch": 38.91724137931035,
+      "grad_norm": 0.5549549460411072,
+      "learning_rate": 2.8146666666666666e-05,
+      "loss": 0.1358,
+      "step": 11286
+    },
+    {
+      "epoch": 38.92068965517242,
+      "grad_norm": 0.79872065782547,
+      "learning_rate": 2.8146206896551725e-05,
+      "loss": 0.1275,
+      "step": 11287
+    },
+    {
+      "epoch": 38.92413793103448,
+      "grad_norm": 0.7759601473808289,
+      "learning_rate": 2.8145747126436783e-05,
+      "loss": 0.1221,
+      "step": 11288
+    },
+    {
+      "epoch": 38.92758620689655,
+      "grad_norm": 1.0268313884735107,
+      "learning_rate": 2.814528735632184e-05,
+      "loss": 0.1338,
+      "step": 11289
+    },
+    {
+      "epoch": 38.93103448275862,
+      "grad_norm": 0.8918996453285217,
+      "learning_rate": 2.8144827586206897e-05,
+      "loss": 0.1302,
+      "step": 11290
+    },
+    {
+      "epoch": 38.93448275862069,
+      "grad_norm": 0.6283668279647827,
+      "learning_rate": 2.8144367816091956e-05,
+      "loss": 0.1472,
+      "step": 11291
+    },
+    {
+      "epoch": 38.93793103448276,
+      "grad_norm": 0.7148279547691345,
+      "learning_rate": 2.814390804597701e-05,
+      "loss": 0.1096,
+      "step": 11292
+    },
+    {
+      "epoch": 38.94137931034483,
+      "grad_norm": 0.8609217405319214,
+      "learning_rate": 2.814344827586207e-05,
+      "loss": 0.1181,
+      "step": 11293
+    },
+    {
+      "epoch": 38.9448275862069,
+      "grad_norm": 2.1705758571624756,
+      "learning_rate": 2.8142988505747125e-05,
+      "loss": 0.1173,
+      "step": 11294
+    },
+    {
+      "epoch": 38.94827586206897,
+      "grad_norm": 1.1649090051651,
+      "learning_rate": 2.8142528735632187e-05,
+      "loss": 0.1902,
+      "step": 11295
+    },
+    {
+      "epoch": 38.95172413793104,
+      "grad_norm": 0.7025328874588013,
+      "learning_rate": 2.8142068965517243e-05,
+      "loss": 0.1935,
+      "step": 11296
+    },
+    {
+      "epoch": 38.9551724137931,
+      "grad_norm": 0.7072032690048218,
+      "learning_rate": 2.8141609195402298e-05,
+      "loss": 0.17,
+      "step": 11297
+    },
+    {
+      "epoch": 38.95862068965517,
+      "grad_norm": 0.5120487809181213,
+      "learning_rate": 2.8141149425287357e-05,
+      "loss": 0.1778,
+      "step": 11298
+    },
+    {
+      "epoch": 38.96206896551724,
+      "grad_norm": 0.7953117489814758,
+      "learning_rate": 2.8140689655172415e-05,
+      "loss": 0.1643,
+      "step": 11299
+    },
+    {
+      "epoch": 38.96551724137931,
+      "grad_norm": 1.2818915843963623,
+      "learning_rate": 2.8140229885057474e-05,
+      "loss": 0.1648,
+      "step": 11300
+    },
+    {
+      "epoch": 38.96896551724138,
+      "grad_norm": 0.9333861470222473,
+      "learning_rate": 2.813977011494253e-05,
+      "loss": 0.1882,
+      "step": 11301
+    },
+    {
+      "epoch": 38.97241379310345,
+      "grad_norm": 0.685040295124054,
+      "learning_rate": 2.8139310344827585e-05,
+      "loss": 0.1538,
+      "step": 11302
+    },
+    {
+      "epoch": 38.97586206896552,
+      "grad_norm": 0.8104397058486938,
+      "learning_rate": 2.8138850574712647e-05,
+      "loss": 0.1505,
+      "step": 11303
+    },
+    {
+      "epoch": 38.97931034482759,
+      "grad_norm": 0.7120717763900757,
+      "learning_rate": 2.8138390804597702e-05,
+      "loss": 0.1573,
+      "step": 11304
+    },
+    {
+      "epoch": 38.98275862068966,
+      "grad_norm": 0.8509507179260254,
+      "learning_rate": 2.813793103448276e-05,
+      "loss": 0.1634,
+      "step": 11305
+    },
+    {
+      "epoch": 38.98620689655172,
+      "grad_norm": 1.565614104270935,
+      "learning_rate": 2.8137471264367816e-05,
+      "loss": 0.1192,
+      "step": 11306
+    },
+    {
+      "epoch": 38.98965517241379,
+      "grad_norm": 1.0979952812194824,
+      "learning_rate": 2.8137011494252875e-05,
+      "loss": 0.1307,
+      "step": 11307
+    },
+    {
+      "epoch": 38.99310344827586,
+      "grad_norm": 0.9961676001548767,
+      "learning_rate": 2.8136551724137933e-05,
+      "loss": 0.1287,
+      "step": 11308
+    },
+    {
+      "epoch": 38.99655172413793,
+      "grad_norm": 1.2895593643188477,
+      "learning_rate": 2.813609195402299e-05,
+      "loss": 0.1135,
+      "step": 11309
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 1.1720943450927734,
+      "learning_rate": 2.8135632183908044e-05,
+      "loss": 0.1684,
+      "step": 11310
+    },
+    {
+      "epoch": 39.00344827586207,
+      "grad_norm": 1.0253268480300903,
+      "learning_rate": 2.8135172413793106e-05,
+      "loss": 0.2464,
+      "step": 11311
+    },
+    {
+      "epoch": 39.00689655172414,
+      "grad_norm": 0.639373242855072,
+      "learning_rate": 2.813471264367816e-05,
+      "loss": 0.1996,
+      "step": 11312
+    },
+    {
+      "epoch": 39.01034482758621,
+      "grad_norm": 1.7809005975723267,
+      "learning_rate": 2.813425287356322e-05,
+      "loss": 0.1662,
+      "step": 11313
+    },
+    {
+      "epoch": 39.01379310344828,
+      "grad_norm": 0.6196741461753845,
+      "learning_rate": 2.8133793103448275e-05,
+      "loss": 0.1673,
+      "step": 11314
+    },
+    {
+      "epoch": 39.01724137931034,
+      "grad_norm": 0.8031917214393616,
+      "learning_rate": 2.8133333333333334e-05,
+      "loss": 0.1571,
+      "step": 11315
+    },
+    {
+      "epoch": 39.02068965517241,
+      "grad_norm": 0.6652750968933105,
+      "learning_rate": 2.8132873563218393e-05,
+      "loss": 0.1812,
+      "step": 11316
+    },
+    {
+      "epoch": 39.02413793103448,
+      "grad_norm": 0.6250159740447998,
+      "learning_rate": 2.8132413793103448e-05,
+      "loss": 0.1665,
+      "step": 11317
+    },
+    {
+      "epoch": 39.02758620689655,
+      "grad_norm": 0.5972087383270264,
+      "learning_rate": 2.8131954022988507e-05,
+      "loss": 0.1731,
+      "step": 11318
+    },
+    {
+      "epoch": 39.03103448275862,
+      "grad_norm": 1.7323660850524902,
+      "learning_rate": 2.8131494252873565e-05,
+      "loss": 0.1429,
+      "step": 11319
+    },
+    {
+      "epoch": 39.03448275862069,
+      "grad_norm": 0.7930275797843933,
+      "learning_rate": 2.813103448275862e-05,
+      "loss": 0.1575,
+      "step": 11320
+    },
+    {
+      "epoch": 39.03793103448276,
+      "grad_norm": 0.691792905330658,
+      "learning_rate": 2.813057471264368e-05,
+      "loss": 0.1434,
+      "step": 11321
+    },
+    {
+      "epoch": 39.04137931034483,
+      "grad_norm": 0.5114172101020813,
+      "learning_rate": 2.8130114942528735e-05,
+      "loss": 0.1478,
+      "step": 11322
+    },
+    {
+      "epoch": 39.0448275862069,
+      "grad_norm": 0.8113369941711426,
+      "learning_rate": 2.8129655172413797e-05,
+      "loss": 0.1446,
+      "step": 11323
+    },
+    {
+      "epoch": 39.04827586206896,
+      "grad_norm": 0.5466988682746887,
+      "learning_rate": 2.8129195402298852e-05,
+      "loss": 0.1445,
+      "step": 11324
+    },
+    {
+      "epoch": 39.05172413793103,
+      "grad_norm": 0.8558183312416077,
+      "learning_rate": 2.8128735632183907e-05,
+      "loss": 0.1358,
+      "step": 11325
+    },
+    {
+      "epoch": 39.0551724137931,
+      "grad_norm": 0.6201282739639282,
+      "learning_rate": 2.8128275862068966e-05,
+      "loss": 0.1212,
+      "step": 11326
+    },
+    {
+      "epoch": 39.05862068965517,
+      "grad_norm": 0.5900123119354248,
+      "learning_rate": 2.8127816091954025e-05,
+      "loss": 0.1322,
+      "step": 11327
+    },
+    {
+      "epoch": 39.06206896551724,
+      "grad_norm": 0.5589483976364136,
+      "learning_rate": 2.8127356321839083e-05,
+      "loss": 0.1284,
+      "step": 11328
+    },
+    {
+      "epoch": 39.06551724137931,
+      "grad_norm": 1.0415207147598267,
+      "learning_rate": 2.812689655172414e-05,
+      "loss": 0.1082,
+      "step": 11329
+    },
+    {
+      "epoch": 39.06896551724138,
+      "grad_norm": 1.2627681493759155,
+      "learning_rate": 2.8126436781609194e-05,
+      "loss": 0.1178,
+      "step": 11330
+    },
+    {
+      "epoch": 39.07241379310345,
+      "grad_norm": 0.8590753078460693,
+      "learning_rate": 2.8125977011494256e-05,
+      "loss": 0.1139,
+      "step": 11331
+    },
+    {
+      "epoch": 39.07586206896552,
+      "grad_norm": 0.6881011128425598,
+      "learning_rate": 2.812551724137931e-05,
+      "loss": 0.0885,
+      "step": 11332
+    },
+    {
+      "epoch": 39.07931034482758,
+      "grad_norm": 0.6686098575592041,
+      "learning_rate": 2.812505747126437e-05,
+      "loss": 0.0989,
+      "step": 11333
+    },
+    {
+      "epoch": 39.08275862068965,
+      "grad_norm": 1.1820886135101318,
+      "learning_rate": 2.8124597701149425e-05,
+      "loss": 0.1161,
+      "step": 11334
+    },
+    {
+      "epoch": 39.08620689655172,
+      "grad_norm": 1.5159265995025635,
+      "learning_rate": 2.8124137931034484e-05,
+      "loss": 0.1628,
+      "step": 11335
+    },
+    {
+      "epoch": 39.08965517241379,
+      "grad_norm": 0.8943869471549988,
+      "learning_rate": 2.8123678160919543e-05,
+      "loss": 0.2254,
+      "step": 11336
+    },
+    {
+      "epoch": 39.09310344827586,
+      "grad_norm": 0.6968327760696411,
+      "learning_rate": 2.8123218390804598e-05,
+      "loss": 0.2095,
+      "step": 11337
+    },
+    {
+      "epoch": 39.09655172413793,
+      "grad_norm": 1.3079099655151367,
+      "learning_rate": 2.8122758620689653e-05,
+      "loss": 0.1567,
+      "step": 11338
+    },
+    {
+      "epoch": 39.1,
+      "grad_norm": 1.5072187185287476,
+      "learning_rate": 2.8122298850574715e-05,
+      "loss": 0.1649,
+      "step": 11339
+    },
+    {
+      "epoch": 39.10344827586207,
+      "grad_norm": 0.6269375085830688,
+      "learning_rate": 2.812183908045977e-05,
+      "loss": 0.1536,
+      "step": 11340
+    },
+    {
+      "epoch": 39.10689655172414,
+      "grad_norm": 0.9921271204948425,
+      "learning_rate": 2.812137931034483e-05,
+      "loss": 0.175,
+      "step": 11341
+    },
+    {
+      "epoch": 39.110344827586204,
+      "grad_norm": 1.8879188299179077,
+      "learning_rate": 2.8120919540229884e-05,
+      "loss": 0.167,
+      "step": 11342
+    },
+    {
+      "epoch": 39.11379310344827,
+      "grad_norm": 0.7134392261505127,
+      "learning_rate": 2.8120459770114943e-05,
+      "loss": 0.1619,
+      "step": 11343
+    },
+    {
+      "epoch": 39.11724137931034,
+      "grad_norm": 0.651421844959259,
+      "learning_rate": 2.8120000000000002e-05,
+      "loss": 0.1367,
+      "step": 11344
+    },
+    {
+      "epoch": 39.12068965517241,
+      "grad_norm": 0.7347873449325562,
+      "learning_rate": 2.8119540229885057e-05,
+      "loss": 0.1416,
+      "step": 11345
+    },
+    {
+      "epoch": 39.12413793103448,
+      "grad_norm": 0.6486946940422058,
+      "learning_rate": 2.8119080459770116e-05,
+      "loss": 0.1481,
+      "step": 11346
+    },
+    {
+      "epoch": 39.12758620689655,
+      "grad_norm": 1.0482981204986572,
+      "learning_rate": 2.8118620689655174e-05,
+      "loss": 0.1384,
+      "step": 11347
+    },
+    {
+      "epoch": 39.13103448275862,
+      "grad_norm": 1.5896856784820557,
+      "learning_rate": 2.811816091954023e-05,
+      "loss": 0.148,
+      "step": 11348
+    },
+    {
+      "epoch": 39.13448275862069,
+      "grad_norm": 0.5301230549812317,
+      "learning_rate": 2.811770114942529e-05,
+      "loss": 0.1105,
+      "step": 11349
+    },
+    {
+      "epoch": 39.13793103448276,
+      "grad_norm": 0.8420941233634949,
+      "learning_rate": 2.8117241379310344e-05,
+      "loss": 0.1358,
+      "step": 11350
+    },
+    {
+      "epoch": 39.141379310344824,
+      "grad_norm": 1.6486687660217285,
+      "learning_rate": 2.8116781609195406e-05,
+      "loss": 0.1194,
+      "step": 11351
+    },
+    {
+      "epoch": 39.144827586206894,
+      "grad_norm": 0.5849422812461853,
+      "learning_rate": 2.811632183908046e-05,
+      "loss": 0.1264,
+      "step": 11352
+    },
+    {
+      "epoch": 39.148275862068964,
+      "grad_norm": 0.7122141718864441,
+      "learning_rate": 2.8115862068965516e-05,
+      "loss": 0.1453,
+      "step": 11353
+    },
+    {
+      "epoch": 39.15172413793103,
+      "grad_norm": 1.420987844467163,
+      "learning_rate": 2.8115402298850575e-05,
+      "loss": 0.1343,
+      "step": 11354
+    },
+    {
+      "epoch": 39.1551724137931,
+      "grad_norm": 2.1680684089660645,
+      "learning_rate": 2.8114942528735634e-05,
+      "loss": 0.1274,
+      "step": 11355
+    },
+    {
+      "epoch": 39.15862068965517,
+      "grad_norm": 0.5910596251487732,
+      "learning_rate": 2.8114482758620692e-05,
+      "loss": 0.1037,
+      "step": 11356
+    },
+    {
+      "epoch": 39.16206896551724,
+      "grad_norm": 1.0430805683135986,
+      "learning_rate": 2.8114022988505748e-05,
+      "loss": 0.1164,
+      "step": 11357
+    },
+    {
+      "epoch": 39.16551724137931,
+      "grad_norm": 1.087209939956665,
+      "learning_rate": 2.8113563218390803e-05,
+      "loss": 0.092,
+      "step": 11358
+    },
+    {
+      "epoch": 39.16896551724138,
+      "grad_norm": 0.7241266965866089,
+      "learning_rate": 2.8113103448275865e-05,
+      "loss": 0.1064,
+      "step": 11359
+    },
+    {
+      "epoch": 39.172413793103445,
+      "grad_norm": 2.1909291744232178,
+      "learning_rate": 2.811264367816092e-05,
+      "loss": 0.1136,
+      "step": 11360
+    },
+    {
+      "epoch": 39.175862068965515,
+      "grad_norm": 0.7654843926429749,
+      "learning_rate": 2.811218390804598e-05,
+      "loss": 0.2248,
+      "step": 11361
+    },
+    {
+      "epoch": 39.179310344827584,
+      "grad_norm": 1.2009152173995972,
+      "learning_rate": 2.8111724137931034e-05,
+      "loss": 0.1706,
+      "step": 11362
+    },
+    {
+      "epoch": 39.182758620689654,
+      "grad_norm": 0.9743526577949524,
+      "learning_rate": 2.8111264367816093e-05,
+      "loss": 0.1729,
+      "step": 11363
+    },
+    {
+      "epoch": 39.186206896551724,
+      "grad_norm": 0.7971171140670776,
+      "learning_rate": 2.8110804597701152e-05,
+      "loss": 0.154,
+      "step": 11364
+    },
+    {
+      "epoch": 39.189655172413794,
+      "grad_norm": 0.5320194363594055,
+      "learning_rate": 2.8110344827586207e-05,
+      "loss": 0.1562,
+      "step": 11365
+    },
+    {
+      "epoch": 39.19310344827586,
+      "grad_norm": 1.0084235668182373,
+      "learning_rate": 2.8109885057471262e-05,
+      "loss": 0.1559,
+      "step": 11366
+    },
+    {
+      "epoch": 39.19655172413793,
+      "grad_norm": 0.5684787034988403,
+      "learning_rate": 2.8109425287356324e-05,
+      "loss": 0.1559,
+      "step": 11367
+    },
+    {
+      "epoch": 39.2,
+      "grad_norm": 0.8705762624740601,
+      "learning_rate": 2.810896551724138e-05,
+      "loss": 0.1604,
+      "step": 11368
+    },
+    {
+      "epoch": 39.203448275862065,
+      "grad_norm": 1.0085281133651733,
+      "learning_rate": 2.810850574712644e-05,
+      "loss": 0.1464,
+      "step": 11369
+    },
+    {
+      "epoch": 39.206896551724135,
+      "grad_norm": 0.5042669773101807,
+      "learning_rate": 2.8108045977011494e-05,
+      "loss": 0.1336,
+      "step": 11370
+    },
+    {
+      "epoch": 39.210344827586205,
+      "grad_norm": 0.614955484867096,
+      "learning_rate": 2.8107586206896552e-05,
+      "loss": 0.1266,
+      "step": 11371
+    },
+    {
+      "epoch": 39.213793103448275,
+      "grad_norm": 1.8639941215515137,
+      "learning_rate": 2.810712643678161e-05,
+      "loss": 0.176,
+      "step": 11372
+    },
+    {
+      "epoch": 39.217241379310344,
+      "grad_norm": 0.504321277141571,
+      "learning_rate": 2.8106666666666666e-05,
+      "loss": 0.1206,
+      "step": 11373
+    },
+    {
+      "epoch": 39.220689655172414,
+      "grad_norm": 0.8796669840812683,
+      "learning_rate": 2.8106206896551725e-05,
+      "loss": 0.1282,
+      "step": 11374
+    },
+    {
+      "epoch": 39.224137931034484,
+      "grad_norm": 0.7915198802947998,
+      "learning_rate": 2.8105747126436784e-05,
+      "loss": 0.1222,
+      "step": 11375
