diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.20012108536518244,
+  "epoch": 0.3001816280477737,
   "eval_steps": 500,
-  "global_step": 3636,
+  "global_step": 5454,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -25459,6 +25459,12732 @@
       "learning_rate": 9.766905396590851e-06,
       "loss": 0.8658,
       "step": 3636
+    },
+    {
+      "epoch": 0.2001761241675381,
+      "grad_norm": 0.8647506237030029,
+      "learning_rate": 9.766774571898516e-06,
+      "loss": 0.84,
+      "step": 3637
+    },
+    {
+      "epoch": 0.20023116296989377,
+      "grad_norm": 0.8545078635215759,
+      "learning_rate": 9.766643711380378e-06,
+      "loss": 0.8455,
+      "step": 3638
+    },
+    {
+      "epoch": 0.20028620177224943,
+      "grad_norm": 0.924404501914978,
+      "learning_rate": 9.766512815037424e-06,
+      "loss": 0.6954,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2003412405746051,
+      "grad_norm": 0.8077614903450012,
+      "learning_rate": 9.766381882870635e-06,
+      "loss": 0.7724,
+      "step": 3640
+    },
+    {
+      "epoch": 0.20039627937696075,
+      "grad_norm": 0.8886739015579224,
+      "learning_rate": 9.766250914880994e-06,
+      "loss": 0.8318,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2004513181793164,
+      "grad_norm": 0.8086267113685608,
+      "learning_rate": 9.76611991106949e-06,
+      "loss": 0.8494,
+      "step": 3642
+    },
+    {
+      "epoch": 0.20050635698167207,
+      "grad_norm": 0.8606873750686646,
+      "learning_rate": 9.765988871437101e-06,
+      "loss": 0.8488,
+      "step": 3643
+    },
+    {
+      "epoch": 0.20056139578402774,
+      "grad_norm": 0.6966355443000793,
+      "learning_rate": 9.76585779598482e-06,
+      "loss": 0.7361,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2006164345863834,
+      "grad_norm": 0.8474385738372803,
+      "learning_rate": 9.765726684713623e-06,
+      "loss": 0.8354,
+      "step": 3645
+    },
+    {
+      "epoch": 0.20067147338873906,
+      "grad_norm": 0.7609736919403076,
+      "learning_rate": 9.765595537624502e-06,
+      "loss": 0.7297,
+      "step": 3646
+    },
+    {
+      "epoch": 0.20072651219109472,
+      "grad_norm": 1.08648681640625,
+      "learning_rate": 9.76546435471844e-06,
+      "loss": 0.7534,
+      "step": 3647
+    },
+    {
+      "epoch": 0.20078155099345038,
+      "grad_norm": 0.7437332272529602,
+      "learning_rate": 9.765333135996425e-06,
+      "loss": 0.8532,
+      "step": 3648
+    },
+    {
+      "epoch": 0.20083658979580604,
+      "grad_norm": 0.9016552567481995,
+      "learning_rate": 9.76520188145944e-06,
+      "loss": 0.7968,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2008916285981617,
+      "grad_norm": 0.8916428089141846,
+      "learning_rate": 9.765070591108473e-06,
+      "loss": 0.9601,
+      "step": 3650
+    },
+    {
+      "epoch": 0.20094666740051736,
+      "grad_norm": 0.7679058313369751,
+      "learning_rate": 9.764939264944512e-06,
+      "loss": 0.816,
+      "step": 3651
+    },
+    {
+      "epoch": 0.20100170620287303,
+      "grad_norm": 0.7716549634933472,
+      "learning_rate": 9.764807902968543e-06,
+      "loss": 0.876,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2010567450052287,
+      "grad_norm": 0.8288074731826782,
+      "learning_rate": 9.764676505181554e-06,
+      "loss": 0.8054,
+      "step": 3653
+    },
+    {
+      "epoch": 0.20111178380758435,
+      "grad_norm": 0.7906842827796936,
+      "learning_rate": 9.76454507158453e-06,
+      "loss": 0.8026,
+      "step": 3654
+    },
+    {
+      "epoch": 0.20116682260994,
+      "grad_norm": 0.8093311190605164,
+      "learning_rate": 9.764413602178461e-06,
+      "loss": 0.8093,
+      "step": 3655
+    },
+    {
+      "epoch": 0.20122186141229567,
+      "grad_norm": 0.7234730124473572,
+      "learning_rate": 9.764282096964335e-06,
+      "loss": 0.7194,
+      "step": 3656
+    },
+    {
+      "epoch": 0.20127690021465133,
+      "grad_norm": 0.9048555493354797,
+      "learning_rate": 9.76415055594314e-06,
+      "loss": 0.8996,
+      "step": 3657
+    },
+    {
+      "epoch": 0.201331939017007,
+      "grad_norm": 0.7630691528320312,
+      "learning_rate": 9.764018979115864e-06,
+      "loss": 0.7876,
+      "step": 3658
+    },
+    {
+      "epoch": 0.20138697781936266,
+      "grad_norm": 0.9551032781600952,
+      "learning_rate": 9.763887366483498e-06,
+      "loss": 0.8249,
+      "step": 3659
+    },
+    {
+      "epoch": 0.20144201662171832,
+      "grad_norm": 0.6988314986228943,
+      "learning_rate": 9.76375571804703e-06,
+      "loss": 0.8011,
+      "step": 3660
+    },
+    {
+      "epoch": 0.20149705542407398,
+      "grad_norm": 0.7790704369544983,
+      "learning_rate": 9.763624033807448e-06,
+      "loss": 0.8287,
+      "step": 3661
+    },
+    {
+      "epoch": 0.20155209422642964,
+      "grad_norm": 0.7201293706893921,
+      "learning_rate": 9.763492313765743e-06,
+      "loss": 0.7854,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2016071330287853,
+      "grad_norm": 0.8691730499267578,
+      "learning_rate": 9.763360557922905e-06,
+      "loss": 0.8348,
+      "step": 3663
+    },
+    {
+      "epoch": 0.20166217183114096,
+      "grad_norm": 0.7660881876945496,
+      "learning_rate": 9.763228766279924e-06,
+      "loss": 0.7686,
+      "step": 3664
+    },
+    {
+      "epoch": 0.20171721063349662,
+      "grad_norm": 1.083796501159668,
+      "learning_rate": 9.76309693883779e-06,
+      "loss": 0.8848,
+      "step": 3665
+    },
+    {
+      "epoch": 0.20177224943585229,
+      "grad_norm": 0.7892678380012512,
+      "learning_rate": 9.762965075597496e-06,
+      "loss": 0.7804,
+      "step": 3666
+    },
+    {
+      "epoch": 0.20182728823820795,
+      "grad_norm": 0.7166122198104858,
+      "learning_rate": 9.762833176560031e-06,
+      "loss": 0.761,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2018823270405636,
+      "grad_norm": 0.8187084794044495,
+      "learning_rate": 9.762701241726386e-06,
+      "loss": 0.8251,
+      "step": 3668
+    },
+    {
+      "epoch": 0.20193736584291927,
+      "grad_norm": 0.6930577158927917,
+      "learning_rate": 9.762569271097556e-06,
+      "loss": 0.6795,
+      "step": 3669
+    },
+    {
+      "epoch": 0.20199240464527493,
+      "grad_norm": 0.8085465431213379,
+      "learning_rate": 9.762437264674527e-06,
+      "loss": 0.8415,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2020474434476306,
+      "grad_norm": 0.8111084699630737,
+      "learning_rate": 9.762305222458294e-06,
+      "loss": 0.792,
+      "step": 3671
+    },
+    {
+      "epoch": 0.20210248224998625,
+      "grad_norm": 0.8200401067733765,
+      "learning_rate": 9.762173144449852e-06,
+      "loss": 0.8224,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2021575210523419,
+      "grad_norm": 0.8460109233856201,
+      "learning_rate": 9.762041030650192e-06,
+      "loss": 0.9025,
+      "step": 3673
+    },
+    {
+      "epoch": 0.20221255985469755,
+      "grad_norm": 0.8152671456336975,
+      "learning_rate": 9.761908881060303e-06,
+      "loss": 0.9002,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2022675986570532,
+      "grad_norm": 0.8204773664474487,
+      "learning_rate": 9.761776695681185e-06,
+      "loss": 0.8324,
+      "step": 3675
+    },
+    {
+      "epoch": 0.20232263745940887,
+      "grad_norm": 0.8121044039726257,
+      "learning_rate": 9.761644474513825e-06,
+      "loss": 0.855,
+      "step": 3676
+    },
+    {
+      "epoch": 0.20237767626176453,
+      "grad_norm": 0.79920494556427,
+      "learning_rate": 9.76151221755922e-06,
+      "loss": 0.7837,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2024327150641202,
+      "grad_norm": 0.862808346748352,
+      "learning_rate": 9.761379924818367e-06,
+      "loss": 0.8714,
+      "step": 3678
+    },
+    {
+      "epoch": 0.20248775386647586,
+      "grad_norm": 0.7135004997253418,
+      "learning_rate": 9.761247596292254e-06,
+      "loss": 0.774,
+      "step": 3679
+    },
+    {
+      "epoch": 0.20254279266883152,
+      "grad_norm": 0.7967603802680969,
+      "learning_rate": 9.761115231981878e-06,
+      "loss": 0.919,
+      "step": 3680
+    },
+    {
+      "epoch": 0.20259783147118718,
+      "grad_norm": 0.7425099611282349,
+      "learning_rate": 9.760982831888236e-06,
+      "loss": 0.819,
+      "step": 3681
+    },
+    {
+      "epoch": 0.20265287027354284,
+      "grad_norm": 0.7631763815879822,
+      "learning_rate": 9.760850396012323e-06,
+      "loss": 0.816,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2027079090758985,
+      "grad_norm": 0.7931755185127258,
+      "learning_rate": 9.76071792435513e-06,
+      "loss": 0.8299,
+      "step": 3683
+    },
+    {
+      "epoch": 0.20276294787825416,
+      "grad_norm": 0.8409438729286194,
+      "learning_rate": 9.760585416917657e-06,
+      "loss": 0.8503,
+      "step": 3684
+    },
+    {
+      "epoch": 0.20281798668060982,
+      "grad_norm": 0.7632728815078735,
+      "learning_rate": 9.760452873700898e-06,
+      "loss": 0.8394,
+      "step": 3685
+    },
+    {
+      "epoch": 0.20287302548296549,
+      "grad_norm": 0.7765083312988281,
+      "learning_rate": 9.76032029470585e-06,
+      "loss": 0.8879,
+      "step": 3686
+    },
+    {
+      "epoch": 0.20292806428532115,
+      "grad_norm": 0.7736936807632446,
+      "learning_rate": 9.760187679933507e-06,
+      "loss": 0.7987,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2029831030876768,
+      "grad_norm": 0.8270270824432373,
+      "learning_rate": 9.760055029384869e-06,
+      "loss": 0.8267,
+      "step": 3688
+    },
+    {
+      "epoch": 0.20303814189003247,
+      "grad_norm": 0.7742369174957275,
+      "learning_rate": 9.759922343060932e-06,
+      "loss": 0.8447,
+      "step": 3689
+    },
+    {
+      "epoch": 0.20309318069238813,
+      "grad_norm": 0.7543869018554688,
+      "learning_rate": 9.759789620962692e-06,
+      "loss": 0.7325,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2031482194947438,
+      "grad_norm": 0.7913174033164978,
+      "learning_rate": 9.759656863091147e-06,
+      "loss": 0.8622,
+      "step": 3691
+    },
+    {
+      "epoch": 0.20320325829709945,
+      "grad_norm": 0.7445376515388489,
+      "learning_rate": 9.759524069447296e-06,
+      "loss": 0.7115,
+      "step": 3692
+    },
+    {
+      "epoch": 0.20325829709945512,
+      "grad_norm": 0.7744696140289307,
+      "learning_rate": 9.759391240032136e-06,
+      "loss": 0.8437,
+      "step": 3693
+    },
+    {
+      "epoch": 0.20331333590181078,
+      "grad_norm": 0.6984724998474121,
+      "learning_rate": 9.759258374846665e-06,
+      "loss": 0.7415,
+      "step": 3694
+    },
+    {
+      "epoch": 0.20336837470416644,
+      "grad_norm": 0.7453249096870422,
+      "learning_rate": 9.759125473891882e-06,
+      "loss": 0.7708,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2034234135065221,
+      "grad_norm": 0.7459438443183899,
+      "learning_rate": 9.758992537168787e-06,
+      "loss": 0.7961,
+      "step": 3696
+    },
+    {
+      "epoch": 0.20347845230887776,
+      "grad_norm": 0.808944582939148,
+      "learning_rate": 9.758859564678377e-06,
+      "loss": 0.8875,
+      "step": 3697
+    },
+    {
+      "epoch": 0.20353349111123342,
+      "grad_norm": 0.7202889323234558,
+      "learning_rate": 9.758726556421652e-06,
+      "loss": 0.8064,
+      "step": 3698
+    },
+    {
+      "epoch": 0.20358852991358908,
+      "grad_norm": 0.7874952554702759,
+      "learning_rate": 9.758593512399613e-06,
+      "loss": 0.7881,
+      "step": 3699
+    },
+    {
+      "epoch": 0.20364356871594474,
+      "grad_norm": 0.771300733089447,
+      "learning_rate": 9.758460432613259e-06,
+      "loss": 0.8938,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2036986075183004,
+      "grad_norm": 0.7332000136375427,
+      "learning_rate": 9.758327317063589e-06,
+      "loss": 0.7369,
+      "step": 3701
+    },
+    {
+      "epoch": 0.20375364632065607,
+      "grad_norm": 0.8206236958503723,
+      "learning_rate": 9.758194165751604e-06,
+      "loss": 0.8727,
+      "step": 3702
+    },
+    {
+      "epoch": 0.20380868512301173,
+      "grad_norm": 0.7750238180160522,
+      "learning_rate": 9.758060978678308e-06,
+      "loss": 0.8013,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2038637239253674,
+      "grad_norm": 0.7213704586029053,
+      "learning_rate": 9.757927755844698e-06,
+      "loss": 0.7413,
+      "step": 3704
+    },
+    {
+      "epoch": 0.20391876272772305,
+      "grad_norm": 0.8982640504837036,
+      "learning_rate": 9.757794497251776e-06,
+      "loss": 0.9077,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2039738015300787,
+      "grad_norm": 0.8439363241195679,
+      "learning_rate": 9.757661202900544e-06,
+      "loss": 0.7887,
+      "step": 3706
+    },
+    {
+      "epoch": 0.20402884033243437,
+      "grad_norm": 0.7700560688972473,
+      "learning_rate": 9.757527872792005e-06,
+      "loss": 0.8677,
+      "step": 3707
+    },
+    {
+      "epoch": 0.20408387913479004,
+      "grad_norm": 0.7462438941001892,
+      "learning_rate": 9.75739450692716e-06,
+      "loss": 0.7937,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2041389179371457,
+      "grad_norm": 0.9125999808311462,
+      "learning_rate": 9.75726110530701e-06,
+      "loss": 0.9374,
+      "step": 3709
+    },
+    {
+      "epoch": 0.20419395673950136,
+      "grad_norm": 0.8949875831604004,
+      "learning_rate": 9.75712766793256e-06,
+      "loss": 0.8586,
+      "step": 3710
+    },
+    {
+      "epoch": 0.20424899554185702,
+      "grad_norm": 0.9042442440986633,
+      "learning_rate": 9.756994194804812e-06,
+      "loss": 0.9411,
+      "step": 3711
+    },
+    {
+      "epoch": 0.20430403434421268,
+      "grad_norm": 0.7646238207817078,
+      "learning_rate": 9.756860685924769e-06,
+      "loss": 0.8353,
+      "step": 3712
+    },
+    {
+      "epoch": 0.20435907314656834,
+      "grad_norm": 0.7551934123039246,
+      "learning_rate": 9.756727141293434e-06,
+      "loss": 0.8109,
+      "step": 3713
+    },
+    {
+      "epoch": 0.204414111948924,
+      "grad_norm": 0.7526532411575317,
+      "learning_rate": 9.756593560911811e-06,
+      "loss": 0.8509,
+      "step": 3714
+    },
+    {
+      "epoch": 0.20446915075127967,
+      "grad_norm": 0.8423319458961487,
+      "learning_rate": 9.756459944780903e-06,
+      "loss": 0.9003,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2045241895536353,
+      "grad_norm": 0.7966015934944153,
+      "learning_rate": 9.756326292901716e-06,
+      "loss": 0.7606,
+      "step": 3716
+    },
+    {
+      "epoch": 0.20457922835599096,
+      "grad_norm": 0.7642805576324463,
+      "learning_rate": 9.756192605275256e-06,
+      "loss": 0.8321,
+      "step": 3717
+    },
+    {
+      "epoch": 0.20463426715834662,
+      "grad_norm": 0.7285729646682739,
+      "learning_rate": 9.756058881902524e-06,
+      "loss": 0.7375,
+      "step": 3718
+    },
+    {
+      "epoch": 0.20468930596070228,
+      "grad_norm": 0.852020263671875,
+      "learning_rate": 9.755925122784525e-06,
+      "loss": 0.8207,
+      "step": 3719
+    },
+    {
+      "epoch": 0.20474434476305794,
+      "grad_norm": 0.8227072358131409,
+      "learning_rate": 9.755791327922268e-06,
+      "loss": 0.872,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2047993835654136,
+      "grad_norm": 1.0128127336502075,
+      "learning_rate": 9.755657497316755e-06,
+      "loss": 0.9186,
+      "step": 3721
+    },
+    {
+      "epoch": 0.20485442236776927,
+      "grad_norm": 0.8208017349243164,
+      "learning_rate": 9.755523630968994e-06,
+      "loss": 0.6968,
+      "step": 3722
+    },
+    {
+      "epoch": 0.20490946117012493,
+      "grad_norm": 0.7716407179832458,
+      "learning_rate": 9.75538972887999e-06,
+      "loss": 0.8068,
+      "step": 3723
+    },
+    {
+      "epoch": 0.2049644999724806,
+      "grad_norm": 0.779608964920044,
+      "learning_rate": 9.75525579105075e-06,
+      "loss": 0.6968,
+      "step": 3724
+    },
+    {
+      "epoch": 0.20501953877483625,
+      "grad_norm": 0.7463479042053223,
+      "learning_rate": 9.75512181748228e-06,
+      "loss": 0.7581,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2050745775771919,
+      "grad_norm": 0.8104956150054932,
+      "learning_rate": 9.754987808175587e-06,
+      "loss": 0.7838,
+      "step": 3726
+    },
+    {
+      "epoch": 0.20512961637954757,
+      "grad_norm": 0.7911564707756042,
+      "learning_rate": 9.75485376313168e-06,
+      "loss": 0.848,
+      "step": 3727
+    },
+    {
+      "epoch": 0.20518465518190324,
+      "grad_norm": 0.8340871334075928,
+      "learning_rate": 9.754719682351564e-06,
+      "loss": 0.7879,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2052396939842589,
+      "grad_norm": 1.5543067455291748,
+      "learning_rate": 9.754585565836247e-06,
+      "loss": 0.8091,
+      "step": 3729
+    },
+    {
+      "epoch": 0.20529473278661456,
+      "grad_norm": 0.8262580633163452,
+      "learning_rate": 9.754451413586739e-06,
+      "loss": 0.9076,
+      "step": 3730
+    },
+    {
+      "epoch": 0.20534977158897022,
+      "grad_norm": 0.7558280825614929,
+      "learning_rate": 9.754317225604045e-06,
+      "loss": 0.7781,
+      "step": 3731
+    },
+    {
+      "epoch": 0.20540481039132588,
+      "grad_norm": 0.7197710275650024,
+      "learning_rate": 9.754183001889177e-06,
+      "loss": 0.765,
+      "step": 3732
+    },
+    {
+      "epoch": 0.20545984919368154,
+      "grad_norm": 0.8053440451622009,
+      "learning_rate": 9.754048742443141e-06,
+      "loss": 0.7986,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2055148879960372,
+      "grad_norm": 0.9183983206748962,
+      "learning_rate": 9.753914447266947e-06,
+      "loss": 0.8522,
+      "step": 3734
+    },
+    {
+      "epoch": 0.20556992679839287,
+      "grad_norm": 0.8095504641532898,
+      "learning_rate": 9.753780116361607e-06,
+      "loss": 0.7243,
+      "step": 3735
+    },
+    {
+      "epoch": 0.20562496560074853,
+      "grad_norm": 0.816818356513977,
+      "learning_rate": 9.753645749728127e-06,
+      "loss": 0.8262,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2056800044031042,
+      "grad_norm": 0.8425988554954529,
+      "learning_rate": 9.753511347367516e-06,
+      "loss": 0.8142,
+      "step": 3737
+    },
+    {
+      "epoch": 0.20573504320545985,
+      "grad_norm": 0.7719724178314209,
+      "learning_rate": 9.753376909280789e-06,
+      "loss": 0.8444,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2057900820078155,
+      "grad_norm": 0.877646803855896,
+      "learning_rate": 9.753242435468952e-06,
+      "loss": 0.8515,
+      "step": 3739
+    },
+    {
+      "epoch": 0.20584512081017117,
+      "grad_norm": 0.9261211156845093,
+      "learning_rate": 9.753107925933017e-06,
+      "loss": 0.7605,
+      "step": 3740
+    },
+    {
+      "epoch": 0.20590015961252683,
+      "grad_norm": 0.7790889739990234,
+      "learning_rate": 9.752973380673995e-06,
+      "loss": 0.7911,
+      "step": 3741
+    },
+    {
+      "epoch": 0.2059551984148825,
+      "grad_norm": 0.7112367153167725,
+      "learning_rate": 9.752838799692899e-06,
+      "loss": 0.8212,
+      "step": 3742
+    },
+    {
+      "epoch": 0.20601023721723816,
+      "grad_norm": 0.7568365335464478,
+      "learning_rate": 9.752704182990736e-06,
+      "loss": 0.8505,
+      "step": 3743
+    },
+    {
+      "epoch": 0.20606527601959382,
+      "grad_norm": 0.7501981258392334,
+      "learning_rate": 9.752569530568523e-06,
+      "loss": 0.8191,
+      "step": 3744
+    },
+    {
+      "epoch": 0.20612031482194948,
+      "grad_norm": 0.7822220325469971,
+      "learning_rate": 9.752434842427268e-06,
+      "loss": 0.8032,
+      "step": 3745
+    },
+    {
+      "epoch": 0.20617535362430514,
+      "grad_norm": 0.810197114944458,
+      "learning_rate": 9.752300118567987e-06,
+      "loss": 0.7789,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2062303924266608,
+      "grad_norm": 0.7386943101882935,
+      "learning_rate": 9.752165358991688e-06,
+      "loss": 0.7733,
+      "step": 3747
+    },
+    {
+      "epoch": 0.20628543122901646,
+      "grad_norm": 0.7086807489395142,
+      "learning_rate": 9.75203056369939e-06,
+      "loss": 0.6328,
+      "step": 3748
+    },
+    {
+      "epoch": 0.20634047003137213,
+      "grad_norm": 0.9881154894828796,
+      "learning_rate": 9.751895732692099e-06,
+      "loss": 0.8515,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2063955088337278,
+      "grad_norm": 0.813521683216095,
+      "learning_rate": 9.751760865970831e-06,
+      "loss": 0.8438,
+      "step": 3750
+    },
+    {
+      "epoch": 0.20645054763608345,
+      "grad_norm": 0.8357470631599426,
+      "learning_rate": 9.751625963536602e-06,
+      "loss": 0.7635,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2065055864384391,
+      "grad_norm": 0.8629693388938904,
+      "learning_rate": 9.751491025390423e-06,
+      "loss": 0.888,
+      "step": 3752
+    },
+    {
+      "epoch": 0.20656062524079477,
+      "grad_norm": 0.8844664096832275,
+      "learning_rate": 9.751356051533311e-06,
+      "loss": 0.7654,
+      "step": 3753
+    },
+    {
+      "epoch": 0.20661566404315043,
+      "grad_norm": 0.7006319165229797,
+      "learning_rate": 9.751221041966276e-06,
+      "loss": 0.7618,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2066707028455061,
+      "grad_norm": 0.9291046261787415,
+      "learning_rate": 9.75108599669034e-06,
+      "loss": 0.8485,
+      "step": 3755
+    },
+    {
+      "epoch": 0.20672574164786175,
+      "grad_norm": 0.7670828700065613,
+      "learning_rate": 9.75095091570651e-06,
+      "loss": 0.7856,
+      "step": 3756
+    },
+    {
+      "epoch": 0.20678078045021742,
+      "grad_norm": 0.8709883689880371,
+      "learning_rate": 9.750815799015804e-06,
+      "loss": 0.7983,
+      "step": 3757
+    },
+    {
+      "epoch": 0.20683581925257308,
+      "grad_norm": 0.7688055634498596,
+      "learning_rate": 9.750680646619241e-06,
+      "loss": 0.8064,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2068908580549287,
+      "grad_norm": 0.9492738246917725,
+      "learning_rate": 9.750545458517832e-06,
+      "loss": 0.8256,
+      "step": 3759
+    },
+    {
+      "epoch": 0.20694589685728437,
+      "grad_norm": 0.9685352444648743,
+      "learning_rate": 9.750410234712596e-06,
+      "loss": 0.839,
+      "step": 3760
+    },
+    {
+      "epoch": 0.20700093565964003,
+      "grad_norm": 0.788577139377594,
+      "learning_rate": 9.750274975204547e-06,
+      "loss": 0.8743,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2070559744619957,
+      "grad_norm": 0.8496370315551758,
+      "learning_rate": 9.750139679994703e-06,
+      "loss": 0.9286,
+      "step": 3762
+    },
+    {
+      "epoch": 0.20711101326435136,
+      "grad_norm": 0.9539788961410522,
+      "learning_rate": 9.750004349084083e-06,
+      "loss": 0.7568,
+      "step": 3763
+    },
+    {
+      "epoch": 0.20716605206670702,
+      "grad_norm": 0.8825643062591553,
+      "learning_rate": 9.7498689824737e-06,
+      "loss": 0.9339,
+      "step": 3764
+    },
+    {
+      "epoch": 0.20722109086906268,
+      "grad_norm": 0.7771373391151428,
+      "learning_rate": 9.749733580164573e-06,
+      "loss": 0.851,
+      "step": 3765
+    },
+    {
+      "epoch": 0.20727612967141834,
+      "grad_norm": 0.7460281252861023,
+      "learning_rate": 9.749598142157721e-06,
+      "loss": 0.8208,
+      "step": 3766
+    },
+    {
+      "epoch": 0.207331168473774,
+      "grad_norm": 0.8370739817619324,
+      "learning_rate": 9.74946266845416e-06,
+      "loss": 0.8634,
+      "step": 3767
+    },
+    {
+      "epoch": 0.20738620727612966,
+      "grad_norm": 0.7770463228225708,
+      "learning_rate": 9.749327159054907e-06,
+      "loss": 0.7955,
+      "step": 3768
+    },
+    {
+      "epoch": 0.20744124607848533,
+      "grad_norm": 0.8048208355903625,