+    },
+    {
+      "epoch": 39.227586206896554,
+      "grad_norm": 0.6976243257522583,
+      "learning_rate": 2.810528735632184e-05,
+      "loss": 0.1445,
+      "step": 11376
+    },
+    {
+      "epoch": 39.23103448275862,
+      "grad_norm": 0.7113633751869202,
+      "learning_rate": 2.8104827586206898e-05,
+      "loss": 0.1292,
+      "step": 11377
+    },
+    {
+      "epoch": 39.234482758620686,
+      "grad_norm": 1.2345213890075684,
+      "learning_rate": 2.8104367816091953e-05,
+      "loss": 0.1336,
+      "step": 11378
+    },
+    {
+      "epoch": 39.237931034482756,
+      "grad_norm": 0.7949514389038086,
+      "learning_rate": 2.8103908045977015e-05,
+      "loss": 0.1245,
+      "step": 11379
+    },
+    {
+      "epoch": 39.241379310344826,
+      "grad_norm": 0.6540402173995972,
+      "learning_rate": 2.810344827586207e-05,
+      "loss": 0.1139,
+      "step": 11380
+    },
+    {
+      "epoch": 39.244827586206895,
+      "grad_norm": 0.9986185431480408,
+      "learning_rate": 2.8102988505747126e-05,
+      "loss": 0.124,
+      "step": 11381
+    },
+    {
+      "epoch": 39.248275862068965,
+      "grad_norm": 0.6179977059364319,
+      "learning_rate": 2.8102528735632184e-05,
+      "loss": 0.1214,
+      "step": 11382
+    },
+    {
+      "epoch": 39.251724137931035,
+      "grad_norm": 2.956439733505249,
+      "learning_rate": 2.8102068965517243e-05,
+      "loss": 0.1219,
+      "step": 11383
+    },
+    {
+      "epoch": 39.255172413793105,
+      "grad_norm": 2.9646363258361816,
+      "learning_rate": 2.8101609195402302e-05,
+      "loss": 0.108,
+      "step": 11384
+    },
+    {
+      "epoch": 39.258620689655174,
+      "grad_norm": 1.1277996301651,
+      "learning_rate": 2.8101149425287357e-05,
+      "loss": 0.1663,
+      "step": 11385
+    },
+    {
+      "epoch": 39.262068965517244,
+      "grad_norm": 0.8954567313194275,
+      "learning_rate": 2.8100689655172412e-05,
+      "loss": 0.2158,
+      "step": 11386
+    },
+    {
+      "epoch": 39.265517241379314,
+      "grad_norm": 0.5649714469909668,
+      "learning_rate": 2.8100229885057474e-05,
+      "loss": 0.1986,
+      "step": 11387
+    },
+    {
+      "epoch": 39.26896551724138,
+      "grad_norm": 0.4658292233943939,
+      "learning_rate": 2.809977011494253e-05,
+      "loss": 0.176,
+      "step": 11388
+    },
+    {
+      "epoch": 39.272413793103446,
+      "grad_norm": 0.6606435179710388,
+      "learning_rate": 2.809931034482759e-05,
+      "loss": 0.157,
+      "step": 11389
+    },
+    {
+      "epoch": 39.275862068965516,
+      "grad_norm": 0.6109241247177124,
+      "learning_rate": 2.8098850574712644e-05,
+      "loss": 0.1645,
+      "step": 11390
+    },
+    {
+      "epoch": 39.279310344827586,
+      "grad_norm": 0.71871417760849,
+      "learning_rate": 2.8098390804597702e-05,
+      "loss": 0.1765,
+      "step": 11391
+    },
+    {
+      "epoch": 39.282758620689656,
+      "grad_norm": 0.9430921077728271,
+      "learning_rate": 2.809793103448276e-05,
+      "loss": 0.1626,
+      "step": 11392
+    },
+    {
+      "epoch": 39.286206896551725,
+      "grad_norm": 1.2927865982055664,
+      "learning_rate": 2.8097471264367816e-05,
+      "loss": 0.1439,
+      "step": 11393
+    },
+    {
+      "epoch": 39.289655172413795,
+      "grad_norm": 0.7112196683883667,
+      "learning_rate": 2.8097011494252875e-05,
+      "loss": 0.131,
+      "step": 11394
+    },
+    {
+      "epoch": 39.293103448275865,
+      "grad_norm": 0.6491581201553345,
+      "learning_rate": 2.8096551724137934e-05,
+      "loss": 0.1389,
+      "step": 11395
+    },
+    {
+      "epoch": 39.296551724137935,
+      "grad_norm": 0.8715078830718994,
+      "learning_rate": 2.809609195402299e-05,
+      "loss": 0.1482,
+      "step": 11396
+    },
+    {
+      "epoch": 39.3,
+      "grad_norm": 1.048891544342041,
+      "learning_rate": 2.8095632183908048e-05,
+      "loss": 0.1618,
+      "step": 11397
+    },
+    {
+      "epoch": 39.30344827586207,
+      "grad_norm": 1.1776371002197266,
+      "learning_rate": 2.8095172413793103e-05,
+      "loss": 0.1521,
+      "step": 11398
+    },
+    {
+      "epoch": 39.30689655172414,
+      "grad_norm": 1.1293679475784302,
+      "learning_rate": 2.809471264367816e-05,
+      "loss": 0.143,
+      "step": 11399
+    },
+    {
+      "epoch": 39.310344827586206,
+      "grad_norm": 0.6758058071136475,
+      "learning_rate": 2.809425287356322e-05,
+      "loss": 0.1583,
+      "step": 11400
+    },
+    {
+      "epoch": 39.313793103448276,
+      "grad_norm": 2.2403564453125,
+      "learning_rate": 2.8093793103448276e-05,
+      "loss": 0.1384,
+      "step": 11401
+    },
+    {
+      "epoch": 39.317241379310346,
+      "grad_norm": 0.687269926071167,
+      "learning_rate": 2.8093333333333334e-05,
+      "loss": 0.1387,
+      "step": 11402
+    },
+    {
+      "epoch": 39.320689655172416,
+      "grad_norm": 1.706756830215454,
+      "learning_rate": 2.8092873563218393e-05,
+      "loss": 0.1237,
+      "step": 11403
+    },
+    {
+      "epoch": 39.324137931034485,
+      "grad_norm": 1.2068983316421509,
+      "learning_rate": 2.8092413793103448e-05,
+      "loss": 0.1263,
+      "step": 11404
+    },
+    {
+      "epoch": 39.327586206896555,
+      "grad_norm": 0.5400568842887878,
+      "learning_rate": 2.8091954022988507e-05,
+      "loss": 0.1114,
+      "step": 11405
+    },
+    {
+      "epoch": 39.33103448275862,
+      "grad_norm": 1.2803703546524048,
+      "learning_rate": 2.8091494252873562e-05,
+      "loss": 0.1056,
+      "step": 11406
+    },
+    {
+      "epoch": 39.33448275862069,
+      "grad_norm": 0.6489802002906799,
+      "learning_rate": 2.8091034482758624e-05,
+      "loss": 0.1099,
+      "step": 11407
+    },
+    {
+      "epoch": 39.33793103448276,
+      "grad_norm": 0.6319431066513062,
+      "learning_rate": 2.809057471264368e-05,
+      "loss": 0.1008,
+      "step": 11408
+    },
+    {
+      "epoch": 39.34137931034483,
+      "grad_norm": 0.856367826461792,
+      "learning_rate": 2.8090114942528735e-05,
+      "loss": 0.1146,
+      "step": 11409
+    },
+    {
+      "epoch": 39.3448275862069,
+      "grad_norm": 1.344766616821289,
+      "learning_rate": 2.8089655172413794e-05,
+      "loss": 0.1619,
+      "step": 11410
+    },
+    {
+      "epoch": 39.34827586206897,
+      "grad_norm": 0.6225053668022156,
+      "learning_rate": 2.8089195402298852e-05,
+      "loss": 0.2241,
+      "step": 11411
+    },
+    {
+      "epoch": 39.351724137931036,
+      "grad_norm": 0.4927051067352295,
+      "learning_rate": 2.808873563218391e-05,
+      "loss": 0.1713,
+      "step": 11412
+    },
+    {
+      "epoch": 39.355172413793106,
+      "grad_norm": 0.4878547191619873,
+      "learning_rate": 2.8088275862068966e-05,
+      "loss": 0.1898,
+      "step": 11413
+    },
+    {
+      "epoch": 39.358620689655176,
+      "grad_norm": 0.453934907913208,
+      "learning_rate": 2.808781609195402e-05,
+      "loss": 0.1788,
+      "step": 11414
+    },
+    {
+      "epoch": 39.36206896551724,
+      "grad_norm": 0.5316328406333923,
+      "learning_rate": 2.8087356321839084e-05,
+      "loss": 0.1842,
+      "step": 11415
+    },
+    {
+      "epoch": 39.36551724137931,
+      "grad_norm": 1.2046620845794678,
+      "learning_rate": 2.808689655172414e-05,
+      "loss": 0.139,
+      "step": 11416
+    },
+    {
+      "epoch": 39.36896551724138,
+      "grad_norm": 0.4694715142250061,
+      "learning_rate": 2.8086436781609198e-05,
+      "loss": 0.1581,
+      "step": 11417
+    },
+    {
+      "epoch": 39.37241379310345,
+      "grad_norm": 0.6577185392379761,
+      "learning_rate": 2.8085977011494253e-05,
+      "loss": 0.1464,
+      "step": 11418
+    },
+    {
+      "epoch": 39.37586206896552,
+      "grad_norm": 1.0235077142715454,
+      "learning_rate": 2.808551724137931e-05,
+      "loss": 0.1544,
+      "step": 11419
+    },
+    {
+      "epoch": 39.37931034482759,
+      "grad_norm": 1.173883318901062,
+      "learning_rate": 2.808505747126437e-05,
+      "loss": 0.1544,
+      "step": 11420
+    },
+    {
+      "epoch": 39.38275862068966,
+      "grad_norm": 0.7688548564910889,
+      "learning_rate": 2.8084597701149426e-05,
+      "loss": 0.1423,
+      "step": 11421
+    },
+    {
+      "epoch": 39.38620689655173,
+      "grad_norm": 0.7255688309669495,
+      "learning_rate": 2.8084137931034484e-05,
+      "loss": 0.1305,
+      "step": 11422
+    },
+    {
+      "epoch": 39.389655172413796,
+      "grad_norm": 1.0433151721954346,
+      "learning_rate": 2.8083678160919543e-05,
+      "loss": 0.1377,
+      "step": 11423
+    },
+    {
+      "epoch": 39.39310344827586,
+      "grad_norm": 0.4874139726161957,
+      "learning_rate": 2.8083218390804598e-05,
+      "loss": 0.1235,
+      "step": 11424
+    },
+    {
+      "epoch": 39.39655172413793,
+      "grad_norm": 0.7005029320716858,
+      "learning_rate": 2.8082758620689657e-05,
+      "loss": 0.1524,
+      "step": 11425
+    },
+    {
+      "epoch": 39.4,
+      "grad_norm": 0.920814573764801,
+      "learning_rate": 2.8082298850574712e-05,
+      "loss": 0.1193,
+      "step": 11426
+    },
+    {
+      "epoch": 39.40344827586207,
+      "grad_norm": 0.5916476249694824,
+      "learning_rate": 2.808183908045977e-05,
+      "loss": 0.1169,
+      "step": 11427
+    },
+    {
+      "epoch": 39.40689655172414,
+      "grad_norm": 1.0757548809051514,
+      "learning_rate": 2.808137931034483e-05,
+      "loss": 0.1067,
+      "step": 11428
+    },
+    {
+      "epoch": 39.41034482758621,
+      "grad_norm": 0.9174897074699402,
+      "learning_rate": 2.8080919540229885e-05,
+      "loss": 0.104,
+      "step": 11429
+    },
+    {
+      "epoch": 39.41379310344828,
+      "grad_norm": 0.9467368721961975,
+      "learning_rate": 2.8080459770114944e-05,
+      "loss": 0.1335,
+      "step": 11430
+    },
+    {
+      "epoch": 39.41724137931035,
+      "grad_norm": 0.710720419883728,
+      "learning_rate": 2.8080000000000002e-05,
+      "loss": 0.1208,
+      "step": 11431
+    },
+    {
+      "epoch": 39.42068965517242,
+      "grad_norm": 0.6420710682868958,
+      "learning_rate": 2.8079540229885057e-05,
+      "loss": 0.1098,
+      "step": 11432
+    },
+    {
+      "epoch": 39.42413793103448,
+      "grad_norm": 1.1723577976226807,
+      "learning_rate": 2.8079080459770116e-05,
+      "loss": 0.1107,
+      "step": 11433
+    },
+    {
+      "epoch": 39.42758620689655,
+      "grad_norm": 1.0184645652770996,
+      "learning_rate": 2.807862068965517e-05,
+      "loss": 0.1288,
+      "step": 11434
+    },
+    {
+      "epoch": 39.43103448275862,
+      "grad_norm": 2.0558419227600098,
+      "learning_rate": 2.8078160919540234e-05,
+      "loss": 0.1802,
+      "step": 11435
+    },
+    {
+      "epoch": 39.43448275862069,
+      "grad_norm": 0.6640236973762512,
+      "learning_rate": 2.807770114942529e-05,
+      "loss": 0.2107,
+      "step": 11436
+    },
+    {
+      "epoch": 39.43793103448276,
+      "grad_norm": 0.520627498626709,
+      "learning_rate": 2.8077241379310344e-05,
+      "loss": 0.1906,
+      "step": 11437
+    },
+    {
+      "epoch": 39.44137931034483,
+      "grad_norm": 0.627945065498352,
+      "learning_rate": 2.8076781609195403e-05,
+      "loss": 0.1929,
+      "step": 11438
+    },
+    {
+      "epoch": 39.4448275862069,
+      "grad_norm": 0.6437147855758667,
+      "learning_rate": 2.807632183908046e-05,
+      "loss": 0.1545,
+      "step": 11439
+    },
+    {
+      "epoch": 39.44827586206897,
+      "grad_norm": 0.6573736667633057,
+      "learning_rate": 2.807586206896552e-05,
+      "loss": 0.1771,
+      "step": 11440
+    },
+    {
+      "epoch": 39.45172413793104,
+      "grad_norm": 0.9056336283683777,
+      "learning_rate": 2.8075402298850575e-05,
+      "loss": 0.1801,
+      "step": 11441
+    },
+    {
+      "epoch": 39.4551724137931,
+      "grad_norm": 0.6939224600791931,
+      "learning_rate": 2.807494252873563e-05,
+      "loss": 0.1693,
+      "step": 11442
+    },
+    {
+      "epoch": 39.45862068965517,
+      "grad_norm": 0.6820400357246399,
+      "learning_rate": 2.8074482758620693e-05,
+      "loss": 0.1579,
+      "step": 11443
+    },
+    {
+      "epoch": 39.46206896551724,
+      "grad_norm": 0.7561465501785278,
+      "learning_rate": 2.8074022988505748e-05,
+      "loss": 0.1601,
+      "step": 11444
+    },
+    {
+      "epoch": 39.46551724137931,
+      "grad_norm": 0.6759411692619324,
+      "learning_rate": 2.8073563218390807e-05,
+      "loss": 0.1404,
+      "step": 11445
+    },
+    {
+      "epoch": 39.46896551724138,
+      "grad_norm": 0.7630171775817871,
+      "learning_rate": 2.8073103448275862e-05,
+      "loss": 0.1649,
+      "step": 11446
+    },
+    {
+      "epoch": 39.47241379310345,
+      "grad_norm": 0.9129201769828796,
+      "learning_rate": 2.8072643678160917e-05,
+      "loss": 0.151,
+      "step": 11447
+    },
+    {
+      "epoch": 39.47586206896552,
+      "grad_norm": 1.050868034362793,
+      "learning_rate": 2.807218390804598e-05,
+      "loss": 0.1401,
+      "step": 11448
+    },
+    {
+      "epoch": 39.47931034482759,
+      "grad_norm": 0.9194563031196594,
+      "learning_rate": 2.8071724137931035e-05,
+      "loss": 0.1595,
+      "step": 11449
+    },
+    {
+      "epoch": 39.48275862068966,
+      "grad_norm": 0.6321210861206055,
+      "learning_rate": 2.8071264367816093e-05,
+      "loss": 0.134,
+      "step": 11450
+    },
+    {
+      "epoch": 39.48620689655172,
+      "grad_norm": 5.3997721672058105,
+      "learning_rate": 2.807080459770115e-05,
+      "loss": 0.1179,
+      "step": 11451
+    },
+    {
+      "epoch": 39.48965517241379,
+      "grad_norm": 0.8643282651901245,
+      "learning_rate": 2.8070344827586207e-05,
+      "loss": 0.1468,
+      "step": 11452
+    },
+    {
+      "epoch": 39.49310344827586,
+      "grad_norm": 0.9580731391906738,
+      "learning_rate": 2.8069885057471266e-05,
+      "loss": 0.1253,
+      "step": 11453
+    },
+    {
+      "epoch": 39.49655172413793,
+      "grad_norm": 0.7964051961898804,
+      "learning_rate": 2.806942528735632e-05,
+      "loss": 0.1304,
+      "step": 11454
+    },
+    {
+      "epoch": 39.5,
+      "grad_norm": 1.4830408096313477,
+      "learning_rate": 2.8068965517241377e-05,
+      "loss": 0.1212,
+      "step": 11455
+    },
+    {
+      "epoch": 39.50344827586207,
+      "grad_norm": 0.704849362373352,
+      "learning_rate": 2.806850574712644e-05,
+      "loss": 0.1382,
+      "step": 11456
+    },
+    {
+      "epoch": 39.50689655172414,
+      "grad_norm": 0.8388257622718811,
+      "learning_rate": 2.8068045977011494e-05,
+      "loss": 0.1024,
+      "step": 11457
+    },
+    {
+      "epoch": 39.51034482758621,
+      "grad_norm": 1.078876256942749,
+      "learning_rate": 2.8067586206896553e-05,
+      "loss": 0.1232,
+      "step": 11458
+    },
+    {
+      "epoch": 39.51379310344828,
+      "grad_norm": 0.810330331325531,
+      "learning_rate": 2.8067126436781608e-05,
+      "loss": 0.134,
+      "step": 11459
+    },
+    {
+      "epoch": 39.51724137931034,
+      "grad_norm": 1.557500958442688,
+      "learning_rate": 2.8066666666666667e-05,
+      "loss": 0.1544,
+      "step": 11460
+    },
+    {
+      "epoch": 39.52068965517241,
+      "grad_norm": 0.7689520716667175,
+      "learning_rate": 2.8066206896551725e-05,
+      "loss": 0.2147,
+      "step": 11461
+    },
+    {
+      "epoch": 39.52413793103448,
+      "grad_norm": 0.766836404800415,
+      "learning_rate": 2.806574712643678e-05,
+      "loss": 0.1814,
+      "step": 11462
+    },
+    {
+      "epoch": 39.52758620689655,
+      "grad_norm": 0.7409343123435974,
+      "learning_rate": 2.806528735632184e-05,
+      "loss": 0.1986,
+      "step": 11463
+    },
+    {
+      "epoch": 39.53103448275862,
+      "grad_norm": 0.6706486940383911,
+      "learning_rate": 2.8064827586206898e-05,
+      "loss": 0.1728,
+      "step": 11464
+    },
+    {
+      "epoch": 39.53448275862069,
+      "grad_norm": 0.7322885394096375,
+      "learning_rate": 2.8064367816091953e-05,
+      "loss": 0.1596,
+      "step": 11465
+    },
+    {
+      "epoch": 39.53793103448276,
+      "grad_norm": 0.8368232846260071,
+      "learning_rate": 2.8063908045977012e-05,
+      "loss": 0.1491,
+      "step": 11466
+    },
+    {
+      "epoch": 39.54137931034483,
+      "grad_norm": 0.537189245223999,
+      "learning_rate": 2.8063448275862067e-05,
+      "loss": 0.1282,
+      "step": 11467
+    },
+    {
+      "epoch": 39.5448275862069,
+      "grad_norm": 0.7506726980209351,
+      "learning_rate": 2.806298850574713e-05,
+      "loss": 0.1538,
+      "step": 11468
+    },
+    {
+      "epoch": 39.54827586206896,
+      "grad_norm": 1.7588565349578857,
+      "learning_rate": 2.8062528735632185e-05,
+      "loss": 0.1761,
+      "step": 11469
+    },
+    {
+      "epoch": 39.55172413793103,
+      "grad_norm": 0.7439810633659363,
+      "learning_rate": 2.806206896551724e-05,
+      "loss": 0.1501,
+      "step": 11470
+    },
+    {
+      "epoch": 39.5551724137931,
+      "grad_norm": 1.13484787940979,
+      "learning_rate": 2.80616091954023e-05,
+      "loss": 0.1675,
+      "step": 11471
+    },
+    {
+      "epoch": 39.55862068965517,
+      "grad_norm": 0.8381078243255615,
+      "learning_rate": 2.8061149425287357e-05,
+      "loss": 0.1375,
+      "step": 11472
+    },
+    {
+      "epoch": 39.56206896551724,
+      "grad_norm": 0.8601353764533997,
+      "learning_rate": 2.8060689655172416e-05,
+      "loss": 0.127,
+      "step": 11473
+    },
+    {
+      "epoch": 39.56551724137931,
+      "grad_norm": 0.8690395951271057,
+      "learning_rate": 2.806022988505747e-05,
+      "loss": 0.1342,
+      "step": 11474
+    },
+    {
+      "epoch": 39.56896551724138,
+      "grad_norm": 0.6036515235900879,
+      "learning_rate": 2.8059770114942527e-05,
+      "loss": 0.1372,
+      "step": 11475
+    },
+    {
+      "epoch": 39.57241379310345,
+      "grad_norm": 0.6451120972633362,
+      "learning_rate": 2.805931034482759e-05,
+      "loss": 0.1363,
+      "step": 11476
+    },
+    {
+      "epoch": 39.57586206896552,
+      "grad_norm": 0.5364387631416321,
+      "learning_rate": 2.8058850574712644e-05,
+      "loss": 0.1125,
+      "step": 11477
+    },
+    {
+      "epoch": 39.57931034482758,
+      "grad_norm": 0.6387764811515808,
+      "learning_rate": 2.8058390804597703e-05,
+      "loss": 0.1368,
+      "step": 11478
+    },
+    {
+      "epoch": 39.58275862068965,
+      "grad_norm": 0.7233579158782959,
+      "learning_rate": 2.8057931034482758e-05,
+      "loss": 0.1279,
+      "step": 11479
+    },
+    {
+      "epoch": 39.58620689655172,
+      "grad_norm": 0.6326417326927185,
+      "learning_rate": 2.8057471264367817e-05,
+      "loss": 0.1115,
+      "step": 11480
+    },
+    {
+      "epoch": 39.58965517241379,
+      "grad_norm": 0.8463090062141418,
+      "learning_rate": 2.8057011494252875e-05,
+      "loss": 0.1636,
+      "step": 11481
+    },
+    {
+      "epoch": 39.59310344827586,
+      "grad_norm": 0.6153095364570618,
+      "learning_rate": 2.805655172413793e-05,
+      "loss": 0.1024,
+      "step": 11482
+    },
+    {
+      "epoch": 39.59655172413793,
+      "grad_norm": 0.5719016790390015,
+      "learning_rate": 2.805609195402299e-05,
+      "loss": 0.1013,
+      "step": 11483
+    },
+    {
+      "epoch": 39.6,
+      "grad_norm": 0.774377703666687,
+      "learning_rate": 2.8055632183908048e-05,
+      "loss": 0.1067,
+      "step": 11484
+    },
+    {
+      "epoch": 39.60344827586207,
+      "grad_norm": 1.4463481903076172,
+      "learning_rate": 2.8055172413793103e-05,
+      "loss": 0.1607,
+      "step": 11485
+    },
+    {
+      "epoch": 39.60689655172414,
+      "grad_norm": 1.2848585844039917,
+      "learning_rate": 2.8054712643678162e-05,
+      "loss": 0.2361,
+      "step": 11486
+    },
+    {
+      "epoch": 39.610344827586204,
+      "grad_norm": 0.6632060408592224,
+      "learning_rate": 2.8054252873563217e-05,
+      "loss": 0.189,
+      "step": 11487
+    },
+    {
+      "epoch": 39.61379310344827,
+      "grad_norm": 0.523815929889679,
+      "learning_rate": 2.8053793103448276e-05,
+      "loss": 0.1843,
+      "step": 11488
+    },
+    {
+      "epoch": 39.61724137931034,
+      "grad_norm": 0.5185151100158691,
+      "learning_rate": 2.8053333333333335e-05,
+      "loss": 0.1866,
+      "step": 11489
+    },
+    {
+      "epoch": 39.62068965517241,
+      "grad_norm": 0.7114824056625366,
+      "learning_rate": 2.805287356321839e-05,
+      "loss": 0.164,
+      "step": 11490
+    },
+    {
+      "epoch": 39.62413793103448,
+      "grad_norm": 0.8949483633041382,
+      "learning_rate": 2.805241379310345e-05,
+      "loss": 0.1449,
+      "step": 11491
+    },
+    {
+      "epoch": 39.62758620689655,
+      "grad_norm": 0.8068279027938843,
+      "learning_rate": 2.8051954022988507e-05,
+      "loss": 0.1588,
+      "step": 11492
+    },
+    {
+      "epoch": 39.63103448275862,
+      "grad_norm": 0.6915461421012878,
+      "learning_rate": 2.8051494252873563e-05,
+      "loss": 0.1476,
+      "step": 11493
+    },
+    {
+      "epoch": 39.63448275862069,
+      "grad_norm": 1.9260233640670776,
+      "learning_rate": 2.805103448275862e-05,
+      "loss": 0.133,
+      "step": 11494
+    },
+    {
+      "epoch": 39.63793103448276,
+      "grad_norm": 0.8284426331520081,
+      "learning_rate": 2.8050574712643677e-05,
+      "loss": 0.1522,
+      "step": 11495
+    },
+    {
+      "epoch": 39.641379310344824,
+      "grad_norm": 0.6056115031242371,
+      "learning_rate": 2.805011494252874e-05,
+      "loss": 0.1417,
+      "step": 11496
+    },
+    {
+      "epoch": 39.644827586206894,
+      "grad_norm": 0.6201438903808594,
+      "learning_rate": 2.8049655172413794e-05,
+      "loss": 0.1842,
+      "step": 11497
+    },
+    {
+      "epoch": 39.648275862068964,
+      "grad_norm": 0.7673490643501282,
+      "learning_rate": 2.804919540229885e-05,
+      "loss": 0.1357,
+      "step": 11498
+    },
+    {
+      "epoch": 39.65172413793103,
+      "grad_norm": 0.7154965996742249,
+      "learning_rate": 2.8048735632183908e-05,
+      "loss": 0.1425,
+      "step": 11499
+    },
+    {
+      "epoch": 39.6551724137931,
+      "grad_norm": 1.130608320236206,
+      "learning_rate": 2.8048275862068967e-05,
+      "loss": 0.1604,
+      "step": 11500
+    },
+    {
+      "epoch": 39.65862068965517,
+      "grad_norm": 1.0151304006576538,
+      "learning_rate": 2.8047816091954025e-05,
+      "loss": 0.142,
+      "step": 11501
+    },
+    {
+      "epoch": 39.66206896551724,
+      "grad_norm": 0.6416845917701721,
+      "learning_rate": 2.804735632183908e-05,
+      "loss": 0.1101,
+      "step": 11502
+    },
+    {
+      "epoch": 39.66551724137931,
+      "grad_norm": 0.7699695825576782,
+      "learning_rate": 2.8046896551724136e-05,
+      "loss": 0.1133,
+      "step": 11503
+    },
+    {
+      "epoch": 39.66896551724138,
+      "grad_norm": 0.8104886412620544,
+      "learning_rate": 2.8046436781609198e-05,
+      "loss": 0.1232,
+      "step": 11504
+    },
+    {
+      "epoch": 39.672413793103445,
+      "grad_norm": 0.553966760635376,
+      "learning_rate": 2.8045977011494253e-05,
+      "loss": 0.1005,
+      "step": 11505
+    },
+    {
+      "epoch": 39.675862068965515,
+      "grad_norm": 1.0513066053390503,
+      "learning_rate": 2.8045517241379312e-05,
+      "loss": 0.1107,
+      "step": 11506
+    },
+    {
+      "epoch": 39.679310344827584,
+      "grad_norm": 0.8021120429039001,
+      "learning_rate": 2.8045057471264367e-05,
+      "loss": 0.1073,
+      "step": 11507
+    },
+    {
+      "epoch": 39.682758620689654,
+      "grad_norm": 0.7450590133666992,
+      "learning_rate": 2.8044597701149426e-05,
+      "loss": 0.109,
+      "step": 11508
+    },
+    {
+      "epoch": 39.686206896551724,
+      "grad_norm": 1.2773017883300781,
+      "learning_rate": 2.8044137931034485e-05,
+      "loss": 0.1083,
+      "step": 11509
+    },
+    {
+      "epoch": 39.689655172413794,
+      "grad_norm": 1.550302267074585,
+      "learning_rate": 2.804367816091954e-05,
+      "loss": 0.1349,
+      "step": 11510
+    },
+    {
+      "epoch": 39.69310344827586,
+      "grad_norm": 0.984873354434967,
+      "learning_rate": 2.80432183908046e-05,
+      "loss": 0.2145,
+      "step": 11511
+    },
+    {
+      "epoch": 39.69655172413793,
+      "grad_norm": 0.6027946472167969,
+      "learning_rate": 2.8042758620689657e-05,
+      "loss": 0.1714,
+      "step": 11512
+    },
+    {
+      "epoch": 39.7,
+      "grad_norm": 0.580236554145813,
+      "learning_rate": 2.8042298850574713e-05,
+      "loss": 0.1769,
+      "step": 11513
+    },
+    {
+      "epoch": 39.703448275862065,
+      "grad_norm": 0.5701853036880493,
+      "learning_rate": 2.804183908045977e-05,
+      "loss": 0.1754,
+      "step": 11514
+    },
+    {
+      "epoch": 39.706896551724135,
+      "grad_norm": 0.6304711699485779,
+      "learning_rate": 2.8041379310344827e-05,
+      "loss": 0.2023,
+      "step": 11515
+    },
+    {
+      "epoch": 39.710344827586205,
+      "grad_norm": 0.8770312666893005,
+      "learning_rate": 2.8040919540229885e-05,
+      "loss": 0.1734,
+      "step": 11516
+    },
+    {
+      "epoch": 39.713793103448275,
+      "grad_norm": 1.2147611379623413,
+      "learning_rate": 2.8040459770114944e-05,
+      "loss": 0.1856,
+      "step": 11517
+    },
+    {
+      "epoch": 39.717241379310344,
+      "grad_norm": 2.212428331375122,
+      "learning_rate": 2.804e-05,
+      "loss": 0.1781,
+      "step": 11518
+    },
+    {
+      "epoch": 39.720689655172414,
+      "grad_norm": 0.8960394859313965,
+      "learning_rate": 2.8039540229885058e-05,
+      "loss": 0.1572,
+      "step": 11519
+    },
+    {
+      "epoch": 39.724137931034484,
+      "grad_norm": 0.5783913731575012,
+      "learning_rate": 2.8039080459770117e-05,
+      "loss": 0.1532,
+      "step": 11520
+    },
+    {
+      "epoch": 39.727586206896554,
+      "grad_norm": 0.7316312193870544,
+      "learning_rate": 2.8038620689655172e-05,
+      "loss": 0.1415,
+      "step": 11521
+    },
+    {
+      "epoch": 39.73103448275862,
+      "grad_norm": 0.8523666858673096,
+      "learning_rate": 2.803816091954023e-05,
+      "loss": 0.1455,
+      "step": 11522
+    },
+    {
+      "epoch": 39.734482758620686,
+      "grad_norm": 0.8479415774345398,
+      "learning_rate": 2.8037701149425286e-05,
+      "loss": 0.1312,
+      "step": 11523
+    },
+    {
+      "epoch": 39.737931034482756,
+      "grad_norm": 2.0010015964508057,
+      "learning_rate": 2.8037241379310348e-05,
+      "loss": 0.1433,
+      "step": 11524
+    },
+    {
+      "epoch": 39.741379310344826,
+      "grad_norm": 0.8007112145423889,
+      "learning_rate": 2.8036781609195403e-05,
+      "loss": 0.1605,
+      "step": 11525
+    },
+    {
+      "epoch": 39.744827586206895,
+      "grad_norm": 0.9121559858322144,
+      "learning_rate": 2.803632183908046e-05,
+      "loss": 0.1324,
+      "step": 11526
+    },
+    {
+      "epoch": 39.748275862068965,
+      "grad_norm": 0.8926279544830322,
+      "learning_rate": 2.8035862068965517e-05,
+      "loss": 0.125,
+      "step": 11527
+    },
+    {
+      "epoch": 39.751724137931035,
+      "grad_norm": 0.7816924452781677,
+      "learning_rate": 2.8035402298850576e-05,
+      "loss": 0.1186,
+      "step": 11528
+    },
+    {
+      "epoch": 39.755172413793105,
+      "grad_norm": 1.2233219146728516,
+      "learning_rate": 2.8034942528735635e-05,
+      "loss": 0.1261,
+      "step": 11529
+    },
+    {
+      "epoch": 39.758620689655174,
+      "grad_norm": 0.7805418968200684,
+      "learning_rate": 2.803448275862069e-05,
+      "loss": 0.1014,
+      "step": 11530
+    },
+    {
+      "epoch": 39.762068965517244,
+      "grad_norm": 1.45162034034729,
+      "learning_rate": 2.8034022988505745e-05,
+      "loss": 0.1391,
+      "step": 11531
+    },
+    {
+      "epoch": 39.765517241379314,
+      "grad_norm": 1.2744754552841187,
+      "learning_rate": 2.8033563218390807e-05,
+      "loss": 0.1247,
+      "step": 11532
+    },
+    {
+      "epoch": 39.76896551724138,
+      "grad_norm": 1.2398245334625244,
+      "learning_rate": 2.8033103448275863e-05,
+      "loss": 0.132,
+      "step": 11533
+    },
+    {
+      "epoch": 39.772413793103446,
+      "grad_norm": 1.2867166996002197,
+      "learning_rate": 2.803264367816092e-05,
+      "loss": 0.1537,
+      "step": 11534
+    },
+    {
+      "epoch": 39.775862068965516,
+      "grad_norm": 0.988305926322937,
+      "learning_rate": 2.8032183908045976e-05,
+      "loss": 0.1809,
+      "step": 11535
+    },
+    {
+      "epoch": 39.779310344827586,
+      "grad_norm": 2.060601234436035,
+      "learning_rate": 2.8031724137931035e-05,
+      "loss": 0.2203,
+      "step": 11536
+    },
+    {
+      "epoch": 39.782758620689656,
+      "grad_norm": 0.9540982246398926,
+      "learning_rate": 2.8031264367816094e-05,
+      "loss": 0.1739,
+      "step": 11537
+    },
+    {
+      "epoch": 39.786206896551725,
+      "grad_norm": 0.5264396667480469,
+      "learning_rate": 2.803080459770115e-05,
+      "loss": 0.2009,
+      "step": 11538
+    },
+    {
+      "epoch": 39.789655172413795,
+      "grad_norm": 0.5899782776832581,
+      "learning_rate": 2.8030344827586208e-05,
+      "loss": 0.1785,
+      "step": 11539
+    },
+    {
+      "epoch": 39.793103448275865,
+      "grad_norm": 1.5404542684555054,
+      "learning_rate": 2.8029885057471267e-05,
+      "loss": 0.1523,
+      "step": 11540
+    },
+    {
+      "epoch": 39.796551724137935,
+      "grad_norm": 0.8499155044555664,
+      "learning_rate": 2.8029425287356322e-05,
+      "loss": 0.1857,
+      "step": 11541
+    },
+    {
+      "epoch": 39.8,
+      "grad_norm": 0.8888062238693237,
+      "learning_rate": 2.802896551724138e-05,
+      "loss": 0.1869,
+      "step": 11542
+    },
+    {
+      "epoch": 39.80344827586207,
+      "grad_norm": 0.5029839277267456,
+      "learning_rate": 2.8028505747126436e-05,
+      "loss": 0.1654,
+      "step": 11543
+    },
+    {
+      "epoch": 39.80689655172414,
+      "grad_norm": 0.7781130075454712,
+      "learning_rate": 2.8028045977011494e-05,
+      "loss": 0.1656,
+      "step": 11544
+    },
+    {
+      "epoch": 39.810344827586206,
+      "grad_norm": 0.8604413866996765,
+      "learning_rate": 2.8027586206896553e-05,
+      "loss": 0.1457,
+      "step": 11545
+    },
+    {
+      "epoch": 39.813793103448276,
+      "grad_norm": 1.0199977159500122,
+      "learning_rate": 2.802712643678161e-05,
+      "loss": 0.1512,
+      "step": 11546
+    },
+    {
+      "epoch": 39.817241379310346,
+      "grad_norm": 0.7425589561462402,
+      "learning_rate": 2.8026666666666667e-05,
+      "loss": 0.1566,
+      "step": 11547
+    },
+    {
+      "epoch": 39.820689655172416,
+      "grad_norm": 0.622673511505127,
+      "learning_rate": 2.8026206896551726e-05,
+      "loss": 0.1635,
+      "step": 11548
+    },
+    {
+      "epoch": 39.824137931034485,
+      "grad_norm": 0.5950734615325928,
+      "learning_rate": 2.802574712643678e-05,
+      "loss": 0.1699,
+      "step": 11549
+    },
+    {
+      "epoch": 39.827586206896555,
+      "grad_norm": 0.6048804521560669,
+      "learning_rate": 2.802528735632184e-05,
+      "loss": 0.1232,
+      "step": 11550
+    },
+    {
+      "epoch": 39.83103448275862,
+      "grad_norm": 0.9689356088638306,
+      "learning_rate": 2.8024827586206895e-05,
+      "loss": 0.1377,
+      "step": 11551
+    },
+    {
+      "epoch": 39.83448275862069,
+      "grad_norm": 0.9682617783546448,
+      "learning_rate": 2.8024367816091957e-05,
+      "loss": 0.1387,
+      "step": 11552
+    },
+    {
+      "epoch": 39.83793103448276,
+      "grad_norm": 1.2814329862594604,
+      "learning_rate": 2.8023908045977012e-05,
+      "loss": 0.1251,
+      "step": 11553
+    },
+    {
+      "epoch": 39.84137931034483,
+      "grad_norm": 0.9389997124671936,
+      "learning_rate": 2.8023448275862068e-05,
+      "loss": 0.1181,
+      "step": 11554
+    },
+    {
+      "epoch": 39.8448275862069,
+      "grad_norm": 0.8210214376449585,