+      "learning_rate": 9.749191613960985e-06,
+      "loss": 0.7736,
+      "step": 3769
+    },
+    {
+      "epoch": 0.207496284880841,
+      "grad_norm": 0.9187547564506531,
+      "learning_rate": 9.74905603317341e-06,
+      "loss": 0.8534,
+      "step": 3770
+    },
+    {
+      "epoch": 0.20755132368319665,
+      "grad_norm": 0.7304024696350098,
+      "learning_rate": 9.7489204166932e-06,
+      "loss": 0.72,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2076063624855523,
+      "grad_norm": 0.86177659034729,
+      "learning_rate": 9.748784764521376e-06,
+      "loss": 0.7838,
+      "step": 3772
+    },
+    {
+      "epoch": 0.20766140128790797,
+      "grad_norm": 0.7988011837005615,
+      "learning_rate": 9.748649076658956e-06,
+      "loss": 0.7776,
+      "step": 3773
+    },
+    {
+      "epoch": 0.20771644009026363,
+      "grad_norm": 0.706099808216095,
+      "learning_rate": 9.74851335310696e-06,
+      "loss": 0.759,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2077714788926193,
+      "grad_norm": 0.8125914931297302,
+      "learning_rate": 9.748377593866412e-06,
+      "loss": 0.8155,
+      "step": 3775
+    },
+    {
+      "epoch": 0.20782651769497495,
+      "grad_norm": 0.8603429794311523,
+      "learning_rate": 9.748241798938326e-06,
+      "loss": 0.8018,
+      "step": 3776
+    },
+    {
+      "epoch": 0.20788155649733062,
+      "grad_norm": 0.7735254764556885,
+      "learning_rate": 9.748105968323726e-06,
+      "loss": 0.7788,
+      "step": 3777
+    },
+    {
+      "epoch": 0.20793659529968628,
+      "grad_norm": 0.9037501811981201,
+      "learning_rate": 9.747970102023635e-06,
+      "loss": 0.8907,
+      "step": 3778
+    },
+    {
+      "epoch": 0.20799163410204194,
+      "grad_norm": 0.8781846761703491,
+      "learning_rate": 9.74783420003907e-06,
+      "loss": 0.867,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2080466729043976,
+      "grad_norm": 0.8486423492431641,
+      "learning_rate": 9.747698262371052e-06,
+      "loss": 0.817,
+      "step": 3780
+    },
+    {
+      "epoch": 0.20810171170675326,
+      "grad_norm": 0.8242751359939575,
+      "learning_rate": 9.747562289020607e-06,
+      "loss": 0.7385,
+      "step": 3781
+    },
+    {
+      "epoch": 0.20815675050910892,
+      "grad_norm": 0.8776529431343079,
+      "learning_rate": 9.747426279988754e-06,
+      "loss": 0.8222,
+      "step": 3782
+    },
+    {
+      "epoch": 0.20821178931146458,
+      "grad_norm": 0.7428975105285645,
+      "learning_rate": 9.747290235276517e-06,
+      "loss": 0.6954,
+      "step": 3783
+    },
+    {
+      "epoch": 0.20826682811382025,
+      "grad_norm": 0.8631997108459473,
+      "learning_rate": 9.747154154884917e-06,
+      "loss": 0.7956,
+      "step": 3784
+    },
+    {
+      "epoch": 0.2083218669161759,
+      "grad_norm": 0.7819229364395142,
+      "learning_rate": 9.747018038814976e-06,
+      "loss": 0.778,
+      "step": 3785
+    },
+    {
+      "epoch": 0.20837690571853157,
+      "grad_norm": 0.7770963311195374,
+      "learning_rate": 9.746881887067718e-06,
+      "loss": 0.8055,
+      "step": 3786
+    },
+    {
+      "epoch": 0.20843194452088723,
+      "grad_norm": 0.7168729305267334,
+      "learning_rate": 9.746745699644169e-06,
+      "loss": 0.7476,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2084869833232429,
+      "grad_norm": 0.7963632941246033,
+      "learning_rate": 9.746609476545348e-06,
+      "loss": 0.8083,
+      "step": 3788
+    },
+    {
+      "epoch": 0.20854202212559855,
+      "grad_norm": 0.6689679026603699,
+      "learning_rate": 9.746473217772281e-06,
+      "loss": 0.6687,
+      "step": 3789
+    },
+    {
+      "epoch": 0.20859706092795421,
+      "grad_norm": 0.8085560202598572,
+      "learning_rate": 9.746336923325991e-06,
+      "loss": 0.8221,
+      "step": 3790
+    },
+    {
+      "epoch": 0.20865209973030988,
+      "grad_norm": 0.7215744256973267,
+      "learning_rate": 9.746200593207505e-06,
+      "loss": 0.7261,
+      "step": 3791
+    },
+    {
+      "epoch": 0.20870713853266554,
+      "grad_norm": 0.7821729779243469,
+      "learning_rate": 9.746064227417844e-06,
+      "loss": 0.7683,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2087621773350212,
+      "grad_norm": 1.0014925003051758,
+      "learning_rate": 9.745927825958036e-06,
+      "loss": 0.7485,
+      "step": 3793
+    },
+    {
+      "epoch": 0.20881721613737686,
+      "grad_norm": 0.9447367787361145,
+      "learning_rate": 9.745791388829102e-06,
+      "loss": 0.835,
+      "step": 3794
+    },
+    {
+      "epoch": 0.20887225493973252,
+      "grad_norm": 0.7333751916885376,
+      "learning_rate": 9.745654916032073e-06,
+      "loss": 0.811,
+      "step": 3795
+    },
+    {
+      "epoch": 0.20892729374208818,
+      "grad_norm": 0.7516912221908569,
+      "learning_rate": 9.745518407567973e-06,
+      "loss": 0.7669,
+      "step": 3796
+    },
+    {
+      "epoch": 0.20898233254444384,
+      "grad_norm": 0.7826053500175476,
+      "learning_rate": 9.745381863437824e-06,
+      "loss": 0.7963,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2090373713467995,
+      "grad_norm": 0.8258751630783081,
+      "learning_rate": 9.745245283642658e-06,
+      "loss": 0.7929,
+      "step": 3798
+    },
+    {
+      "epoch": 0.20909241014915517,
+      "grad_norm": 0.7990522980690002,
+      "learning_rate": 9.745108668183497e-06,
+      "loss": 0.8518,
+      "step": 3799
+    },
+    {
+      "epoch": 0.20914744895151083,
+      "grad_norm": 1.3855403661727905,
+      "learning_rate": 9.744972017061369e-06,
+      "loss": 0.7768,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2092024877538665,
+      "grad_norm": 0.8456707000732422,
+      "learning_rate": 9.744835330277302e-06,
+      "loss": 0.7629,
+      "step": 3801
+    },
+    {
+      "epoch": 0.20925752655622212,
+      "grad_norm": 0.8992564678192139,
+      "learning_rate": 9.744698607832323e-06,
+      "loss": 0.8991,
+      "step": 3802
+    },
+    {
+      "epoch": 0.20931256535857778,
+      "grad_norm": 0.8533509969711304,
+      "learning_rate": 9.744561849727459e-06,
+      "loss": 0.8883,
+      "step": 3803
+    },
+    {
+      "epoch": 0.20936760416093345,
+      "grad_norm": 0.8363122940063477,
+      "learning_rate": 9.744425055963739e-06,
+      "loss": 0.8537,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2094226429632891,
+      "grad_norm": 0.7462213039398193,
+      "learning_rate": 9.744288226542189e-06,
+      "loss": 0.7713,
+      "step": 3805
+    },
+    {
+      "epoch": 0.20947768176564477,
+      "grad_norm": 0.8148539066314697,
+      "learning_rate": 9.744151361463841e-06,
+      "loss": 0.7887,
+      "step": 3806
+    },
+    {
+      "epoch": 0.20953272056800043,
+      "grad_norm": 0.7504319548606873,
+      "learning_rate": 9.744014460729718e-06,
+      "loss": 0.7385,
+      "step": 3807
+    },
+    {
+      "epoch": 0.2095877593703561,
+      "grad_norm": 0.9291114807128906,
+      "learning_rate": 9.743877524340854e-06,
+      "loss": 0.9886,
+      "step": 3808
+    },
+    {
+      "epoch": 0.20964279817271175,
+      "grad_norm": 0.7747925519943237,
+      "learning_rate": 9.743740552298276e-06,
+      "loss": 0.8772,
+      "step": 3809
+    },
+    {
+      "epoch": 0.20969783697506741,
+      "grad_norm": 0.7283097505569458,
+      "learning_rate": 9.743603544603016e-06,
+      "loss": 0.7403,
+      "step": 3810
+    },
+    {
+      "epoch": 0.20975287577742308,
+      "grad_norm": 0.8403457999229431,
+      "learning_rate": 9.743466501256098e-06,
+      "loss": 0.7998,
+      "step": 3811
+    },
+    {
+      "epoch": 0.20980791457977874,
+      "grad_norm": 0.8218665719032288,
+      "learning_rate": 9.743329422258557e-06,
+      "loss": 0.8019,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2098629533821344,
+      "grad_norm": 0.6991317868232727,
+      "learning_rate": 9.743192307611423e-06,
+      "loss": 0.743,
+      "step": 3813
+    },
+    {
+      "epoch": 0.20991799218449006,
+      "grad_norm": 0.767295241355896,
+      "learning_rate": 9.743055157315725e-06,
+      "loss": 0.8003,
+      "step": 3814
+    },
+    {
+      "epoch": 0.20997303098684572,
+      "grad_norm": 0.9457303285598755,
+      "learning_rate": 9.742917971372492e-06,
+      "loss": 0.8448,
+      "step": 3815
+    },
+    {
+      "epoch": 0.21002806978920138,
+      "grad_norm": 0.7839058637619019,
+      "learning_rate": 9.742780749782758e-06,
+      "loss": 0.8828,
+      "step": 3816
+    },
+    {
+      "epoch": 0.21008310859155704,
+      "grad_norm": 0.7831344604492188,
+      "learning_rate": 9.742643492547553e-06,
+      "loss": 0.7714,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2101381473939127,
+      "grad_norm": 0.7637175917625427,
+      "learning_rate": 9.74250619966791e-06,
+      "loss": 0.7508,
+      "step": 3818
+    },
+    {
+      "epoch": 0.21019318619626837,
+      "grad_norm": 0.8830221891403198,
+      "learning_rate": 9.74236887114486e-06,
+      "loss": 0.8508,
+      "step": 3819
+    },
+    {
+      "epoch": 0.21024822499862403,
+      "grad_norm": 0.7803365588188171,
+      "learning_rate": 9.742231506979434e-06,
+      "loss": 0.8094,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2103032638009797,
+      "grad_norm": 0.7701493501663208,
+      "learning_rate": 9.742094107172666e-06,
+      "loss": 0.8851,
+      "step": 3821
+    },
+    {
+      "epoch": 0.21035830260333535,
+      "grad_norm": 0.6434544324874878,
+      "learning_rate": 9.741956671725588e-06,
+      "loss": 0.7015,
+      "step": 3822
+    },
+    {
+      "epoch": 0.210413341405691,
+      "grad_norm": 0.7294684052467346,
+      "learning_rate": 9.741819200639233e-06,
+      "loss": 0.7357,
+      "step": 3823
+    },
+    {
+      "epoch": 0.21046838020804667,
+      "grad_norm": 0.702367901802063,
+      "learning_rate": 9.741681693914635e-06,
+      "loss": 0.7518,
+      "step": 3824
+    },
+    {
+      "epoch": 0.21052341901040234,
+      "grad_norm": 0.7567246556282043,
+      "learning_rate": 9.741544151552826e-06,
+      "loss": 0.8259,
+      "step": 3825
+    },
+    {
+      "epoch": 0.210578457812758,
+      "grad_norm": 0.7478607892990112,
+      "learning_rate": 9.741406573554841e-06,
+      "loss": 0.81,
+      "step": 3826
+    },
+    {
+      "epoch": 0.21063349661511366,
+      "grad_norm": 0.7270129323005676,
+      "learning_rate": 9.741268959921712e-06,
+      "loss": 0.8201,
+      "step": 3827
+    },
+    {
+      "epoch": 0.21068853541746932,
+      "grad_norm": 0.8108176589012146,
+      "learning_rate": 9.741131310654475e-06,
+      "loss": 0.8425,
+      "step": 3828
+    },
+    {
+      "epoch": 0.21074357421982498,
+      "grad_norm": 0.7773691415786743,
+      "learning_rate": 9.740993625754165e-06,
+      "loss": 0.8372,
+      "step": 3829
+    },
+    {
+      "epoch": 0.21079861302218064,
+      "grad_norm": 0.8988421559333801,
+      "learning_rate": 9.740855905221816e-06,
+      "loss": 0.8285,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2108536518245363,
+      "grad_norm": 0.7339534759521484,
+      "learning_rate": 9.740718149058462e-06,
+      "loss": 0.7567,
+      "step": 3831
+    },
+    {
+      "epoch": 0.21090869062689196,
+      "grad_norm": 0.8465108275413513,
+      "learning_rate": 9.740580357265141e-06,
+      "loss": 0.8747,
+      "step": 3832
+    },
+    {
+      "epoch": 0.21096372942924763,
+      "grad_norm": 0.7956714034080505,
+      "learning_rate": 9.740442529842885e-06,
+      "loss": 0.7665,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2110187682316033,
+      "grad_norm": 0.96270751953125,
+      "learning_rate": 9.740304666792733e-06,
+      "loss": 0.8338,
+      "step": 3834
+    },
+    {
+      "epoch": 0.21107380703395895,
+      "grad_norm": 0.812329113483429,
+      "learning_rate": 9.74016676811572e-06,
+      "loss": 0.8407,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2111288458363146,
+      "grad_norm": 0.7975192070007324,
+      "learning_rate": 9.740028833812882e-06,
+      "loss": 0.7836,
+      "step": 3836
+    },
+    {
+      "epoch": 0.21118388463867027,
+      "grad_norm": 0.826621949672699,
+      "learning_rate": 9.739890863885258e-06,
+      "loss": 0.732,
+      "step": 3837
+    },
+    {
+      "epoch": 0.21123892344102593,
+      "grad_norm": 0.9015662670135498,
+      "learning_rate": 9.73975285833388e-06,
+      "loss": 0.8837,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2112939622433816,
+      "grad_norm": 0.7641518712043762,
+      "learning_rate": 9.73961481715979e-06,
+      "loss": 0.7334,
+      "step": 3839
+    },
+    {
+      "epoch": 0.21134900104573726,
+      "grad_norm": 0.8062206506729126,
+      "learning_rate": 9.739476740364023e-06,
+      "loss": 0.8381,
+      "step": 3840
+    },
+    {
+      "epoch": 0.21140403984809292,
+      "grad_norm": 0.7301875352859497,
+      "learning_rate": 9.739338627947619e-06,
+      "loss": 0.7389,
+      "step": 3841
+    },
+    {
+      "epoch": 0.21145907865044858,
+      "grad_norm": 0.8995181322097778,
+      "learning_rate": 9.739200479911612e-06,
+      "loss": 0.8111,
+      "step": 3842
+    },
+    {
+      "epoch": 0.21151411745280424,
+      "grad_norm": 0.7154433131217957,
+      "learning_rate": 9.739062296257045e-06,
+      "loss": 0.7501,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2115691562551599,
+      "grad_norm": 0.8403087854385376,
+      "learning_rate": 9.738924076984954e-06,
+      "loss": 0.8212,
+      "step": 3844
+    },
+    {
+      "epoch": 0.21162419505751554,
+      "grad_norm": 0.7616639137268066,
+      "learning_rate": 9.738785822096377e-06,
+      "loss": 0.82,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2116792338598712,
+      "grad_norm": 0.7897970080375671,
+      "learning_rate": 9.738647531592356e-06,
+      "loss": 0.7972,
+      "step": 3846
+    },
+    {
+      "epoch": 0.21173427266222686,
+      "grad_norm": 0.7909015417098999,
+      "learning_rate": 9.738509205473928e-06,
+      "loss": 0.7939,
+      "step": 3847
+    },
+    {
+      "epoch": 0.21178931146458252,
+      "grad_norm": 0.9553212523460388,
+      "learning_rate": 9.73837084374213e-06,
+      "loss": 0.8672,
+      "step": 3848
+    },
+    {
+      "epoch": 0.21184435026693818,
+      "grad_norm": 0.9558283686637878,
+      "learning_rate": 9.73823244639801e-06,
+      "loss": 0.897,
+      "step": 3849
+    },
+    {
+      "epoch": 0.21189938906929384,
+      "grad_norm": 0.819530725479126,
+      "learning_rate": 9.7380940134426e-06,
+      "loss": 0.86,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2119544278716495,
+      "grad_norm": 0.7301751971244812,
+      "learning_rate": 9.737955544876945e-06,
+      "loss": 0.8265,
+      "step": 3851
+    },
+    {
+      "epoch": 0.21200946667400516,
+      "grad_norm": 0.8564972281455994,
+      "learning_rate": 9.737817040702085e-06,
+      "loss": 0.8253,
+      "step": 3852
+    },
+    {
+      "epoch": 0.21206450547636083,
+      "grad_norm": 0.7715204358100891,
+      "learning_rate": 9.737678500919059e-06,
+      "loss": 0.7779,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2121195442787165,
+      "grad_norm": 0.7296929955482483,
+      "learning_rate": 9.73753992552891e-06,
+      "loss": 0.787,
+      "step": 3854
+    },
+    {
+      "epoch": 0.21217458308107215,
+      "grad_norm": 0.8574217557907104,
+      "learning_rate": 9.73740131453268e-06,
+      "loss": 0.797,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2122296218834278,
+      "grad_norm": 0.8320643901824951,
+      "learning_rate": 9.737262667931409e-06,
+      "loss": 0.876,
+      "step": 3856
+    },
+    {
+      "epoch": 0.21228466068578347,
+      "grad_norm": 0.7313587069511414,
+      "learning_rate": 9.73712398572614e-06,
+      "loss": 0.7151,
+      "step": 3857
+    },
+    {
+      "epoch": 0.21233969948813913,
+      "grad_norm": 0.7039312720298767,
+      "learning_rate": 9.736985267917916e-06,
+      "loss": 0.7353,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2123947382904948,
+      "grad_norm": 0.7893409132957458,
+      "learning_rate": 9.736846514507776e-06,
+      "loss": 0.8383,
+      "step": 3859
+    },
+    {
+      "epoch": 0.21244977709285046,
+      "grad_norm": 0.8771371245384216,
+      "learning_rate": 9.736707725496767e-06,
+      "loss": 0.7543,
+      "step": 3860
+    },
+    {
+      "epoch": 0.21250481589520612,
+      "grad_norm": 1.0067707300186157,
+      "learning_rate": 9.736568900885932e-06,
+      "loss": 0.796,
+      "step": 3861
+    },
+    {
+      "epoch": 0.21255985469756178,
+      "grad_norm": 0.9171931743621826,
+      "learning_rate": 9.736430040676312e-06,
+      "loss": 0.8174,
+      "step": 3862
+    },
+    {
+      "epoch": 0.21261489349991744,
+      "grad_norm": 0.7616068720817566,
+      "learning_rate": 9.736291144868952e-06,
+      "loss": 0.7762,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2126699323022731,
+      "grad_norm": 0.789010226726532,
+      "learning_rate": 9.736152213464895e-06,
+      "loss": 0.7749,
+      "step": 3864
+    },
+    {
+      "epoch": 0.21272497110462876,
+      "grad_norm": 0.7943348288536072,
+      "learning_rate": 9.736013246465187e-06,
+      "loss": 0.6687,
+      "step": 3865
+    },
+    {
+      "epoch": 0.21278000990698442,
+      "grad_norm": 0.8351758718490601,
+      "learning_rate": 9.73587424387087e-06,
+      "loss": 0.9201,
+      "step": 3866
+    },
+    {
+      "epoch": 0.21283504870934009,
+      "grad_norm": 0.7710975408554077,
+      "learning_rate": 9.735735205682991e-06,
+      "loss": 0.8357,
+      "step": 3867
+    },
+    {
+      "epoch": 0.21289008751169575,
+      "grad_norm": 0.8955768942832947,
+      "learning_rate": 9.73559613190259e-06,
+      "loss": 0.8396,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2129451263140514,
+      "grad_norm": 0.8664666414260864,
+      "learning_rate": 9.735457022530722e-06,
+      "loss": 0.8176,
+      "step": 3869
+    },
+    {
+      "epoch": 0.21300016511640707,
+      "grad_norm": 0.7955949902534485,
+      "learning_rate": 9.735317877568424e-06,
+      "loss": 0.8421,
+      "step": 3870
+    },
+    {
+      "epoch": 0.21305520391876273,
+      "grad_norm": 0.849866509437561,
+      "learning_rate": 9.735178697016742e-06,
+      "loss": 0.7677,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2131102427211184,
+      "grad_norm": 0.7782625555992126,
+      "learning_rate": 9.735039480876727e-06,
+      "loss": 0.7838,
+      "step": 3872
+    },
+    {
+      "epoch": 0.21316528152347405,
+      "grad_norm": 0.7734919190406799,
+      "learning_rate": 9.734900229149423e-06,
+      "loss": 0.757,
+      "step": 3873
+    },
+    {
+      "epoch": 0.21322032032582972,
+      "grad_norm": 0.8462040424346924,
+      "learning_rate": 9.734760941835876e-06,
+      "loss": 0.8841,
+      "step": 3874
+    },
+    {
+      "epoch": 0.21327535912818538,
+      "grad_norm": 0.7219869494438171,
+      "learning_rate": 9.734621618937133e-06,
+      "loss": 0.7651,
+      "step": 3875
+    },
+    {
+      "epoch": 0.21333039793054104,
+      "grad_norm": 0.7550874352455139,
+      "learning_rate": 9.734482260454241e-06,
+      "loss": 0.8032,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2133854367328967,
+      "grad_norm": 0.7504588961601257,
+      "learning_rate": 9.734342866388247e-06,
+      "loss": 0.7923,
+      "step": 3877
+    },
+    {
+      "epoch": 0.21344047553525236,
+      "grad_norm": 0.7407390475273132,
+      "learning_rate": 9.7342034367402e-06,
+      "loss": 0.7569,
+      "step": 3878
+    },
+    {
+      "epoch": 0.21349551433760802,
+      "grad_norm": 0.7911562323570251,
+      "learning_rate": 9.734063971511147e-06,
+      "loss": 0.8726,
+      "step": 3879
+    },
+    {
+      "epoch": 0.21355055313996368,
+      "grad_norm": 0.9132450819015503,
+      "learning_rate": 9.733924470702139e-06,
+      "loss": 0.9445,
+      "step": 3880
+    },
+    {
+      "epoch": 0.21360559194231934,
+      "grad_norm": 0.9639442563056946,
+      "learning_rate": 9.733784934314218e-06,
+      "loss": 0.7307,
+      "step": 3881
+    },
+    {
+      "epoch": 0.213660630744675,
+      "grad_norm": 0.7724352478981018,
+      "learning_rate": 9.73364536234844e-06,
+      "loss": 0.8337,
+      "step": 3882
+    },
+    {
+      "epoch": 0.21371566954703067,
+      "grad_norm": 0.9643296599388123,
+      "learning_rate": 9.733505754805848e-06,
+      "loss": 0.8277,
+      "step": 3883
+    },
+    {
+      "epoch": 0.21377070834938633,
+      "grad_norm": 0.8135218620300293,
+      "learning_rate": 9.733366111687494e-06,
+      "loss": 0.7933,
+      "step": 3884
+    },
+    {
+      "epoch": 0.213825747151742,
+      "grad_norm": 0.7527105212211609,
+      "learning_rate": 9.733226432994426e-06,
+      "loss": 0.7302,
+      "step": 3885
+    },
+    {
+      "epoch": 0.21388078595409765,
+      "grad_norm": 1.090550184249878,
+      "learning_rate": 9.733086718727698e-06,
+      "loss": 0.8646,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2139358247564533,
+      "grad_norm": 0.9227491617202759,
+      "learning_rate": 9.732946968888358e-06,
+      "loss": 0.8525,
+      "step": 3887
+    },
+    {
+      "epoch": 0.21399086355880895,
+      "grad_norm": 0.7781830430030823,
+      "learning_rate": 9.732807183477454e-06,
+      "loss": 0.8757,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2140459023611646,
+      "grad_norm": 0.7740090489387512,
+      "learning_rate": 9.732667362496036e-06,
+      "loss": 0.7557,
+      "step": 3889
+    },
+    {
+      "epoch": 0.21410094116352027,
+      "grad_norm": 0.7341694831848145,
+      "learning_rate": 9.732527505945159e-06,
+      "loss": 0.7481,
+      "step": 3890
+    },
+    {
+      "epoch": 0.21415597996587593,
+      "grad_norm": 0.8691402673721313,
+      "learning_rate": 9.732387613825872e-06,
+      "loss": 0.8395,
+      "step": 3891
+    },
+    {
+      "epoch": 0.2142110187682316,
+      "grad_norm": 0.7845497131347656,
+      "learning_rate": 9.732247686139227e-06,
+      "loss": 0.6999,
+      "step": 3892
+    },
+    {
+      "epoch": 0.21426605757058725,
+      "grad_norm": 0.7944281697273254,
+      "learning_rate": 9.732107722886275e-06,
+      "loss": 0.7677,
+      "step": 3893
+    },
+    {
+      "epoch": 0.21432109637294292,
+      "grad_norm": 0.904195249080658,
+      "learning_rate": 9.731967724068065e-06,
+      "loss": 0.8429,
+      "step": 3894
+    },
+    {
+      "epoch": 0.21437613517529858,
+      "grad_norm": 0.7968988418579102,
+      "learning_rate": 9.731827689685655e-06,
+      "loss": 0.8224,
+      "step": 3895
+    },
+    {
+      "epoch": 0.21443117397765424,
+      "grad_norm": 0.773674726486206,
+      "learning_rate": 9.731687619740095e-06,
+      "loss": 0.7743,
+      "step": 3896
+    },
+    {
+      "epoch": 0.2144862127800099,
+      "grad_norm": 0.7873631715774536,
+      "learning_rate": 9.731547514232439e-06,
+      "loss": 0.8581,
+      "step": 3897
+    },
+    {
+      "epoch": 0.21454125158236556,
+      "grad_norm": 0.7989653944969177,
+      "learning_rate": 9.731407373163735e-06,
+      "loss": 0.8447,
+      "step": 3898
+    },
+    {
+      "epoch": 0.21459629038472122,
+      "grad_norm": 0.74820876121521,
+      "learning_rate": 9.73126719653504e-06,
+      "loss": 0.8745,
+      "step": 3899
+    },
+    {
+      "epoch": 0.21465132918707688,
+      "grad_norm": 0.7191246747970581,
+      "learning_rate": 9.731126984347408e-06,
+      "loss": 0.7533,
+      "step": 3900
+    },
+    {
+      "epoch": 0.21470636798943254,
+      "grad_norm": 0.7718465328216553,
+      "learning_rate": 9.730986736601893e-06,
+      "loss": 0.8184,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2147614067917882,
+      "grad_norm": 0.7055066823959351,
+      "learning_rate": 9.730846453299547e-06,
+      "loss": 0.7352,
+      "step": 3902
+    },
+    {
+      "epoch": 0.21481644559414387,
+      "grad_norm": 0.7500855326652527,
+      "learning_rate": 9.730706134441425e-06,
+      "loss": 0.8111,
+      "step": 3903
+    },
+    {
+      "epoch": 0.21487148439649953,
+      "grad_norm": 0.7568232417106628,
+      "learning_rate": 9.730565780028583e-06,
+      "loss": 0.8126,
+      "step": 3904
+    },
+    {
+      "epoch": 0.2149265231988552,
+      "grad_norm": 0.7418738007545471,
+      "learning_rate": 9.730425390062075e-06,
+      "loss": 0.8014,
+      "step": 3905
+    },
+    {
+      "epoch": 0.21498156200121085,
+      "grad_norm": 0.7967441082000732,
+      "learning_rate": 9.730284964542955e-06,
+      "loss": 0.7965,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2150366008035665,
+      "grad_norm": 0.7444791197776794,
+      "learning_rate": 9.730144503472281e-06,
+      "loss": 0.7113,
+      "step": 3907
+    },
+    {
+      "epoch": 0.21509163960592217,
+      "grad_norm": 0.8372869491577148,
+      "learning_rate": 9.730004006851107e-06,
+      "loss": 0.838,
+      "step": 3908
+    },
+    {
+      "epoch": 0.21514667840827784,
+      "grad_norm": 0.7984300851821899,
+      "learning_rate": 9.729863474680488e-06,
+      "loss": 0.856,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2152017172106335,
+      "grad_norm": 0.7508612871170044,
+      "learning_rate": 9.72972290696148e-06,
+      "loss": 0.7947,
+      "step": 3910
+    },
+    {
+      "epoch": 0.21525675601298916,
+      "grad_norm": 0.7559992074966431,
+      "learning_rate": 9.729582303695142e-06,
+      "loss": 0.785,
+      "step": 3911
+    },
+    {
+      "epoch": 0.21531179481534482,
+      "grad_norm": 0.7764164209365845,
+      "learning_rate": 9.729441664882531e-06,
+      "loss": 0.8297,
+      "step": 3912
+    },
+    {
+      "epoch": 0.21536683361770048,
+      "grad_norm": 0.8112726211547852,
+      "learning_rate": 9.7293009905247e-06,
+      "loss": 0.8073,
+      "step": 3913
+    },
+    {
+      "epoch": 0.21542187242005614,
+      "grad_norm": 0.9748952388763428,
+      "learning_rate": 9.729160280622709e-06,
+      "loss": 0.7584,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2154769112224118,
+      "grad_norm": 0.789191484451294,
+      "learning_rate": 9.729019535177617e-06,
+      "loss": 0.7568,
+      "step": 3915
+    },
+    {
+      "epoch": 0.21553195002476747,
+      "grad_norm": 0.7300963401794434,
+      "learning_rate": 9.728878754190478e-06,
+      "loss": 0.8029,
+      "step": 3916
+    },
+    {
+      "epoch": 0.21558698882712313,
+      "grad_norm": 0.9201067686080933,
+      "learning_rate": 9.728737937662354e-06,
+      "loss": 0.8665,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2156420276294788,
+      "grad_norm": 0.8820425271987915,
+      "learning_rate": 9.728597085594301e-06,
+      "loss": 0.8378,
+      "step": 3918
+    },
+    {
+      "epoch": 0.21569706643183445,
+      "grad_norm": 0.7762684226036072,
+      "learning_rate": 9.728456197987376e-06,
+      "loss": 0.8005,
+      "step": 3919
+    },
+    {
+      "epoch": 0.2157521052341901,
+      "grad_norm": 0.8429732918739319,
+      "learning_rate": 9.728315274842641e-06,
+      "loss": 0.8337,
+      "step": 3920
+    },
+    {
+      "epoch": 0.21580714403654577,
+      "grad_norm": 0.7820748090744019,
+      "learning_rate": 9.728174316161156e-06,
+      "loss": 0.8085,
+      "step": 3921
+    },
+    {
+      "epoch": 0.21586218283890143,
+      "grad_norm": 0.8748064637184143,
+      "learning_rate": 9.728033321943977e-06,
+      "loss": 0.7734,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2159172216412571,
+      "grad_norm": 0.8878050446510315,
+      "learning_rate": 9.727892292192166e-06,
+      "loss": 0.9226,
+      "step": 3923
+    },
+    {
+      "epoch": 0.21597226044361276,
+      "grad_norm": 0.8156047463417053,
+      "learning_rate": 9.72775122690678e-06,
+      "loss": 0.8111,
+      "step": 3924
+    },
+    {
+      "epoch": 0.21602729924596842,
+      "grad_norm": 0.7392945885658264,
+      "learning_rate": 9.727610126088883e-06,
+      "loss": 0.758,
+      "step": 3925
+    },
+    {
+      "epoch": 0.21608233804832408,
+      "grad_norm": 0.7573148608207703,
+      "learning_rate": 9.727468989739532e-06,
+      "loss": 0.8142,
+      "step": 3926
+    },
+    {
+      "epoch": 0.21613737685067974,
+      "grad_norm": 0.831847608089447,
+      "learning_rate": 9.727327817859792e-06,
+      "loss": 0.7337,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2161924156530354,
+      "grad_norm": 0.8012371063232422,
+      "learning_rate": 9.72718661045072e-06,
+      "loss": 0.8128,
+      "step": 3928
+    },
+    {
+      "epoch": 0.21624745445539106,
+      "grad_norm": 0.7985890507698059,
+      "learning_rate": 9.72704536751338e-06,
+      "loss": 0.8549,
+      "step": 3929
+    },
+    {
+      "epoch": 0.21630249325774673,
+      "grad_norm": 0.7194695472717285,
+      "learning_rate": 9.726904089048832e-06,
+      "loss": 0.775,
+      "step": 3930
+    },
+    {
+      "epoch": 0.21635753206010236,
+      "grad_norm": 0.8029330968856812,
+      "learning_rate": 9.726762775058138e-06,
+      "loss": 0.9167,
+      "step": 3931
+    },
+    {
+      "epoch": 0.21641257086245802,
+      "grad_norm": 0.7388954162597656,
+      "learning_rate": 9.72662142554236e-06,
+      "loss": 0.7295,
+      "step": 3932
+    },
+    {
+      "epoch": 0.21646760966481368,
+      "grad_norm": 0.798796534538269,
+      "learning_rate": 9.726480040502559e-06,
+      "loss": 0.8686,
+      "step": 3933
+    },
+    {
+      "epoch": 0.21652264846716934,
+      "grad_norm": 0.9977202415466309,
+      "learning_rate": 9.726338619939802e-06,
+      "loss": 0.8387,
+      "step": 3934
+    },
+    {
+      "epoch": 0.216577687269525,
+      "grad_norm": 0.8173295855522156,
+      "learning_rate": 9.726197163855148e-06,
+      "loss": 0.7773,
+      "step": 3935
+    },
+    {
+      "epoch": 0.21663272607188067,
+      "grad_norm": 0.6519538760185242,
+      "learning_rate": 9.72605567224966e-06,
+      "loss": 0.6319,
+      "step": 3936
+    },
+    {
+      "epoch": 0.21668776487423633,
+      "grad_norm": 0.8004894852638245,
+      "learning_rate": 9.725914145124404e-06,
+      "loss": 0.8281,
+      "step": 3937
+    },
+    {
+      "epoch": 0.216742803676592,
+      "grad_norm": 0.7327558398246765,
+      "learning_rate": 9.725772582480442e-06,
+      "loss": 0.7105,
+      "step": 3938
+    },
+    {
+      "epoch": 0.21679784247894765,
+      "grad_norm": 0.7624199986457825,
+      "learning_rate": 9.725630984318839e-06,
+      "loss": 0.7823,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2168528812813033,
+      "grad_norm": 0.7750238180160522,
+      "learning_rate": 9.725489350640658e-06,
+      "loss": 0.8147,
+      "step": 3940
+    },
+    {
+      "epoch": 0.21690792008365897,
+      "grad_norm": 0.6886566877365112,
+      "learning_rate": 9.725347681446964e-06,
+      "loss": 0.7263,
+      "step": 3941
+    },
+    {
+      "epoch": 0.21696295888601463,
+      "grad_norm": 0.882060170173645,
+      "learning_rate": 9.725205976738821e-06,
+      "loss": 0.8931,
+      "step": 3942
+    },
+    {
+      "epoch": 0.2170179976883703,
+      "grad_norm": 0.7946881055831909,
+      "learning_rate": 9.725064236517297e-06,
+      "loss": 0.8036,
+      "step": 3943
+    },
+    {
+      "epoch": 0.21707303649072596,
+      "grad_norm": 0.7062187194824219,
+      "learning_rate": 9.724922460783453e-06,
+      "loss": 0.6915,
+      "step": 3944
+    },
+    {
+      "epoch": 0.21712807529308162,
+      "grad_norm": 0.7978640794754028,
+      "learning_rate": 9.724780649538356e-06,
+      "loss": 0.8873,
+      "step": 3945
+    },
+    {
+      "epoch": 0.21718311409543728,
+      "grad_norm": 0.8828096389770508,
+      "learning_rate": 9.724638802783073e-06,
+      "loss": 0.7114,
+      "step": 3946
+    },
+    {
+      "epoch": 0.21723815289779294,
+      "grad_norm": 0.7301073670387268,
+      "learning_rate": 9.724496920518672e-06,
+      "loss": 0.8107,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2172931917001486,
+      "grad_norm": 0.7944212555885315,
+      "learning_rate": 9.724355002746213e-06,
+      "loss": 0.8135,
+      "step": 3948
+    },
+    {
+      "epoch": 0.21734823050250426,
+      "grad_norm": 0.7988898754119873,
+      "learning_rate": 9.724213049466768e-06,
+      "loss": 0.7173,
+      "step": 3949
+    },
+    {
+      "epoch": 0.21740326930485993,
+      "grad_norm": 0.7734915018081665,
+      "learning_rate": 9.724071060681401e-06,
+      "loss": 0.8131,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2174583081072156,
+      "grad_norm": 0.6856646537780762,
+      "learning_rate": 9.723929036391183e-06,
+      "loss": 0.6873,
+      "step": 3951
+    },
+    {
+      "epoch": 0.21751334690957125,
+      "grad_norm": 0.8652976751327515,
+      "learning_rate": 9.723786976597179e-06,
+      "loss": 0.7908,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2175683857119269,
+      "grad_norm": 0.7325445413589478,
+      "learning_rate": 9.723644881300453e-06,
+      "loss": 0.7389,
+      "step": 3953
+    },
+    {
+      "epoch": 0.21762342451428257,
+      "grad_norm": 0.8596270084381104,
+      "learning_rate": 9.723502750502079e-06,
+      "loss": 0.7785,
+      "step": 3954
+    },
+    {
+      "epoch": 0.21767846331663823,
+      "grad_norm": 0.739248514175415,
+      "learning_rate": 9.723360584203123e-06,
+      "loss": 0.8125,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2177335021189939,
+      "grad_norm": 0.815617561340332,
+      "learning_rate": 9.723218382404652e-06,
+      "loss": 0.8682,
+      "step": 3956
+    },
+    {
+      "epoch": 0.21778854092134955,
+      "grad_norm": 0.758756160736084,
+      "learning_rate": 9.723076145107738e-06,
+      "loss": 0.7717,
+      "step": 3957
+    },
+    {
+      "epoch": 0.21784357972370522,
+      "grad_norm": 0.9007643461227417,
+      "learning_rate": 9.722933872313445e-06,
+      "loss": 0.7901,
+      "step": 3958
+    },
+    {
+      "epoch": 0.21789861852606088,
+      "grad_norm": 0.781548023223877,
+      "learning_rate": 9.722791564022846e-06,
+      "loss": 0.8338,
+      "step": 3959
+    },
+    {
+      "epoch": 0.21795365732841654,
+      "grad_norm": 0.7730190753936768,
+      "learning_rate": 9.722649220237011e-06,
+      "loss": 0.8032,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2180086961307722,
+      "grad_norm": 0.8737791776657104,
+      "learning_rate": 9.722506840957009e-06,
+      "loss": 0.8436,
+      "step": 3961
+    },
+    {
+      "epoch": 0.21806373493312786,
+      "grad_norm": 0.8151329159736633,
+      "learning_rate": 9.722364426183908e-06,
+      "loss": 0.8115,
+      "step": 3962
+    },
+    {
+      "epoch": 0.21811877373548352,
+      "grad_norm": 0.7852860689163208,
+      "learning_rate": 9.722221975918782e-06,
+      "loss": 0.7977,
+      "step": 3963
+    },
+    {
+      "epoch": 0.21817381253783918,
+      "grad_norm": 0.9064140319824219,
+      "learning_rate": 9.722079490162698e-06,
+      "loss": 0.8799,
+      "step": 3964
+    },
+    {
+      "epoch": 0.21822885134019485,
+      "grad_norm": 0.8579906821250916,
+      "learning_rate": 9.72193696891673e-06,
+      "loss": 0.7825,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2182838901425505,
+      "grad_norm": 0.8005900382995605,
+      "learning_rate": 9.721794412181946e-06,
+      "loss": 0.8601,
+      "step": 3966
+    },
+    {
+      "epoch": 0.21833892894490617,
+      "grad_norm": 0.7661529183387756,
+      "learning_rate": 9.721651819959421e-06,
+      "loss": 0.7446,
+      "step": 3967
+    },
+    {
+      "epoch": 0.21839396774726183,
+      "grad_norm": 0.7558436989784241,
+      "learning_rate": 9.721509192250224e-06,
+      "loss": 0.7484,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2184490065496175,
+      "grad_norm": 0.765446126461029,
+      "learning_rate": 9.721366529055427e-06,
+      "loss": 0.7727,
+      "step": 3969
+    },
+    {
+      "epoch": 0.21850404535197315,
+      "grad_norm": 0.7329973578453064,
+      "learning_rate": 9.721223830376103e-06,
+      "loss": 0.797,
+      "step": 3970
+    },
+    {
+      "epoch": 0.21855908415432881,
+      "grad_norm": 0.8881974220275879,
+      "learning_rate": 9.721081096213324e-06,
+      "loss": 0.9199,
+      "step": 3971
+    },
+    {
+      "epoch": 0.21861412295668448,
+      "grad_norm": 0.8246786594390869,
+      "learning_rate": 9.720938326568165e-06,
+      "loss": 0.9108,
+      "step": 3972
+    },
+    {
+      "epoch": 0.21866916175904014,
+      "grad_norm": 0.7187291979789734,
+      "learning_rate": 9.720795521441697e-06,
+      "loss": 0.7756,
+      "step": 3973
+    },
+    {
+      "epoch": 0.21872420056139577,
+      "grad_norm": 0.7880695462226868,
+      "learning_rate": 9.720652680834995e-06,
+      "loss": 0.8548,
+      "step": 3974
+    },
+    {
+      "epoch": 0.21877923936375143,
+      "grad_norm": 0.8841108679771423,
+      "learning_rate": 9.720509804749128e-06,
+      "loss": 0.8477,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2188342781661071,
+      "grad_norm": 0.9061402678489685,
+      "learning_rate": 9.720366893185173e-06,
+      "loss": 0.8235,
+      "step": 3976
+    },
+    {
+      "epoch": 0.21888931696846275,
+      "grad_norm": 0.8342392444610596,
+      "learning_rate": 9.720223946144206e-06,
+      "loss": 0.7777,
+      "step": 3977
+    },
+    {
+      "epoch": 0.21894435577081842,
+      "grad_norm": 0.7933762073516846,
+      "learning_rate": 9.720080963627299e-06,
+      "loss": 0.7943,
+      "step": 3978
+    },
+    {
+      "epoch": 0.21899939457317408,
+      "grad_norm": 0.8358896374702454,
+      "learning_rate": 9.719937945635527e-06,
+      "loss": 0.8932,
+      "step": 3979
+    },
+    {
+      "epoch": 0.21905443337552974,
+      "grad_norm": 0.7479808926582336,
+      "learning_rate": 9.719794892169964e-06,
+      "loss": 0.7446,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2191094721778854,
+      "grad_norm": 0.7920958399772644,
+      "learning_rate": 9.719651803231685e-06,
+      "loss": 0.7489,
+      "step": 3981
+    },
+    {
+      "epoch": 0.21916451098024106,
+      "grad_norm": 0.7098824977874756,
+      "learning_rate": 9.719508678821768e-06,
+      "loss": 0.7763,
+      "step": 3982
+    },
+    {
+      "epoch": 0.21921954978259672,
+      "grad_norm": 0.8733491897583008,
+      "learning_rate": 9.719365518941288e-06,
+      "loss": 0.7325,
+      "step": 3983
+    },
+    {
+      "epoch": 0.21927458858495238,
+      "grad_norm": 0.8328796029090881,
+      "learning_rate": 9.719222323591318e-06,
+      "loss": 0.9097,
+      "step": 3984
+    },
+    {
+      "epoch": 0.21932962738730805,
+      "grad_norm": 0.7869352698326111,
+      "learning_rate": 9.719079092772936e-06,
+      "loss": 0.759,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2193846661896637,
+      "grad_norm": 0.8278539180755615,
+      "learning_rate": 9.718935826487221e-06,
+      "loss": 0.8545,
+      "step": 3986
+    },
+    {
+      "epoch": 0.21943970499201937,
+      "grad_norm": 0.8122449517250061,
+      "learning_rate": 9.718792524735246e-06,
+      "loss": 0.7646,
+      "step": 3987
+    },
+    {
+      "epoch": 0.21949474379437503,
+      "grad_norm": 1.072253942489624,
+      "learning_rate": 9.71864918751809e-06,
+      "loss": 0.915,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2195497825967307,
+      "grad_norm": 0.7770013213157654,
+      "learning_rate": 9.718505814836829e-06,
+      "loss": 0.7561,
+      "step": 3989
+    },
+    {
+      "epoch": 0.21960482139908635,
+      "grad_norm": 0.9011678695678711,
+      "learning_rate": 9.718362406692544e-06,
+      "loss": 0.7532,
+      "step": 3990
+    },
+    {
+      "epoch": 0.21965986020144201,
+      "grad_norm": 0.8867584466934204,
+      "learning_rate": 9.718218963086307e-06,
+      "loss": 0.8732,
+      "step": 3991
+    },
+    {
+      "epoch": 0.21971489900379768,
+      "grad_norm": 0.8884773850440979,
+      "learning_rate": 9.718075484019201e-06,
+      "loss": 0.7403,
+      "step": 3992
+    },
+    {
+      "epoch": 0.21976993780615334,
+      "grad_norm": 0.8995673060417175,
+      "learning_rate": 9.7179319694923e-06,
+      "loss": 0.9283,
+      "step": 3993
+    },
+    {
+      "epoch": 0.219824976608509,
+      "grad_norm": 0.7875818014144897,
+      "learning_rate": 9.717788419506688e-06,
+      "loss": 0.8633,
+      "step": 3994
+    },
+    {
+      "epoch": 0.21988001541086466,
+      "grad_norm": 0.7693219184875488,
+      "learning_rate": 9.71764483406344e-06,
+      "loss": 0.8073,
+      "step": 3995
+    },
+    {
+      "epoch": 0.21993505421322032,
+      "grad_norm": 0.7932817339897156,
+      "learning_rate": 9.717501213163636e-06,
+      "loss": 0.7537,
+      "step": 3996
+    },
+    {
+      "epoch": 0.21999009301557598,
+      "grad_norm": 0.8274912238121033,
+      "learning_rate": 9.717357556808358e-06,
+      "loss": 0.7715,
+      "step": 3997
+    },
+    {
+      "epoch": 0.22004513181793164,
+      "grad_norm": 0.7533993124961853,
+      "learning_rate": 9.71721386499868e-06,
+      "loss": 0.7482,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2201001706202873,
+      "grad_norm": 1.028228759765625,
+      "learning_rate": 9.717070137735687e-06,
+      "loss": 0.9897,
+      "step": 3999
+    },
+    {
+      "epoch": 0.22015520942264297,
+      "grad_norm": 1.1093978881835938,
+      "learning_rate": 9.716926375020457e-06,
+      "loss": 0.8701,
+      "step": 4000
+    },
+    {
+      "epoch": 0.22021024822499863,
+      "grad_norm": 0.7891124486923218,
+      "learning_rate": 9.716782576854073e-06,
+      "loss": 0.8533,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2202652870273543,
+      "grad_norm": 1.1783788204193115,
+      "learning_rate": 9.716638743237611e-06,
+      "loss": 0.8088,
+      "step": 4002
+    },
+    {
+      "epoch": 0.22032032582970995,
+      "grad_norm": 0.8713383078575134,
+      "learning_rate": 9.716494874172157e-06,
+      "loss": 0.8382,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2203753646320656,
+      "grad_norm": 0.7821565270423889,
+      "learning_rate": 9.716350969658787e-06,
+      "loss": 0.8168,
+      "step": 4004
+    },
+    {
+      "epoch": 0.22043040343442127,
+      "grad_norm": 0.7642589211463928,
+      "learning_rate": 9.716207029698589e-06,
+      "loss": 0.7209,
+      "step": 4005
+    },
+    {
+      "epoch": 0.22048544223677694,
+      "grad_norm": 0.935625433921814,
+      "learning_rate": 9.716063054292639e-06,
+      "loss": 0.8436,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2205404810391326,
+      "grad_norm": 0.7064627408981323,
+      "learning_rate": 9.715919043442024e-06,
+      "loss": 0.7651,
+      "step": 4007
+    },
+    {
+      "epoch": 0.22059551984148826,
+      "grad_norm": 0.6980876326560974,
+      "learning_rate": 9.715774997147823e-06,
+      "loss": 0.7842,
+      "step": 4008
+    },
+    {
+      "epoch": 0.22065055864384392,
+      "grad_norm": 0.7691119313240051,
+      "learning_rate": 9.715630915411118e-06,
+      "loss": 0.7345,
+      "step": 4009
+    },
+    {
+      "epoch": 0.22070559744619958,
+      "grad_norm": 0.8870186805725098,
+      "learning_rate": 9.715486798232994e-06,
+      "loss": 0.7531,
+      "step": 4010
+    },
+    {
+      "epoch": 0.22076063624855524,
+      "grad_norm": 0.7225383520126343,
+      "learning_rate": 9.715342645614533e-06,
+      "loss": 0.8543,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2208156750509109,
+      "grad_norm": 0.7517428994178772,
+      "learning_rate": 9.71519845755682e-06,
+      "loss": 0.84,
+      "step": 4012
+    },
+    {
+      "epoch": 0.22087071385326656,
+      "grad_norm": 0.8115549087524414,
+      "learning_rate": 9.715054234060937e-06,
+      "loss": 0.7823,
+      "step": 4013
+    },
+    {
+      "epoch": 0.22092575265562223,
+      "grad_norm": 1.6656148433685303,
+      "learning_rate": 9.714909975127968e-06,
+      "loss": 0.8951,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2209807914579779,
+      "grad_norm": 0.906508207321167,
+      "learning_rate": 9.714765680758997e-06,
+      "loss": 0.8599,
+      "step": 4015
+    },
+    {
+      "epoch": 0.22103583026033355,
+      "grad_norm": 0.8274093866348267,
+      "learning_rate": 9.71462135095511e-06,
+      "loss": 0.9568,
+      "step": 4016
+    },
+    {
+      "epoch": 0.22109086906268918,
+      "grad_norm": 0.7745386958122253,
+      "learning_rate": 9.714476985717393e-06,
+      "loss": 0.8641,
+      "step": 4017
+    },
+    {
+      "epoch": 0.22114590786504484,
+      "grad_norm": 0.8112689256668091,
+      "learning_rate": 9.714332585046928e-06,
+      "loss": 0.834,
+      "step": 4018
+    },
+    {
+      "epoch": 0.2212009466674005,
+      "grad_norm": 0.916847825050354,
+      "learning_rate": 9.714188148944799e-06,
+      "loss": 0.8546,
+      "step": 4019
+    },
+    {
+      "epoch": 0.22125598546975617,
+      "grad_norm": 0.8595414161682129,
+      "learning_rate": 9.714043677412096e-06,
+      "loss": 0.9388,
+      "step": 4020
+    },
+    {
+      "epoch": 0.22131102427211183,
+      "grad_norm": 0.8672438263893127,
+      "learning_rate": 9.713899170449901e-06,
+      "loss": 0.8151,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2213660630744675,
+      "grad_norm": 0.699749767780304,
+      "learning_rate": 9.713754628059304e-06,
+      "loss": 0.7433,
+      "step": 4022
+    },
+    {
+      "epoch": 0.22142110187682315,
+      "grad_norm": 0.8071898818016052,
+      "learning_rate": 9.713610050241387e-06,
+      "loss": 0.7663,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2214761406791788,
+      "grad_norm": 0.745030403137207,
+      "learning_rate": 9.713465436997239e-06,
+      "loss": 0.7733,
+      "step": 4024
+    },
+    {
+      "epoch": 0.22153117948153447,
+      "grad_norm": 0.8034930229187012,
+      "learning_rate": 9.713320788327947e-06,
+      "loss": 0.9015,
+      "step": 4025
+    },
+    {
+      "epoch": 0.22158621828389014,
+      "grad_norm": 0.8549708724021912,
+      "learning_rate": 9.713176104234597e-06,
+      "loss": 0.7127,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2216412570862458,
+      "grad_norm": 0.8432256579399109,
+      "learning_rate": 9.713031384718277e-06,
+      "loss": 0.8163,
+      "step": 4027
+    },
+    {
+      "epoch": 0.22169629588860146,
+      "grad_norm": 0.7623703479766846,
+      "learning_rate": 9.712886629780075e-06,
+      "loss": 0.8272,
+      "step": 4028
+    },
+    {
+      "epoch": 0.22175133469095712,
+      "grad_norm": 0.8425806760787964,
+      "learning_rate": 9.712741839421079e-06,
+      "loss": 0.7907,
+      "step": 4029
+    },
+    {
+      "epoch": 0.22180637349331278,
+      "grad_norm": 0.7477750778198242,
+      "learning_rate": 9.712597013642376e-06,
+      "loss": 0.7662,
+      "step": 4030
+    },
+    {
+      "epoch": 0.22186141229566844,
+      "grad_norm": 0.7761805057525635,
+      "learning_rate": 9.712452152445056e-06,
+      "loss": 0.7999,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2219164510980241,
+      "grad_norm": 0.8604531288146973,
+      "learning_rate": 9.712307255830207e-06,
+      "loss": 0.812,
+      "step": 4032
+    },
+    {
+      "epoch": 0.22197148990037976,
+      "grad_norm": 0.8113332986831665,
+      "learning_rate": 9.712162323798918e-06,
+      "loss": 0.8092,
+      "step": 4033
+    },
+    {
+      "epoch": 0.22202652870273543,
+      "grad_norm": 0.7980128526687622,
+      "learning_rate": 9.71201735635228e-06,
+      "loss": 0.6934,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2220815675050911,
+      "grad_norm": 0.7819470763206482,
+      "learning_rate": 9.711872353491377e-06,
+      "loss": 0.8531,
+      "step": 4035
+    },
+    {
+      "epoch": 0.22213660630744675,
+      "grad_norm": 0.8283445835113525,
+      "learning_rate": 9.711727315217305e-06,
+      "loss": 0.8594,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2221916451098024,
+      "grad_norm": 0.7282612919807434,
+      "learning_rate": 9.711582241531153e-06,
+      "loss": 0.7374,
+      "step": 4037
+    },