+      "learning_rate": 2.8022988505747126e-05,
+      "loss": 0.1261,
+      "step": 11555
+    },
+    {
+      "epoch": 39.84827586206897,
+      "grad_norm": 0.7662495970726013,
+      "learning_rate": 2.8022528735632185e-05,
+      "loss": 0.1318,
+      "step": 11556
+    },
+    {
+      "epoch": 39.851724137931036,
+      "grad_norm": 0.8982017636299133,
+      "learning_rate": 2.8022068965517244e-05,
+      "loss": 0.0962,
+      "step": 11557
+    },
+    {
+      "epoch": 39.855172413793106,
+      "grad_norm": 1.0896496772766113,
+      "learning_rate": 2.80216091954023e-05,
+      "loss": 0.1334,
+      "step": 11558
+    },
+    {
+      "epoch": 39.858620689655176,
+      "grad_norm": 0.8507948517799377,
+      "learning_rate": 2.8021149425287354e-05,
+      "loss": 0.1229,
+      "step": 11559
+    },
+    {
+      "epoch": 39.86206896551724,
+      "grad_norm": 1.1775119304656982,
+      "learning_rate": 2.8020689655172416e-05,
+      "loss": 0.15,
+      "step": 11560
+    },
+    {
+      "epoch": 39.86551724137931,
+      "grad_norm": 1.5011897087097168,
+      "learning_rate": 2.8020229885057472e-05,
+      "loss": 0.2115,
+      "step": 11561
+    },
+    {
+      "epoch": 39.86896551724138,
+      "grad_norm": 0.7803739309310913,
+      "learning_rate": 2.801977011494253e-05,
+      "loss": 0.201,
+      "step": 11562
+    },
+    {
+      "epoch": 39.87241379310345,
+      "grad_norm": 0.5539829730987549,
+      "learning_rate": 2.8019310344827586e-05,
+      "loss": 0.1819,
+      "step": 11563
+    },
+    {
+      "epoch": 39.87586206896552,
+      "grad_norm": 0.5139386653900146,
+      "learning_rate": 2.8018850574712644e-05,
+      "loss": 0.1932,
+      "step": 11564
+    },
+    {
+      "epoch": 39.87931034482759,
+      "grad_norm": 1.0221246480941772,
+      "learning_rate": 2.8018390804597703e-05,
+      "loss": 0.162,
+      "step": 11565
+    },
+    {
+      "epoch": 39.88275862068966,
+      "grad_norm": 0.8142642974853516,
+      "learning_rate": 2.801793103448276e-05,
+      "loss": 0.1623,
+      "step": 11566
+    },
+    {
+      "epoch": 39.88620689655173,
+      "grad_norm": 1.0034339427947998,
+      "learning_rate": 2.8017471264367817e-05,
+      "loss": 0.1818,
+      "step": 11567
+    },
+    {
+      "epoch": 39.889655172413796,
+      "grad_norm": 0.7470972537994385,
+      "learning_rate": 2.8017011494252876e-05,
+      "loss": 0.1666,
+      "step": 11568
+    },
+    {
+      "epoch": 39.89310344827586,
+      "grad_norm": 0.6089482307434082,
+      "learning_rate": 2.801655172413793e-05,
+      "loss": 0.1771,
+      "step": 11569
+    },
+    {
+      "epoch": 39.89655172413793,
+      "grad_norm": 0.8579623699188232,
+      "learning_rate": 2.801609195402299e-05,
+      "loss": 0.1618,
+      "step": 11570
+    },
+    {
+      "epoch": 39.9,
+      "grad_norm": 0.6752385497093201,
+      "learning_rate": 2.8015632183908045e-05,
+      "loss": 0.1445,
+      "step": 11571
+    },
+    {
+      "epoch": 39.90344827586207,
+      "grad_norm": 0.662358283996582,
+      "learning_rate": 2.8015172413793104e-05,
+      "loss": 0.1567,
+      "step": 11572
+    },
+    {
+      "epoch": 39.90689655172414,
+      "grad_norm": 0.7682713866233826,
+      "learning_rate": 2.8014712643678162e-05,
+      "loss": 0.1512,
+      "step": 11573
+    },
+    {
+      "epoch": 39.91034482758621,
+      "grad_norm": 0.8414244651794434,
+      "learning_rate": 2.8014252873563218e-05,
+      "loss": 0.1561,
+      "step": 11574
+    },
+    {
+      "epoch": 39.91379310344828,
+      "grad_norm": 0.9113763570785522,
+      "learning_rate": 2.8013793103448276e-05,
+      "loss": 0.1399,
+      "step": 11575
+    },
+    {
+      "epoch": 39.91724137931035,
+      "grad_norm": 0.8288007378578186,
+      "learning_rate": 2.8013333333333335e-05,
+      "loss": 0.1484,
+      "step": 11576
+    },
+    {
+      "epoch": 39.92068965517242,
+      "grad_norm": 1.0269458293914795,
+      "learning_rate": 2.801287356321839e-05,
+      "loss": 0.1322,
+      "step": 11577
+    },
+    {
+      "epoch": 39.92413793103448,
+      "grad_norm": 0.6786368489265442,
+      "learning_rate": 2.801241379310345e-05,
+      "loss": 0.1202,
+      "step": 11578
+    },
+    {
+      "epoch": 39.92758620689655,
+      "grad_norm": 1.5795526504516602,
+      "learning_rate": 2.8011954022988504e-05,
+      "loss": 0.1332,
+      "step": 11579
+    },
+    {
+      "epoch": 39.93103448275862,
+      "grad_norm": 0.8821081519126892,
+      "learning_rate": 2.8011494252873566e-05,
+      "loss": 0.1117,
+      "step": 11580
+    },
+    {
+      "epoch": 39.93448275862069,
+      "grad_norm": 0.7510960102081299,
+      "learning_rate": 2.801103448275862e-05,
+      "loss": 0.121,
+      "step": 11581
+    },
+    {
+      "epoch": 39.93793103448276,
+      "grad_norm": 0.9274259209632874,
+      "learning_rate": 2.8010574712643677e-05,
+      "loss": 0.1234,
+      "step": 11582
+    },
+    {
+      "epoch": 39.94137931034483,
+      "grad_norm": 0.8250483274459839,
+      "learning_rate": 2.8010114942528736e-05,
+      "loss": 0.114,
+      "step": 11583
+    },
+    {
+      "epoch": 39.9448275862069,
+      "grad_norm": 0.865753173828125,
+      "learning_rate": 2.8009655172413794e-05,
+      "loss": 0.1243,
+      "step": 11584
+    },
+    {
+      "epoch": 39.94827586206897,
+      "grad_norm": 1.1547582149505615,
+      "learning_rate": 2.8009195402298853e-05,
+      "loss": 0.1898,
+      "step": 11585
+    },
+    {
+      "epoch": 39.95172413793104,
+      "grad_norm": 1.0319417715072632,
+      "learning_rate": 2.800873563218391e-05,
+      "loss": 0.2102,
+      "step": 11586
+    },
+    {
+      "epoch": 39.9551724137931,
+      "grad_norm": 0.5368159413337708,
+      "learning_rate": 2.8008275862068964e-05,
+      "loss": 0.1818,
+      "step": 11587
+    },
+    {
+      "epoch": 39.95862068965517,
+      "grad_norm": 0.5381379723548889,
+      "learning_rate": 2.8007816091954026e-05,
+      "loss": 0.1835,
+      "step": 11588
+    },
+    {
+      "epoch": 39.96206896551724,
+      "grad_norm": 1.0764422416687012,
+      "learning_rate": 2.800735632183908e-05,
+      "loss": 0.172,
+      "step": 11589
+    },
+    {
+      "epoch": 39.96551724137931,
+      "grad_norm": 0.5636780858039856,
+      "learning_rate": 2.800689655172414e-05,
+      "loss": 0.1434,
+      "step": 11590
+    },
+    {
+      "epoch": 39.96896551724138,
+      "grad_norm": 1.5343387126922607,
+      "learning_rate": 2.8006436781609195e-05,
+      "loss": 0.1636,
+      "step": 11591
+    },
+    {
+      "epoch": 39.97241379310345,
+      "grad_norm": 0.7999375462532043,
+      "learning_rate": 2.8005977011494254e-05,
+      "loss": 0.1614,
+      "step": 11592
+    },
+    {
+      "epoch": 39.97586206896552,
+      "grad_norm": 1.0915523767471313,
+      "learning_rate": 2.8005517241379312e-05,
+      "loss": 0.1361,
+      "step": 11593
+    },
+    {
+      "epoch": 39.97931034482759,
+      "grad_norm": 0.6097292304039001,
+      "learning_rate": 2.8005057471264368e-05,
+      "loss": 0.1432,
+      "step": 11594
+    },
+    {
+      "epoch": 39.98275862068966,
+      "grad_norm": 0.9179913997650146,
+      "learning_rate": 2.8004597701149426e-05,
+      "loss": 0.1488,
+      "step": 11595
+    },
+    {
+      "epoch": 39.98620689655172,
+      "grad_norm": 0.8879362940788269,
+      "learning_rate": 2.8004137931034485e-05,
+      "loss": 0.1519,
+      "step": 11596
+    },
+    {
+      "epoch": 39.98965517241379,
+      "grad_norm": 0.8876436948776245,
+      "learning_rate": 2.800367816091954e-05,
+      "loss": 0.1216,
+      "step": 11597
+    },
+    {
+      "epoch": 39.99310344827586,
+      "grad_norm": 1.6118890047073364,
+      "learning_rate": 2.80032183908046e-05,
+      "loss": 0.116,
+      "step": 11598
+    },
+    {
+      "epoch": 39.99655172413793,
+      "grad_norm": 1.871911644935608,
+      "learning_rate": 2.8002758620689654e-05,
+      "loss": 0.1281,
+      "step": 11599
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.9541428089141846,
+      "learning_rate": 2.8002298850574713e-05,
+      "loss": 0.2223,
+      "step": 11600
+    },
+    {
+      "epoch": 40.00344827586207,
+      "grad_norm": 0.5181897282600403,
+      "learning_rate": 2.800183908045977e-05,
+      "loss": 0.2095,
+      "step": 11601
+    },
+    {
+      "epoch": 40.00689655172414,
+      "grad_norm": 0.5921854972839355,
+      "learning_rate": 2.8001379310344827e-05,
+      "loss": 0.1867,
+      "step": 11602
+    },
+    {
+      "epoch": 40.01034482758621,
+      "grad_norm": 0.48063069581985474,
+      "learning_rate": 2.8000919540229886e-05,
+      "loss": 0.1516,
+      "step": 11603
+    },
+    {
+      "epoch": 40.01379310344828,
+      "grad_norm": 0.5444591641426086,
+      "learning_rate": 2.8000459770114944e-05,
+      "loss": 0.1567,
+      "step": 11604
+    },
+    {
+      "epoch": 40.01724137931034,
+      "grad_norm": 1.1245945692062378,
+      "learning_rate": 2.8e-05,
+      "loss": 0.1698,
+      "step": 11605
+    },
+    {
+      "epoch": 40.02068965517241,
+      "grad_norm": 0.5311099886894226,
+      "learning_rate": 2.7999540229885058e-05,
+      "loss": 0.1673,
+      "step": 11606
+    },
+    {
+      "epoch": 40.02413793103448,
+      "grad_norm": 0.8394598960876465,
+      "learning_rate": 2.7999080459770114e-05,
+      "loss": 0.1573,
+      "step": 11607
+    },
+    {
+      "epoch": 40.02758620689655,
+      "grad_norm": 0.6501821279525757,
+      "learning_rate": 2.7998620689655176e-05,
+      "loss": 0.1354,
+      "step": 11608
+    },
+    {
+      "epoch": 40.03103448275862,
+      "grad_norm": 0.5301545858383179,
+      "learning_rate": 2.799816091954023e-05,
+      "loss": 0.1545,
+      "step": 11609
+    },
+    {
+      "epoch": 40.03448275862069,
+      "grad_norm": 0.547879695892334,
+      "learning_rate": 2.7997701149425286e-05,
+      "loss": 0.1451,
+      "step": 11610
+    },
+    {
+      "epoch": 40.03793103448276,
+      "grad_norm": 0.5859290361404419,
+      "learning_rate": 2.7997241379310345e-05,
+      "loss": 0.135,
+      "step": 11611
+    },
+    {
+      "epoch": 40.04137931034483,
+      "grad_norm": 0.8631305694580078,
+      "learning_rate": 2.7996781609195404e-05,
+      "loss": 0.1417,
+      "step": 11612
+    },
+    {
+      "epoch": 40.0448275862069,
+      "grad_norm": 0.6690996885299683,
+      "learning_rate": 2.7996321839080462e-05,
+      "loss": 0.1219,
+      "step": 11613
+    },
+    {
+      "epoch": 40.04827586206896,
+      "grad_norm": 0.5393000841140747,
+      "learning_rate": 2.7995862068965518e-05,
+      "loss": 0.1277,
+      "step": 11614
+    },
+    {
+      "epoch": 40.05172413793103,
+      "grad_norm": 0.7986444234848022,
+      "learning_rate": 2.7995402298850573e-05,
+      "loss": 0.121,
+      "step": 11615
+    },
+    {
+      "epoch": 40.0551724137931,
+      "grad_norm": 0.9650202989578247,
+      "learning_rate": 2.7994942528735635e-05,
+      "loss": 0.1264,
+      "step": 11616
+    },
+    {
+      "epoch": 40.05862068965517,
+      "grad_norm": 1.3113383054733276,
+      "learning_rate": 2.799448275862069e-05,
+      "loss": 0.1108,
+      "step": 11617
+    },
+    {
+      "epoch": 40.06206896551724,
+      "grad_norm": 1.6393826007843018,
+      "learning_rate": 2.799402298850575e-05,
+      "loss": 0.1163,
+      "step": 11618
+    },
+    {
+      "epoch": 40.06551724137931,
+      "grad_norm": 0.690750002861023,
+      "learning_rate": 2.7993563218390804e-05,
+      "loss": 0.1079,
+      "step": 11619
+    },
+    {
+      "epoch": 40.06896551724138,
+      "grad_norm": 0.5626112222671509,
+      "learning_rate": 2.7993103448275863e-05,
+      "loss": 0.1003,
+      "step": 11620
+    },
+    {
+      "epoch": 40.07241379310345,
+      "grad_norm": 0.7423434257507324,
+      "learning_rate": 2.799264367816092e-05,
+      "loss": 0.117,
+      "step": 11621
+    },
+    {
+      "epoch": 40.07586206896552,
+      "grad_norm": 0.6133210062980652,
+      "learning_rate": 2.7992183908045977e-05,
+      "loss": 0.0907,
+      "step": 11622
+    },
+    {
+      "epoch": 40.07931034482758,
+      "grad_norm": 0.8096251487731934,
+      "learning_rate": 2.7991724137931036e-05,
+      "loss": 0.0993,
+      "step": 11623
+    },
+    {
+      "epoch": 40.08275862068965,
+      "grad_norm": 1.315021276473999,
+      "learning_rate": 2.7991264367816094e-05,
+      "loss": 0.1343,
+      "step": 11624
+    },
+    {
+      "epoch": 40.08620689655172,
+      "grad_norm": 1.6839079856872559,
+      "learning_rate": 2.799080459770115e-05,
+      "loss": 0.1622,
+      "step": 11625
+    },
+    {
+      "epoch": 40.08965517241379,
+      "grad_norm": 0.5504493713378906,
+      "learning_rate": 2.7990344827586208e-05,
+      "loss": 0.1995,
+      "step": 11626
+    },
+    {
+      "epoch": 40.09310344827586,
+      "grad_norm": 0.6462870836257935,
+      "learning_rate": 2.7989885057471263e-05,
+      "loss": 0.1762,
+      "step": 11627
+    },
+    {
+      "epoch": 40.09655172413793,
+      "grad_norm": 0.6583945751190186,
+      "learning_rate": 2.7989425287356326e-05,
+      "loss": 0.1676,
+      "step": 11628
+    },
+    {
+      "epoch": 40.1,
+      "grad_norm": 0.4450218975543976,
+      "learning_rate": 2.798896551724138e-05,
+      "loss": 0.1519,
+      "step": 11629
+    },
+    {
+      "epoch": 40.10344827586207,
+      "grad_norm": 0.5877785682678223,
+      "learning_rate": 2.7988505747126436e-05,
+      "loss": 0.1409,
+      "step": 11630
+    },
+    {
+      "epoch": 40.10689655172414,
+      "grad_norm": 0.7537900805473328,
+      "learning_rate": 2.7988045977011495e-05,
+      "loss": 0.1532,
+      "step": 11631
+    },
+    {
+      "epoch": 40.110344827586204,
+      "grad_norm": 0.9696903228759766,
+      "learning_rate": 2.7987586206896554e-05,
+      "loss": 0.1627,
+      "step": 11632
+    },
+    {
+      "epoch": 40.11379310344827,
+      "grad_norm": 0.7010102868080139,
+      "learning_rate": 2.798712643678161e-05,
+      "loss": 0.1424,
+      "step": 11633
+    },
+    {
+      "epoch": 40.11724137931034,
+      "grad_norm": 0.8179371356964111,
+      "learning_rate": 2.7986666666666668e-05,
+      "loss": 0.1398,
+      "step": 11634
+    },
+    {
+      "epoch": 40.12068965517241,
+      "grad_norm": 0.5714012980461121,
+      "learning_rate": 2.7986206896551723e-05,
+      "loss": 0.1534,
+      "step": 11635
+    },
+    {
+      "epoch": 40.12413793103448,
+      "grad_norm": 0.6157788038253784,
+      "learning_rate": 2.7985747126436785e-05,
+      "loss": 0.1157,
+      "step": 11636
+    },
+    {
+      "epoch": 40.12758620689655,
+      "grad_norm": 0.6693355441093445,
+      "learning_rate": 2.798528735632184e-05,
+      "loss": 0.1279,
+      "step": 11637
+    },
+    {
+      "epoch": 40.13103448275862,
+      "grad_norm": 0.7183413505554199,
+      "learning_rate": 2.7984827586206895e-05,
+      "loss": 0.1077,
+      "step": 11638
+    },
+    {
+      "epoch": 40.13448275862069,
+      "grad_norm": 0.5054282546043396,
+      "learning_rate": 2.7984367816091954e-05,
+      "loss": 0.1306,
+      "step": 11639
+    },
+    {
+      "epoch": 40.13793103448276,
+      "grad_norm": 0.5936710834503174,
+      "learning_rate": 2.7983908045977013e-05,
+      "loss": 0.1341,
+      "step": 11640
+    },
+    {
+      "epoch": 40.141379310344824,
+      "grad_norm": 0.6644858121871948,
+      "learning_rate": 2.798344827586207e-05,
+      "loss": 0.1316,
+      "step": 11641
+    },
+    {
+      "epoch": 40.144827586206894,
+      "grad_norm": 0.8058156967163086,
+      "learning_rate": 2.7982988505747127e-05,
+      "loss": 0.1149,
+      "step": 11642
+    },
+    {
+      "epoch": 40.148275862068964,
+      "grad_norm": 1.2202357053756714,
+      "learning_rate": 2.7982528735632182e-05,
+      "loss": 0.1192,
+      "step": 11643
+    },
+    {
+      "epoch": 40.15172413793103,
+      "grad_norm": 0.7190808653831482,
+      "learning_rate": 2.7982068965517244e-05,
+      "loss": 0.1127,
+      "step": 11644
+    },
+    {
+      "epoch": 40.1551724137931,
+      "grad_norm": 1.0018106698989868,