+    {
+      "epoch": 0.22224668391215807,
+      "grad_norm": 0.9564353823661804,
+      "learning_rate": 9.711437132434007e-06,
+      "loss": 0.7996,
+      "step": 4038
+    },
+    {
+      "epoch": 0.22230172271451373,
+      "grad_norm": 0.8559701442718506,
+      "learning_rate": 9.711291987926963e-06,
+      "loss": 0.949,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2223567615168694,
+      "grad_norm": 0.7515334486961365,
+      "learning_rate": 9.71114680801111e-06,
+      "loss": 0.7188,
+      "step": 4040
+    },
+    {
+      "epoch": 0.22241180031922506,
+      "grad_norm": 0.7685608863830566,
+      "learning_rate": 9.711001592687537e-06,
+      "loss": 0.7679,
+      "step": 4041
+    },
+    {
+      "epoch": 0.22246683912158072,
+      "grad_norm": 0.6848913431167603,
+      "learning_rate": 9.710856341957337e-06,
+      "loss": 0.7666,
+      "step": 4042
+    },
+    {
+      "epoch": 0.22252187792393638,
+      "grad_norm": 0.7270542979240417,
+      "learning_rate": 9.710711055821602e-06,
+      "loss": 0.7563,
+      "step": 4043
+    },
+    {
+      "epoch": 0.22257691672629204,
+      "grad_norm": 0.7965164184570312,
+      "learning_rate": 9.710565734281424e-06,
+      "loss": 0.7586,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2226319555286477,
+      "grad_norm": 0.7872949242591858,
+      "learning_rate": 9.710420377337895e-06,
+      "loss": 0.8423,
+      "step": 4045
+    },
+    {
+      "epoch": 0.22268699433100336,
+      "grad_norm": 0.7466526627540588,
+      "learning_rate": 9.710274984992107e-06,
+      "loss": 0.7578,
+      "step": 4046
+    },
+    {
+      "epoch": 0.22274203313335902,
+      "grad_norm": 0.7208731770515442,
+      "learning_rate": 9.710129557245154e-06,
+      "loss": 0.7019,
+      "step": 4047
+    },
+    {
+      "epoch": 0.22279707193571469,
+      "grad_norm": 0.6953400373458862,
+      "learning_rate": 9.709984094098127e-06,
+      "loss": 0.7234,
+      "step": 4048
+    },
+    {
+      "epoch": 0.22285211073807035,
+      "grad_norm": 0.7866283059120178,
+      "learning_rate": 9.709838595552122e-06,
+      "loss": 0.785,
+      "step": 4049
+    },
+    {
+      "epoch": 0.222907149540426,
+      "grad_norm": 0.7404114007949829,
+      "learning_rate": 9.709693061608227e-06,
+      "loss": 0.7706,
+      "step": 4050
+    },
+    {
+      "epoch": 0.22296218834278167,
+      "grad_norm": 0.8788254857063293,
+      "learning_rate": 9.709547492267544e-06,
+      "loss": 0.8392,
+      "step": 4051
+    },
+    {
+      "epoch": 0.22301722714513733,
+      "grad_norm": 0.7493161559104919,
+      "learning_rate": 9.70940188753116e-06,
+      "loss": 0.8346,
+      "step": 4052
+    },
+    {
+      "epoch": 0.223072265947493,
+      "grad_norm": 0.7340379357337952,
+      "learning_rate": 9.709256247400174e-06,
+      "loss": 0.7715,
+      "step": 4053
+    },
+    {
+      "epoch": 0.22312730474984865,
+      "grad_norm": 0.7291178107261658,
+      "learning_rate": 9.709110571875677e-06,
+      "loss": 0.866,
+      "step": 4054
+    },
+    {
+      "epoch": 0.22318234355220432,
+      "grad_norm": 0.8046013712882996,
+      "learning_rate": 9.708964860958765e-06,
+      "loss": 0.7885,
+      "step": 4055
+    },
+    {
+      "epoch": 0.22323738235455998,
+      "grad_norm": 0.832941472530365,
+      "learning_rate": 9.708819114650535e-06,
+      "loss": 0.873,
+      "step": 4056
+    },
+    {
+      "epoch": 0.22329242115691564,
+      "grad_norm": 0.6933377981185913,
+      "learning_rate": 9.70867333295208e-06,
+      "loss": 0.7944,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2233474599592713,
+      "grad_norm": 0.7976044416427612,
+      "learning_rate": 9.708527515864499e-06,
+      "loss": 0.72,
+      "step": 4058
+    },
+    {
+      "epoch": 0.22340249876162696,
+      "grad_norm": 0.7698904871940613,
+      "learning_rate": 9.708381663388884e-06,
+      "loss": 0.7603,
+      "step": 4059
+    },
+    {
+      "epoch": 0.2234575375639826,
+      "grad_norm": 0.7554401159286499,
+      "learning_rate": 9.708235775526331e-06,
+      "loss": 0.7488,
+      "step": 4060
+    },
+    {
+      "epoch": 0.22351257636633826,
+      "grad_norm": 0.7382954359054565,
+      "learning_rate": 9.70808985227794e-06,
+      "loss": 0.7418,
+      "step": 4061
+    },
+    {
+      "epoch": 0.22356761516869392,
+      "grad_norm": 0.7220499515533447,
+      "learning_rate": 9.707943893644806e-06,
+      "loss": 0.7691,
+      "step": 4062
+    },
+    {
+      "epoch": 0.22362265397104958,
+      "grad_norm": 0.727542519569397,
+      "learning_rate": 9.707797899628027e-06,
+      "loss": 0.7603,
+      "step": 4063
+    },
+    {
+      "epoch": 0.22367769277340524,
+      "grad_norm": 0.7857500910758972,
+      "learning_rate": 9.707651870228697e-06,
+      "loss": 0.8633,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2237327315757609,
+      "grad_norm": 0.7975600361824036,
+      "learning_rate": 9.707505805447917e-06,
+      "loss": 0.8591,
+      "step": 4065
+    },
+    {
+      "epoch": 0.22378777037811656,
+      "grad_norm": 1.0063475370407104,
+      "learning_rate": 9.707359705286784e-06,
+      "loss": 0.7935,
+      "step": 4066
+    },
+    {
+      "epoch": 0.22384280918047222,
+      "grad_norm": 0.7307062745094299,
+      "learning_rate": 9.707213569746393e-06,
+      "loss": 0.797,
+      "step": 4067
+    },
+    {
+      "epoch": 0.22389784798282789,
+      "grad_norm": 0.7891914248466492,
+      "learning_rate": 9.707067398827847e-06,
+      "loss": 0.853,
+      "step": 4068
+    },
+    {
+      "epoch": 0.22395288678518355,
+      "grad_norm": 0.7479422092437744,
+      "learning_rate": 9.706921192532242e-06,
+      "loss": 0.7359,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2240079255875392,
+      "grad_norm": 0.8436065912246704,
+      "learning_rate": 9.706774950860676e-06,
+      "loss": 0.7916,
+      "step": 4070
+    },
+    {
+      "epoch": 0.22406296438989487,
+      "grad_norm": 0.7586960196495056,
+      "learning_rate": 9.706628673814252e-06,
+      "loss": 0.7871,
+      "step": 4071
+    },
+    {
+      "epoch": 0.22411800319225053,
+      "grad_norm": 0.8181111812591553,
+      "learning_rate": 9.706482361394064e-06,
+      "loss": 0.7782,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2241730419946062,
+      "grad_norm": 0.7205253839492798,
+      "learning_rate": 9.706336013601217e-06,
+      "loss": 0.7912,
+      "step": 4073
+    },
+    {
+      "epoch": 0.22422808079696185,
+      "grad_norm": 0.9823397397994995,
+      "learning_rate": 9.706189630436806e-06,
+      "loss": 0.8393,
+      "step": 4074
+    },
+    {
+      "epoch": 0.22428311959931752,
+      "grad_norm": 0.7360854148864746,
+      "learning_rate": 9.706043211901935e-06,
+      "loss": 0.8239,
+      "step": 4075
+    },
+    {
+      "epoch": 0.22433815840167318,
+      "grad_norm": 0.7590144872665405,
+      "learning_rate": 9.705896757997701e-06,
+      "loss": 0.7177,
+      "step": 4076
+    },
+    {
+      "epoch": 0.22439319720402884,
+      "grad_norm": 0.7691343426704407,
+      "learning_rate": 9.70575026872521e-06,
+      "loss": 0.7731,
+      "step": 4077
+    },
+    {
+      "epoch": 0.2244482360063845,
+      "grad_norm": 0.7057286500930786,
+      "learning_rate": 9.705603744085556e-06,
+      "loss": 0.7746,
+      "step": 4078
+    },
+    {
+      "epoch": 0.22450327480874016,
+      "grad_norm": 0.7954769134521484,
+      "learning_rate": 9.705457184079847e-06,
+      "loss": 0.8215,
+      "step": 4079
+    },
+    {
+      "epoch": 0.22455831361109582,
+      "grad_norm": 0.7089072465896606,
+      "learning_rate": 9.70531058870918e-06,
+      "loss": 0.7263,
+      "step": 4080
+    },
+    {
+      "epoch": 0.22461335241345148,
+      "grad_norm": 0.9847552180290222,
+      "learning_rate": 9.705163957974657e-06,
+      "loss": 0.8948,
+      "step": 4081
+    },
+    {
+      "epoch": 0.22466839121580715,
+      "grad_norm": 0.7977012395858765,
+      "learning_rate": 9.705017291877383e-06,
+      "loss": 0.7518,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2247234300181628,
+      "grad_norm": 0.8084518909454346,
+      "learning_rate": 9.704870590418458e-06,
+      "loss": 0.8711,
+      "step": 4083
+    },
+    {
+      "epoch": 0.22477846882051847,
+      "grad_norm": 0.9151536822319031,
+      "learning_rate": 9.704723853598986e-06,
+      "loss": 0.8217,
+      "step": 4084
+    },
+    {
+      "epoch": 0.22483350762287413,
+      "grad_norm": 0.908136248588562,
+      "learning_rate": 9.704577081420065e-06,
+      "loss": 0.6961,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2248885464252298,
+      "grad_norm": 0.8569996953010559,
+      "learning_rate": 9.704430273882806e-06,
+      "loss": 0.8405,
+      "step": 4086
+    },
+    {
+      "epoch": 0.22494358522758545,
+      "grad_norm": 0.7687774300575256,
+      "learning_rate": 9.704283430988307e-06,
+      "loss": 0.6903,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2249986240299411,
+      "grad_norm": 0.863203763961792,
+      "learning_rate": 9.704136552737673e-06,
+      "loss": 0.8927,
+      "step": 4088
+    },
+    {
+      "epoch": 0.22505366283229677,
+      "grad_norm": 1.252581238746643,
+      "learning_rate": 9.703989639132008e-06,
+      "loss": 0.8792,
+      "step": 4089
+    },
+    {
+      "epoch": 0.22510870163465244,
+      "grad_norm": 0.7844160795211792,
+      "learning_rate": 9.703842690172415e-06,
+      "loss": 0.844,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2251637404370081,
+      "grad_norm": 0.8669766187667847,
+      "learning_rate": 9.703695705860002e-06,
+      "loss": 0.7008,
+      "step": 4091
+    },
+    {
+      "epoch": 0.22521877923936376,
+      "grad_norm": 0.7180137634277344,
+      "learning_rate": 9.703548686195869e-06,
+      "loss": 0.8242,
+      "step": 4092
+    },
+    {
+      "epoch": 0.22527381804171942,
+      "grad_norm": 0.7225000858306885,
+      "learning_rate": 9.703401631181124e-06,
+      "loss": 0.724,
+      "step": 4093
+    },
+    {
+      "epoch": 0.22532885684407508,
+      "grad_norm": 0.8348065614700317,
+      "learning_rate": 9.70325454081687e-06,
+      "loss": 0.7996,
+      "step": 4094
+    },
+    {
+      "epoch": 0.22538389564643074,
+      "grad_norm": 0.8099488019943237,
+      "learning_rate": 9.703107415104216e-06,
+      "loss": 0.7498,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2254389344487864,
+      "grad_norm": 0.7051188945770264,
+      "learning_rate": 9.702960254044264e-06,
+      "loss": 0.7322,
+      "step": 4096
+    },
+    {
+      "epoch": 0.22549397325114207,
+      "grad_norm": 0.742859423160553,
+      "learning_rate": 9.702813057638122e-06,
+      "loss": 0.746,
+      "step": 4097
+    },
+    {
+      "epoch": 0.22554901205349773,
+      "grad_norm": 0.7981536984443665,
+      "learning_rate": 9.702665825886897e-06,
+      "loss": 0.8705,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2256040508558534,
+      "grad_norm": 1.0317178964614868,
+      "learning_rate": 9.702518558791693e-06,
+      "loss": 0.8261,
+      "step": 4099
+    },
+    {
+      "epoch": 0.22565908965820905,
+      "grad_norm": 0.7811983823776245,
+      "learning_rate": 9.702371256353618e-06,
+      "loss": 0.7633,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2257141284605647,
+      "grad_norm": 0.8288078308105469,
+      "learning_rate": 9.702223918573782e-06,
+      "loss": 0.7974,
+      "step": 4101
+    },
+    {
+      "epoch": 0.22576916726292034,
+      "grad_norm": 0.8932577967643738,
+      "learning_rate": 9.702076545453286e-06,
+      "loss": 0.7517,
+      "step": 4102
+    },
+    {
+      "epoch": 0.225824206065276,
+      "grad_norm": 0.8342248201370239,
+      "learning_rate": 9.701929136993243e-06,
+      "loss": 0.8634,
+      "step": 4103
+    },
+    {
+      "epoch": 0.22587924486763167,
+      "grad_norm": 0.790392279624939,
+      "learning_rate": 9.701781693194761e-06,
+      "loss": 0.7705,
+      "step": 4104
+    },
+    {
+      "epoch": 0.22593428366998733,
+      "grad_norm": 0.824691891670227,
+      "learning_rate": 9.701634214058944e-06,
+      "loss": 0.877,
+      "step": 4105
+    },
+    {
+      "epoch": 0.225989322472343,
+      "grad_norm": 0.9237051010131836,
+      "learning_rate": 9.701486699586904e-06,
+      "loss": 0.842,
+      "step": 4106
+    },
+    {
+      "epoch": 0.22604436127469865,
+      "grad_norm": 0.7453535199165344,
+      "learning_rate": 9.701339149779747e-06,
+      "loss": 0.8217,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2260994000770543,
+      "grad_norm": 0.727872371673584,
+      "learning_rate": 9.701191564638586e-06,
+      "loss": 0.849,
+      "step": 4108
+    },
+    {
+      "epoch": 0.22615443887940997,
+      "grad_norm": 0.966585636138916,
+      "learning_rate": 9.701043944164526e-06,
+      "loss": 0.7742,
+      "step": 4109
+    },
+    {
+      "epoch": 0.22620947768176564,
+      "grad_norm": 0.7556117177009583,
+      "learning_rate": 9.700896288358678e-06,
+      "loss": 0.7498,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2262645164841213,
+      "grad_norm": 0.848143458366394,
+      "learning_rate": 9.700748597222151e-06,
+      "loss": 0.7237,
+      "step": 4111
+    },
+    {
+      "epoch": 0.22631955528647696,
+      "grad_norm": 0.9046787619590759,
+      "learning_rate": 9.700600870756056e-06,
+      "loss": 0.8066,
+      "step": 4112
+    },
+    {
+      "epoch": 0.22637459408883262,
+      "grad_norm": 0.923159658908844,
+      "learning_rate": 9.700453108961505e-06,
+      "loss": 0.8404,
+      "step": 4113
+    },
+    {
+      "epoch": 0.22642963289118828,
+      "grad_norm": 0.8697664737701416,
+      "learning_rate": 9.700305311839606e-06,
+      "loss": 0.7269,
+      "step": 4114
+    },
+    {
+      "epoch": 0.22648467169354394,
+      "grad_norm": 0.8179994821548462,
+      "learning_rate": 9.70015747939147e-06,
+      "loss": 0.8083,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2265397104958996,
+      "grad_norm": 0.7961694002151489,
+      "learning_rate": 9.700009611618208e-06,
+      "loss": 0.7327,
+      "step": 4116
+    },
+    {
+      "epoch": 0.22659474929825527,
+      "grad_norm": 0.7317802309989929,
+      "learning_rate": 9.699861708520934e-06,
+      "loss": 0.8273,
+      "step": 4117
+    },
+    {
+      "epoch": 0.22664978810061093,
+      "grad_norm": 0.9190557599067688,
+      "learning_rate": 9.699713770100757e-06,
+      "loss": 0.8027,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2267048269029666,
+      "grad_norm": 0.7618072628974915,
+      "learning_rate": 9.699565796358788e-06,
+      "loss": 0.7669,
+      "step": 4119
+    },
+    {
+      "epoch": 0.22675986570532225,
+      "grad_norm": 1.0236154794692993,
+      "learning_rate": 9.699417787296139e-06,
+      "loss": 0.7511,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2268149045076779,
+      "grad_norm": 0.8011670708656311,
+      "learning_rate": 9.699269742913927e-06,
+      "loss": 0.7644,
+      "step": 4121
+    },
+    {
+      "epoch": 0.22686994331003357,
+      "grad_norm": 0.7808024287223816,
+      "learning_rate": 9.69912166321326e-06,
+      "loss": 0.7894,
+      "step": 4122
+    },
+    {
+      "epoch": 0.22692498211238923,
+      "grad_norm": 0.8645655512809753,
+      "learning_rate": 9.698973548195252e-06,
+      "loss": 0.7989,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2269800209147449,
+      "grad_norm": 0.7478770613670349,
+      "learning_rate": 9.698825397861017e-06,
+      "loss": 0.7758,
+      "step": 4124
+    },
+    {
+      "epoch": 0.22703505971710056,
+      "grad_norm": 0.8988361954689026,
+      "learning_rate": 9.698677212211668e-06,
+      "loss": 0.8312,
+      "step": 4125
+    },
+    {
+      "epoch": 0.22709009851945622,
+      "grad_norm": 0.773028552532196,
+      "learning_rate": 9.69852899124832e-06,
+      "loss": 0.7415,
+      "step": 4126
+    },
+    {
+      "epoch": 0.22714513732181188,
+      "grad_norm": 0.8173778653144836,
+      "learning_rate": 9.698380734972085e-06,
+      "loss": 0.8241,
+      "step": 4127
+    },
+    {
+      "epoch": 0.22720017612416754,
+      "grad_norm": 0.7868672013282776,
+      "learning_rate": 9.698232443384078e-06,
+      "loss": 0.7294,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2272552149265232,
+      "grad_norm": 0.8662189841270447,
+      "learning_rate": 9.698084116485413e-06,
+      "loss": 0.9307,
+      "step": 4129
+    },
+    {
+      "epoch": 0.22731025372887886,
+      "grad_norm": 0.7571321129798889,
+      "learning_rate": 9.697935754277207e-06,
+      "loss": 0.7756,
+      "step": 4130
+    },
+    {
+      "epoch": 0.22736529253123453,
+      "grad_norm": 0.8222649097442627,
+      "learning_rate": 9.697787356760574e-06,
+      "loss": 0.8689,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2274203313335902,
+      "grad_norm": 0.8302241563796997,
+      "learning_rate": 9.697638923936626e-06,
+      "loss": 0.8139,
+      "step": 4132
+    },
+    {
+      "epoch": 0.22747537013594585,
+      "grad_norm": 0.779951274394989,
+      "learning_rate": 9.697490455806482e-06,
+      "loss": 0.7493,
+      "step": 4133
+    },
+    {
+      "epoch": 0.2275304089383015,
+      "grad_norm": 0.8409813046455383,
+      "learning_rate": 9.697341952371257e-06,
+      "loss": 0.777,
+      "step": 4134
+    },
+    {
+      "epoch": 0.22758544774065717,
+      "grad_norm": 0.8599729537963867,
+      "learning_rate": 9.697193413632068e-06,
+      "loss": 0.7678,
+      "step": 4135
+    },
+    {
+      "epoch": 0.22764048654301283,
+      "grad_norm": 0.7505115270614624,
+      "learning_rate": 9.69704483959003e-06,
+      "loss": 0.787,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2276955253453685,
+      "grad_norm": 0.7326868176460266,
+      "learning_rate": 9.696896230246262e-06,
+      "loss": 0.7066,
+      "step": 4137
+    },
+    {
+      "epoch": 0.22775056414772415,
+      "grad_norm": 0.8269753456115723,
+      "learning_rate": 9.696747585601878e-06,
+      "loss": 0.7379,
+      "step": 4138
+    },
+    {
+      "epoch": 0.22780560295007982,
+      "grad_norm": 0.7841970324516296,
+      "learning_rate": 9.696598905657997e-06,
+      "loss": 0.764,
+      "step": 4139
+    },
+    {
+      "epoch": 0.22786064175243548,
+      "grad_norm": 0.7131417989730835,
+      "learning_rate": 9.696450190415735e-06,
+      "loss": 0.7629,
+      "step": 4140
+    },
+    {
+      "epoch": 0.22791568055479114,
+      "grad_norm": 0.7922703623771667,
+      "learning_rate": 9.69630143987621e-06,
+      "loss": 0.8354,
+      "step": 4141
+    },
+    {
+      "epoch": 0.2279707193571468,
+      "grad_norm": 0.9628629684448242,
+      "learning_rate": 9.696152654040543e-06,
+      "loss": 0.8077,
+      "step": 4142
+    },
+    {
+      "epoch": 0.22802575815950246,
+      "grad_norm": 0.8566663265228271,
+      "learning_rate": 9.696003832909847e-06,
+      "loss": 0.685,
+      "step": 4143
+    },
+    {
+      "epoch": 0.22808079696185812,
+      "grad_norm": 0.7181339859962463,
+      "learning_rate": 9.695854976485244e-06,
+      "loss": 0.8135,
+      "step": 4144
+    },
+    {
+      "epoch": 0.22813583576421376,
+      "grad_norm": 0.9119813442230225,
+      "learning_rate": 9.695706084767853e-06,
+      "loss": 0.7276,
+      "step": 4145
+    },
+    {
+      "epoch": 0.22819087456656942,
+      "grad_norm": 0.8547400832176208,
+      "learning_rate": 9.69555715775879e-06,
+      "loss": 0.8656,
+      "step": 4146
+    },
+    {
+      "epoch": 0.22824591336892508,
+      "grad_norm": 0.77585768699646,
+      "learning_rate": 9.695408195459179e-06,
+      "loss": 0.8218,
+      "step": 4147
+    },
+    {
+      "epoch": 0.22830095217128074,
+      "grad_norm": 0.7832447290420532,
+      "learning_rate": 9.695259197870135e-06,
+      "loss": 0.8002,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2283559909736364,
+      "grad_norm": 0.9184865355491638,
+      "learning_rate": 9.69511016499278e-06,
+      "loss": 0.8651,
+      "step": 4149
+    },
+    {
+      "epoch": 0.22841102977599206,
+      "grad_norm": 0.8663797974586487,
+      "learning_rate": 9.694961096828235e-06,
+      "loss": 0.7381,
+      "step": 4150
+    },
+    {
+      "epoch": 0.22846606857834773,
+      "grad_norm": 0.843265950679779,
+      "learning_rate": 9.694811993377617e-06,
+      "loss": 0.8546,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2285211073807034,
+      "grad_norm": 0.8021818399429321,
+      "learning_rate": 9.694662854642049e-06,
+      "loss": 0.9166,
+      "step": 4152
+    },
+    {
+      "epoch": 0.22857614618305905,
+      "grad_norm": 0.7762879729270935,
+      "learning_rate": 9.694513680622653e-06,
+      "loss": 0.7055,
+      "step": 4153
+    },
+    {
+      "epoch": 0.2286311849854147,
+      "grad_norm": 0.809352457523346,
+      "learning_rate": 9.694364471320548e-06,
+      "loss": 0.7988,
+      "step": 4154
+    },
+    {
+      "epoch": 0.22868622378777037,
+      "grad_norm": 0.7239902019500732,
+      "learning_rate": 9.694215226736858e-06,
+      "loss": 0.7783,
+      "step": 4155
+    },
+    {
+      "epoch": 0.22874126259012603,
+      "grad_norm": 0.7072625160217285,
+      "learning_rate": 9.694065946872702e-06,
+      "loss": 0.7607,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2287963013924817,
+      "grad_norm": 0.7696169018745422,
+      "learning_rate": 9.693916631729201e-06,
+      "loss": 0.7519,
+      "step": 4157
+    },
+    {
+      "epoch": 0.22885134019483735,
+      "grad_norm": 0.9198557734489441,
+      "learning_rate": 9.69376728130748e-06,
+      "loss": 0.7754,
+      "step": 4158
+    },
+    {
+      "epoch": 0.22890637899719302,
+      "grad_norm": 0.7589097619056702,
+      "learning_rate": 9.693617895608662e-06,
+      "loss": 0.7258,
+      "step": 4159
+    },
+    {
+      "epoch": 0.22896141779954868,
+      "grad_norm": 0.8351333141326904,
+      "learning_rate": 9.693468474633867e-06,
+      "loss": 0.8633,
+      "step": 4160
+    },
+    {
+      "epoch": 0.22901645660190434,
+      "grad_norm": 0.8331828713417053,
+      "learning_rate": 9.69331901838422e-06,
+      "loss": 0.7361,
+      "step": 4161
+    },
+    {
+      "epoch": 0.22907149540426,
+      "grad_norm": 0.8810774087905884,
+      "learning_rate": 9.693169526860843e-06,
+      "loss": 0.7651,
+      "step": 4162
+    },
+    {
+      "epoch": 0.22912653420661566,
+      "grad_norm": 0.8151684999465942,
+      "learning_rate": 9.69302000006486e-06,
+      "loss": 0.8533,
+      "step": 4163
+    },
+    {
+      "epoch": 0.22918157300897132,
+      "grad_norm": 0.8683320879936218,
+      "learning_rate": 9.692870437997394e-06,
+      "loss": 0.8323,
+      "step": 4164
+    },
+    {
+      "epoch": 0.22923661181132698,
+      "grad_norm": 0.7488875389099121,
+      "learning_rate": 9.692720840659572e-06,
+      "loss": 0.8414,
+      "step": 4165
+    },
+    {
+      "epoch": 0.22929165061368265,
+      "grad_norm": 0.7916452288627625,
+      "learning_rate": 9.692571208052515e-06,
+      "loss": 0.7058,
+      "step": 4166
+    },
+    {
+      "epoch": 0.2293466894160383,
+      "grad_norm": 0.8228384256362915,
+      "learning_rate": 9.69242154017735e-06,
+      "loss": 0.7667,
+      "step": 4167
+    },
+    {
+      "epoch": 0.22940172821839397,
+      "grad_norm": 0.7395613789558411,
+      "learning_rate": 9.692271837035202e-06,
+      "loss": 0.7649,
+      "step": 4168
+    },
+    {
+      "epoch": 0.22945676702074963,
+      "grad_norm": 0.7187666893005371,
+      "learning_rate": 9.692122098627192e-06,
+      "loss": 0.7575,
+      "step": 4169
+    },
+    {
+      "epoch": 0.2295118058231053,
+      "grad_norm": 0.7060030102729797,
+      "learning_rate": 9.691972324954449e-06,
+      "loss": 0.8309,
+      "step": 4170
+    },
+    {
+      "epoch": 0.22956684462546095,
+      "grad_norm": 0.7807210087776184,
+      "learning_rate": 9.691822516018099e-06,
+      "loss": 0.8185,
+      "step": 4171
+    },
+    {
+      "epoch": 0.22962188342781661,