+      "learning_rate": 2.79816091954023e-05,
+      "loss": 0.0967,
+      "step": 11645
+    },
+    {
+      "epoch": 40.15862068965517,
+      "grad_norm": 0.6785632967948914,
+      "learning_rate": 2.7981149425287358e-05,
+      "loss": 0.1221,
+      "step": 11646
+    },
+    {
+      "epoch": 40.16206896551724,
+      "grad_norm": 0.8175433278083801,
+      "learning_rate": 2.7980689655172413e-05,
+      "loss": 0.1115,
+      "step": 11647
+    },
+    {
+      "epoch": 40.16551724137931,
+      "grad_norm": 1.1953258514404297,
+      "learning_rate": 2.7980229885057472e-05,
+      "loss": 0.0901,
+      "step": 11648
+    },
+    {
+      "epoch": 40.16896551724138,
+      "grad_norm": 0.7493051290512085,
+      "learning_rate": 2.797977011494253e-05,
+      "loss": 0.1049,
+      "step": 11649
+    },
+    {
+      "epoch": 40.172413793103445,
+      "grad_norm": 0.9021804332733154,
+      "learning_rate": 2.7979310344827586e-05,
+      "loss": 0.1565,
+      "step": 11650
+    },
+    {
+      "epoch": 40.175862068965515,
+      "grad_norm": 0.6075196266174316,
+      "learning_rate": 2.7978850574712645e-05,
+      "loss": 0.221,
+      "step": 11651
+    },
+    {
+      "epoch": 40.179310344827584,
+      "grad_norm": 0.6591607332229614,
+      "learning_rate": 2.7978390804597703e-05,
+      "loss": 0.1739,
+      "step": 11652
+    },
+    {
+      "epoch": 40.182758620689654,
+      "grad_norm": 0.555267333984375,
+      "learning_rate": 2.797793103448276e-05,
+      "loss": 0.1842,
+      "step": 11653
+    },
+    {
+      "epoch": 40.186206896551724,
+      "grad_norm": 0.9013941884040833,
+      "learning_rate": 2.7977471264367817e-05,
+      "loss": 0.1607,
+      "step": 11654
+    },
+    {
+      "epoch": 40.189655172413794,
+      "grad_norm": 0.6267364025115967,
+      "learning_rate": 2.7977011494252873e-05,
+      "loss": 0.1621,
+      "step": 11655
+    },
+    {
+      "epoch": 40.19310344827586,
+      "grad_norm": 0.5677948594093323,
+      "learning_rate": 2.7976551724137935e-05,
+      "loss": 0.1452,
+      "step": 11656
+    },
+    {
+      "epoch": 40.19655172413793,
+      "grad_norm": 0.502986490726471,
+      "learning_rate": 2.797609195402299e-05,
+      "loss": 0.1669,
+      "step": 11657
+    },
+    {
+      "epoch": 40.2,
+      "grad_norm": 0.6021277904510498,
+      "learning_rate": 2.7975632183908045e-05,
+      "loss": 0.1313,
+      "step": 11658
+    },
+    {
+      "epoch": 40.203448275862065,
+      "grad_norm": 0.6180835962295532,
+      "learning_rate": 2.7975172413793104e-05,
+      "loss": 0.1562,
+      "step": 11659
+    },
+    {
+      "epoch": 40.206896551724135,
+      "grad_norm": 1.2324230670928955,
+      "learning_rate": 2.7974712643678163e-05,
+      "loss": 0.1376,
+      "step": 11660
+    },
+    {
+      "epoch": 40.210344827586205,
+      "grad_norm": 0.6263248920440674,
+      "learning_rate": 2.7974252873563218e-05,
+      "loss": 0.1569,
+      "step": 11661
+    },
+    {
+      "epoch": 40.213793103448275,
+      "grad_norm": 0.7797892093658447,
+      "learning_rate": 2.7973793103448277e-05,
+      "loss": 0.1315,
+      "step": 11662
+    },
+    {
+      "epoch": 40.217241379310344,
+      "grad_norm": 0.5494678616523743,
+      "learning_rate": 2.7973333333333332e-05,
+      "loss": 0.1314,
+      "step": 11663
+    },
+    {
+      "epoch": 40.220689655172414,
+      "grad_norm": 0.5808693766593933,
+      "learning_rate": 2.7972873563218394e-05,
+      "loss": 0.1028,
+      "step": 11664
+    },
+    {
+      "epoch": 40.224137931034484,
+      "grad_norm": 0.5643265247344971,
+      "learning_rate": 2.797241379310345e-05,
+      "loss": 0.1202,
+      "step": 11665
+    },
+    {
+      "epoch": 40.227586206896554,
+      "grad_norm": 0.7029027342796326,
+      "learning_rate": 2.7971954022988505e-05,
+      "loss": 0.1092,
+      "step": 11666
+    },
+    {
+      "epoch": 40.23103448275862,
+      "grad_norm": 0.7036227583885193,
+      "learning_rate": 2.7971494252873563e-05,
+      "loss": 0.1246,
+      "step": 11667
+    },
+    {
+      "epoch": 40.234482758620686,
+      "grad_norm": 0.5897785425186157,
+      "learning_rate": 2.7971034482758622e-05,
+      "loss": 0.1088,
+      "step": 11668
+    },
+    {
+      "epoch": 40.237931034482756,
+      "grad_norm": 0.7039844393730164,
+      "learning_rate": 2.797057471264368e-05,
+      "loss": 0.1092,
+      "step": 11669
+    },
+    {
+      "epoch": 40.241379310344826,
+      "grad_norm": 1.0365760326385498,
+      "learning_rate": 2.7970114942528736e-05,
+      "loss": 0.111,
+      "step": 11670
+    },
+    {
+      "epoch": 40.244827586206895,
+      "grad_norm": 1.865059733390808,
+      "learning_rate": 2.796965517241379e-05,
+      "loss": 0.1308,
+      "step": 11671
+    },
+    {
+      "epoch": 40.248275862068965,
+      "grad_norm": 0.6405381560325623,
+      "learning_rate": 2.7969195402298853e-05,
+      "loss": 0.0902,
+      "step": 11672
+    },
+    {
+      "epoch": 40.251724137931035,
+      "grad_norm": 2.853517532348633,
+      "learning_rate": 2.796873563218391e-05,
+      "loss": 0.092,
+      "step": 11673
+    },
+    {
+      "epoch": 40.255172413793105,
+      "grad_norm": 1.1927621364593506,
+      "learning_rate": 2.7968275862068967e-05,
+      "loss": 0.1035,
+      "step": 11674
+    },
+    {
+      "epoch": 40.258620689655174,
+      "grad_norm": 1.0837582349777222,
+      "learning_rate": 2.7967816091954023e-05,
+      "loss": 0.1199,
+      "step": 11675
+    },
+    {
+      "epoch": 40.262068965517244,
+      "grad_norm": 0.8633678555488586,
+      "learning_rate": 2.796735632183908e-05,
+      "loss": 0.227,
+      "step": 11676
+    },
+    {
+      "epoch": 40.265517241379314,
+      "grad_norm": 0.7727140784263611,
+      "learning_rate": 2.796689655172414e-05,
+      "loss": 0.165,
+      "step": 11677
+    },
+    {
+      "epoch": 40.26896551724138,
+      "grad_norm": 0.794918954372406,
+      "learning_rate": 2.7966436781609195e-05,
+      "loss": 0.1749,
+      "step": 11678
+    },
+    {
+      "epoch": 40.272413793103446,
+      "grad_norm": 0.49571534991264343,
+      "learning_rate": 2.7965977011494254e-05,
+      "loss": 0.1662,
+      "step": 11679
+    },
+    {
+      "epoch": 40.275862068965516,
+      "grad_norm": 0.8848534822463989,
+      "learning_rate": 2.7965517241379313e-05,
+      "loss": 0.148,
+      "step": 11680
+    },
+    {
+      "epoch": 40.279310344827586,
+      "grad_norm": 0.9959067702293396,
+      "learning_rate": 2.7965057471264368e-05,
+      "loss": 0.1543,
+      "step": 11681
+    },
+    {
+      "epoch": 40.282758620689656,
+      "grad_norm": 2.4409101009368896,
+      "learning_rate": 2.7964597701149427e-05,
+      "loss": 0.1506,
+      "step": 11682
+    },
+    {
+      "epoch": 40.286206896551725,
+      "grad_norm": 0.6493564248085022,
+      "learning_rate": 2.7964137931034482e-05,
+      "loss": 0.1648,
+      "step": 11683
+    },
+    {
+      "epoch": 40.289655172413795,
+      "grad_norm": 0.7148909568786621,
+      "learning_rate": 2.7963678160919544e-05,
+      "loss": 0.1441,
+      "step": 11684
+    },
+    {
+      "epoch": 40.293103448275865,
+      "grad_norm": 0.6141997575759888,
+      "learning_rate": 2.79632183908046e-05,
+      "loss": 0.1395,
+      "step": 11685
+    },
+    {
+      "epoch": 40.296551724137935,
+      "grad_norm": 0.6709399819374084,
+      "learning_rate": 2.7962758620689655e-05,
+      "loss": 0.15,
+      "step": 11686
+    },
+    {
+      "epoch": 40.3,
+      "grad_norm": 1.1454461812973022,
+      "learning_rate": 2.7962298850574713e-05,
+      "loss": 0.1721,
+      "step": 11687
+    },
+    {
+      "epoch": 40.30344827586207,
+      "grad_norm": 1.5469913482666016,
+      "learning_rate": 2.7961839080459772e-05,
+      "loss": 0.1346,
+      "step": 11688
+    },
+    {
+      "epoch": 40.30689655172414,
+      "grad_norm": 0.8761661648750305,
+      "learning_rate": 2.7961379310344827e-05,
+      "loss": 0.1239,
+      "step": 11689
+    },
+    {
+      "epoch": 40.310344827586206,
+      "grad_norm": 0.8666542172431946,
+      "learning_rate": 2.7960919540229886e-05,
+      "loss": 0.1427,
+      "step": 11690
+    },
+    {
+      "epoch": 40.313793103448276,
+      "grad_norm": 0.6497369408607483,
+      "learning_rate": 2.796045977011494e-05,
+      "loss": 0.1076,
+      "step": 11691
+    },
+    {
+      "epoch": 40.317241379310346,
+      "grad_norm": 0.7148862481117249,
+      "learning_rate": 2.7960000000000003e-05,
+      "loss": 0.1267,
+      "step": 11692
+    },
+    {
+      "epoch": 40.320689655172416,
+      "grad_norm": 0.9966643452644348,
+      "learning_rate": 2.795954022988506e-05,
+      "loss": 0.1176,
+      "step": 11693
+    },
+    {
+      "epoch": 40.324137931034485,
+      "grad_norm": 0.8974887132644653,
+      "learning_rate": 2.7959080459770114e-05,
+      "loss": 0.1207,
+      "step": 11694
+    },
+    {
+      "epoch": 40.327586206896555,
+      "grad_norm": 0.629341185092926,
+      "learning_rate": 2.7958620689655173e-05,
+      "loss": 0.1185,
+      "step": 11695
+    },
+    {
+      "epoch": 40.33103448275862,
+      "grad_norm": 0.8130760788917542,
+      "learning_rate": 2.795816091954023e-05,
+      "loss": 0.1262,
+      "step": 11696
+    },
+    {
+      "epoch": 40.33448275862069,
+      "grad_norm": 0.7954714894294739,
+      "learning_rate": 2.795770114942529e-05,
+      "loss": 0.0981,
+      "step": 11697
+    },
+    {
+      "epoch": 40.33793103448276,
+      "grad_norm": 0.5662274956703186,
+      "learning_rate": 2.7957241379310345e-05,
+      "loss": 0.1151,
+      "step": 11698
+    },
+    {
+      "epoch": 40.34137931034483,
+      "grad_norm": 0.8570734858512878,
+      "learning_rate": 2.79567816091954e-05,
+      "loss": 0.124,
+      "step": 11699
+    },
+    {
+      "epoch": 40.3448275862069,
+      "grad_norm": 1.1559773683547974,
+      "learning_rate": 2.7956321839080463e-05,
+      "loss": 0.1374,
+      "step": 11700
+    },
+    {
+      "epoch": 40.34827586206897,
+      "grad_norm": 0.6159823536872864,
+      "learning_rate": 2.7955862068965518e-05,
+      "loss": 0.2043,
+      "step": 11701
+    },
+    {
+      "epoch": 40.351724137931036,
+      "grad_norm": 0.5950391292572021,
+      "learning_rate": 2.7955402298850577e-05,
+      "loss": 0.1681,
+      "step": 11702
+    },
+    {
+      "epoch": 40.355172413793106,
+      "grad_norm": 0.4927203059196472,
+      "learning_rate": 2.7954942528735632e-05,
+      "loss": 0.167,
+      "step": 11703
+    },
+    {
+      "epoch": 40.358620689655176,
+      "grad_norm": 1.321279764175415,
+      "learning_rate": 2.795448275862069e-05,
+      "loss": 0.1613,
+      "step": 11704
+    },
+    {
+      "epoch": 40.36206896551724,
+      "grad_norm": 0.8601357936859131,
+      "learning_rate": 2.795402298850575e-05,
+      "loss": 0.1641,
+      "step": 11705
+    },
+    {
+      "epoch": 40.36551724137931,
+      "grad_norm": 1.159642219543457,
+      "learning_rate": 2.7953563218390805e-05,
+      "loss": 0.1678,
+      "step": 11706
+    },
+    {
+      "epoch": 40.36896551724138,
+      "grad_norm": 0.8981372714042664,
+      "learning_rate": 2.7953103448275863e-05,
+      "loss": 0.1828,
+      "step": 11707
+    },
+    {
+      "epoch": 40.37241379310345,
+      "grad_norm": 0.6486225724220276,
+      "learning_rate": 2.7952643678160922e-05,
+      "loss": 0.178,
+      "step": 11708
+    },
+    {
+      "epoch": 40.37586206896552,
+      "grad_norm": 0.7072604298591614,
+      "learning_rate": 2.7952183908045977e-05,
+      "loss": 0.1514,
+      "step": 11709
+    },
+    {
+      "epoch": 40.37931034482759,
+      "grad_norm": 0.5991432070732117,
+      "learning_rate": 2.7951724137931036e-05,
+      "loss": 0.1413,
+      "step": 11710
+    },
+    {
+      "epoch": 40.38275862068966,
+      "grad_norm": 0.9521961212158203,
+      "learning_rate": 2.795126436781609e-05,
+      "loss": 0.1206,
+      "step": 11711
+    },
+    {
+      "epoch": 40.38620689655173,
+      "grad_norm": 1.9931718111038208,
+      "learning_rate": 2.7950804597701153e-05,
+      "loss": 0.177,
+      "step": 11712
+    },
+    {
+      "epoch": 40.389655172413796,
+      "grad_norm": 1.1548880338668823,
+      "learning_rate": 2.795034482758621e-05,
+      "loss": 0.1368,
+      "step": 11713
+    },
+    {
+      "epoch": 40.39310344827586,
+      "grad_norm": 0.6284099817276001,
+      "learning_rate": 2.7949885057471264e-05,
+      "loss": 0.1328,
+      "step": 11714
+    },
+    {
+      "epoch": 40.39655172413793,
+      "grad_norm": 0.8119350671768188,
+      "learning_rate": 2.7949425287356323e-05,
+      "loss": 0.1034,
+      "step": 11715
+    },
+    {
+      "epoch": 40.4,
+      "grad_norm": 0.7650372982025146,
+      "learning_rate": 2.794896551724138e-05,
+      "loss": 0.1515,
+      "step": 11716
+    },
+    {
+      "epoch": 40.40344827586207,
+      "grad_norm": 0.6360500454902649,
+      "learning_rate": 2.794850574712644e-05,
+      "loss": 0.1262,
+      "step": 11717
+    },
+    {
+      "epoch": 40.40689655172414,
+      "grad_norm": 0.731512725353241,
+      "learning_rate": 2.7948045977011495e-05,
+      "loss": 0.1173,
+      "step": 11718
+    },
+    {
+      "epoch": 40.41034482758621,
+      "grad_norm": 0.5277930498123169,
+      "learning_rate": 2.794758620689655e-05,
+      "loss": 0.1159,
+      "step": 11719
+    },
+    {
+      "epoch": 40.41379310344828,
+      "grad_norm": 0.7175533771514893,
+      "learning_rate": 2.7947126436781613e-05,
+      "loss": 0.1156,
+      "step": 11720
+    },
+    {
+      "epoch": 40.41724137931035,
+      "grad_norm": 0.6196317076683044,
+      "learning_rate": 2.7946666666666668e-05,
+      "loss": 0.1016,
+      "step": 11721
+    },
+    {
+      "epoch": 40.42068965517242,
+      "grad_norm": 0.7266805768013,
+      "learning_rate": 2.7946206896551723e-05,
+      "loss": 0.1029,
+      "step": 11722
+    },
+    {
+      "epoch": 40.42413793103448,
+      "grad_norm": 2.7246503829956055,
+      "learning_rate": 2.7945747126436782e-05,
+      "loss": 0.1045,
+      "step": 11723
+    },
+    {
+      "epoch": 40.42758620689655,
+      "grad_norm": 0.7863669395446777,
+      "learning_rate": 2.794528735632184e-05,
+      "loss": 0.099,
+      "step": 11724
+    },
+    {
+      "epoch": 40.43103448275862,
+      "grad_norm": 1.2347368001937866,
+      "learning_rate": 2.79448275862069e-05,
+      "loss": 0.1559,
+      "step": 11725
+    },
+    {
+      "epoch": 40.43448275862069,
+      "grad_norm": 1.5440723896026611,
+      "learning_rate": 2.7944367816091955e-05,
+      "loss": 0.2245,
+      "step": 11726
+    },
+    {
+      "epoch": 40.43793103448276,
+      "grad_norm": 0.5833970308303833,
+      "learning_rate": 2.794390804597701e-05,
+      "loss": 0.1639,
+      "step": 11727
+    },
+    {
+      "epoch": 40.44137931034483,
+      "grad_norm": 1.124950885772705,
+      "learning_rate": 2.7943448275862072e-05,
+      "loss": 0.1765,
+      "step": 11728
+    },
+    {
+      "epoch": 40.4448275862069,
+      "grad_norm": 0.7326246500015259,
+      "learning_rate": 2.7942988505747127e-05,
+      "loss": 0.1669,
+      "step": 11729
+    },
+    {
+      "epoch": 40.44827586206897,
+      "grad_norm": 0.7290887832641602,
+      "learning_rate": 2.7942528735632186e-05,
+      "loss": 0.1556,
+      "step": 11730
+    },
+    {
+      "epoch": 40.45172413793104,
+      "grad_norm": 1.0433979034423828,
+      "learning_rate": 2.794206896551724e-05,
+      "loss": 0.1513,
+      "step": 11731
+    },
+    {
+      "epoch": 40.4551724137931,
+      "grad_norm": 0.5453125834465027,
+      "learning_rate": 2.79416091954023e-05,
+      "loss": 0.1647,
+      "step": 11732
+    },
+    {
+      "epoch": 40.45862068965517,
+      "grad_norm": 0.9280747771263123,
+      "learning_rate": 2.794114942528736e-05,
+      "loss": 0.1586,
+      "step": 11733
+    },
+    {
+      "epoch": 40.46206896551724,
+      "grad_norm": 1.8095715045928955,
+      "learning_rate": 2.7940689655172414e-05,
+      "loss": 0.134,
+      "step": 11734
+    },
+    {
+      "epoch": 40.46551724137931,
+      "grad_norm": 0.9350835084915161,
+      "learning_rate": 2.7940229885057473e-05,
+      "loss": 0.1379,
+      "step": 11735
+    },
+    {
+      "epoch": 40.46896551724138,
+      "grad_norm": 1.448866605758667,
+      "learning_rate": 2.793977011494253e-05,
+      "loss": 0.1383,
+      "step": 11736
+    },
+    {
+      "epoch": 40.47241379310345,
+      "grad_norm": 0.825294017791748,
+      "learning_rate": 2.7939310344827586e-05,
+      "loss": 0.1382,
+      "step": 11737
+    },
+    {
+      "epoch": 40.47586206896552,
+      "grad_norm": 1.2684051990509033,
+      "learning_rate": 2.7938850574712645e-05,
+      "loss": 0.1382,
+      "step": 11738
+    },
+    {
+      "epoch": 40.47931034482759,
+      "grad_norm": 0.5286739468574524,
+      "learning_rate": 2.79383908045977e-05,
+      "loss": 0.1308,
+      "step": 11739
+    },
+    {
+      "epoch": 40.48275862068966,
+      "grad_norm": 0.7574636340141296,
+      "learning_rate": 2.7937931034482763e-05,
+      "loss": 0.141,
+      "step": 11740
+    },
+    {
+      "epoch": 40.48620689655172,
+      "grad_norm": 0.7143364548683167,
+      "learning_rate": 2.7937471264367818e-05,
+      "loss": 0.1255,
+      "step": 11741
+    },
+    {
+      "epoch": 40.48965517241379,
+      "grad_norm": 0.687987208366394,
+      "learning_rate": 2.7937011494252873e-05,
+      "loss": 0.1132,
+      "step": 11742
+    },
+    {
+      "epoch": 40.49310344827586,
+      "grad_norm": 0.7487812638282776,
+      "learning_rate": 2.7936551724137932e-05,
+      "loss": 0.1155,
+      "step": 11743
+    },
+    {
+      "epoch": 40.49655172413793,
+      "grad_norm": 1.6839125156402588,
+      "learning_rate": 2.793609195402299e-05,
+      "loss": 0.11,
+      "step": 11744
+    },
+    {
+      "epoch": 40.5,
+      "grad_norm": 0.7727195024490356,
+      "learning_rate": 2.793563218390805e-05,
+      "loss": 0.1211,
+      "step": 11745
+    },
+    {
+      "epoch": 40.50344827586207,
+      "grad_norm": 0.6054051518440247,
+      "learning_rate": 2.7935172413793104e-05,
+      "loss": 0.0913,
+      "step": 11746
+    },
+    {
+      "epoch": 40.50689655172414,
+      "grad_norm": 0.6255416870117188,
+      "learning_rate": 2.793471264367816e-05,
+      "loss": 0.0832,
+      "step": 11747
+    },
+    {
+      "epoch": 40.51034482758621,
+      "grad_norm": 1.1705973148345947,
+      "learning_rate": 2.793425287356322e-05,
+      "loss": 0.112,
+      "step": 11748
+    },
+    {
+      "epoch": 40.51379310344828,
+      "grad_norm": 1.0925915241241455,
+      "learning_rate": 2.7933793103448277e-05,
+      "loss": 0.1019,
+      "step": 11749
+    },
+    {
+      "epoch": 40.51724137931034,
+      "grad_norm": 0.877522349357605,
+      "learning_rate": 2.7933333333333332e-05,
+      "loss": 0.1196,
+      "step": 11750
+    },
+    {
+      "epoch": 40.52068965517241,
+      "grad_norm": 0.7351245880126953,
+      "learning_rate": 2.793287356321839e-05,
+      "loss": 0.2012,
+      "step": 11751
+    },
+    {
+      "epoch": 40.52413793103448,
+      "grad_norm": 0.7945444583892822,
+      "learning_rate": 2.7932413793103446e-05,
+      "loss": 0.1697,
+      "step": 11752
+    },
+    {
+      "epoch": 40.52758620689655,
+      "grad_norm": 0.6955210566520691,
+      "learning_rate": 2.793195402298851e-05,
+      "loss": 0.1793,
+      "step": 11753
+    },
+    {
+      "epoch": 40.53103448275862,
+      "grad_norm": 0.8090256452560425,
+      "learning_rate": 2.7931494252873564e-05,
+      "loss": 0.1758,
+      "step": 11754
+    },
+    {
+      "epoch": 40.53448275862069,
+      "grad_norm": 0.7219394445419312,
+      "learning_rate": 2.793103448275862e-05,
+      "loss": 0.18,
+      "step": 11755
+    },
+    {
+      "epoch": 40.53793103448276,
+      "grad_norm": 1.8151274919509888,
+      "learning_rate": 2.7930574712643678e-05,
+      "loss": 0.1468,
+      "step": 11756
+    },
+    {
+      "epoch": 40.54137931034483,
+      "grad_norm": 1.0861194133758545,
+      "learning_rate": 2.7930114942528736e-05,
+      "loss": 0.1517,
+      "step": 11757
+    },
+    {
+      "epoch": 40.5448275862069,
+      "grad_norm": 0.5736615061759949,
+      "learning_rate": 2.7929655172413795e-05,
+      "loss": 0.1491,
+      "step": 11758
+    },
+    {
+      "epoch": 40.54827586206896,
+      "grad_norm": 0.6556744575500488,
+      "learning_rate": 2.792919540229885e-05,
+      "loss": 0.1436,
+      "step": 11759
+    },
+    {
+      "epoch": 40.55172413793103,
+      "grad_norm": 0.641986072063446,
+      "learning_rate": 2.7928735632183906e-05,
+      "loss": 0.137,
+      "step": 11760
+    },
+    {
+      "epoch": 40.5551724137931,
+      "grad_norm": 1.0249788761138916,
+      "learning_rate": 2.7928275862068968e-05,
+      "loss": 0.1429,
+      "step": 11761
+    },
+    {
+      "epoch": 40.55862068965517,
+      "grad_norm": 0.5966471433639526,
+      "learning_rate": 2.7927816091954023e-05,
+      "loss": 0.1416,
+      "step": 11762
+    },
+    {
+      "epoch": 40.56206896551724,
+      "grad_norm": 0.5168145895004272,
+      "learning_rate": 2.7927356321839082e-05,
+      "loss": 0.1242,
+      "step": 11763
+    },
+    {
+      "epoch": 40.56551724137931,
+      "grad_norm": 0.6422237157821655,
+      "learning_rate": 2.7926896551724137e-05,
+      "loss": 0.1205,
+      "step": 11764
+    },
+    {
+      "epoch": 40.56896551724138,
+      "grad_norm": 0.8198093175888062,
+      "learning_rate": 2.7926436781609196e-05,
+      "loss": 0.1216,
+      "step": 11765
+    },
+    {
+      "epoch": 40.57241379310345,
+      "grad_norm": 0.7240015268325806,
+      "learning_rate": 2.7925977011494254e-05,
+      "loss": 0.1243,
+      "step": 11766
+    },
+    {
+      "epoch": 40.57586206896552,
+      "grad_norm": 0.6955036520957947,
+      "learning_rate": 2.792551724137931e-05,
+      "loss": 0.1144,
+      "step": 11767
+    },
+    {
+      "epoch": 40.57931034482758,
+      "grad_norm": 0.9881756901741028,
+      "learning_rate": 2.792505747126437e-05,
+      "loss": 0.1179,
+      "step": 11768
+    },
+    {
+      "epoch": 40.58275862068965,
+      "grad_norm": 0.5950742959976196,
+      "learning_rate": 2.7924597701149427e-05,
+      "loss": 0.1188,
+      "step": 11769
+    },
+    {
+      "epoch": 40.58620689655172,
+      "grad_norm": 0.7534858584403992,
+      "learning_rate": 2.7924137931034482e-05,
+      "loss": 0.1104,
+      "step": 11770
+    },
+    {
+      "epoch": 40.58965517241379,
+      "grad_norm": 1.7280205488204956,
+      "learning_rate": 2.792367816091954e-05,
+      "loss": 0.1291,
+      "step": 11771
+    },
+    {
+      "epoch": 40.59310344827586,
+      "grad_norm": 2.9699957370758057,
+      "learning_rate": 2.7923218390804596e-05,
+      "loss": 0.1021,
+      "step": 11772
+    },
+    {
+      "epoch": 40.59655172413793,
+      "grad_norm": 2.6579055786132812,
+      "learning_rate": 2.792275862068966e-05,
+      "loss": 0.1267,
+      "step": 11773
+    },
+    {
+      "epoch": 40.6,
+      "grad_norm": 0.9990864992141724,
+      "learning_rate": 2.7922298850574714e-05,
+      "loss": 0.1184,
+      "step": 11774
+    },
+    {
+      "epoch": 40.60344827586207,
+      "grad_norm": 1.852492094039917,
+      "learning_rate": 2.792183908045977e-05,
+      "loss": 0.1501,
+      "step": 11775
+    },
+    {
+      "epoch": 40.60689655172414,
+      "grad_norm": 0.7869579792022705,
+      "learning_rate": 2.7921379310344828e-05,
+      "loss": 0.2018,
+      "step": 11776
+    },
+    {
+      "epoch": 40.610344827586204,
+      "grad_norm": 1.1157070398330688,
+      "learning_rate": 2.7920919540229886e-05,
+      "loss": 0.1974,
+      "step": 11777
+    },
+    {
+      "epoch": 40.61379310344827,
+      "grad_norm": 1.480584740638733,
+      "learning_rate": 2.792045977011494e-05,
+      "loss": 0.1659,
+      "step": 11778
+    },
+    {
+      "epoch": 40.61724137931034,
+      "grad_norm": 1.0910156965255737,
+      "learning_rate": 2.792e-05,
+      "loss": 0.1818,
+      "step": 11779
+    },
+    {
+      "epoch": 40.62068965517241,
+      "grad_norm": 0.8567787408828735,
+      "learning_rate": 2.7919540229885056e-05,
+      "loss": 0.1565,
+      "step": 11780
+    },
+    {
+      "epoch": 40.62413793103448,
+      "grad_norm": 2.324859619140625,
+      "learning_rate": 2.7919080459770118e-05,
+      "loss": 0.1597,
+      "step": 11781
+    },
+    {
+      "epoch": 40.62758620689655,
+      "grad_norm": 0.73204106092453,
+      "learning_rate": 2.7918620689655173e-05,
+      "loss": 0.1511,
+      "step": 11782
+    },
+    {
+      "epoch": 40.63103448275862,
+      "grad_norm": 1.0159928798675537,
+      "learning_rate": 2.7918160919540228e-05,
+      "loss": 0.1498,
+      "step": 11783
+    },
+    {
+      "epoch": 40.63448275862069,
+      "grad_norm": 0.4772260785102844,
+      "learning_rate": 2.7917701149425287e-05,
+      "loss": 0.1266,
+      "step": 11784
+    },
+    {
+      "epoch": 40.63793103448276,
+      "grad_norm": 0.6090050339698792,
+      "learning_rate": 2.7917241379310346e-05,
+      "loss": 0.1433,
+      "step": 11785
+    },
+    {
+      "epoch": 40.641379310344824,
+      "grad_norm": 0.9100318551063538,
+      "learning_rate": 2.7916781609195404e-05,
+      "loss": 0.1569,
+      "step": 11786
+    },
+    {
+      "epoch": 40.644827586206894,
+      "grad_norm": 1.195310354232788,
+      "learning_rate": 2.791632183908046e-05,
+      "loss": 0.1397,
+      "step": 11787
+    },
+    {
+      "epoch": 40.648275862068964,
+      "grad_norm": 0.9153019189834595,
+      "learning_rate": 2.7915862068965515e-05,
+      "loss": 0.124,
+      "step": 11788
+    },
+    {
+      "epoch": 40.65172413793103,
+      "grad_norm": 0.8968881368637085,
+      "learning_rate": 2.7915402298850577e-05,
+      "loss": 0.1596,
+      "step": 11789
+    },
+    {
+      "epoch": 40.6551724137931,
+      "grad_norm": 0.8315351605415344,
+      "learning_rate": 2.7914942528735632e-05,
+      "loss": 0.151,
+      "step": 11790
+    },
+    {
+      "epoch": 40.65862068965517,
+      "grad_norm": 0.9347983002662659,
+      "learning_rate": 2.791448275862069e-05,
+      "loss": 0.1404,
+      "step": 11791
+    },
+    {
+      "epoch": 40.66206896551724,
+      "grad_norm": 0.9890848398208618,
+      "learning_rate": 2.7914022988505746e-05,
+      "loss": 0.1286,
+      "step": 11792
+    },
+    {
+      "epoch": 40.66551724137931,
+      "grad_norm": 0.6763122081756592,
+      "learning_rate": 2.7913563218390805e-05,
+      "loss": 0.1172,
+      "step": 11793
+    },
+    {
+      "epoch": 40.66896551724138,
+      "grad_norm": 0.9448462724685669,
+      "learning_rate": 2.7913103448275864e-05,
+      "loss": 0.1068,
+      "step": 11794
+    },
+    {
+      "epoch": 40.672413793103445,
+      "grad_norm": 1.370289921760559,
+      "learning_rate": 2.791264367816092e-05,
+      "loss": 0.1139,
+      "step": 11795
+    },
+    {
+      "epoch": 40.675862068965515,
+      "grad_norm": 0.844764232635498,
+      "learning_rate": 2.7912183908045978e-05,
+      "loss": 0.115,
+      "step": 11796
+    },
+    {
+      "epoch": 40.679310344827584,
+      "grad_norm": 0.6647254824638367,
+      "learning_rate": 2.7911724137931036e-05,
+      "loss": 0.1034,
+      "step": 11797
+    },
+    {
+      "epoch": 40.682758620689654,
+      "grad_norm": 3.673616409301758,
+      "learning_rate": 2.791126436781609e-05,
+      "loss": 0.1151,
+      "step": 11798
+    },
+    {
+      "epoch": 40.686206896551724,
+      "grad_norm": 0.7570379376411438,
+      "learning_rate": 2.791080459770115e-05,
+      "loss": 0.1299,
+      "step": 11799
+    },
+    {
+      "epoch": 40.689655172413794,
+      "grad_norm": 1.274048089981079,
+      "learning_rate": 2.7910344827586206e-05,
+      "loss": 0.1861,
+      "step": 11800
+    },
+    {
+      "epoch": 40.69310344827586,
+      "grad_norm": 0.7875194549560547,
+      "learning_rate": 2.7909885057471268e-05,
+      "loss": 0.2268,
+      "step": 11801
+    },
+    {
+      "epoch": 40.69655172413793,
+      "grad_norm": 0.8849730491638184,
+      "learning_rate": 2.7909425287356323e-05,
+      "loss": 0.1776,
+      "step": 11802
+    },
+    {
+      "epoch": 40.7,
+      "grad_norm": 1.3349769115447998,
+      "learning_rate": 2.7908965517241378e-05,
+      "loss": 0.184,
+      "step": 11803
+    },
+    {
+      "epoch": 40.703448275862065,
+      "grad_norm": 0.6739205718040466,
+      "learning_rate": 2.7908505747126437e-05,
+      "loss": 0.1648,
+      "step": 11804
+    },
+    {
+      "epoch": 40.706896551724135,
+      "grad_norm": 0.7420412302017212,
+      "learning_rate": 2.7908045977011496e-05,
+      "loss": 0.1723,
+      "step": 11805
+    },
+    {
+      "epoch": 40.710344827586205,
+      "grad_norm": 0.8694465160369873,
+      "learning_rate": 2.7907586206896554e-05,
+      "loss": 0.1563,
+      "step": 11806
+    },
+    {
+      "epoch": 40.713793103448275,
+      "grad_norm": 0.6973437070846558,
+      "learning_rate": 2.790712643678161e-05,
+      "loss": 0.1582,
+      "step": 11807
+    },
+    {
+      "epoch": 40.717241379310344,
+      "grad_norm": 0.7746396064758301,
+      "learning_rate": 2.7906666666666665e-05,
+      "loss": 0.1501,
+      "step": 11808
+    },
+    {
+      "epoch": 40.720689655172414,
+      "grad_norm": 1.4606819152832031,
+      "learning_rate": 2.7906206896551727e-05,
+      "loss": 0.1417,
+      "step": 11809
+    },
+    {
+      "epoch": 40.724137931034484,
+      "grad_norm": 0.6521008610725403,
+      "learning_rate": 2.7905747126436782e-05,
+      "loss": 0.1438,
+      "step": 11810
+    },
+    {
+      "epoch": 40.727586206896554,
+      "grad_norm": 0.7490999698638916,
+      "learning_rate": 2.7905287356321838e-05,
+      "loss": 0.1611,
+      "step": 11811
+    },
+    {
+      "epoch": 40.73103448275862,
+      "grad_norm": 0.7601701617240906,
+      "learning_rate": 2.7904827586206896e-05,
+      "loss": 0.1282,
+      "step": 11812
+    },
+    {
+      "epoch": 40.734482758620686,
+      "grad_norm": 0.6846688389778137,
+      "learning_rate": 2.7904367816091955e-05,
+      "loss": 0.1358,
+      "step": 11813
+    },
+    {
+      "epoch": 40.737931034482756,
+      "grad_norm": 0.7778067588806152,
+      "learning_rate": 2.7903908045977014e-05,
+      "loss": 0.1286,
+      "step": 11814
+    },
+    {
+      "epoch": 40.741379310344826,
+      "grad_norm": 1.0610971450805664,
+      "learning_rate": 2.790344827586207e-05,
+      "loss": 0.1493,
+      "step": 11815
+    },
+    {
+      "epoch": 40.744827586206895,
+      "grad_norm": 1.0356992483139038,
+      "learning_rate": 2.7902988505747124e-05,
+      "loss": 0.1177,
+      "step": 11816
+    },
+    {
+      "epoch": 40.748275862068965,
+      "grad_norm": 1.1441752910614014,
+      "learning_rate": 2.7902528735632186e-05,
+      "loss": 0.116,
+      "step": 11817
+    },
+    {
+      "epoch": 40.751724137931035,
+      "grad_norm": 0.7899928092956543,
+      "learning_rate": 2.790206896551724e-05,
+      "loss": 0.1192,
+      "step": 11818
+    },
+    {
+      "epoch": 40.755172413793105,
+      "grad_norm": 0.6208645105361938,
+      "learning_rate": 2.79016091954023e-05,
+      "loss": 0.122,
+      "step": 11819
+    },
+    {
+      "epoch": 40.758620689655174,
+      "grad_norm": 1.048755168914795,
+      "learning_rate": 2.7901149425287356e-05,
+      "loss": 0.1028,
+      "step": 11820
+    },
+    {
+      "epoch": 40.762068965517244,
+      "grad_norm": 1.3539106845855713,
+      "learning_rate": 2.7900689655172414e-05,
+      "loss": 0.1193,
+      "step": 11821
+    },
+    {
+      "epoch": 40.765517241379314,
+      "grad_norm": 0.6610104441642761,
+      "learning_rate": 2.7900229885057473e-05,
+      "loss": 0.1115,
+      "step": 11822
+    },
+    {
+      "epoch": 40.76896551724138,
+      "grad_norm": 0.8307573199272156,
+      "learning_rate": 2.7899770114942528e-05,
+      "loss": 0.108,
+      "step": 11823
+    },
+    {
+      "epoch": 40.772413793103446,
+      "grad_norm": 0.88006192445755,
+      "learning_rate": 2.7899310344827587e-05,
+      "loss": 0.1279,
+      "step": 11824
+    },
+    {