+      "grad_norm": 0.6918593645095825,
+      "learning_rate": 9.691672671819265e-06,
+      "loss": 0.6983,
+      "step": 4172
+    },
+    {
+      "epoch": 0.22967692223017228,
+      "grad_norm": 0.7947858572006226,
+      "learning_rate": 9.691522792359077e-06,
+      "loss": 0.8098,
+      "step": 4173
+    },
+    {
+      "epoch": 0.22973196103252794,
+      "grad_norm": 0.7907306551933289,
+      "learning_rate": 9.691372877638658e-06,
+      "loss": 0.8,
+      "step": 4174
+    },
+    {
+      "epoch": 0.2297869998348836,
+      "grad_norm": 0.7669435739517212,
+      "learning_rate": 9.691222927659137e-06,
+      "loss": 0.8121,
+      "step": 4175
+    },
+    {
+      "epoch": 0.22984203863723926,
+      "grad_norm": 0.8128299117088318,
+      "learning_rate": 9.691072942421642e-06,
+      "loss": 0.7554,
+      "step": 4176
+    },
+    {
+      "epoch": 0.22989707743959492,
+      "grad_norm": 0.9043960571289062,
+      "learning_rate": 9.690922921927295e-06,
+      "loss": 0.8601,
+      "step": 4177
+    },
+    {
+      "epoch": 0.22995211624195058,
+      "grad_norm": 0.835445761680603,
+      "learning_rate": 9.690772866177229e-06,
+      "loss": 0.8185,
+      "step": 4178
+    },
+    {
+      "epoch": 0.23000715504430624,
+      "grad_norm": 0.734601616859436,
+      "learning_rate": 9.69062277517257e-06,
+      "loss": 0.6486,
+      "step": 4179
+    },
+    {
+      "epoch": 0.2300621938466619,
+      "grad_norm": 0.8252671957015991,
+      "learning_rate": 9.690472648914445e-06,
+      "loss": 0.8455,
+      "step": 4180
+    },
+    {
+      "epoch": 0.23011723264901757,
+      "grad_norm": 0.8266329169273376,
+      "learning_rate": 9.690322487403984e-06,
+      "loss": 0.7348,
+      "step": 4181
+    },
+    {
+      "epoch": 0.23017227145137323,
+      "grad_norm": 0.8280256390571594,
+      "learning_rate": 9.690172290642314e-06,
+      "loss": 0.8191,
+      "step": 4182
+    },
+    {
+      "epoch": 0.2302273102537289,
+      "grad_norm": 0.8854276537895203,
+      "learning_rate": 9.690022058630564e-06,
+      "loss": 0.9327,
+      "step": 4183
+    },
+    {
+      "epoch": 0.23028234905608455,
+      "grad_norm": 0.7308807969093323,
+      "learning_rate": 9.689871791369865e-06,
+      "loss": 0.8144,
+      "step": 4184
+    },
+    {
+      "epoch": 0.2303373878584402,
+      "grad_norm": 0.7171719670295715,
+      "learning_rate": 9.689721488861344e-06,
+      "loss": 0.8265,
+      "step": 4185
+    },
+    {
+      "epoch": 0.23039242666079587,
+      "grad_norm": 0.7955548763275146,
+      "learning_rate": 9.689571151106131e-06,
+      "loss": 0.7313,
+      "step": 4186
+    },
+    {
+      "epoch": 0.23044746546315154,
+      "grad_norm": 0.8218876123428345,
+      "learning_rate": 9.689420778105359e-06,
+      "loss": 0.883,
+      "step": 4187
+    },
+    {
+      "epoch": 0.23050250426550717,
+      "grad_norm": 0.79570072889328,
+      "learning_rate": 9.689270369860154e-06,
+      "loss": 0.8898,
+      "step": 4188
+    },
+    {
+      "epoch": 0.23055754306786283,
+      "grad_norm": 0.8163344264030457,
+      "learning_rate": 9.689119926371649e-06,
+      "loss": 0.8638,
+      "step": 4189
+    },
+    {
+      "epoch": 0.2306125818702185,
+      "grad_norm": 0.7767764329910278,
+      "learning_rate": 9.688969447640972e-06,
+      "loss": 0.7822,
+      "step": 4190
+    },
+    {
+      "epoch": 0.23066762067257415,
+      "grad_norm": 0.9357114434242249,
+      "learning_rate": 9.688818933669258e-06,
+      "loss": 0.8031,
+      "step": 4191
+    },
+    {
+      "epoch": 0.23072265947492981,
+      "grad_norm": 0.8340080380439758,
+      "learning_rate": 9.688668384457635e-06,
+      "loss": 0.8947,
+      "step": 4192
+    },
+    {
+      "epoch": 0.23077769827728548,
+      "grad_norm": 0.8187471628189087,
+      "learning_rate": 9.688517800007235e-06,
+      "loss": 0.7989,
+      "step": 4193
+    },
+    {
+      "epoch": 0.23083273707964114,
+      "grad_norm": 0.8131871819496155,
+      "learning_rate": 9.688367180319191e-06,
+      "loss": 0.8377,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2308877758819968,
+      "grad_norm": 0.7933448553085327,
+      "learning_rate": 9.688216525394634e-06,
+      "loss": 0.8723,
+      "step": 4195
+    },
+    {
+      "epoch": 0.23094281468435246,
+      "grad_norm": 0.7262325286865234,
+      "learning_rate": 9.688065835234695e-06,
+      "loss": 0.7802,
+      "step": 4196
+    },
+    {
+      "epoch": 0.23099785348670812,
+      "grad_norm": 0.8289293050765991,
+      "learning_rate": 9.68791510984051e-06,
+      "loss": 0.642,
+      "step": 4197
+    },
+    {
+      "epoch": 0.23105289228906378,
+      "grad_norm": 0.8835988640785217,
+      "learning_rate": 9.687764349213211e-06,
+      "loss": 0.9002,
+      "step": 4198
+    },
+    {
+      "epoch": 0.23110793109141944,
+      "grad_norm": 0.9478649497032166,
+      "learning_rate": 9.687613553353927e-06,
+      "loss": 0.8668,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2311629698937751,
+      "grad_norm": 0.872936487197876,
+      "learning_rate": 9.687462722263796e-06,
+      "loss": 0.8312,
+      "step": 4200
+    },
+    {
+      "epoch": 0.23121800869613077,
+      "grad_norm": 0.7073879241943359,
+      "learning_rate": 9.68731185594395e-06,
+      "loss": 0.776,
+      "step": 4201
+    },
+    {
+      "epoch": 0.23127304749848643,
+      "grad_norm": 0.8265218734741211,
+      "learning_rate": 9.687160954395522e-06,
+      "loss": 0.8152,
+      "step": 4202
+    },
+    {
+      "epoch": 0.2313280863008421,
+      "grad_norm": 0.8027207255363464,
+      "learning_rate": 9.687010017619649e-06,
+      "loss": 0.9514,
+      "step": 4203
+    },
+    {
+      "epoch": 0.23138312510319775,
+      "grad_norm": 0.7416790127754211,
+      "learning_rate": 9.68685904561746e-06,
+      "loss": 0.7708,
+      "step": 4204
+    },
+    {
+      "epoch": 0.2314381639055534,
+      "grad_norm": 0.7916150689125061,
+      "learning_rate": 9.686708038390096e-06,
+      "loss": 0.7753,
+      "step": 4205
+    },
+    {
+      "epoch": 0.23149320270790907,
+      "grad_norm": 0.7213300466537476,
+      "learning_rate": 9.686556995938688e-06,
+      "loss": 0.83,
+      "step": 4206
+    },
+    {
+      "epoch": 0.23154824151026474,
+      "grad_norm": 0.7595892548561096,
+      "learning_rate": 9.68640591826437e-06,
+      "loss": 0.8186,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2316032803126204,
+      "grad_norm": 0.7042104601860046,
+      "learning_rate": 9.686254805368282e-06,
+      "loss": 0.7126,
+      "step": 4208
+    },
+    {
+      "epoch": 0.23165831911497606,
+      "grad_norm": 0.7416805028915405,
+      "learning_rate": 9.686103657251558e-06,
+      "loss": 0.7791,
+      "step": 4209
+    },
+    {
+      "epoch": 0.23171335791733172,
+      "grad_norm": 0.9868568181991577,
+      "learning_rate": 9.685952473915333e-06,
+      "loss": 0.8453,
+      "step": 4210
+    },
+    {
+      "epoch": 0.23176839671968738,
+      "grad_norm": 0.7133191823959351,
+      "learning_rate": 9.68580125536074e-06,
+      "loss": 0.6061,
+      "step": 4211
+    },
+    {
+      "epoch": 0.23182343552204304,
+      "grad_norm": 0.8307366967201233,
+      "learning_rate": 9.685650001588921e-06,
+      "loss": 0.8403,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2318784743243987,
+      "grad_norm": 0.8395226001739502,
+      "learning_rate": 9.685498712601014e-06,
+      "loss": 0.7945,
+      "step": 4213
+    },
+    {
+      "epoch": 0.23193351312675436,
+      "grad_norm": 0.7557219862937927,
+      "learning_rate": 9.68534738839815e-06,
+      "loss": 0.7765,
+      "step": 4214
+    },
+    {
+      "epoch": 0.23198855192911003,
+      "grad_norm": 0.7003554105758667,
+      "learning_rate": 9.68519602898147e-06,
+      "loss": 0.7228,
+      "step": 4215
+    },
+    {
+      "epoch": 0.2320435907314657,
+      "grad_norm": 0.8422999382019043,
+      "learning_rate": 9.68504463435211e-06,
+      "loss": 0.8524,
+      "step": 4216
+    },
+    {
+      "epoch": 0.23209862953382135,
+      "grad_norm": 0.9369016289710999,
+      "learning_rate": 9.68489320451121e-06,
+      "loss": 0.7646,
+      "step": 4217
+    },
+    {
+      "epoch": 0.232153668336177,
+      "grad_norm": 0.8456607460975647,
+      "learning_rate": 9.684741739459905e-06,
+      "loss": 0.7481,
+      "step": 4218
+    },
+    {
+      "epoch": 0.23220870713853267,
+      "grad_norm": 0.9284812211990356,
+      "learning_rate": 9.684590239199336e-06,
+      "loss": 0.8192,
+      "step": 4219
+    },
+    {
+      "epoch": 0.23226374594088833,
+      "grad_norm": 0.8474242687225342,
+      "learning_rate": 9.68443870373064e-06,
+      "loss": 0.7143,
+      "step": 4220
+    },
+    {
+      "epoch": 0.232318784743244,
+      "grad_norm": 0.8259334564208984,
+      "learning_rate": 9.684287133054957e-06,
+      "loss": 0.8667,
+      "step": 4221
+    },
+    {
+      "epoch": 0.23237382354559966,
+      "grad_norm": 0.8016416430473328,
+      "learning_rate": 9.684135527173427e-06,
+      "loss": 0.8694,
+      "step": 4222
+    },
+    {
+      "epoch": 0.23242886234795532,
+      "grad_norm": 0.7575937509536743,
+      "learning_rate": 9.683983886087186e-06,
+      "loss": 0.7591,
+      "step": 4223
+    },
+    {
+      "epoch": 0.23248390115031098,
+      "grad_norm": 0.7004683613777161,
+      "learning_rate": 9.683832209797377e-06,
+      "loss": 0.739,
+      "step": 4224
+    },
+    {
+      "epoch": 0.23253893995266664,
+      "grad_norm": 0.8265832662582397,
+      "learning_rate": 9.68368049830514e-06,
+      "loss": 0.7705,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2325939787550223,
+      "grad_norm": 0.7705711722373962,
+      "learning_rate": 9.683528751611612e-06,
+      "loss": 0.7896,
+      "step": 4226
+    },
+    {
+      "epoch": 0.23264901755737796,
+      "grad_norm": 0.7426978349685669,
+      "learning_rate": 9.683376969717937e-06,
+      "loss": 0.8217,
+      "step": 4227
+    },
+    {
+      "epoch": 0.23270405635973362,
+      "grad_norm": 0.7425839304924011,
+      "learning_rate": 9.683225152625255e-06,
+      "loss": 0.7426,
+      "step": 4228
+    },
+    {
+      "epoch": 0.23275909516208929,
+      "grad_norm": 1.0415440797805786,
+      "learning_rate": 9.683073300334705e-06,
+      "loss": 0.8585,
+      "step": 4229
+    },
+    {
+      "epoch": 0.23281413396444495,
+      "grad_norm": 0.7706055045127869,
+      "learning_rate": 9.68292141284743e-06,
+      "loss": 0.8349,
+      "step": 4230
+    },
+    {
+      "epoch": 0.23286917276680058,
+      "grad_norm": 0.8407607674598694,
+      "learning_rate": 9.682769490164572e-06,
+      "loss": 0.8592,
+      "step": 4231
+    },
+    {
+      "epoch": 0.23292421156915624,
+      "grad_norm": 0.6830767393112183,
+      "learning_rate": 9.68261753228727e-06,
+      "loss": 0.6773,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2329792503715119,
+      "grad_norm": 1.6661429405212402,
+      "learning_rate": 9.68246553921667e-06,
+      "loss": 1.005,
+      "step": 4233
+    },
+    {
+      "epoch": 0.23303428917386756,
+      "grad_norm": 0.7677092552185059,
+      "learning_rate": 9.682313510953912e-06,
+      "loss": 0.7689,
+      "step": 4234
+    },
+    {
+      "epoch": 0.23308932797622323,
+      "grad_norm": 0.7232248187065125,
+      "learning_rate": 9.682161447500139e-06,
+      "loss": 0.7765,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2331443667785789,
+      "grad_norm": 0.8667388558387756,
+      "learning_rate": 9.682009348856494e-06,
+      "loss": 0.8099,
+      "step": 4236
+    },
+    {
+      "epoch": 0.23319940558093455,
+      "grad_norm": 0.8220446705818176,
+      "learning_rate": 9.68185721502412e-06,
+      "loss": 0.8078,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2332544443832902,
+      "grad_norm": 0.9670323133468628,
+      "learning_rate": 9.68170504600416e-06,
+      "loss": 0.8912,
+      "step": 4238
+    },
+    {
+      "epoch": 0.23330948318564587,
+      "grad_norm": 0.7950771450996399,
+      "learning_rate": 9.68155284179776e-06,
+      "loss": 0.8165,
+      "step": 4239
+    },
+    {
+      "epoch": 0.23336452198800153,
+      "grad_norm": 0.7606233358383179,
+      "learning_rate": 9.68140060240606e-06,
+      "loss": 0.7795,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2334195607903572,
+      "grad_norm": 0.9580656886100769,
+      "learning_rate": 9.681248327830205e-06,
+      "loss": 0.7949,
+      "step": 4241
+    },
+    {
+      "epoch": 0.23347459959271286,
+      "grad_norm": 0.6878347992897034,
+      "learning_rate": 9.681096018071341e-06,
+      "loss": 0.7776,
+      "step": 4242
+    },
+    {
+      "epoch": 0.23352963839506852,
+      "grad_norm": 0.8449816107749939,
+      "learning_rate": 9.680943673130614e-06,
+      "loss": 0.8456,
+      "step": 4243
+    },
+    {
+      "epoch": 0.23358467719742418,
+      "grad_norm": 0.77314692735672,
+      "learning_rate": 9.680791293009167e-06,
+      "loss": 0.7915,
+      "step": 4244
+    },
+    {
+      "epoch": 0.23363971599977984,
+      "grad_norm": 0.8034142255783081,
+      "learning_rate": 9.680638877708146e-06,
+      "loss": 0.7377,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2336947548021355,
+      "grad_norm": 0.8754952549934387,
+      "learning_rate": 9.680486427228695e-06,
+      "loss": 0.8072,
+      "step": 4246
+    },
+    {
+      "epoch": 0.23374979360449116,
+      "grad_norm": 0.8169820308685303,
+      "learning_rate": 9.680333941571963e-06,
+      "loss": 0.8253,
+      "step": 4247
+    },
+    {
+      "epoch": 0.23380483240684682,
+      "grad_norm": 0.7848341464996338,
+      "learning_rate": 9.680181420739092e-06,
+      "loss": 0.8243,
+      "step": 4248
+    },
+    {
+      "epoch": 0.23385987120920249,
+      "grad_norm": 0.7599799036979675,
+      "learning_rate": 9.68002886473123e-06,
+      "loss": 0.781,
+      "step": 4249
+    },
+    {
+      "epoch": 0.23391491001155815,
+      "grad_norm": 0.8920254707336426,
+      "learning_rate": 9.679876273549524e-06,
+      "loss": 0.8199,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2339699488139138,
+      "grad_norm": 0.7813586592674255,
+      "learning_rate": 9.679723647195121e-06,
+      "loss": 0.7758,
+      "step": 4251
+    },
+    {
+      "epoch": 0.23402498761626947,
+      "grad_norm": 0.735282838344574,
+      "learning_rate": 9.679570985669168e-06,
+      "loss": 0.7651,
+      "step": 4252
+    },
+    {
+      "epoch": 0.23408002641862513,
+      "grad_norm": 0.7305853962898254,
+      "learning_rate": 9.679418288972813e-06,
+      "loss": 0.8202,
+      "step": 4253
+    },
+    {
+      "epoch": 0.2341350652209808,
+      "grad_norm": 0.8331005573272705,
+      "learning_rate": 9.6792655571072e-06,
+      "loss": 0.8784,
+      "step": 4254
+    },
+    {
+      "epoch": 0.23419010402333645,
+      "grad_norm": 0.8526305556297302,
+      "learning_rate": 9.679112790073481e-06,
+      "loss": 0.8116,
+      "step": 4255
+    },
+    {
+      "epoch": 0.23424514282569212,
+      "grad_norm": 0.741073489189148,
+      "learning_rate": 9.678959987872805e-06,
+      "loss": 0.6928,
+      "step": 4256
+    },
+    {
+      "epoch": 0.23430018162804778,
+      "grad_norm": 0.727859616279602,
+      "learning_rate": 9.678807150506315e-06,
+      "loss": 0.7571,
+      "step": 4257
+    },
+    {
+      "epoch": 0.23435522043040344,
+      "grad_norm": 0.8890698552131653,
+      "learning_rate": 9.678654277975165e-06,
+      "loss": 0.8145,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2344102592327591,
+      "grad_norm": 0.7372937798500061,
+      "learning_rate": 9.6785013702805e-06,
+      "loss": 0.7104,
+      "step": 4259
+    },
+    {
+      "epoch": 0.23446529803511476,
+      "grad_norm": 0.7205008268356323,
+      "learning_rate": 9.678348427423472e-06,
+      "loss": 0.7498,
+      "step": 4260
+    },
+    {
+      "epoch": 0.23452033683747042,
+      "grad_norm": 0.7766392230987549,
+      "learning_rate": 9.67819544940523e-06,
+      "loss": 0.7814,
+      "step": 4261
+    },
+    {
+      "epoch": 0.23457537563982608,
+      "grad_norm": 0.7441498637199402,
+      "learning_rate": 9.678042436226922e-06,
+      "loss": 0.7429,
+      "step": 4262
+    },
+    {
+      "epoch": 0.23463041444218175,
+      "grad_norm": 0.8838522434234619,
+      "learning_rate": 9.677889387889701e-06,
+      "loss": 0.8719,
+      "step": 4263
+    },
+    {
+      "epoch": 0.2346854532445374,
+      "grad_norm": 1.2349655628204346,
+      "learning_rate": 9.677736304394716e-06,
+      "loss": 0.8491,
+      "step": 4264
+    },
+    {
+      "epoch": 0.23474049204689307,
+      "grad_norm": 0.8050087690353394,
+      "learning_rate": 9.677583185743116e-06,
+      "loss": 0.795,
+      "step": 4265
+    },
+    {
+      "epoch": 0.23479553084924873,
+      "grad_norm": 0.7885709404945374,
+      "learning_rate": 9.677430031936051e-06,
+      "loss": 0.8594,
+      "step": 4266
+    },
+    {
+      "epoch": 0.2348505696516044,
+      "grad_norm": 0.7753557562828064,
+      "learning_rate": 9.677276842974676e-06,
+      "loss": 0.8196,
+      "step": 4267
+    },
+    {
+      "epoch": 0.23490560845396005,
+      "grad_norm": 0.7325392961502075,
+      "learning_rate": 9.67712361886014e-06,
+      "loss": 0.7905,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2349606472563157,
+      "grad_norm": 0.7925617694854736,
+      "learning_rate": 9.676970359593594e-06,
+      "loss": 0.7416,
+      "step": 4269
+    },
+    {
+      "epoch": 0.23501568605867137,
+      "grad_norm": 0.7981371283531189,
+      "learning_rate": 9.676817065176192e-06,
+      "loss": 0.81,
+      "step": 4270
+    },
+    {
+      "epoch": 0.23507072486102704,
+      "grad_norm": 0.7490524053573608,
+      "learning_rate": 9.676663735609084e-06,
+      "loss": 0.8347,
+      "step": 4271
+    },
+    {
+      "epoch": 0.2351257636633827,
+      "grad_norm": 1.000349521636963,
+      "learning_rate": 9.676510370893424e-06,
+      "loss": 0.7469,
+      "step": 4272
+    },
+    {
+      "epoch": 0.23518080246573836,
+      "grad_norm": 0.9310774207115173,
+      "learning_rate": 9.676356971030364e-06,
+      "loss": 0.8088,
+      "step": 4273
+    },
+    {
+      "epoch": 0.235235841268094,
+      "grad_norm": 0.8868544101715088,
+      "learning_rate": 9.676203536021055e-06,
+      "loss": 0.7472,
+      "step": 4274
+    },
+    {
+      "epoch": 0.23529088007044965,
+      "grad_norm": 0.7702255845069885,
+      "learning_rate": 9.676050065866653e-06,
+      "loss": 0.8395,
+      "step": 4275
+    },
+    {
+      "epoch": 0.23534591887280532,
+      "grad_norm": 0.7138833999633789,
+      "learning_rate": 9.675896560568311e-06,
+      "loss": 0.8529,
+      "step": 4276
+    },
+    {
+      "epoch": 0.23540095767516098,
+      "grad_norm": 0.8399729132652283,
+      "learning_rate": 9.675743020127182e-06,
+      "loss": 0.7844,
+      "step": 4277
+    },
+    {
+      "epoch": 0.23545599647751664,
+      "grad_norm": 0.8500726819038391,
+      "learning_rate": 9.67558944454442e-06,
+      "loss": 0.8209,
+      "step": 4278
+    },
+    {
+      "epoch": 0.2355110352798723,
+      "grad_norm": 0.766638994216919,
+      "learning_rate": 9.675435833821178e-06,
+      "loss": 0.7834,
+      "step": 4279
+    },
+    {
+      "epoch": 0.23556607408222796,
+      "grad_norm": 0.9121370315551758,
+      "learning_rate": 9.675282187958613e-06,
+      "loss": 0.8697,
+      "step": 4280
+    },
+    {
+      "epoch": 0.23562111288458362,
+      "grad_norm": 0.7862319946289062,
+      "learning_rate": 9.675128506957879e-06,
+      "loss": 0.8262,
+      "step": 4281
+    },
+    {
+      "epoch": 0.23567615168693928,
+      "grad_norm": 1.072777509689331,
+      "learning_rate": 9.67497479082013e-06,
+      "loss": 0.7963,
+      "step": 4282
+    },
+    {
+      "epoch": 0.23573119048929495,
+      "grad_norm": 0.7574695944786072,
+      "learning_rate": 9.67482103954652e-06,
+      "loss": 0.8178,
+      "step": 4283
+    },
+    {
+      "epoch": 0.2357862292916506,
+      "grad_norm": 0.7996877431869507,
+      "learning_rate": 9.674667253138209e-06,
+      "loss": 0.8465,
+      "step": 4284
+    },
+    {
+      "epoch": 0.23584126809400627,
+      "grad_norm": 0.711513340473175,
+      "learning_rate": 9.674513431596349e-06,
+      "loss": 0.7445,
+      "step": 4285
+    },
+    {
+      "epoch": 0.23589630689636193,
+      "grad_norm": 0.7431296706199646,
+      "learning_rate": 9.674359574922098e-06,
+      "loss": 0.8102,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2359513456987176,
+      "grad_norm": 0.7745676040649414,
+      "learning_rate": 9.674205683116612e-06,
+      "loss": 0.8733,
+      "step": 4287
+    },
+    {
+      "epoch": 0.23600638450107325,
+      "grad_norm": 1.0117937326431274,
+      "learning_rate": 9.674051756181046e-06,
+      "loss": 0.9035,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2360614233034289,
+      "grad_norm": 0.7848078608512878,
+      "learning_rate": 9.67389779411656e-06,
+      "loss": 0.8486,
+      "step": 4289
+    },
+    {
+      "epoch": 0.23611646210578457,
+      "grad_norm": 0.8439378142356873,
+      "learning_rate": 9.673743796924307e-06,
+      "loss": 0.8032,
+      "step": 4290
+    },
+    {
+      "epoch": 0.23617150090814024,
+      "grad_norm": 0.8268104791641235,
+      "learning_rate": 9.673589764605449e-06,
+      "loss": 0.8182,
+      "step": 4291
+    },
+    {
+      "epoch": 0.2362265397104959,
+      "grad_norm": 0.8896234631538391,
+      "learning_rate": 9.67343569716114e-06,
+      "loss": 0.8081,
+      "step": 4292
+    },
+    {
+      "epoch": 0.23628157851285156,
+      "grad_norm": 0.8515019416809082,
+      "learning_rate": 9.67328159459254e-06,
+      "loss": 0.8239,
+      "step": 4293
+    },
+    {
+      "epoch": 0.23633661731520722,
+      "grad_norm": 0.7779792547225952,
+      "learning_rate": 9.673127456900806e-06,
+      "loss": 0.8437,
+      "step": 4294
+    },
+    {
+      "epoch": 0.23639165611756288,
+      "grad_norm": 0.7782402634620667,
+      "learning_rate": 9.672973284087097e-06,
+      "loss": 0.8498,
+      "step": 4295
+    },
+    {
+      "epoch": 0.23644669491991854,
+      "grad_norm": 0.7588973641395569,
+      "learning_rate": 9.67281907615257e-06,
+      "loss": 0.7034,
+      "step": 4296
+    },
+    {
+      "epoch": 0.2365017337222742,
+      "grad_norm": 0.8426640629768372,
+      "learning_rate": 9.67266483309839e-06,
+      "loss": 0.803,
+      "step": 4297
+    },
+    {
+      "epoch": 0.23655677252462987,
+      "grad_norm": 0.8945889472961426,
+      "learning_rate": 9.672510554925707e-06,
+      "loss": 0.8971,
+      "step": 4298
+    },
+    {
+      "epoch": 0.23661181132698553,
+      "grad_norm": 0.8604227304458618,
+      "learning_rate": 9.672356241635688e-06,
+      "loss": 0.7548,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2366668501293412,
+      "grad_norm": 0.7277490496635437,
+      "learning_rate": 9.672201893229489e-06,
+      "loss": 0.8083,
+      "step": 4300
+    },
+    {
+      "epoch": 0.23672188893169685,
+      "grad_norm": 0.9089379906654358,
+      "learning_rate": 9.672047509708273e-06,
+      "loss": 0.9717,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2367769277340525,
+      "grad_norm": 0.7207155823707581,
+      "learning_rate": 9.671893091073198e-06,
+      "loss": 0.6794,
+      "step": 4302
+    },
+    {
+      "epoch": 0.23683196653640817,
+      "grad_norm": 0.7319806814193726,
+      "learning_rate": 9.671738637325425e-06,
+      "loss": 0.6821,
+      "step": 4303
+    },
+    {
+      "epoch": 0.23688700533876383,
+      "grad_norm": 0.7339589595794678,
+      "learning_rate": 9.671584148466112e-06,
+      "loss": 0.7895,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2369420441411195,
+      "grad_norm": 0.7725476622581482,
+      "learning_rate": 9.671429624496428e-06,
+      "loss": 0.7414,
+      "step": 4305
+    },
+    {
+      "epoch": 0.23699708294347516,
+      "grad_norm": 0.7040137648582458,
+      "learning_rate": 9.671275065417527e-06,
+      "loss": 0.696,
+      "step": 4306
+    },
+    {
+      "epoch": 0.23705212174583082,
+      "grad_norm": 0.8804189562797546,
+      "learning_rate": 9.671120471230572e-06,
+      "loss": 0.8184,
+      "step": 4307
+    },
+    {
+      "epoch": 0.23710716054818648,
+      "grad_norm": 0.8062872886657715,
+      "learning_rate": 9.670965841936728e-06,
+      "loss": 0.7856,
+      "step": 4308
+    },
+    {
+      "epoch": 0.23716219935054214,
+      "grad_norm": 0.7537097930908203,
+      "learning_rate": 9.670811177537154e-06,
+      "loss": 0.7562,
+      "step": 4309
+    },
+    {
+      "epoch": 0.2372172381528978,
+      "grad_norm": 0.8168618083000183,
+      "learning_rate": 9.670656478033013e-06,
+      "loss": 0.7416,
+      "step": 4310
+    },
+    {
+      "epoch": 0.23727227695525346,
+      "grad_norm": 0.8367040157318115,
+      "learning_rate": 9.670501743425469e-06,
+      "loss": 0.7759,
+      "step": 4311
+    },
+    {
+      "epoch": 0.23732731575760913,
+      "grad_norm": 0.860418975353241,
+      "learning_rate": 9.670346973715683e-06,
+      "loss": 0.9013,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2373823545599648,
+      "grad_norm": 0.8736678957939148,
+      "learning_rate": 9.67019216890482e-06,
+      "loss": 0.8677,
+      "step": 4313
+    },
+    {
+      "epoch": 0.23743739336232045,
+      "grad_norm": 0.8258964419364929,
+      "learning_rate": 9.670037328994044e-06,
+      "loss": 0.8208,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2374924321646761,
+      "grad_norm": 0.7936292886734009,
+      "learning_rate": 9.669882453984516e-06,
+      "loss": 0.8643,
+      "step": 4315
+    },
+    {
+      "epoch": 0.23754747096703177,
+      "grad_norm": 0.805500864982605,
+      "learning_rate": 9.669727543877401e-06,
+      "loss": 0.779,
+      "step": 4316
+    },
+    {
+      "epoch": 0.2376025097693874,
+      "grad_norm": 0.8072311282157898,
+      "learning_rate": 9.669572598673866e-06,
+      "loss": 0.8258,
+      "step": 4317
+    },
+    {
+      "epoch": 0.23765754857174307,
+      "grad_norm": 0.8917607665061951,
+      "learning_rate": 9.669417618375072e-06,
+      "loss": 0.7528,
+      "step": 4318
+    },
+    {
+      "epoch": 0.23771258737409873,
+      "grad_norm": 0.7054246068000793,
+      "learning_rate": 9.669262602982186e-06,
+      "loss": 0.86,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2377676261764544,
+      "grad_norm": 0.8600299954414368,
+      "learning_rate": 9.66910755249637e-06,
+      "loss": 0.8165,
+      "step": 4320
+    },
+    {
+      "epoch": 0.23782266497881005,
+      "grad_norm": 0.8685561418533325,
+      "learning_rate": 9.668952466918793e-06,
+      "loss": 0.8129,
+      "step": 4321
+    },
+    {
+      "epoch": 0.2378777037811657,
+      "grad_norm": 0.7859770655632019,
+      "learning_rate": 9.668797346250618e-06,
+      "loss": 0.8703,
+      "step": 4322
+    },
+    {
+      "epoch": 0.23793274258352137,
+      "grad_norm": 0.8128730058670044,
+      "learning_rate": 9.668642190493015e-06,
+      "loss": 0.7595,
+      "step": 4323
+    },
+    {
+      "epoch": 0.23798778138587703,
+      "grad_norm": 0.8223204612731934,
+      "learning_rate": 9.668486999647143e-06,
+      "loss": 0.825,
+      "step": 4324
+    },
+    {
+      "epoch": 0.2380428201882327,
+      "grad_norm": 0.859619677066803,
+      "learning_rate": 9.668331773714175e-06,
+      "loss": 0.8239,
+      "step": 4325
+    },
+    {
+      "epoch": 0.23809785899058836,
+      "grad_norm": 0.9861679673194885,
+      "learning_rate": 9.668176512695273e-06,
+      "loss": 0.8409,
+      "step": 4326
+    },
+    {
+      "epoch": 0.23815289779294402,
+      "grad_norm": 0.7178627252578735,
+      "learning_rate": 9.668021216591607e-06,
+      "loss": 0.818,
+      "step": 4327
+    },
+    {
+      "epoch": 0.23820793659529968,
+      "grad_norm": 0.9160923957824707,
+      "learning_rate": 9.667865885404343e-06,
+      "loss": 0.8703,
+      "step": 4328
+    },
+    {
+      "epoch": 0.23826297539765534,
+      "grad_norm": 0.7043942213058472,
+      "learning_rate": 9.667710519134648e-06,
+      "loss": 0.6884,
+      "step": 4329
+    },
+    {
+      "epoch": 0.238318014200011,
+      "grad_norm": 1.213121771812439,
+      "learning_rate": 9.667555117783691e-06,
+      "loss": 0.7843,
+      "step": 4330
+    },
+    {
+      "epoch": 0.23837305300236666,
+      "grad_norm": 0.8008033037185669,
+      "learning_rate": 9.66739968135264e-06,
+      "loss": 0.9312,
+      "step": 4331
+    },
+    {
+      "epoch": 0.23842809180472233,
+      "grad_norm": 0.7862009406089783,
+      "learning_rate": 9.667244209842662e-06,
+      "loss": 0.6965,
+      "step": 4332
+    },
+    {
+      "epoch": 0.238483130607078,
+      "grad_norm": 1.081398844718933,
+      "learning_rate": 9.667088703254923e-06,
+      "loss": 0.9793,
+      "step": 4333
+    },
+    {
+      "epoch": 0.23853816940943365,
+      "grad_norm": 0.7672395706176758,
+      "learning_rate": 9.666933161590597e-06,
+      "loss": 0.813,
+      "step": 4334
+    },
+    {
+      "epoch": 0.2385932082117893,
+      "grad_norm": 0.6955493092536926,
+      "learning_rate": 9.66677758485085e-06,
+      "loss": 0.7778,
+      "step": 4335
+    },
+    {
+      "epoch": 0.23864824701414497,
+      "grad_norm": 0.8609682321548462,
+      "learning_rate": 9.666621973036854e-06,
+      "loss": 0.7817,
+      "step": 4336
+    },
+    {
+      "epoch": 0.23870328581650063,
+      "grad_norm": 0.7312196493148804,
+      "learning_rate": 9.666466326149774e-06,
+      "loss": 0.7368,
+      "step": 4337
+    },
+    {
+      "epoch": 0.2387583246188563,
+      "grad_norm": 0.7964538931846619,
+      "learning_rate": 9.666310644190782e-06,
+      "loss": 0.8124,
+      "step": 4338
+    },
+    {
+      "epoch": 0.23881336342121195,
+      "grad_norm": 1.1138910055160522,
+      "learning_rate": 9.66615492716105e-06,
+      "loss": 0.8886,
+      "step": 4339
+    },
+    {
+      "epoch": 0.23886840222356762,
+      "grad_norm": 0.8789949417114258,
+      "learning_rate": 9.665999175061747e-06,
+      "loss": 0.7854,
+      "step": 4340
+    },
+    {
+      "epoch": 0.23892344102592328,
+      "grad_norm": 0.7761380076408386,
+      "learning_rate": 9.665843387894041e-06,
+      "loss": 0.7915,
+      "step": 4341
+    },
+    {
+      "epoch": 0.23897847982827894,
+      "grad_norm": 0.888482928276062,
+      "learning_rate": 9.665687565659106e-06,
+      "loss": 0.8799,
+      "step": 4342
+    },
+    {
+      "epoch": 0.2390335186306346,
+      "grad_norm": 0.7799200415611267,
+      "learning_rate": 9.665531708358111e-06,
+      "loss": 0.8519,
+      "step": 4343
+    },
+    {
+      "epoch": 0.23908855743299026,
+      "grad_norm": 0.7407697439193726,
+      "learning_rate": 9.665375815992231e-06,
+      "loss": 0.7637,
+      "step": 4344
+    },
+    {
+      "epoch": 0.23914359623534592,
+      "grad_norm": 0.8098278045654297,
+      "learning_rate": 9.665219888562634e-06,
+      "loss": 0.7991,
+      "step": 4345
+    },
+    {
+      "epoch": 0.23919863503770158,
+      "grad_norm": 0.7585136294364929,
+      "learning_rate": 9.665063926070493e-06,
+      "loss": 0.8478,
+      "step": 4346
+    },
+    {
+      "epoch": 0.23925367384005725,
+      "grad_norm": 0.7294817566871643,
+      "learning_rate": 9.66490792851698e-06,
+      "loss": 0.8312,
+      "step": 4347
+    },
+    {
+      "epoch": 0.2393087126424129,
+      "grad_norm": 0.8325762748718262,
+      "learning_rate": 9.664751895903269e-06,
+      "loss": 0.9365,
+      "step": 4348
+    },
+    {
+      "epoch": 0.23936375144476857,
+      "grad_norm": 0.9992470741271973,
+      "learning_rate": 9.66459582823053e-06,
+      "loss": 0.8649,
+      "step": 4349
+    },
+    {
+      "epoch": 0.23941879024712423,
+      "grad_norm": 0.7206875681877136,
+      "learning_rate": 9.664439725499938e-06,
+      "loss": 0.7013,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2394738290494799,
+      "grad_norm": 0.946657657623291,
+      "learning_rate": 9.664283587712665e-06,
+      "loss": 0.7953,
+      "step": 4351
+    },
+    {
+      "epoch": 0.23952886785183555,
+      "grad_norm": 0.7684911489486694,
+      "learning_rate": 9.664127414869887e-06,
+      "loss": 0.8403,
+      "step": 4352
+    },
+    {
+      "epoch": 0.23958390665419121,
+      "grad_norm": 0.7875770926475525,
+      "learning_rate": 9.663971206972773e-06,
+      "loss": 0.7961,
+      "step": 4353
+    },
+    {
+      "epoch": 0.23963894545654688,
+      "grad_norm": 0.7387273907661438,
+      "learning_rate": 9.663814964022502e-06,
+      "loss": 0.8265,
+      "step": 4354
+    },
+    {
+      "epoch": 0.23969398425890254,
+      "grad_norm": 0.7413492202758789,
+      "learning_rate": 9.663658686020245e-06,
+      "loss": 0.8458,
+      "step": 4355
+    },
+    {
+      "epoch": 0.2397490230612582,
+      "grad_norm": 0.7563235759735107,
+      "learning_rate": 9.663502372967177e-06,
+      "loss": 0.8498,
+      "step": 4356
+    },
+    {
+      "epoch": 0.23980406186361386,
+      "grad_norm": 0.7529472708702087,
+      "learning_rate": 9.663346024864475e-06,
+      "loss": 0.7597,
+      "step": 4357
+    },
+    {
+      "epoch": 0.23985910066596952,
+      "grad_norm": 0.7582191824913025,
+      "learning_rate": 9.663189641713314e-06,
+      "loss": 0.804,
+      "step": 4358
+    },
+    {
+      "epoch": 0.23991413946832518,
+      "grad_norm": 0.8394485712051392,
+      "learning_rate": 9.663033223514865e-06,
+      "loss": 0.8329,
+      "step": 4359
+    },
+    {
+      "epoch": 0.23996917827068082,
+      "grad_norm": 0.7088292241096497,
+      "learning_rate": 9.662876770270308e-06,
+      "loss": 0.7131,
+      "step": 4360
+    },
+    {
+      "epoch": 0.24002421707303648,
+      "grad_norm": 0.8548080325126648,
+      "learning_rate": 9.662720281980817e-06,
+      "loss": 0.8925,
+      "step": 4361
+    },
+    {
+      "epoch": 0.24007925587539214,
+      "grad_norm": 0.8027567267417908,
+      "learning_rate": 9.662563758647568e-06,
+      "loss": 0.8652,
+      "step": 4362
+    },
+    {
+      "epoch": 0.2401342946777478,
+      "grad_norm": 0.7471736669540405,
+      "learning_rate": 9.662407200271738e-06,
+      "loss": 0.7722,
+      "step": 4363
+    },
+    {
+      "epoch": 0.24018933348010346,
+      "grad_norm": 0.7358804941177368,
+      "learning_rate": 9.662250606854504e-06,
+      "loss": 0.767,
+      "step": 4364
+    },
+    {
+      "epoch": 0.24024437228245912,
+      "grad_norm": 0.7948476672172546,
+      "learning_rate": 9.662093978397042e-06,
+      "loss": 0.961,
+      "step": 4365
+    },
+    {
+      "epoch": 0.24029941108481478,
+      "grad_norm": 0.7030961513519287,
+      "learning_rate": 9.66193731490053e-06,
+      "loss": 0.7826,
+      "step": 4366
+    },
+    {
+      "epoch": 0.24035444988717045,
+      "grad_norm": 0.8376098871231079,
+      "learning_rate": 9.661780616366145e-06,
+      "loss": 0.7697,
+      "step": 4367
+    },
+    {
+      "epoch": 0.2404094886895261,
+      "grad_norm": 0.7449594140052795,
+      "learning_rate": 9.661623882795065e-06,
+      "loss": 0.7944,
+      "step": 4368
+    },
+    {
+      "epoch": 0.24046452749188177,
+      "grad_norm": 0.7317184805870056,
+      "learning_rate": 9.661467114188468e-06,
+      "loss": 0.7059,
+      "step": 4369
+    },
+    {
+      "epoch": 0.24051956629423743,
+      "grad_norm": 0.843912661075592,
+      "learning_rate": 9.661310310547531e-06,
+      "loss": 0.7889,
+      "step": 4370
+    },
+    {
+      "epoch": 0.2405746050965931,
+      "grad_norm": 0.8673211336135864,
+      "learning_rate": 9.661153471873435e-06,
+      "loss": 0.7234,
+      "step": 4371
+    },
+    {
+      "epoch": 0.24062964389894875,
+      "grad_norm": 0.8179688453674316,
+      "learning_rate": 9.660996598167354e-06,
+      "loss": 0.8937,
+      "step": 4372
+    },
+    {
+      "epoch": 0.24068468270130441,
+      "grad_norm": 0.7800211906433105,
+      "learning_rate": 9.660839689430473e-06,
+      "loss": 0.8596,
+      "step": 4373
+    },
+    {
+      "epoch": 0.24073972150366008,
+      "grad_norm": 0.8781671524047852,
+      "learning_rate": 9.660682745663967e-06,
+      "loss": 0.8507,
+      "step": 4374
+    },
+    {
+      "epoch": 0.24079476030601574,
+      "grad_norm": 0.7701708674430847,
+      "learning_rate": 9.660525766869019e-06,
+      "loss": 0.8212,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2408497991083714,
+      "grad_norm": 0.7721084356307983,
+      "learning_rate": 9.660368753046806e-06,
+      "loss": 0.7493,
+      "step": 4376
+    },
+    {
+      "epoch": 0.24090483791072706,
+      "grad_norm": 0.8126489520072937,
+      "learning_rate": 9.660211704198508e-06,
+      "loss": 0.8527,
+      "step": 4377
+    },
+    {
+      "epoch": 0.24095987671308272,
+      "grad_norm": 0.8172717690467834,
+      "learning_rate": 9.660054620325307e-06,
+      "loss": 0.8448,
+      "step": 4378
+    },
+    {
+      "epoch": 0.24101491551543838,
+      "grad_norm": 0.8293611407279968,
+      "learning_rate": 9.659897501428384e-06,
+      "loss": 0.9318,
+      "step": 4379
+    },
+    {
+      "epoch": 0.24106995431779404,
+      "grad_norm": 0.7445098161697388,
+      "learning_rate": 9.659740347508917e-06,
+      "loss": 0.7358,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2411249931201497,
+      "grad_norm": 0.7778907418251038,
+      "learning_rate": 9.659583158568088e-06,
+      "loss": 0.7671,
+      "step": 4381
+    },
+    {
+      "epoch": 0.24118003192250537,
+      "grad_norm": 0.7828608751296997,
+      "learning_rate": 9.659425934607082e-06,
+      "loss": 0.8141,
+      "step": 4382
+    },
+    {
+      "epoch": 0.24123507072486103,
+      "grad_norm": 0.9433113932609558,
+      "learning_rate": 9.659268675627075e-06,
+      "loss": 0.7904,
+      "step": 4383
+    },
+    {
+      "epoch": 0.2412901095272167,
+      "grad_norm": 0.7097491025924683,
+      "learning_rate": 9.659111381629255e-06,
+      "loss": 0.7445,
+      "step": 4384
+    },
+    {
+      "epoch": 0.24134514832957235,
+      "grad_norm": 0.7450230717658997,
+      "learning_rate": 9.6589540526148e-06,
+      "loss": 0.6869,
+      "step": 4385
+    },
+    {
+      "epoch": 0.241400187131928,
+      "grad_norm": 0.7429760694503784,
+      "learning_rate": 9.658796688584893e-06,
+      "loss": 0.7367,
+      "step": 4386
+    },
+    {
+      "epoch": 0.24145522593428367,
+      "grad_norm": 0.7250030040740967,
+      "learning_rate": 9.658639289540716e-06,
+      "loss": 0.7502,
+      "step": 4387
+    },
+    {
+      "epoch": 0.24151026473663934,
+      "grad_norm": 0.6577159762382507,
+      "learning_rate": 9.658481855483455e-06,
+      "loss": 0.5785,
+      "step": 4388
+    },
+    {
+      "epoch": 0.241565303538995,
+      "grad_norm": 0.7846524119377136,
+      "learning_rate": 9.65832438641429e-06,
+      "loss": 0.7435,
+      "step": 4389
+    },
+    {
+      "epoch": 0.24162034234135066,
+      "grad_norm": 0.8370404839515686,
+      "learning_rate": 9.658166882334408e-06,
+      "loss": 0.8536,
+      "step": 4390
+    },
+    {
+      "epoch": 0.24167538114370632,
+      "grad_norm": 0.7451018691062927,
+      "learning_rate": 9.658009343244987e-06,
+      "loss": 0.8443,
+      "step": 4391
+    },
+    {
+      "epoch": 0.24173041994606198,
+      "grad_norm": 0.7629074454307556,
+      "learning_rate": 9.657851769147218e-06,
+      "loss": 0.7394,
+      "step": 4392
+    },
+    {
+      "epoch": 0.24178545874841764,
+      "grad_norm": 0.7767705321311951,
+      "learning_rate": 9.657694160042282e-06,
+      "loss": 0.8497,
+      "step": 4393
+    },
+    {
+      "epoch": 0.2418404975507733,
+      "grad_norm": 0.8635357022285461,
+      "learning_rate": 9.65753651593136e-06,
+      "loss": 0.8495,
+      "step": 4394
+    },
+    {
+      "epoch": 0.24189553635312896,
+      "grad_norm": 0.7652365565299988,
+      "learning_rate": 9.657378836815643e-06,
+      "loss": 0.7967,
+      "step": 4395
+    },
+    {
+      "epoch": 0.24195057515548463,
+      "grad_norm": 0.7721680402755737,
+      "learning_rate": 9.657221122696313e-06,
+      "loss": 0.8227,
+      "step": 4396
+    },
+    {
+      "epoch": 0.2420056139578403,
+      "grad_norm": 1.016366720199585,
+      "learning_rate": 9.657063373574555e-06,
+      "loss": 0.8291,
+      "step": 4397
+    },
+    {
+      "epoch": 0.24206065276019595,
+      "grad_norm": 0.7770145535469055,
+      "learning_rate": 9.656905589451555e-06,
+      "loss": 0.8335,
+      "step": 4398
+    },
+    {
+      "epoch": 0.2421156915625516,
+      "grad_norm": 0.812882125377655,
+      "learning_rate": 9.6567477703285e-06,
+      "loss": 0.8189,
+      "step": 4399
+    },
+    {
+      "epoch": 0.24217073036490727,
+      "grad_norm": 0.7253247499465942,
+      "learning_rate": 9.656589916206576e-06,
+      "loss": 0.8418,
+      "step": 4400
+    },
+    {
+      "epoch": 0.24222576916726293,
+      "grad_norm": 0.7784958481788635,
+      "learning_rate": 9.656432027086969e-06,
+      "loss": 0.8541,
+      "step": 4401
+    },
+    {
+      "epoch": 0.2422808079696186,
+      "grad_norm": 0.8001978397369385,
+      "learning_rate": 9.656274102970865e-06,
+      "loss": 0.8888,
+      "step": 4402
+    },
+    {
+      "epoch": 0.24233584677197423,
+      "grad_norm": 0.7535765767097473,
+      "learning_rate": 9.656116143859448e-06,
+      "loss": 0.7691,
+      "step": 4403
+    },
+    {
+      "epoch": 0.2423908855743299,
+      "grad_norm": 0.6554346680641174,
+      "learning_rate": 9.655958149753913e-06,
+      "loss": 0.7592,
+      "step": 4404
+    },
+    {
+      "epoch": 0.24244592437668555,
+      "grad_norm": 0.8599995374679565,
+      "learning_rate": 9.655800120655439e-06,
+      "loss": 0.8396,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2425009631790412,
+      "grad_norm": 0.8172232508659363,
+      "learning_rate": 9.65564205656522e-06,
+      "loss": 0.6931,
+      "step": 4406
+    },
+    {
+      "epoch": 0.24255600198139687,
+      "grad_norm": 0.8005852699279785,
+      "learning_rate": 9.65548395748444e-06,
+      "loss": 0.8344,
+      "step": 4407
+    },
+    {
+      "epoch": 0.24261104078375254,
+      "grad_norm": 0.7823762893676758,
+      "learning_rate": 9.65532582341429e-06,
+      "loss": 0.7991,
+      "step": 4408
+    },
+    {
+      "epoch": 0.2426660795861082,
+      "grad_norm": 0.7743250727653503,
+      "learning_rate": 9.655167654355957e-06,
+      "loss": 0.9048,
+      "step": 4409
+    },
+    {
+      "epoch": 0.24272111838846386,
+      "grad_norm": 0.9825221300125122,
+      "learning_rate": 9.655009450310629e-06,
+      "loss": 0.7491,
+      "step": 4410
+    },
+    {
+      "epoch": 0.24277615719081952,
+      "grad_norm": 1.2921068668365479,
+      "learning_rate": 9.654851211279496e-06,
+      "loss": 0.8175,
+      "step": 4411
+    },
+    {
+      "epoch": 0.24283119599317518,
+      "grad_norm": 0.8267684578895569,
+      "learning_rate": 9.65469293726375e-06,
+      "loss": 0.8896,
+      "step": 4412
+    },
+    {
+      "epoch": 0.24288623479553084,
+      "grad_norm": 0.8020186424255371,
+      "learning_rate": 9.654534628264576e-06,
+      "loss": 0.7145,
+      "step": 4413
+    },
+    {
+      "epoch": 0.2429412735978865,
+      "grad_norm": 0.8192574977874756,
+      "learning_rate": 9.654376284283166e-06,
+      "loss": 0.7451,
+      "step": 4414
+    },
+    {
+      "epoch": 0.24299631240024216,
+      "grad_norm": 0.7733662128448486,
+      "learning_rate": 9.65421790532071e-06,
+      "loss": 0.768,
+      "step": 4415
+    },
+    {
+      "epoch": 0.24305135120259783,
+      "grad_norm": 0.8342406153678894,
+      "learning_rate": 9.654059491378396e-06,
+      "loss": 0.8137,
+      "step": 4416
+    },
+    {
+      "epoch": 0.2431063900049535,
+      "grad_norm": 1.014755368232727,
+      "learning_rate": 9.653901042457418e-06,
+      "loss": 0.8922,
+      "step": 4417
+    },
+    {
+      "epoch": 0.24316142880730915,
+      "grad_norm": 0.864608645439148,
+      "learning_rate": 9.653742558558967e-06,
+      "loss": 0.9412,
+      "step": 4418
+    },
+    {
+      "epoch": 0.2432164676096648,
+      "grad_norm": 0.7383908033370972,
+      "learning_rate": 9.65358403968423e-06,
+      "loss": 0.8261,
+      "step": 4419
+    },
+    {
+      "epoch": 0.24327150641202047,
+      "grad_norm": 0.7464672923088074,
+      "learning_rate": 9.653425485834403e-06,
+      "loss": 0.7074,
+      "step": 4420
+    },
+    {
+      "epoch": 0.24332654521437613,
+      "grad_norm": 0.7010141611099243,
+      "learning_rate": 9.653266897010676e-06,
+      "loss": 0.6849,
+      "step": 4421
+    },
+    {
+      "epoch": 0.2433815840167318,
+      "grad_norm": 0.7135268449783325,
+      "learning_rate": 9.653108273214239e-06,
+      "loss": 0.8228,
+      "step": 4422
+    },
+    {
+      "epoch": 0.24343662281908746,
+      "grad_norm": 0.8061006665229797,
+      "learning_rate": 9.652949614446287e-06,
+      "loss": 0.8345,
+      "step": 4423
+    },
+    {
+      "epoch": 0.24349166162144312,
+      "grad_norm": 0.6954759955406189,
+      "learning_rate": 9.652790920708011e-06,
+      "loss": 0.7189,
+      "step": 4424
+    },
+    {
+      "epoch": 0.24354670042379878,
+      "grad_norm": 0.8669333457946777,
+      "learning_rate": 9.652632192000603e-06,
+      "loss": 0.8872,
+      "step": 4425
+    },
+    {
+      "epoch": 0.24360173922615444,
+      "grad_norm": 0.7445051670074463,
+      "learning_rate": 9.652473428325258e-06,
+      "loss": 0.826,
+      "step": 4426
+    },
+    {
+      "epoch": 0.2436567780285101,
+      "grad_norm": 0.7444632649421692,
+      "learning_rate": 9.652314629683165e-06,
+      "loss": 0.8568,
+      "step": 4427
+    },
+    {
+      "epoch": 0.24371181683086576,
+      "grad_norm": 0.7160165309906006,
+      "learning_rate": 9.652155796075524e-06,
+      "loss": 0.799,
+      "step": 4428
+    },
+    {
+      "epoch": 0.24376685563322142,
+      "grad_norm": 0.7098904252052307,
+      "learning_rate": 9.651996927503526e-06,
+      "loss": 0.8148,
+      "step": 4429
+    },
+    {
+      "epoch": 0.24382189443557709,
+      "grad_norm": 0.7911115288734436,
+      "learning_rate": 9.651838023968363e-06,
+      "loss": 0.8279,
+      "step": 4430
+    },
+    {
+      "epoch": 0.24387693323793275,
+      "grad_norm": 0.8887501955032349,
+      "learning_rate": 9.651679085471229e-06,
+      "loss": 0.8464,
+      "step": 4431
+    },
+    {
+      "epoch": 0.2439319720402884,
+      "grad_norm": 0.8343196511268616,
+      "learning_rate": 9.651520112013321e-06,
+      "loss": 0.7364,
+      "step": 4432
+    },
+    {
+      "epoch": 0.24398701084264407,
+      "grad_norm": 0.7279361486434937,
+      "learning_rate": 9.651361103595835e-06,
+      "loss": 0.7958,
+      "step": 4433
+    },
+    {
+      "epoch": 0.24404204964499973,
+      "grad_norm": 0.8221089243888855,
+      "learning_rate": 9.651202060219962e-06,
+      "loss": 0.7753,
+      "step": 4434
+    },
+    {
+      "epoch": 0.2440970884473554,
+      "grad_norm": 0.7205086350440979,
+      "learning_rate": 9.6510429818869e-06,
+      "loss": 0.7411,
+      "step": 4435
+    },
+    {
+      "epoch": 0.24415212724971105,
+      "grad_norm": 0.854967474937439,
+      "learning_rate": 9.650883868597845e-06,
+      "loss": 0.8192,
+      "step": 4436
+    },
+    {
+      "epoch": 0.24420716605206672,
+      "grad_norm": 0.7622473835945129,
+      "learning_rate": 9.65072472035399e-06,
+      "loss": 0.7645,
+      "step": 4437
+    },
+    {
+      "epoch": 0.24426220485442238,
+      "grad_norm": 0.7430302500724792,
+      "learning_rate": 9.650565537156533e-06,
+      "loss": 0.7817,
+      "step": 4438
+    },
+    {
+      "epoch": 0.24431724365677804,
+      "grad_norm": 0.8022677898406982,
+      "learning_rate": 9.650406319006672e-06,
+      "loss": 0.8035,
+      "step": 4439
+    },
+    {
+      "epoch": 0.2443722824591337,
+      "grad_norm": 0.7346476912498474,
+      "learning_rate": 9.6502470659056e-06,