+      "epoch": 40.775862068965516,
+      "grad_norm": 1.4205797910690308,
+      "learning_rate": 2.7898850574712646e-05,
+      "loss": 0.1211,
+      "step": 11825
+    },
+    {
+      "epoch": 40.779310344827586,
+      "grad_norm": 1.053330659866333,
+      "learning_rate": 2.78983908045977e-05,
+      "loss": 0.2244,
+      "step": 11826
+    },
+    {
+      "epoch": 40.782758620689656,
+      "grad_norm": 0.5796409249305725,
+      "learning_rate": 2.789793103448276e-05,
+      "loss": 0.1967,
+      "step": 11827
+    },
+    {
+      "epoch": 40.786206896551725,
+      "grad_norm": 0.5129582285881042,
+      "learning_rate": 2.7897471264367815e-05,
+      "loss": 0.172,
+      "step": 11828
+    },
+    {
+      "epoch": 40.789655172413795,
+      "grad_norm": 0.5083187818527222,
+      "learning_rate": 2.7897011494252877e-05,
+      "loss": 0.1812,
+      "step": 11829
+    },
+    {
+      "epoch": 40.793103448275865,
+      "grad_norm": 0.7066628336906433,
+      "learning_rate": 2.7896551724137932e-05,
+      "loss": 0.1873,
+      "step": 11830
+    },
+    {
+      "epoch": 40.796551724137935,
+      "grad_norm": 0.5763672590255737,
+      "learning_rate": 2.7896091954022987e-05,
+      "loss": 0.1519,
+      "step": 11831
+    },
+    {
+      "epoch": 40.8,
+      "grad_norm": 2.805004119873047,
+      "learning_rate": 2.7895632183908046e-05,
+      "loss": 0.1396,
+      "step": 11832
+    },
+    {
+      "epoch": 40.80344827586207,
+      "grad_norm": 0.6230461001396179,
+      "learning_rate": 2.7895172413793105e-05,
+      "loss": 0.1459,
+      "step": 11833
+    },
+    {
+      "epoch": 40.80689655172414,
+      "grad_norm": 0.4898602366447449,
+      "learning_rate": 2.7894712643678164e-05,
+      "loss": 0.1503,
+      "step": 11834
+    },
+    {
+      "epoch": 40.810344827586206,
+      "grad_norm": 0.6101446151733398,
+      "learning_rate": 2.789425287356322e-05,
+      "loss": 0.1436,
+      "step": 11835
+    },
+    {
+      "epoch": 40.813793103448276,
+      "grad_norm": 2.803314208984375,
+      "learning_rate": 2.7893793103448274e-05,
+      "loss": 0.1412,
+      "step": 11836
+    },
+    {
+      "epoch": 40.817241379310346,
+      "grad_norm": 1.1435198783874512,
+      "learning_rate": 2.7893333333333336e-05,
+      "loss": 0.1156,
+      "step": 11837
+    },
+    {
+      "epoch": 40.820689655172416,
+      "grad_norm": 0.7559360265731812,
+      "learning_rate": 2.789287356321839e-05,
+      "loss": 0.1354,
+      "step": 11838
+    },
+    {
+      "epoch": 40.824137931034485,
+      "grad_norm": 1.0477510690689087,
+      "learning_rate": 2.7892413793103447e-05,
+      "loss": 0.121,
+      "step": 11839
+    },
+    {
+      "epoch": 40.827586206896555,
+      "grad_norm": 0.6943573355674744,
+      "learning_rate": 2.7891954022988505e-05,
+      "loss": 0.1142,
+      "step": 11840
+    },
+    {
+      "epoch": 40.83103448275862,
+      "grad_norm": 0.6271378993988037,
+      "learning_rate": 2.7891494252873564e-05,
+      "loss": 0.12,
+      "step": 11841
+    },
+    {
+      "epoch": 40.83448275862069,
+      "grad_norm": 0.7589528560638428,
+      "learning_rate": 2.7891034482758623e-05,
+      "loss": 0.123,
+      "step": 11842
+    },
+    {
+      "epoch": 40.83793103448276,
+      "grad_norm": 1.1358497142791748,
+      "learning_rate": 2.7890574712643678e-05,
+      "loss": 0.1257,
+      "step": 11843
+    },
+    {
+      "epoch": 40.84137931034483,
+      "grad_norm": 1.1042550802230835,
+      "learning_rate": 2.7890114942528733e-05,
+      "loss": 0.1318,
+      "step": 11844
+    },
+    {
+      "epoch": 40.8448275862069,
+      "grad_norm": 0.6960496306419373,
+      "learning_rate": 2.7889655172413795e-05,
+      "loss": 0.1177,
+      "step": 11845
+    },
+    {
+      "epoch": 40.84827586206897,
+      "grad_norm": 0.6922822594642639,
+      "learning_rate": 2.788919540229885e-05,
+      "loss": 0.0955,
+      "step": 11846
+    },
+    {
+      "epoch": 40.851724137931036,
+      "grad_norm": 1.0474412441253662,
+      "learning_rate": 2.788873563218391e-05,
+      "loss": 0.102,
+      "step": 11847
+    },
+    {
+      "epoch": 40.855172413793106,
+      "grad_norm": 1.7826316356658936,
+      "learning_rate": 2.7888275862068965e-05,
+      "loss": 0.114,
+      "step": 11848
+    },
+    {
+      "epoch": 40.858620689655176,
+      "grad_norm": 0.9580329060554504,
+      "learning_rate": 2.7887816091954023e-05,
+      "loss": 0.1125,
+      "step": 11849
+    },
+    {
+      "epoch": 40.86206896551724,
+      "grad_norm": 0.9531774520874023,
+      "learning_rate": 2.7887356321839082e-05,
+      "loss": 0.1423,
+      "step": 11850
+    },
+    {
+      "epoch": 40.86551724137931,
+      "grad_norm": 0.659142792224884,
+      "learning_rate": 2.7886896551724137e-05,
+      "loss": 0.2409,
+      "step": 11851
+    },
+    {
+      "epoch": 40.86896551724138,
+      "grad_norm": 0.6826728582382202,
+      "learning_rate": 2.7886436781609196e-05,
+      "loss": 0.1995,
+      "step": 11852
+    },
+    {
+      "epoch": 40.87241379310345,
+      "grad_norm": 0.547243058681488,
+      "learning_rate": 2.7885977011494255e-05,
+      "loss": 0.1736,
+      "step": 11853
+    },
+    {
+      "epoch": 40.87586206896552,
+      "grad_norm": 1.1701080799102783,
+      "learning_rate": 2.788551724137931e-05,
+      "loss": 0.1799,
+      "step": 11854
+    },
+    {
+      "epoch": 40.87931034482759,
+      "grad_norm": 1.5189658403396606,
+      "learning_rate": 2.788505747126437e-05,
+      "loss": 0.1613,
+      "step": 11855
+    },
+    {
+      "epoch": 40.88275862068966,
+      "grad_norm": 1.4116151332855225,
+      "learning_rate": 2.7884597701149424e-05,
+      "loss": 0.1483,
+      "step": 11856
+    },
+    {
+      "epoch": 40.88620689655173,
+      "grad_norm": 0.8265933394432068,
+      "learning_rate": 2.7884137931034486e-05,
+      "loss": 0.1326,
+      "step": 11857
+    },
+    {
+      "epoch": 40.889655172413796,
+      "grad_norm": 2.1229608058929443,
+      "learning_rate": 2.788367816091954e-05,
+      "loss": 0.1705,
+      "step": 11858
+    },
+    {
+      "epoch": 40.89310344827586,
+      "grad_norm": 0.9507710933685303,
+      "learning_rate": 2.7883218390804597e-05,
+      "loss": 0.1535,
+      "step": 11859
+    },
+    {
+      "epoch": 40.89655172413793,
+      "grad_norm": 1.0392764806747437,
+      "learning_rate": 2.7882758620689655e-05,
+      "loss": 0.1404,
+      "step": 11860
+    },
+    {
+      "epoch": 40.9,
+      "grad_norm": 0.7244572639465332,
+      "learning_rate": 2.7882298850574714e-05,
+      "loss": 0.1268,
+      "step": 11861
+    },
+    {
+      "epoch": 40.90344827586207,
+      "grad_norm": 0.5686535239219666,
+      "learning_rate": 2.7881839080459773e-05,
+      "loss": 0.1426,
+      "step": 11862
+    },
+    {
+      "epoch": 40.90689655172414,
+      "grad_norm": 0.5321837067604065,
+      "learning_rate": 2.7881379310344828e-05,
+      "loss": 0.1473,
+      "step": 11863
+    },
+    {
+      "epoch": 40.91034482758621,
+      "grad_norm": 0.9425690770149231,
+      "learning_rate": 2.7880919540229883e-05,
+      "loss": 0.1515,
+      "step": 11864
+    },
+    {
+      "epoch": 40.91379310344828,
+      "grad_norm": 2.6231701374053955,
+      "learning_rate": 2.7880459770114945e-05,
+      "loss": 0.1443,
+      "step": 11865
+    },
+    {
+      "epoch": 40.91724137931035,
+      "grad_norm": 0.6337547302246094,
+      "learning_rate": 2.788e-05,
+      "loss": 0.1374,
+      "step": 11866
+    },
+    {
+      "epoch": 40.92068965517242,
+      "grad_norm": 0.9334180951118469,
+      "learning_rate": 2.7879540229885056e-05,
+      "loss": 0.1141,
+      "step": 11867
+    },
+    {
+      "epoch": 40.92413793103448,
+      "grad_norm": 0.5588310360908508,
+      "learning_rate": 2.7879080459770115e-05,
+      "loss": 0.1245,
+      "step": 11868
+    },
+    {
+      "epoch": 40.92758620689655,
+      "grad_norm": 0.7732442617416382,
+      "learning_rate": 2.7878620689655173e-05,
+      "loss": 0.1248,
+      "step": 11869
+    },
+    {
+      "epoch": 40.93103448275862,
+      "grad_norm": 0.9320520758628845,
+      "learning_rate": 2.7878160919540232e-05,
+      "loss": 0.1183,
+      "step": 11870
+    },
+    {
+      "epoch": 40.93448275862069,
+      "grad_norm": 4.59244966506958,
+      "learning_rate": 2.7877701149425287e-05,
+      "loss": 0.1158,
+      "step": 11871
+    },
+    {
+      "epoch": 40.93793103448276,
+      "grad_norm": 0.7224825024604797,
+      "learning_rate": 2.7877241379310343e-05,
+      "loss": 0.1088,
+      "step": 11872
+    },
+    {
+      "epoch": 40.94137931034483,
+      "grad_norm": 0.9052416086196899,
+      "learning_rate": 2.7876781609195405e-05,
+      "loss": 0.0993,
+      "step": 11873
+    },
+    {
+      "epoch": 40.9448275862069,
+      "grad_norm": 0.7622255086898804,
+      "learning_rate": 2.787632183908046e-05,
+      "loss": 0.1053,
+      "step": 11874
+    },
+    {
+      "epoch": 40.94827586206897,
+      "grad_norm": 1.82595694065094,
+      "learning_rate": 2.787586206896552e-05,
+      "loss": 0.1727,
+      "step": 11875
+    },
+    {
+      "epoch": 40.95172413793104,
+      "grad_norm": 0.8928139805793762,
+      "learning_rate": 2.7875402298850574e-05,
+      "loss": 0.1884,
+      "step": 11876
+    },
+    {
+      "epoch": 40.9551724137931,
+      "grad_norm": 0.6927770972251892,
+      "learning_rate": 2.7874942528735633e-05,
+      "loss": 0.1922,
+      "step": 11877
+    },
+    {
+      "epoch": 40.95862068965517,
+      "grad_norm": 0.9186509251594543,
+      "learning_rate": 2.787448275862069e-05,
+      "loss": 0.1554,
+      "step": 11878
+    },
+    {
+      "epoch": 40.96206896551724,
+      "grad_norm": 1.2442209720611572,
+      "learning_rate": 2.7874022988505747e-05,
+      "loss": 0.1656,
+      "step": 11879
+    },
+    {
+      "epoch": 40.96551724137931,
+      "grad_norm": 0.7952768206596375,
+      "learning_rate": 2.7873563218390805e-05,
+      "loss": 0.1712,
+      "step": 11880
+    },
+    {
+      "epoch": 40.96896551724138,
+      "grad_norm": 4.886902809143066,
+      "learning_rate": 2.7873103448275864e-05,
+      "loss": 0.1549,
+      "step": 11881
+    },
+    {
+      "epoch": 40.97241379310345,
+      "grad_norm": 0.9248790144920349,
+      "learning_rate": 2.787264367816092e-05,
+      "loss": 0.1605,
+      "step": 11882
+    },
+    {
+      "epoch": 40.97586206896552,
+      "grad_norm": 0.5161663293838501,
+      "learning_rate": 2.7872183908045978e-05,
+      "loss": 0.1385,
+      "step": 11883
+    },
+    {
+      "epoch": 40.97931034482759,
+      "grad_norm": 1.8625320196151733,
+      "learning_rate": 2.7871724137931033e-05,
+      "loss": 0.1542,
+      "step": 11884
+    },
+    {
+      "epoch": 40.98275862068966,
+      "grad_norm": 0.5619131326675415,
+      "learning_rate": 2.7871264367816095e-05,
+      "loss": 0.1164,
+      "step": 11885
+    },
+    {
+      "epoch": 40.98620689655172,
+      "grad_norm": 1.0469545125961304,
+      "learning_rate": 2.787080459770115e-05,
+      "loss": 0.1263,
+      "step": 11886
+    },
+    {
+      "epoch": 40.98965517241379,
+      "grad_norm": 0.8741365075111389,
+      "learning_rate": 2.7870344827586206e-05,
+      "loss": 0.1191,
+      "step": 11887
+    },
+    {
+      "epoch": 40.99310344827586,
+      "grad_norm": 0.9119535684585571,
+      "learning_rate": 2.7869885057471265e-05,
+      "loss": 0.1188,
+      "step": 11888
+    },
+    {
+      "epoch": 40.99655172413793,
+      "grad_norm": 0.720704972743988,
+      "learning_rate": 2.7869425287356323e-05,
+      "loss": 0.0982,
+      "step": 11889
+    },
+    {
+      "epoch": 41.0,
+      "grad_norm": 6.535160541534424,
+      "learning_rate": 2.7868965517241382e-05,
+      "loss": 0.1581,
+      "step": 11890
+    },
+    {
+      "epoch": 41.00344827586207,
+      "grad_norm": 0.6805986166000366,
+      "learning_rate": 2.7868505747126437e-05,
+      "loss": 0.2226,
+      "step": 11891
+    },
+    {
+      "epoch": 41.00689655172414,
+      "grad_norm": 0.572300910949707,
+      "learning_rate": 2.7868045977011493e-05,
+      "loss": 0.1827,
+      "step": 11892
+    },
+    {
+      "epoch": 41.01034482758621,
+      "grad_norm": 0.7962373495101929,
+      "learning_rate": 2.7867586206896555e-05,
+      "loss": 0.1738,
+      "step": 11893
+    },
+    {
+      "epoch": 41.01379310344828,
+      "grad_norm": 0.5130933523178101,
+      "learning_rate": 2.786712643678161e-05,
+      "loss": 0.1571,
+      "step": 11894
+    },
+    {
+      "epoch": 41.01724137931034,
+      "grad_norm": 1.1399163007736206,
+      "learning_rate": 2.7866666666666665e-05,
+      "loss": 0.1429,
+      "step": 11895
+    },
+    {
+      "epoch": 41.02068965517241,
+      "grad_norm": 0.642897367477417,
+      "learning_rate": 2.7866206896551724e-05,
+      "loss": 0.1567,
+      "step": 11896
+    },
+    {
+      "epoch": 41.02413793103448,
+      "grad_norm": 0.725182294845581,
+      "learning_rate": 2.7865747126436783e-05,
+      "loss": 0.1543,
+      "step": 11897
+    },
+    {
+      "epoch": 41.02758620689655,
+      "grad_norm": 0.6084275841712952,
+      "learning_rate": 2.786528735632184e-05,
+      "loss": 0.1556,
+      "step": 11898
+    },
+    {
+      "epoch": 41.03103448275862,
+      "grad_norm": 0.4611421823501587,
+      "learning_rate": 2.7864827586206897e-05,
+      "loss": 0.1344,
+      "step": 11899
+    },
+    {
+      "epoch": 41.03448275862069,
+      "grad_norm": 2.6376240253448486,
+      "learning_rate": 2.7864367816091952e-05,
+      "loss": 0.1328,
+      "step": 11900
+    },
+    {
+      "epoch": 41.03793103448276,
+      "grad_norm": 0.6717047095298767,
+      "learning_rate": 2.7863908045977014e-05,
+      "loss": 0.1345,
+      "step": 11901
+    },
+    {
+      "epoch": 41.04137931034483,
+      "grad_norm": 0.5581512451171875,
+      "learning_rate": 2.786344827586207e-05,
+      "loss": 0.1254,
+      "step": 11902
+    },
+    {
+      "epoch": 41.0448275862069,
+      "grad_norm": 0.896269679069519,
+      "learning_rate": 2.7862988505747128e-05,
+      "loss": 0.1384,
+      "step": 11903
+    },
+    {
+      "epoch": 41.04827586206896,
+      "grad_norm": 0.5928277969360352,
+      "learning_rate": 2.7862528735632183e-05,
+      "loss": 0.1307,
+      "step": 11904
+    },
+    {
+      "epoch": 41.05172413793103,
+      "grad_norm": 1.040685772895813,
+      "learning_rate": 2.7862068965517242e-05,
+      "loss": 0.123,
+      "step": 11905
+    },
+    {
+      "epoch": 41.0551724137931,
+      "grad_norm": 0.8467941284179688,
+      "learning_rate": 2.78616091954023e-05,
+      "loss": 0.1053,
+      "step": 11906
+    },
+    {
+      "epoch": 41.05862068965517,
+      "grad_norm": 0.6045325994491577,
+      "learning_rate": 2.7861149425287356e-05,
+      "loss": 0.1054,
+      "step": 11907
+    },
+    {
+      "epoch": 41.06206896551724,
+      "grad_norm": 1.0747803449630737,
+      "learning_rate": 2.7860689655172415e-05,
+      "loss": 0.1122,
+      "step": 11908
+    },
+    {
+      "epoch": 41.06551724137931,
+      "grad_norm": 1.2228373289108276,
+      "learning_rate": 2.7860229885057473e-05,
+      "loss": 0.121,
+      "step": 11909
+    },
+    {
+      "epoch": 41.06896551724138,
+      "grad_norm": 1.0814549922943115,
+      "learning_rate": 2.785977011494253e-05,
+      "loss": 0.0961,
+      "step": 11910
+    },
+    {
+      "epoch": 41.07241379310345,
+      "grad_norm": 0.8392794728279114,
+      "learning_rate": 2.7859310344827587e-05,
+      "loss": 0.1144,
+      "step": 11911
+    },
+    {
+      "epoch": 41.07586206896552,
+      "grad_norm": 0.5752587914466858,
+      "learning_rate": 2.7858850574712643e-05,
+      "loss": 0.0908,
+      "step": 11912
+    },
+    {
+      "epoch": 41.07931034482758,
+      "grad_norm": 1.0401684045791626,
+      "learning_rate": 2.7858390804597705e-05,
+      "loss": 0.1066,
+      "step": 11913
+    },
+    {
+      "epoch": 41.08275862068965,
+      "grad_norm": 1.4423011541366577,
+      "learning_rate": 2.785793103448276e-05,
+      "loss": 0.0978,
+      "step": 11914
+    },
+    {
+      "epoch": 41.08620689655172,
+      "grad_norm": 1.0593105554580688,
+      "learning_rate": 2.7857471264367815e-05,
+      "loss": 0.1423,
+      "step": 11915
+    },
+    {
+      "epoch": 41.08965517241379,
+      "grad_norm": 0.804160475730896,
+      "learning_rate": 2.7857011494252874e-05,
+      "loss": 0.2111,
+      "step": 11916
+    },
+    {
+      "epoch": 41.09310344827586,
+      "grad_norm": 0.6861268281936646,
+      "learning_rate": 2.7856551724137933e-05,
+      "loss": 0.1742,
+      "step": 11917
+    },
+    {
+      "epoch": 41.09655172413793,
+      "grad_norm": 0.6653047204017639,
+      "learning_rate": 2.785609195402299e-05,
+      "loss": 0.17,
+      "step": 11918
+    },
+    {
+      "epoch": 41.1,
+      "grad_norm": 0.8181291222572327,
+      "learning_rate": 2.7855632183908047e-05,
+      "loss": 0.1613,
+      "step": 11919
+    },
+    {
+      "epoch": 41.10344827586207,
+      "grad_norm": 0.7577361464500427,
+      "learning_rate": 2.7855172413793102e-05,
+      "loss": 0.1474,
+      "step": 11920
+    },
+    {
+      "epoch": 41.10689655172414,
+      "grad_norm": 0.5354645252227783,
+      "learning_rate": 2.7854712643678164e-05,
+      "loss": 0.1288,
+      "step": 11921
+    },
+    {
+      "epoch": 41.110344827586204,
+      "grad_norm": 0.6103320717811584,
+      "learning_rate": 2.785425287356322e-05,
+      "loss": 0.1479,
+      "step": 11922
+    },
+    {
+      "epoch": 41.11379310344827,
+      "grad_norm": 0.7566019892692566,
+      "learning_rate": 2.7853793103448278e-05,
+      "loss": 0.1577,
+      "step": 11923
+    },
+    {
+      "epoch": 41.11724137931034,
+      "grad_norm": 0.615723192691803,
+      "learning_rate": 2.7853333333333333e-05,
+      "loss": 0.1439,
+      "step": 11924
+    },
+    {
+      "epoch": 41.12068965517241,
+      "grad_norm": 0.5227155685424805,
+      "learning_rate": 2.7852873563218392e-05,
+      "loss": 0.1342,
+      "step": 11925
+    },
+    {
+      "epoch": 41.12413793103448,
+      "grad_norm": 0.954628050327301,
+      "learning_rate": 2.785241379310345e-05,
+      "loss": 0.1435,
+      "step": 11926
+    },
+    {
+      "epoch": 41.12758620689655,
+      "grad_norm": 0.9029730558395386,
+      "learning_rate": 2.7851954022988506e-05,
+      "loss": 0.1175,
+      "step": 11927
+    },
+    {
+      "epoch": 41.13103448275862,
+      "grad_norm": 0.6986355781555176,
+      "learning_rate": 2.785149425287356e-05,
+      "loss": 0.1281,
+      "step": 11928
+    },
+    {
+      "epoch": 41.13448275862069,
+      "grad_norm": 0.48720887303352356,
+      "learning_rate": 2.7851034482758623e-05,
+      "loss": 0.1301,
+      "step": 11929
+    },
+    {
+      "epoch": 41.13793103448276,
+      "grad_norm": 0.5560234785079956,
+      "learning_rate": 2.785057471264368e-05,
+      "loss": 0.12,
+      "step": 11930
+    },
+    {
+      "epoch": 41.141379310344824,
+      "grad_norm": 0.6567030549049377,
+      "learning_rate": 2.7850114942528737e-05,
+      "loss": 0.111,
+      "step": 11931
+    },
+    {
+      "epoch": 41.144827586206894,
+      "grad_norm": 1.221776008605957,
+      "learning_rate": 2.7849655172413792e-05,
+      "loss": 0.1256,
+      "step": 11932
+    },
+    {
+      "epoch": 41.148275862068964,
+      "grad_norm": 1.1641077995300293,
+      "learning_rate": 2.784919540229885e-05,
+      "loss": 0.1234,
+      "step": 11933
+    },
+    {
+      "epoch": 41.15172413793103,
+      "grad_norm": 0.5724689364433289,
+      "learning_rate": 2.784873563218391e-05,
+      "loss": 0.0992,
+      "step": 11934
+    },
+    {
+      "epoch": 41.1551724137931,
+      "grad_norm": 1.264756679534912,
+      "learning_rate": 2.7848275862068965e-05,
+      "loss": 0.1176,
+      "step": 11935
+    },
+    {
+      "epoch": 41.15862068965517,
+      "grad_norm": 1.532914161682129,
+      "learning_rate": 2.7847816091954024e-05,
+      "loss": 0.1023,
+      "step": 11936
+    },
+    {
+      "epoch": 41.16206896551724,
+      "grad_norm": 2.1076996326446533,
+      "learning_rate": 2.7847356321839083e-05,
+      "loss": 0.098,
+      "step": 11937
+    },
+    {
+      "epoch": 41.16551724137931,
+      "grad_norm": 0.7160444855690002,
+      "learning_rate": 2.7846896551724138e-05,
+      "loss": 0.0981,
+      "step": 11938
+    },
+    {
+      "epoch": 41.16896551724138,
+      "grad_norm": 0.6355757117271423,
+      "learning_rate": 2.7846436781609196e-05,
+      "loss": 0.0868,
+      "step": 11939
+    },
+    {
+      "epoch": 41.172413793103445,
+      "grad_norm": 1.03900146484375,
+      "learning_rate": 2.7845977011494252e-05,
+      "loss": 0.1194,
+      "step": 11940
+    },
+    {
+      "epoch": 41.175862068965515,
+      "grad_norm": 0.5686364769935608,
+      "learning_rate": 2.7845517241379314e-05,
+      "loss": 0.2129,
+      "step": 11941
+    },
+    {
+      "epoch": 41.179310344827584,
+      "grad_norm": 0.5388920903205872,
+      "learning_rate": 2.784505747126437e-05,
+      "loss": 0.1381,
+      "step": 11942
+    },
+    {
+      "epoch": 41.182758620689654,
+      "grad_norm": 0.6442224979400635,
+      "learning_rate": 2.7844597701149424e-05,
+      "loss": 0.1695,
+      "step": 11943
+    },
+    {
+      "epoch": 41.186206896551724,
+      "grad_norm": 1.1506837606430054,
+      "learning_rate": 2.7844137931034483e-05,
+      "loss": 0.1684,
+      "step": 11944
+    },
+    {
+      "epoch": 41.189655172413794,
+      "grad_norm": 2.0077507495880127,
+      "learning_rate": 2.7843678160919542e-05,
+      "loss": 0.1428,
+      "step": 11945
+    },
+    {
+      "epoch": 41.19310344827586,
+      "grad_norm": 0.631734311580658,
+      "learning_rate": 2.78432183908046e-05,
+      "loss": 0.1513,
+      "step": 11946
+    },
+    {
+      "epoch": 41.19655172413793,
+      "grad_norm": 0.8695550560951233,
+      "learning_rate": 2.7842758620689656e-05,
+      "loss": 0.1467,
+      "step": 11947
+    },
+    {
+      "epoch": 41.2,
+      "grad_norm": 0.5237994194030762,
+      "learning_rate": 2.784229885057471e-05,
+      "loss": 0.1535,
+      "step": 11948
+    },
+    {
+      "epoch": 41.203448275862065,
+      "grad_norm": 0.706161618232727,
+      "learning_rate": 2.7841839080459773e-05,
+      "loss": 0.1251,
+      "step": 11949
+    },
+    {
+      "epoch": 41.206896551724135,
+      "grad_norm": 0.7336849570274353,
+      "learning_rate": 2.784137931034483e-05,
+      "loss": 0.132,
+      "step": 11950
+    },
+    {
+      "epoch": 41.210344827586205,
+      "grad_norm": 0.5422658324241638,
+      "learning_rate": 2.7840919540229887e-05,
+      "loss": 0.121,
+      "step": 11951
+    },
+    {
+      "epoch": 41.213793103448275,
+      "grad_norm": 0.6456677913665771,
+      "learning_rate": 2.7840459770114942e-05,
+      "loss": 0.1297,
+      "step": 11952
+    },
+    {
+      "epoch": 41.217241379310344,
+      "grad_norm": 0.7811408638954163,
+      "learning_rate": 2.784e-05,
+      "loss": 0.1219,
+      "step": 11953
+    },
+    {
+      "epoch": 41.220689655172414,
+      "grad_norm": 0.8268111348152161,
+      "learning_rate": 2.783954022988506e-05,
+      "loss": 0.1165,
+      "step": 11954
+    },
+    {
+      "epoch": 41.224137931034484,
+      "grad_norm": 0.7123024463653564,
+      "learning_rate": 2.7839080459770115e-05,
+      "loss": 0.1258,
+      "step": 11955
+    },
+    {
+      "epoch": 41.227586206896554,
+      "grad_norm": 0.5659504532814026,
+      "learning_rate": 2.783862068965517e-05,
+      "loss": 0.1172,
+      "step": 11956
+    },
+    {
+      "epoch": 41.23103448275862,
+      "grad_norm": 0.993030309677124,
+      "learning_rate": 2.7838160919540232e-05,
+      "loss": 0.1119,
+      "step": 11957
+    },
+    {
+      "epoch": 41.234482758620686,
+      "grad_norm": 0.6471213698387146,
+      "learning_rate": 2.7837701149425288e-05,
+      "loss": 0.1057,
+      "step": 11958
+    },
+    {
+      "epoch": 41.237931034482756,
+      "grad_norm": 0.6332545876502991,
+      "learning_rate": 2.7837241379310346e-05,
+      "loss": 0.1038,
+      "step": 11959
+    },
+    {
+      "epoch": 41.241379310344826,
+      "grad_norm": 0.6505969762802124,
+      "learning_rate": 2.7836781609195402e-05,
+      "loss": 0.1046,
+      "step": 11960
+    },
+    {
+      "epoch": 41.244827586206895,
+      "grad_norm": 0.7660714387893677,
+      "learning_rate": 2.783632183908046e-05,
+      "loss": 0.1012,
+      "step": 11961
+    },
+    {
+      "epoch": 41.248275862068965,
+      "grad_norm": 0.619316041469574,
+      "learning_rate": 2.783586206896552e-05,
+      "loss": 0.079,
+      "step": 11962
+    },
+    {
+      "epoch": 41.251724137931035,
+      "grad_norm": 0.8008268475532532,
+      "learning_rate": 2.7835402298850574e-05,
+      "loss": 0.1011,
+      "step": 11963
+    },
+    {
+      "epoch": 41.255172413793105,
+      "grad_norm": 0.8644103407859802,
+      "learning_rate": 2.7834942528735633e-05,
+      "loss": 0.0878,
+      "step": 11964
+    },
+    {
+      "epoch": 41.258620689655174,
+      "grad_norm": 1.2630164623260498,
+      "learning_rate": 2.7834482758620692e-05,
+      "loss": 0.1432,
+      "step": 11965
+    },
+    {
+      "epoch": 41.262068965517244,
+      "grad_norm": 0.7756518125534058,
+      "learning_rate": 2.7834022988505747e-05,
+      "loss": 0.2246,
+      "step": 11966
+    },
+    {
+      "epoch": 41.265517241379314,
+      "grad_norm": 0.521003246307373,
+      "learning_rate": 2.7833563218390806e-05,
+      "loss": 0.1681,
+      "step": 11967
+    },
+    {
+      "epoch": 41.26896551724138,
+      "grad_norm": 0.6028465628623962,
+      "learning_rate": 2.783310344827586e-05,
+      "loss": 0.1795,
+      "step": 11968
+    },
+    {
+      "epoch": 41.272413793103446,
+      "grad_norm": 0.5923756957054138,
+      "learning_rate": 2.7832643678160923e-05,
+      "loss": 0.1676,
+      "step": 11969
+    },
+    {
+      "epoch": 41.275862068965516,
+      "grad_norm": 0.8886812925338745,
+      "learning_rate": 2.783218390804598e-05,
+      "loss": 0.1344,
+      "step": 11970
+    },
+    {
+      "epoch": 41.279310344827586,
+      "grad_norm": 1.0262154340744019,
+      "learning_rate": 2.7831724137931034e-05,
+      "loss": 0.1437,
+      "step": 11971
+    },
+    {
+      "epoch": 41.282758620689656,
+      "grad_norm": 1.1608253717422485,
+      "learning_rate": 2.7831264367816092e-05,
+      "loss": 0.1671,
+      "step": 11972
+    },
+    {
+      "epoch": 41.286206896551725,
+      "grad_norm": 0.9611548185348511,
+      "learning_rate": 2.783080459770115e-05,
+      "loss": 0.1403,
+      "step": 11973
+    },
+    {
+      "epoch": 41.289655172413795,
+      "grad_norm": 0.69005286693573,
+      "learning_rate": 2.783034482758621e-05,
+      "loss": 0.1386,
+      "step": 11974
+    },
+    {
+      "epoch": 41.293103448275865,
+      "grad_norm": 0.6114193797111511,
+      "learning_rate": 2.7829885057471265e-05,
+      "loss": 0.1308,
+      "step": 11975
+    },
+    {
+      "epoch": 41.296551724137935,
+      "grad_norm": 0.7856693863868713,
+      "learning_rate": 2.782942528735632e-05,
+      "loss": 0.1193,
+      "step": 11976
+    },
+    {
+      "epoch": 41.3,
+      "grad_norm": 0.793415367603302,
+      "learning_rate": 2.7828965517241382e-05,
+      "loss": 0.1382,
+      "step": 11977
+    },
+    {
+      "epoch": 41.30344827586207,
+      "grad_norm": 0.467740923166275,
+      "learning_rate": 2.7828505747126438e-05,
+      "loss": 0.111,
+      "step": 11978
+    },
+    {
+      "epoch": 41.30689655172414,
+      "grad_norm": 4.855302810668945,
+      "learning_rate": 2.7828045977011496e-05,
+      "loss": 0.1012,
+      "step": 11979
+    },
+    {
+      "epoch": 41.310344827586206,
+      "grad_norm": 0.6347041130065918,
+      "learning_rate": 2.782758620689655e-05,
+      "loss": 0.118,
+      "step": 11980
+    },
+    {
+      "epoch": 41.313793103448276,
+      "grad_norm": 0.6863338351249695,
+      "learning_rate": 2.782712643678161e-05,
+      "loss": 0.1285,
+      "step": 11981
+    },
+    {
+      "epoch": 41.317241379310346,
+      "grad_norm": 0.7415573000907898,
+      "learning_rate": 2.782666666666667e-05,
+      "loss": 0.1262,
+      "step": 11982
+    },
+    {
+      "epoch": 41.320689655172416,
+      "grad_norm": 2.5014076232910156,
+      "learning_rate": 2.7826206896551724e-05,
+      "loss": 0.1116,
+      "step": 11983
+    },
+    {
+      "epoch": 41.324137931034485,
+      "grad_norm": 0.9302916526794434,
+      "learning_rate": 2.782574712643678e-05,
+      "loss": 0.0943,
+      "step": 11984
+    },
+    {
+      "epoch": 41.327586206896555,
+      "grad_norm": 0.982127845287323,
+      "learning_rate": 2.7825287356321842e-05,
+      "loss": 0.1,
+      "step": 11985
+    },
+    {
+      "epoch": 41.33103448275862,
+      "grad_norm": 3.882004976272583,
+      "learning_rate": 2.7824827586206897e-05,
+      "loss": 0.1085,
+      "step": 11986
+    },
+    {
+      "epoch": 41.33448275862069,
+      "grad_norm": 0.8281375169754028,
+      "learning_rate": 2.7824367816091956e-05,
+      "loss": 0.0916,
+      "step": 11987
+    },
+    {
+      "epoch": 41.33793103448276,
+      "grad_norm": 0.8988734483718872,
+      "learning_rate": 2.782390804597701e-05,
+      "loss": 0.0929,
+      "step": 11988
+    },
+    {
+      "epoch": 41.34137931034483,
+      "grad_norm": 0.8919851779937744,
+      "learning_rate": 2.782344827586207e-05,
+      "loss": 0.0939,
+      "step": 11989
+    },
+    {
+      "epoch": 41.3448275862069,
+      "grad_norm": 1.2896007299423218,
+      "learning_rate": 2.782298850574713e-05,
+      "loss": 0.1532,
+      "step": 11990
+    },
+    {
+      "epoch": 41.34827586206897,
+      "grad_norm": 0.6805572509765625,
+      "learning_rate": 2.7822528735632184e-05,
+      "loss": 0.2117,
+      "step": 11991
+    },
+    {
+      "epoch": 41.351724137931036,
+      "grad_norm": 0.5688475370407104,
+      "learning_rate": 2.7822068965517242e-05,
+      "loss": 0.1796,
+      "step": 11992
+    },
+    {
+      "epoch": 41.355172413793106,
+      "grad_norm": 1.2013331651687622,
+      "learning_rate": 2.78216091954023e-05,
+      "loss": 0.1571,
+      "step": 11993
+    },
+    {
+      "epoch": 41.358620689655176,
+      "grad_norm": 1.2206281423568726,
+      "learning_rate": 2.7821149425287356e-05,
+      "loss": 0.1494,
+      "step": 11994
+    },
+    {
+      "epoch": 41.36206896551724,
+      "grad_norm": 0.5915943384170532,
+      "learning_rate": 2.7820689655172415e-05,
+      "loss": 0.1678,
+      "step": 11995
+    },
+    {
+      "epoch": 41.36551724137931,
+      "grad_norm": 0.6025797128677368,
+      "learning_rate": 2.782022988505747e-05,
+      "loss": 0.1413,
+      "step": 11996
+    },
+    {
+      "epoch": 41.36896551724138,
+      "grad_norm": 0.7588165402412415,
+      "learning_rate": 2.7819770114942532e-05,
+      "loss": 0.1628,
+      "step": 11997
+    },
+    {
+      "epoch": 41.37241379310345,
+      "grad_norm": 0.5640226602554321,
+      "learning_rate": 2.7819310344827588e-05,
+      "loss": 0.1445,
+      "step": 11998
+    },
+    {
+      "epoch": 41.37586206896552,
+      "grad_norm": 1.794654369354248,
+      "learning_rate": 2.7818850574712643e-05,
+      "loss": 0.1355,
+      "step": 11999
+    },
+    {
+      "epoch": 41.37931034482759,
+      "grad_norm": 0.6887781620025635,
+      "learning_rate": 2.78183908045977e-05,
+      "loss": 0.1455,
+      "step": 12000
+    },
+    {
+      "epoch": 41.37931034482759,
+      "eval_cer": 0.1347972885899121,
+      "eval_loss": 0.35223913192749023,
+      "eval_runtime": 18.6328,
+      "eval_samples_per_second": 49.751,
+      "eval_steps_per_second": 0.161,
+      "eval_wer": 0.3079710144927536,
+      "step": 12000
     }
   ],
   "logging_steps": 1.0,
@@ -77131,7 +84141,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 3
+        "early_stopping_patience_counter": 4
       }
     },
     "TrainerControl": {
@@ -77145,7 +84155,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.1495567357267096e+20,
+  "total_flos": 4.527891266461646e+20,
   "train_batch_size": 160,
   "trial_name": null,
   "trial_params": null