+      "loss": 0.826,
+      "step": 4440
+    },
+    {
+      "epoch": 0.24442732126148936,
+      "grad_norm": 0.8393376469612122,
+      "learning_rate": 9.650087777854517e-06,
+      "loss": 0.8073,
+      "step": 4441
+    },
+    {
+      "epoch": 0.24448236006384502,
+      "grad_norm": 0.7920215129852295,
+      "learning_rate": 9.649928454854618e-06,
+      "loss": 0.7774,
+      "step": 4442
+    },
+    {
+      "epoch": 0.24453739886620068,
+      "grad_norm": 0.8192804455757141,
+      "learning_rate": 9.649769096907102e-06,
+      "loss": 0.7817,
+      "step": 4443
+    },
+    {
+      "epoch": 0.24459243766855635,
+      "grad_norm": 0.7727654576301575,
+      "learning_rate": 9.649609704013167e-06,
+      "loss": 0.8201,
+      "step": 4444
+    },
+    {
+      "epoch": 0.244647476470912,
+      "grad_norm": 0.8005746603012085,
+      "learning_rate": 9.649450276174008e-06,
+      "loss": 0.8893,
+      "step": 4445
+    },
+    {
+      "epoch": 0.24470251527326764,
+      "grad_norm": 0.9029125571250916,
+      "learning_rate": 9.649290813390828e-06,
+      "loss": 0.7735,
+      "step": 4446
+    },
+    {
+      "epoch": 0.2447575540756233,
+      "grad_norm": 0.8336170315742493,
+      "learning_rate": 9.64913131566482e-06,
+      "loss": 0.7505,
+      "step": 4447
+    },
+    {
+      "epoch": 0.24481259287797896,
+      "grad_norm": 1.0272265672683716,
+      "learning_rate": 9.648971782997188e-06,
+      "loss": 0.8371,
+      "step": 4448
+    },
+    {
+      "epoch": 0.24486763168033462,
+      "grad_norm": 0.8095843195915222,
+      "learning_rate": 9.648812215389128e-06,
+      "loss": 0.7599,
+      "step": 4449
+    },
+    {
+      "epoch": 0.24492267048269029,
+      "grad_norm": 0.7690166234970093,
+      "learning_rate": 9.648652612841837e-06,
+      "loss": 0.8172,
+      "step": 4450
+    },
+    {
+      "epoch": 0.24497770928504595,
+      "grad_norm": 0.8282617926597595,
+      "learning_rate": 9.64849297535652e-06,
+      "loss": 0.8477,
+      "step": 4451
+    },
+    {
+      "epoch": 0.2450327480874016,
+      "grad_norm": 0.8307822346687317,
+      "learning_rate": 9.648333302934373e-06,
+      "loss": 0.7744,
+      "step": 4452
+    },
+    {
+      "epoch": 0.24508778688975727,
+      "grad_norm": 0.7619080543518066,
+      "learning_rate": 9.6481735955766e-06,
+      "loss": 0.8417,
+      "step": 4453
+    },
+    {
+      "epoch": 0.24514282569211293,
+      "grad_norm": 0.7879447937011719,
+      "learning_rate": 9.648013853284396e-06,
+      "loss": 0.7799,
+      "step": 4454
+    },
+    {
+      "epoch": 0.2451978644944686,
+      "grad_norm": 0.7352256774902344,
+      "learning_rate": 9.647854076058965e-06,
+      "loss": 0.8386,
+      "step": 4455
+    },
+    {
+      "epoch": 0.24525290329682425,
+      "grad_norm": 0.8318933248519897,
+      "learning_rate": 9.647694263901507e-06,
+      "loss": 0.7631,
+      "step": 4456
+    },
+    {
+      "epoch": 0.24530794209917992,
+      "grad_norm": 0.8609912395477295,
+      "learning_rate": 9.647534416813221e-06,
+      "loss": 0.7479,
+      "step": 4457
+    },
+    {
+      "epoch": 0.24536298090153558,
+      "grad_norm": 0.9590480327606201,
+      "learning_rate": 9.647374534795311e-06,
+      "loss": 0.8543,
+      "step": 4458
+    },
+    {
+      "epoch": 0.24541801970389124,
+      "grad_norm": 0.7902723550796509,
+      "learning_rate": 9.647214617848979e-06,
+      "loss": 0.6796,
+      "step": 4459
+    },
+    {
+      "epoch": 0.2454730585062469,
+      "grad_norm": 0.7725642919540405,
+      "learning_rate": 9.647054665975427e-06,
+      "loss": 0.7563,
+      "step": 4460
+    },
+    {
+      "epoch": 0.24552809730860256,
+      "grad_norm": 0.8387014269828796,
+      "learning_rate": 9.646894679175853e-06,
+      "loss": 0.8184,
+      "step": 4461
+    },
+    {
+      "epoch": 0.24558313611095822,
+      "grad_norm": 0.9200852513313293,
+      "learning_rate": 9.646734657451464e-06,
+      "loss": 0.8436,
+      "step": 4462
+    },
+    {
+      "epoch": 0.24563817491331388,
+      "grad_norm": 0.7565840482711792,
+      "learning_rate": 9.646574600803462e-06,
+      "loss": 0.7393,
+      "step": 4463
+    },
+    {
+      "epoch": 0.24569321371566955,
+      "grad_norm": 0.7685559988021851,
+      "learning_rate": 9.646414509233048e-06,
+      "loss": 0.7836,
+      "step": 4464
+    },
+    {
+      "epoch": 0.2457482525180252,
+      "grad_norm": 0.8172003030776978,
+      "learning_rate": 9.646254382741428e-06,
+      "loss": 0.787,
+      "step": 4465
+    },
+    {
+      "epoch": 0.24580329132038087,
+      "grad_norm": 0.902632474899292,
+      "learning_rate": 9.646094221329802e-06,
+      "loss": 0.7139,
+      "step": 4466
+    },
+    {
+      "epoch": 0.24585833012273653,
+      "grad_norm": 0.7810692191123962,
+      "learning_rate": 9.645934024999374e-06,
+      "loss": 0.6904,
+      "step": 4467
+    },
+    {
+      "epoch": 0.2459133689250922,
+      "grad_norm": 0.7242134213447571,
+      "learning_rate": 9.645773793751352e-06,
+      "loss": 0.7035,
+      "step": 4468
+    },
+    {
+      "epoch": 0.24596840772744785,
+      "grad_norm": 0.7192920446395874,
+      "learning_rate": 9.645613527586938e-06,
+      "loss": 0.7081,
+      "step": 4469
+    },
+    {
+      "epoch": 0.2460234465298035,
+      "grad_norm": 0.7613840103149414,
+      "learning_rate": 9.645453226507336e-06,
+      "loss": 0.8066,
+      "step": 4470
+    },
+    {
+      "epoch": 0.24607848533215917,
+      "grad_norm": 0.8154922127723694,
+      "learning_rate": 9.64529289051375e-06,
+      "loss": 0.812,
+      "step": 4471
+    },
+    {
+      "epoch": 0.24613352413451484,
+      "grad_norm": 0.9521573185920715,
+      "learning_rate": 9.645132519607387e-06,
+      "loss": 0.7456,
+      "step": 4472
+    },
+    {
+      "epoch": 0.2461885629368705,
+      "grad_norm": 0.785943329334259,
+      "learning_rate": 9.64497211378945e-06,
+      "loss": 0.832,
+      "step": 4473
+    },
+    {
+      "epoch": 0.24624360173922616,
+      "grad_norm": 0.7675127983093262,
+      "learning_rate": 9.644811673061148e-06,
+      "loss": 0.7984,
+      "step": 4474
+    },
+    {
+      "epoch": 0.24629864054158182,
+      "grad_norm": 0.7317580580711365,
+      "learning_rate": 9.644651197423683e-06,
+      "loss": 0.7634,
+      "step": 4475
+    },
+    {
+      "epoch": 0.24635367934393748,
+      "grad_norm": 0.744937539100647,
+      "learning_rate": 9.644490686878265e-06,
+      "loss": 0.729,
+      "step": 4476
+    },
+    {
+      "epoch": 0.24640871814629314,
+      "grad_norm": 0.7472458481788635,
+      "learning_rate": 9.644330141426097e-06,
+      "loss": 0.7517,
+      "step": 4477
+    },
+    {
+      "epoch": 0.2464637569486488,
+      "grad_norm": 0.8379414677619934,
+      "learning_rate": 9.644169561068387e-06,
+      "loss": 0.8008,
+      "step": 4478
+    },
+    {
+      "epoch": 0.24651879575100447,
+      "grad_norm": 0.8845154047012329,
+      "learning_rate": 9.64400894580634e-06,
+      "loss": 0.8135,
+      "step": 4479
+    },
+    {
+      "epoch": 0.24657383455336013,
+      "grad_norm": 0.7394443154335022,
+      "learning_rate": 9.643848295641167e-06,
+      "loss": 0.7697,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2466288733557158,
+      "grad_norm": 0.8840840458869934,
+      "learning_rate": 9.643687610574073e-06,
+      "loss": 0.825,
+      "step": 4481
+    },
+    {
+      "epoch": 0.24668391215807145,
+      "grad_norm": 0.7924874424934387,
+      "learning_rate": 9.643526890606265e-06,
+      "loss": 0.793,
+      "step": 4482
+    },
+    {
+      "epoch": 0.2467389509604271,
+      "grad_norm": 0.7966769933700562,
+      "learning_rate": 9.643366135738951e-06,
+      "loss": 0.8042,
+      "step": 4483
+    },
+    {
+      "epoch": 0.24679398976278277,
+      "grad_norm": 0.911756694316864,
+      "learning_rate": 9.643205345973343e-06,
+      "loss": 0.7801,
+      "step": 4484
+    },
+    {
+      "epoch": 0.24684902856513843,
+      "grad_norm": 0.903378963470459,
+      "learning_rate": 9.643044521310645e-06,
+      "loss": 0.7863,
+      "step": 4485
+    },
+    {
+      "epoch": 0.2469040673674941,
+      "grad_norm": 0.9021226167678833,
+      "learning_rate": 9.642883661752067e-06,
+      "loss": 0.8005,
+      "step": 4486
+    },
+    {
+      "epoch": 0.24695910616984976,
+      "grad_norm": 0.8853413462638855,
+      "learning_rate": 9.64272276729882e-06,
+      "loss": 0.8371,
+      "step": 4487
+    },
+    {
+      "epoch": 0.24701414497220542,
+      "grad_norm": 1.0654630661010742,
+      "learning_rate": 9.642561837952108e-06,
+      "loss": 0.92,
+      "step": 4488
+    },
+    {
+      "epoch": 0.24706918377456105,
+      "grad_norm": 0.8663573265075684,
+      "learning_rate": 9.642400873713146e-06,
+      "loss": 0.8066,
+      "step": 4489
+    },
+    {
+      "epoch": 0.2471242225769167,
+      "grad_norm": 0.7483134269714355,
+      "learning_rate": 9.642239874583143e-06,
+      "loss": 0.9013,
+      "step": 4490
+    },
+    {
+      "epoch": 0.24717926137927237,
+      "grad_norm": 0.7582293748855591,
+      "learning_rate": 9.642078840563306e-06,
+      "loss": 0.7795,
+      "step": 4491
+    },
+    {
+      "epoch": 0.24723430018162804,
+      "grad_norm": 0.8276637196540833,
+      "learning_rate": 9.641917771654848e-06,
+      "loss": 0.7756,
+      "step": 4492
+    },
+    {
+      "epoch": 0.2472893389839837,
+      "grad_norm": 0.697088360786438,
+      "learning_rate": 9.641756667858976e-06,
+      "loss": 0.7092,
+      "step": 4493
+    },
+    {
+      "epoch": 0.24734437778633936,
+      "grad_norm": 0.8960816860198975,
+      "learning_rate": 9.641595529176907e-06,
+      "loss": 0.8835,
+      "step": 4494
+    },
+    {
+      "epoch": 0.24739941658869502,
+      "grad_norm": 0.9210898280143738,
+      "learning_rate": 9.641434355609846e-06,
+      "loss": 0.7881,
+      "step": 4495
+    },
+    {
+      "epoch": 0.24745445539105068,
+      "grad_norm": 0.7205467820167542,
+      "learning_rate": 9.64127314715901e-06,
+      "loss": 0.7204,
+      "step": 4496
+    },
+    {
+      "epoch": 0.24750949419340634,
+      "grad_norm": 0.7313701510429382,
+      "learning_rate": 9.641111903825603e-06,
+      "loss": 0.8296,
+      "step": 4497
+    },
+    {
+      "epoch": 0.247564532995762,
+      "grad_norm": 0.771159827709198,
+      "learning_rate": 9.640950625610845e-06,
+      "loss": 0.7974,
+      "step": 4498
+    },
+    {
+      "epoch": 0.24761957179811767,
+      "grad_norm": 0.9227705597877502,
+      "learning_rate": 9.64078931251594e-06,
+      "loss": 0.9215,
+      "step": 4499
+    },
+    {
+      "epoch": 0.24767461060047333,
+      "grad_norm": 0.7569915652275085,
+      "learning_rate": 9.64062796454211e-06,
+      "loss": 0.83,
+      "step": 4500
+    },
+    {
+      "epoch": 0.247729649402829,
+      "grad_norm": 0.7453131675720215,
+      "learning_rate": 9.64046658169056e-06,
+      "loss": 0.6747,
+      "step": 4501
+    },
+    {
+      "epoch": 0.24778468820518465,
+      "grad_norm": 0.7228132486343384,
+      "learning_rate": 9.640305163962504e-06,
+      "loss": 0.7535,
+      "step": 4502
+    },
+    {
+      "epoch": 0.2478397270075403,
+      "grad_norm": 0.8160690069198608,
+      "learning_rate": 9.640143711359159e-06,
+      "loss": 0.8655,
+      "step": 4503
+    },
+    {
+      "epoch": 0.24789476580989597,
+      "grad_norm": 0.7641691565513611,
+      "learning_rate": 9.639982223881735e-06,
+      "loss": 0.8353,
+      "step": 4504
+    },
+    {
+      "epoch": 0.24794980461225163,
+      "grad_norm": 0.8669107556343079,
+      "learning_rate": 9.639820701531445e-06,
+      "loss": 0.8614,
+      "step": 4505
+    },
+    {
+      "epoch": 0.2480048434146073,
+      "grad_norm": 0.7433111667633057,
+      "learning_rate": 9.639659144309508e-06,
+      "loss": 0.6891,
+      "step": 4506
+    },
+    {
+      "epoch": 0.24805988221696296,
+      "grad_norm": 1.4303346872329712,
+      "learning_rate": 9.639497552217131e-06,
+      "loss": 0.8016,
+      "step": 4507
+    },
+    {
+      "epoch": 0.24811492101931862,
+      "grad_norm": 0.8684772253036499,
+      "learning_rate": 9.639335925255535e-06,
+      "loss": 0.8324,
+      "step": 4508
+    },
+    {
+      "epoch": 0.24816995982167428,
+      "grad_norm": 0.9222162365913391,
+      "learning_rate": 9.639174263425932e-06,
+      "loss": 0.8715,
+      "step": 4509
+    },
+    {
+      "epoch": 0.24822499862402994,
+      "grad_norm": 0.9789180755615234,
+      "learning_rate": 9.639012566729535e-06,
+      "loss": 0.823,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2482800374263856,
+      "grad_norm": 0.8475140333175659,
+      "learning_rate": 9.638850835167564e-06,
+      "loss": 0.768,
+      "step": 4511
+    },
+    {
+      "epoch": 0.24833507622874126,
+      "grad_norm": 0.7943722605705261,
+      "learning_rate": 9.63868906874123e-06,
+      "loss": 0.788,
+      "step": 4512
+    },
+    {
+      "epoch": 0.24839011503109693,
+      "grad_norm": 0.8723915815353394,
+      "learning_rate": 9.63852726745175e-06,
+      "loss": 0.7865,
+      "step": 4513
+    },
+    {
+      "epoch": 0.2484451538334526,
+      "grad_norm": 0.837001383304596,
+      "learning_rate": 9.638365431300342e-06,
+      "loss": 0.7799,
+      "step": 4514
+    },
+    {
+      "epoch": 0.24850019263580825,
+      "grad_norm": 0.7992665767669678,
+      "learning_rate": 9.638203560288222e-06,
+      "loss": 0.8951,
+      "step": 4515
+    },
+    {
+      "epoch": 0.2485552314381639,
+      "grad_norm": 0.8712993264198303,
+      "learning_rate": 9.638041654416603e-06,
+      "loss": 0.8157,
+      "step": 4516
+    },
+    {
+      "epoch": 0.24861027024051957,
+      "grad_norm": 0.7176356911659241,
+      "learning_rate": 9.637879713686706e-06,
+      "loss": 0.8197,
+      "step": 4517
+    },
+    {
+      "epoch": 0.24866530904287523,
+      "grad_norm": 0.7624368071556091,
+      "learning_rate": 9.637717738099747e-06,
+      "loss": 0.7545,
+      "step": 4518
+    },
+    {
+      "epoch": 0.2487203478452309,
+      "grad_norm": 0.857222318649292,
+      "learning_rate": 9.637555727656943e-06,
+      "loss": 0.8146,
+      "step": 4519
+    },
+    {
+      "epoch": 0.24877538664758655,
+      "grad_norm": 0.7461313605308533,
+      "learning_rate": 9.637393682359511e-06,
+      "loss": 0.8569,
+      "step": 4520
+    },
+    {
+      "epoch": 0.24883042544994222,
+      "grad_norm": 0.8491896986961365,
+      "learning_rate": 9.637231602208668e-06,
+      "loss": 0.863,
+      "step": 4521
+    },
+    {
+      "epoch": 0.24888546425229788,
+      "grad_norm": 0.8139386177062988,
+      "learning_rate": 9.637069487205635e-06,
+      "loss": 0.7105,
+      "step": 4522
+    },
+    {
+      "epoch": 0.24894050305465354,
+      "grad_norm": 0.7782894968986511,
+      "learning_rate": 9.636907337351629e-06,
+      "loss": 0.8044,
+      "step": 4523
+    },
+    {
+      "epoch": 0.2489955418570092,
+      "grad_norm": 0.8225486874580383,
+      "learning_rate": 9.636745152647868e-06,
+      "loss": 0.7877,
+      "step": 4524
+    },
+    {
+      "epoch": 0.24905058065936486,
+      "grad_norm": 0.9087927341461182,
+      "learning_rate": 9.636582933095573e-06,
+      "loss": 0.8017,
+      "step": 4525
+    },
+    {
+      "epoch": 0.24910561946172052,
+      "grad_norm": 0.7392508387565613,
+      "learning_rate": 9.636420678695962e-06,
+      "loss": 0.7953,
+      "step": 4526
+    },
+    {
+      "epoch": 0.24916065826407618,
+      "grad_norm": 0.7906273007392883,
+      "learning_rate": 9.636258389450253e-06,
+      "loss": 0.9491,
+      "step": 4527
+    },
+    {
+      "epoch": 0.24921569706643185,
+      "grad_norm": 0.840394139289856,
+      "learning_rate": 9.636096065359666e-06,
+      "loss": 0.8621,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2492707358687875,
+      "grad_norm": 0.7923862934112549,
+      "learning_rate": 9.635933706425424e-06,
+      "loss": 0.8215,
+      "step": 4529
+    },
+    {
+      "epoch": 0.24932577467114317,
+      "grad_norm": 0.8372805714607239,
+      "learning_rate": 9.635771312648744e-06,
+      "loss": 0.8845,
+      "step": 4530
+    },
+    {
+      "epoch": 0.24938081347349883,
+      "grad_norm": 0.7569165229797363,
+      "learning_rate": 9.635608884030848e-06,
+      "loss": 0.8406,
+      "step": 4531
+    },
+    {
+      "epoch": 0.24943585227585446,
+      "grad_norm": 0.8260865807533264,
+      "learning_rate": 9.635446420572956e-06,
+      "loss": 0.8418,
+      "step": 4532
+    },
+    {
+      "epoch": 0.24949089107821013,
+      "grad_norm": 0.6841318607330322,
+      "learning_rate": 9.635283922276291e-06,
+      "loss": 0.6732,
+      "step": 4533
+    },
+    {
+      "epoch": 0.2495459298805658,
+      "grad_norm": 0.7055326104164124,
+      "learning_rate": 9.635121389142072e-06,
+      "loss": 0.7702,
+      "step": 4534
+    },
+    {
+      "epoch": 0.24960096868292145,
+      "grad_norm": 0.7293457388877869,
+      "learning_rate": 9.63495882117152e-06,
+      "loss": 0.6836,
+      "step": 4535
+    },
+    {
+      "epoch": 0.2496560074852771,
+      "grad_norm": 0.7411924004554749,
+      "learning_rate": 9.63479621836586e-06,
+      "loss": 0.8686,
+      "step": 4536
+    },
+    {
+      "epoch": 0.24971104628763277,
+      "grad_norm": 0.7864643931388855,
+      "learning_rate": 9.634633580726313e-06,
+      "loss": 0.7801,
+      "step": 4537
+    },
+    {
+      "epoch": 0.24976608508998843,
+      "grad_norm": 0.9730797410011292,
+      "learning_rate": 9.634470908254099e-06,
+      "loss": 0.8362,
+      "step": 4538
+    },
+    {
+      "epoch": 0.2498211238923441,
+      "grad_norm": 0.8390370011329651,
+      "learning_rate": 9.634308200950442e-06,
+      "loss": 0.8079,
+      "step": 4539
+    },
+    {
+      "epoch": 0.24987616269469975,
+      "grad_norm": 0.8951246738433838,
+      "learning_rate": 9.634145458816566e-06,
+      "loss": 0.7662,
+      "step": 4540
+    },
+    {
+      "epoch": 0.24993120149705542,
+      "grad_norm": 0.7654157280921936,
+      "learning_rate": 9.633982681853693e-06,
+      "loss": 0.8699,
+      "step": 4541
+    },
+    {
+      "epoch": 0.24998624029941108,
+      "grad_norm": 0.8152109980583191,
+      "learning_rate": 9.633819870063046e-06,
+      "loss": 0.7875,
+      "step": 4542
+    },
+    {
+      "epoch": 0.25004127910176677,
+      "grad_norm": 0.9407321214675903,
+      "learning_rate": 9.63365702344585e-06,
+      "loss": 0.7708,
+      "step": 4543
+    },
+    {
+      "epoch": 0.2500963179041224,
+      "grad_norm": 0.8169927597045898,
+      "learning_rate": 9.633494142003327e-06,
+      "loss": 0.8078,
+      "step": 4544
+    },
+    {
+      "epoch": 0.2501513567064781,
+      "grad_norm": 0.7380755543708801,
+      "learning_rate": 9.633331225736704e-06,
+      "loss": 0.7818,
+      "step": 4545
+    },
+    {
+      "epoch": 0.2502063955088337,
+      "grad_norm": 0.8124812841415405,
+      "learning_rate": 9.633168274647203e-06,
+      "loss": 0.8133,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2502614343111894,
+      "grad_norm": 0.8511367440223694,
+      "learning_rate": 9.63300528873605e-06,
+      "loss": 0.7747,
+      "step": 4547
+    },
+    {
+      "epoch": 0.25031647311354505,
+      "grad_norm": 0.7305121421813965,
+      "learning_rate": 9.632842268004469e-06,
+      "loss": 0.8479,
+      "step": 4548
+    },
+    {
+      "epoch": 0.25037151191590074,
+      "grad_norm": 0.7127692103385925,
+      "learning_rate": 9.632679212453686e-06,
+      "loss": 0.8514,
+      "step": 4549
+    },
+    {
+      "epoch": 0.25042655071825637,
+      "grad_norm": 0.8251872062683105,
+      "learning_rate": 9.632516122084926e-06,
+      "loss": 0.7686,
+      "step": 4550
+    },
+    {
+      "epoch": 0.25048158952061206,
+      "grad_norm": 0.6756613850593567,
+      "learning_rate": 9.632352996899413e-06,
+      "loss": 0.5959,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2505366283229677,
+      "grad_norm": 0.9266120791435242,
+      "learning_rate": 9.632189836898377e-06,
+      "loss": 0.7889,
+      "step": 4552
+    },
+    {
+      "epoch": 0.2505916671253233,
+      "grad_norm": 0.769890546798706,
+      "learning_rate": 9.63202664208304e-06,
+      "loss": 0.7864,
+      "step": 4553
+    },
+    {
+      "epoch": 0.250646705927679,
+      "grad_norm": 0.7314025163650513,
+      "learning_rate": 9.631863412454634e-06,
+      "loss": 0.8088,
+      "step": 4554
+    },
+    {
+      "epoch": 0.25070174473003465,
+      "grad_norm": 0.818317711353302,
+      "learning_rate": 9.63170014801438e-06,
+      "loss": 0.7096,
+      "step": 4555
+    },
+    {
+      "epoch": 0.25075678353239034,
+      "grad_norm": 0.7538807392120361,
+      "learning_rate": 9.631536848763508e-06,
+      "loss": 0.7779,
+      "step": 4556
+    },
+    {
+      "epoch": 0.25081182233474597,
+      "grad_norm": 0.7658100128173828,
+      "learning_rate": 9.631373514703247e-06,
+      "loss": 0.8535,
+      "step": 4557
+    },
+    {
+      "epoch": 0.25086686113710166,
+      "grad_norm": 0.8019290566444397,
+      "learning_rate": 9.631210145834819e-06,
+      "loss": 0.8141,
+      "step": 4558
+    },
+    {
+      "epoch": 0.2509218999394573,
+      "grad_norm": 0.7257653474807739,
+      "learning_rate": 9.631046742159456e-06,
+      "loss": 0.7451,
+      "step": 4559
+    },
+    {
+      "epoch": 0.250976938741813,
+      "grad_norm": 0.7546024918556213,
+      "learning_rate": 9.630883303678386e-06,
+      "loss": 0.7707,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2510319775441686,
+      "grad_norm": 0.7288938760757446,
+      "learning_rate": 9.630719830392835e-06,
+      "loss": 0.7362,
+      "step": 4561
+    },
+    {
+      "epoch": 0.2510870163465243,
+      "grad_norm": 0.7814223170280457,
+      "learning_rate": 9.630556322304036e-06,
+      "loss": 0.8514,
+      "step": 4562
+    },
+    {
+      "epoch": 0.25114205514887994,
+      "grad_norm": 0.7561381459236145,
+      "learning_rate": 9.630392779413214e-06,
+      "loss": 0.7659,
+      "step": 4563
+    },
+    {
+      "epoch": 0.25119709395123563,
+      "grad_norm": 0.750641942024231,
+      "learning_rate": 9.6302292017216e-06,
+      "loss": 0.8496,
+      "step": 4564
+    },
+    {
+      "epoch": 0.25125213275359126,
+      "grad_norm": 0.832155704498291,
+      "learning_rate": 9.630065589230422e-06,
+      "loss": 0.7778,
+      "step": 4565
+    },
+    {
+      "epoch": 0.25130717155594695,
+      "grad_norm": 0.8202440142631531,
+      "learning_rate": 9.62990194194091e-06,
+      "loss": 0.8962,
+      "step": 4566
+    },
+    {
+      "epoch": 0.2513622103583026,
+      "grad_norm": 0.8777977824211121,
+      "learning_rate": 9.629738259854295e-06,
+      "loss": 0.7215,
+      "step": 4567
+    },
+    {
+      "epoch": 0.2514172491606583,
+      "grad_norm": 1.1868599653244019,
+      "learning_rate": 9.629574542971806e-06,
+      "loss": 0.8238,
+      "step": 4568
+    },
+    {
+      "epoch": 0.2514722879630139,
+      "grad_norm": 0.9128753542900085,
+      "learning_rate": 9.629410791294675e-06,
+      "loss": 0.7638,
+      "step": 4569
+    },
+    {
+      "epoch": 0.2515273267653696,
+      "grad_norm": 0.7350082993507385,
+      "learning_rate": 9.629247004824132e-06,
+      "loss": 0.8041,
+      "step": 4570
+    },
+    {
+      "epoch": 0.25158236556772523,
+      "grad_norm": 0.7279660701751709,
+      "learning_rate": 9.629083183561407e-06,
+      "loss": 0.7377,
+      "step": 4571
+    },
+    {
+      "epoch": 0.2516374043700809,
+      "grad_norm": 0.8570461273193359,
+      "learning_rate": 9.628919327507732e-06,
+      "loss": 0.8106,
+      "step": 4572
+    },
+    {
+      "epoch": 0.25169244317243655,
+      "grad_norm": 0.8998312950134277,
+      "learning_rate": 9.62875543666434e-06,
+      "loss": 0.8171,
+      "step": 4573
+    },
+    {
+      "epoch": 0.25174748197479224,
+      "grad_norm": 0.7631624937057495,
+      "learning_rate": 9.628591511032456e-06,
+      "loss": 0.7871,
+      "step": 4574
+    },
+    {
+      "epoch": 0.2518025207771479,
+      "grad_norm": 0.7752320766448975,
+      "learning_rate": 9.628427550613322e-06,
+      "loss": 0.8241,
+      "step": 4575
+    },
+    {
+      "epoch": 0.25185755957950356,
+      "grad_norm": 0.8741563558578491,
+      "learning_rate": 9.628263555408163e-06,
+      "loss": 0.7312,
+      "step": 4576
+    },
+    {
+      "epoch": 0.2519125983818592,
+      "grad_norm": 0.8615008592605591,
+      "learning_rate": 9.628099525418216e-06,
+      "loss": 0.8586,
+      "step": 4577
+    },
+    {
+      "epoch": 0.2519676371842149,
+      "grad_norm": 0.8273662328720093,
+      "learning_rate": 9.62793546064471e-06,
+      "loss": 0.7838,
+      "step": 4578
+    },
+    {
+      "epoch": 0.2520226759865705,
+      "grad_norm": 0.7454090118408203,
+      "learning_rate": 9.627771361088882e-06,
+      "loss": 0.8461,
+      "step": 4579
+    },
+    {
+      "epoch": 0.2520777147889262,
+      "grad_norm": 0.8225379586219788,
+      "learning_rate": 9.627607226751962e-06,
+      "loss": 0.7792,
+      "step": 4580
+    },
+    {
+      "epoch": 0.25213275359128184,
+      "grad_norm": 0.8655416369438171,
+      "learning_rate": 9.627443057635184e-06,
+      "loss": 0.8165,
+      "step": 4581
+    },
+    {
+      "epoch": 0.25218779239363753,
+      "grad_norm": 0.7735984921455383,
+      "learning_rate": 9.627278853739783e-06,
+      "loss": 0.8208,
+      "step": 4582
+    },
+    {
+      "epoch": 0.25224283119599317,
+      "grad_norm": 0.8293350338935852,
+      "learning_rate": 9.627114615066994e-06,
+      "loss": 0.7394,
+      "step": 4583
+    },
+    {
+      "epoch": 0.25229786999834886,
+      "grad_norm": 0.7840214371681213,
+      "learning_rate": 9.626950341618048e-06,
+      "loss": 0.8522,
+      "step": 4584
+    },
+    {
+      "epoch": 0.2523529088007045,
+      "grad_norm": 0.7724186182022095,
+      "learning_rate": 9.626786033394185e-06,
+      "loss": 0.8175,
+      "step": 4585
+    },
+    {
+      "epoch": 0.2524079476030602,
+      "grad_norm": 1.0751588344573975,
+      "learning_rate": 9.626621690396634e-06,
+      "loss": 0.9229,
+      "step": 4586
+    },
+    {
+      "epoch": 0.2524629864054158,
+      "grad_norm": 0.7016913294792175,
+      "learning_rate": 9.626457312626634e-06,
+      "loss": 0.6883,
+      "step": 4587
+    },
+    {
+      "epoch": 0.2525180252077715,
+      "grad_norm": 0.918377697467804,
+      "learning_rate": 9.626292900085419e-06,
+      "loss": 0.7889,
+      "step": 4588
+    },
+    {
+      "epoch": 0.25257306401012714,
+      "grad_norm": 1.006564736366272,
+      "learning_rate": 9.626128452774226e-06,
+      "loss": 0.7888,
+      "step": 4589
+    },
+    {
+      "epoch": 0.2526281028124828,
+      "grad_norm": 1.0214998722076416,
+      "learning_rate": 9.625963970694287e-06,
+      "loss": 0.768,
+      "step": 4590
+    },
+    {
+      "epoch": 0.25268314161483846,
+      "grad_norm": 0.7980843186378479,
+      "learning_rate": 9.625799453846844e-06,
+      "loss": 0.8662,
+      "step": 4591
+    },
+    {
+      "epoch": 0.25273818041719415,
+      "grad_norm": 0.734582245349884,
+      "learning_rate": 9.625634902233128e-06,
+      "loss": 0.759,
+      "step": 4592
+    },
+    {
+      "epoch": 0.2527932192195498,
+      "grad_norm": 0.7185904383659363,
+      "learning_rate": 9.62547031585438e-06,
+      "loss": 0.774,
+      "step": 4593
+    },
+    {
+      "epoch": 0.25284825802190547,
+      "grad_norm": 0.7356622219085693,
+      "learning_rate": 9.625305694711835e-06,
+      "loss": 0.7435,
+      "step": 4594
+    },
+    {
+      "epoch": 0.2529032968242611,
+      "grad_norm": 0.7589355707168579,
+      "learning_rate": 9.62514103880673e-06,
+      "loss": 0.807,
+      "step": 4595
+    },
+    {
+      "epoch": 0.25295833562661674,
+      "grad_norm": 0.889228880405426,
+      "learning_rate": 9.624976348140305e-06,
+      "loss": 0.8609,
+      "step": 4596
+    },
+    {
+      "epoch": 0.2530133744289724,
+      "grad_norm": 0.7546125650405884,
+      "learning_rate": 9.624811622713793e-06,
+      "loss": 0.8379,
+      "step": 4597
+    },
+    {
+      "epoch": 0.25306841323132806,
+      "grad_norm": 0.8262770175933838,
+      "learning_rate": 9.624646862528436e-06,
+      "loss": 0.7611,
+      "step": 4598
+    },
+    {
+      "epoch": 0.25312345203368375,
+      "grad_norm": 0.8876076936721802,
+      "learning_rate": 9.624482067585472e-06,
+      "loss": 0.8106,
+      "step": 4599
+    },
+    {
+      "epoch": 0.2531784908360394,
+      "grad_norm": 0.7045544981956482,
+      "learning_rate": 9.624317237886137e-06,
+      "loss": 0.7121,
+      "step": 4600
+    },
+    {
+      "epoch": 0.25323352963839507,
+      "grad_norm": 0.7693355083465576,
+      "learning_rate": 9.624152373431672e-06,
+      "loss": 0.8052,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2532885684407507,
+      "grad_norm": 0.8072683811187744,
+      "learning_rate": 9.623987474223316e-06,
+      "loss": 0.8543,
+      "step": 4602
+    },
+    {
+      "epoch": 0.2533436072431064,
+      "grad_norm": 0.8158687949180603,
+      "learning_rate": 9.62382254026231e-06,
+      "loss": 0.6922,
+      "step": 4603
+    },
+    {
+      "epoch": 0.25339864604546203,
+      "grad_norm": 0.7688641548156738,
+      "learning_rate": 9.623657571549887e-06,
+      "loss": 0.7198,
+      "step": 4604
+    },
+    {
+      "epoch": 0.2534536848478177,
+      "grad_norm": 0.7806578278541565,
+      "learning_rate": 9.623492568087293e-06,
+      "loss": 0.8539,
+      "step": 4605
+    },
+    {
+      "epoch": 0.25350872365017335,
+      "grad_norm": 0.9557347893714905,
+      "learning_rate": 9.623327529875769e-06,
+      "loss": 0.6996,
+      "step": 4606
+    },
+    {
+      "epoch": 0.25356376245252904,
+      "grad_norm": 0.9465067386627197,
+      "learning_rate": 9.62316245691655e-06,
+      "loss": 0.8756,
+      "step": 4607
+    },
+    {
+      "epoch": 0.2536188012548847,
+      "grad_norm": 0.8029165863990784,
+      "learning_rate": 9.62299734921088e-06,
+      "loss": 0.8573,
+      "step": 4608
+    },
+    {
+      "epoch": 0.25367384005724036,
+      "grad_norm": 0.7530128955841064,
+      "learning_rate": 9.62283220676e-06,
+      "loss": 0.7466,
+      "step": 4609
+    },
+    {
+      "epoch": 0.253728878859596,
+      "grad_norm": 0.6704453825950623,
+      "learning_rate": 9.622667029565151e-06,
+      "loss": 0.6512,
+      "step": 4610
+    },
+    {
+      "epoch": 0.2537839176619517,
+      "grad_norm": 0.7162728309631348,
+      "learning_rate": 9.622501817627574e-06,
+      "loss": 0.7615,
+      "step": 4611
+    },
+    {
+      "epoch": 0.2538389564643073,
+      "grad_norm": 0.7599188089370728,
+      "learning_rate": 9.622336570948509e-06,
+      "loss": 0.8463,
+      "step": 4612
+    },
+    {
+      "epoch": 0.253893995266663,
+      "grad_norm": 0.7922326922416687,
+      "learning_rate": 9.6221712895292e-06,
+      "loss": 0.9221,
+      "step": 4613
+    },
+    {
+      "epoch": 0.25394903406901864,
+      "grad_norm": 1.4635218381881714,
+      "learning_rate": 9.622005973370892e-06,
+      "loss": 0.9159,
+      "step": 4614
+    },
+    {
+      "epoch": 0.25400407287137433,
+      "grad_norm": 0.8695057034492493,
+      "learning_rate": 9.62184062247482e-06,
+      "loss": 0.6792,
+      "step": 4615
+    },
+    {
+      "epoch": 0.25405911167372996,
+      "grad_norm": 0.8070930242538452,
+      "learning_rate": 9.621675236842235e-06,
+      "loss": 0.8257,
+      "step": 4616
+    },
+    {
+      "epoch": 0.25411415047608565,
+      "grad_norm": 0.8642075061798096,
+      "learning_rate": 9.621509816474372e-06,
+      "loss": 0.8223,
+      "step": 4617
+    },
+    {
+      "epoch": 0.2541691892784413,
+      "grad_norm": 0.7131080031394958,
+      "learning_rate": 9.621344361372483e-06,
+      "loss": 0.6831,
+      "step": 4618
+    },
+    {
+      "epoch": 0.254224228080797,
+      "grad_norm": 0.7582216262817383,
+      "learning_rate": 9.621178871537804e-06,
+      "loss": 0.8091,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2542792668831526,
+      "grad_norm": 0.7705016732215881,
+      "learning_rate": 9.62101334697158e-06,
+      "loss": 0.7537,
+      "step": 4620
+    },
+    {
+      "epoch": 0.2543343056855083,
+      "grad_norm": 0.7638342976570129,
+      "learning_rate": 9.62084778767506e-06,
+      "loss": 0.7661,
+      "step": 4621
+    },
+    {
+      "epoch": 0.25438934448786393,
+      "grad_norm": 0.9296607971191406,
+      "learning_rate": 9.620682193649482e-06,
+      "loss": 0.8875,
+      "step": 4622
+    },
+    {
+      "epoch": 0.2544443832902196,
+      "grad_norm": 0.795394778251648,
+      "learning_rate": 9.620516564896096e-06,
+      "loss": 0.6884,
+      "step": 4623
+    },
+    {
+      "epoch": 0.25449942209257526,
+      "grad_norm": 0.9164957404136658,
+      "learning_rate": 9.620350901416142e-06,
+      "loss": 0.8693,
+      "step": 4624
+    },
+    {
+      "epoch": 0.25455446089493095,
+      "grad_norm": 0.8306281566619873,
+      "learning_rate": 9.62018520321087e-06,
+      "loss": 0.8972,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2546094996972866,
+      "grad_norm": 0.778831422328949,
+      "learning_rate": 9.620019470281521e-06,
+      "loss": 0.7574,
+      "step": 4626
+    },
+    {
+      "epoch": 0.25466453849964227,
+      "grad_norm": 0.9326225519180298,
+      "learning_rate": 9.619853702629343e-06,
+      "loss": 0.7712,
+      "step": 4627
+    },
+    {
+      "epoch": 0.2547195773019979,
+      "grad_norm": 0.8772255182266235,
+      "learning_rate": 9.619687900255581e-06,
+      "loss": 0.8241,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2547746161043536,
+      "grad_norm": 0.8777550458908081,
+      "learning_rate": 9.619522063161482e-06,
+      "loss": 0.8724,
+      "step": 4629
+    },
+    {
+      "epoch": 0.2548296549067092,
+      "grad_norm": 0.8332602381706238,
+      "learning_rate": 9.61935619134829e-06,
+      "loss": 0.8716,
+      "step": 4630
+    },
+    {
+      "epoch": 0.2548846937090649,
+      "grad_norm": 0.8246355056762695,
+      "learning_rate": 9.619190284817255e-06,
+      "loss": 0.7789,
+      "step": 4631
+    },
+    {
+      "epoch": 0.25493973251142055,
+      "grad_norm": 0.7200644612312317,
+      "learning_rate": 9.61902434356962e-06,
+      "loss": 0.7956,
+      "step": 4632
+    },
+    {
+      "epoch": 0.25499477131377624,
+      "grad_norm": 0.827756404876709,
+      "learning_rate": 9.618858367606638e-06,
+      "loss": 0.7925,
+      "step": 4633
+    },
+    {
+      "epoch": 0.25504981011613187,
+      "grad_norm": 0.7749341726303101,
+      "learning_rate": 9.618692356929551e-06,
+      "loss": 0.8706,
+      "step": 4634
+    },
+    {
+      "epoch": 0.25510484891848756,
+      "grad_norm": 0.7233432531356812,
+      "learning_rate": 9.618526311539608e-06,
+      "loss": 0.7725,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2551598877208432,
+      "grad_norm": 0.846340537071228,
+      "learning_rate": 9.618360231438058e-06,
+      "loss": 0.8758,
+      "step": 4636
+    },
+    {
+      "epoch": 0.2552149265231989,
+      "grad_norm": 0.8262908458709717,
+      "learning_rate": 9.61819411662615e-06,
+      "loss": 0.7758,
+      "step": 4637
+    },
+    {
+      "epoch": 0.2552699653255545,
+      "grad_norm": 0.7829110026359558,
+      "learning_rate": 9.61802796710513e-06,
+      "loss": 0.8494,
+      "step": 4638
+    },
+    {
+      "epoch": 0.25532500412791015,
+      "grad_norm": 0.7480815649032593,
+      "learning_rate": 9.617861782876247e-06,
+      "loss": 0.7639,
+      "step": 4639
+    },
+    {
+      "epoch": 0.25538004293026584,
+      "grad_norm": 0.8782994747161865,
+      "learning_rate": 9.617695563940752e-06,
+      "loss": 0.9651,
+      "step": 4640
+    },
+    {
+      "epoch": 0.25543508173262147,
+      "grad_norm": 0.7215868234634399,
+      "learning_rate": 9.617529310299895e-06,
+      "loss": 0.7833,
+      "step": 4641
+    },
+    {
+      "epoch": 0.25549012053497716,
+      "grad_norm": 0.8287535905838013,
+      "learning_rate": 9.617363021954922e-06,
+      "loss": 0.901,
+      "step": 4642
+    },
+    {
+      "epoch": 0.2555451593373328,
+      "grad_norm": 0.7679935097694397,
+      "learning_rate": 9.617196698907084e-06,
+      "loss": 0.761,
+      "step": 4643
+    },
+    {
+      "epoch": 0.2556001981396885,
+      "grad_norm": 0.7765942811965942,
+      "learning_rate": 9.617030341157632e-06,
+      "loss": 0.7356,
+      "step": 4644
+    },
+    {
+      "epoch": 0.2556552369420441,
+      "grad_norm": 0.6964583396911621,
+      "learning_rate": 9.616863948707816e-06,
+      "loss": 0.7683,
+      "step": 4645
+    },
+    {
+      "epoch": 0.2557102757443998,
+      "grad_norm": 0.8031953573226929,
+      "learning_rate": 9.616697521558886e-06,
+      "loss": 0.7875,
+      "step": 4646
+    },
+    {
+      "epoch": 0.25576531454675544,
+      "grad_norm": 0.7155965566635132,
+      "learning_rate": 9.616531059712094e-06,
+      "loss": 0.6516,
+      "step": 4647
+    },
+    {
+      "epoch": 0.25582035334911113,
+      "grad_norm": 0.6870070099830627,
+      "learning_rate": 9.61636456316869e-06,
+      "loss": 0.7217,
+      "step": 4648
+    },
+    {
+      "epoch": 0.25587539215146676,
+      "grad_norm": 0.7686315774917603,
+      "learning_rate": 9.616198031929926e-06,
+      "loss": 0.8136,
+      "step": 4649
+    },
+    {
+      "epoch": 0.25593043095382245,
+      "grad_norm": 0.7532772421836853,
+      "learning_rate": 9.616031465997054e-06,
+      "loss": 0.696,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2559854697561781,
+      "grad_norm": 0.8111574053764343,
+      "learning_rate": 9.615864865371323e-06,
+      "loss": 0.8501,
+      "step": 4651
+    },
+    {
+      "epoch": 0.2560405085585338,
+      "grad_norm": 0.771065890789032,
+      "learning_rate": 9.615698230053989e-06,
+      "loss": 0.7417,
+      "step": 4652
+    },
+    {
+      "epoch": 0.2560955473608894,
+      "grad_norm": 0.7468003034591675,
+      "learning_rate": 9.6155315600463e-06,
+      "loss": 0.7303,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2561505861632451,
+      "grad_norm": 0.8041057586669922,
+      "learning_rate": 9.615364855349514e-06,
+      "loss": 0.8689,
+      "step": 4654
+    },
+    {
+      "epoch": 0.25620562496560073,
+      "grad_norm": 0.8439033627510071,
+      "learning_rate": 9.61519811596488e-06,
+      "loss": 0.8654,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2562606637679564,
+      "grad_norm": 0.7768430113792419,
+      "learning_rate": 9.615031341893653e-06,
+      "loss": 0.8789,
+      "step": 4656
+    },
+    {
+      "epoch": 0.25631570257031205,
+      "grad_norm": 0.712876558303833,
+      "learning_rate": 9.614864533137086e-06,
+      "loss": 0.7497,
+      "step": 4657
+    },
+    {
+      "epoch": 0.25637074137266774,
+      "grad_norm": 0.7586949467658997,
+      "learning_rate": 9.614697689696431e-06,
+      "loss": 0.81,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2564257801750234,
+      "grad_norm": 0.717078447341919,
+      "learning_rate": 9.614530811572946e-06,
+      "loss": 0.8023,
+      "step": 4659
+    },
+    {
+      "epoch": 0.25648081897737907,
+      "grad_norm": 0.7369407415390015,
+      "learning_rate": 9.61436389876788e-06,
+      "loss": 0.784,
+      "step": 4660
+    },
+    {
+      "epoch": 0.2565358577797347,
+      "grad_norm": 0.7536265850067139,
+      "learning_rate": 9.61419695128249e-06,
+      "loss": 0.7687,
+      "step": 4661
+    },
+    {
+      "epoch": 0.2565908965820904,
+      "grad_norm": 0.9718124866485596,
+      "learning_rate": 9.614029969118033e-06,
+      "loss": 0.8495,
+      "step": 4662
+    },
+    {
+      "epoch": 0.256645935384446,
+      "grad_norm": 1.1578630208969116,
+      "learning_rate": 9.613862952275762e-06,
+      "loss": 0.9189,
+      "step": 4663
+    },
+    {
+      "epoch": 0.2567009741868017,
+      "grad_norm": 0.7752498984336853,
+      "learning_rate": 9.613695900756929e-06,
+      "loss": 0.7677,
+      "step": 4664
+    },
+    {
+      "epoch": 0.25675601298915735,
+      "grad_norm": 0.9640393257141113,
+      "learning_rate": 9.613528814562795e-06,
+      "loss": 0.719,
+      "step": 4665
+    },
+    {
+      "epoch": 0.25681105179151303,
+      "grad_norm": 0.7690972089767456,
+      "learning_rate": 9.613361693694614e-06,
+      "loss": 0.7977,
+      "step": 4666
+    },
+    {
+      "epoch": 0.25686609059386867,
+      "grad_norm": 0.8390190601348877,
+      "learning_rate": 9.61319453815364e-06,
+      "loss": 0.8032,
+      "step": 4667
+    },
+    {
+      "epoch": 0.25692112939622436,
+      "grad_norm": 0.8293220400810242,
+      "learning_rate": 9.613027347941131e-06,
+      "loss": 0.8645,
+      "step": 4668
+    },
+    {
+      "epoch": 0.25697616819858,
+      "grad_norm": 0.8020731210708618,
+      "learning_rate": 9.612860123058344e-06,
+      "loss": 0.8374,
+      "step": 4669
+    },
+    {
+      "epoch": 0.2570312070009357,
+      "grad_norm": 0.7756736278533936,
+      "learning_rate": 9.612692863506534e-06,
+      "loss": 0.7318,
+      "step": 4670
+    },
+    {
+      "epoch": 0.2570862458032913,
+      "grad_norm": 0.895416259765625,
+      "learning_rate": 9.61252556928696e-06,
+      "loss": 0.9654,
+      "step": 4671
+    },
+    {
+      "epoch": 0.257141284605647,
+      "grad_norm": 0.8647375106811523,
+      "learning_rate": 9.61235824040088e-06,
+      "loss": 0.7411,
+      "step": 4672
+    },
+    {
+      "epoch": 0.25719632340800264,
+      "grad_norm": 0.6927250623703003,
+      "learning_rate": 9.612190876849546e-06,
+      "loss": 0.7558,
+      "step": 4673
+    },
+    {
+      "epoch": 0.2572513622103583,
+      "grad_norm": 0.7614898085594177,
+      "learning_rate": 9.612023478634222e-06,
+      "loss": 0.7696,
+      "step": 4674
+    },
+    {
+      "epoch": 0.25730640101271396,
+      "grad_norm": 0.7910586595535278,
+      "learning_rate": 9.611856045756166e-06,
+      "loss": 0.8207,
+      "step": 4675
+    },
+    {
+      "epoch": 0.25736143981506965,
+      "grad_norm": 0.7330125570297241,
+      "learning_rate": 9.611688578216632e-06,
+      "loss": 0.8615,
+      "step": 4676
+    },
+    {
+      "epoch": 0.2574164786174253,
+      "grad_norm": 0.7703417539596558,
+      "learning_rate": 9.611521076016882e-06,
+      "loss": 0.8321,
+      "step": 4677
+    },
+    {
+      "epoch": 0.25747151741978097,
+      "grad_norm": 0.7121796607971191,
+      "learning_rate": 9.611353539158174e-06,
+      "loss": 0.8228,
+      "step": 4678
+    },
+    {
+      "epoch": 0.2575265562221366,
+      "grad_norm": 0.8313117027282715,
+      "learning_rate": 9.611185967641768e-06,
+      "loss": 0.9012,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2575815950244923,
+      "grad_norm": 0.806776225566864,
+      "learning_rate": 9.61101836146892e-06,
+      "loss": 0.769,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2576366338268479,
+      "grad_norm": 0.7049515843391418,
+      "learning_rate": 9.610850720640894e-06,
+      "loss": 0.7938,
+      "step": 4681
+    },
+    {
+      "epoch": 0.25769167262920356,
+      "grad_norm": 0.7286638021469116,
+      "learning_rate": 9.610683045158948e-06,
+      "loss": 0.8168,
+      "step": 4682
+    },
+    {
+      "epoch": 0.25774671143155925,
+      "grad_norm": 0.7916898727416992,
+      "learning_rate": 9.610515335024345e-06,
+      "loss": 0.7681,
+      "step": 4683
+    },
+    {
+      "epoch": 0.2578017502339149,
+      "grad_norm": 0.7649673819541931,
+      "learning_rate": 9.61034759023834e-06,
+      "loss": 0.7273,
+      "step": 4684
+    },
+    {
+      "epoch": 0.2578567890362706,
+      "grad_norm": 0.8280686736106873,
+      "learning_rate": 9.610179810802196e-06,
+      "loss": 0.7968,
+      "step": 4685
+    },
+    {
+      "epoch": 0.2579118278386262,
+      "grad_norm": 0.7206569910049438,
+      "learning_rate": 9.610011996717175e-06,
+      "loss": 0.7359,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2579668666409819,
+      "grad_norm": 0.7365424036979675,
+      "learning_rate": 9.60984414798454e-06,
+      "loss": 0.7962,
+      "step": 4687
+    },
+    {
+      "epoch": 0.25802190544333753,
+      "grad_norm": 0.8030344247817993,
+      "learning_rate": 9.609676264605549e-06,
+      "loss": 0.7931,
+      "step": 4688
+    },
+    {
+      "epoch": 0.2580769442456932,
+      "grad_norm": 0.8812693357467651,
+      "learning_rate": 9.609508346581464e-06,
+      "loss": 0.8493,
+      "step": 4689
+    },
+    {
+      "epoch": 0.25813198304804885,
+      "grad_norm": 0.8026734590530396,
+      "learning_rate": 9.60934039391355e-06,
+      "loss": 0.8368,
+      "step": 4690
+    },
+    {
+      "epoch": 0.25818702185040454,
+      "grad_norm": 0.8270768523216248,
+      "learning_rate": 9.609172406603067e-06,
+      "loss": 0.9077,
+      "step": 4691
+    },
+    {
+      "epoch": 0.2582420606527602,
+      "grad_norm": 0.7362856864929199,
+      "learning_rate": 9.609004384651276e-06,
+      "loss": 0.7384,
+      "step": 4692
+    },
+    {
+      "epoch": 0.25829709945511586,
+      "grad_norm": 0.7195929288864136,
+      "learning_rate": 9.608836328059444e-06,
+      "loss": 0.8475,
+      "step": 4693
+    },
+    {
+      "epoch": 0.2583521382574715,
+      "grad_norm": 0.7653167843818665,
+      "learning_rate": 9.60866823682883e-06,
+      "loss": 0.7704,
+      "step": 4694
+    },
+    {
+      "epoch": 0.2584071770598272,
+      "grad_norm": 0.7056792974472046,
+      "learning_rate": 9.6085001109607e-06,
+      "loss": 0.7835,
+      "step": 4695
+    },
+    {
+      "epoch": 0.2584622158621828,
+      "grad_norm": 0.7299804091453552,
+      "learning_rate": 9.60833195045632e-06,
+      "loss": 0.7894,
+      "step": 4696
+    },
+    {
+      "epoch": 0.2585172546645385,
+      "grad_norm": 0.7235645055770874,
+      "learning_rate": 9.608163755316948e-06,
+      "loss": 0.8113,
+      "step": 4697
+    },
+    {
+      "epoch": 0.25857229346689414,
+      "grad_norm": 0.7066782116889954,
+      "learning_rate": 9.60799552554385e-06,
+      "loss": 0.739,
+      "step": 4698
+    },
+    {
+      "epoch": 0.25862733226924983,
+      "grad_norm": 0.769930362701416,
+      "learning_rate": 9.607827261138291e-06,
+      "loss": 0.7565,
+      "step": 4699
+    },
+    {
+      "epoch": 0.25868237107160547,
+      "grad_norm": 0.8875935077667236,
+      "learning_rate": 9.607658962101538e-06,
+      "loss": 0.849,
+      "step": 4700
+    },
+    {
+      "epoch": 0.25873740987396115,
+      "grad_norm": 0.7887380123138428,
+      "learning_rate": 9.60749062843485e-06,
+      "loss": 0.8795,
+      "step": 4701
+    },
+    {
+      "epoch": 0.2587924486763168,
+      "grad_norm": 0.7600420117378235,
+      "learning_rate": 9.607322260139499e-06,
+      "loss": 0.7581,
+      "step": 4702
+    },
+    {
+      "epoch": 0.2588474874786725,
+      "grad_norm": 0.7431491017341614,
+      "learning_rate": 9.607153857216746e-06,
+      "loss": 0.7119,
+      "step": 4703
+    },
+    {
+      "epoch": 0.2589025262810281,
+      "grad_norm": 0.7444193363189697,
+      "learning_rate": 9.606985419667858e-06,
+      "loss": 0.7492,
+      "step": 4704
+    },
+    {
+      "epoch": 0.2589575650833838,
+      "grad_norm": 0.8348917365074158,
+      "learning_rate": 9.6068169474941e-06,
+      "loss": 0.7656,
+      "step": 4705
+    },
+    {
+      "epoch": 0.25901260388573943,
+      "grad_norm": 0.6790240406990051,
+      "learning_rate": 9.60664844069674e-06,
+      "loss": 0.6354,
+      "step": 4706
+    },
+    {
+      "epoch": 0.2590676426880951,
+      "grad_norm": 0.8425769805908203,
+      "learning_rate": 9.606479899277044e-06,
+      "loss": 0.7927,
+      "step": 4707
+    },
+    {
+      "epoch": 0.25912268149045076,
+      "grad_norm": 0.7234740853309631,
+      "learning_rate": 9.606311323236277e-06,
+      "loss": 0.8122,
+      "step": 4708
+    },
+    {
+      "epoch": 0.25917772029280645,
+      "grad_norm": 0.839507520198822,
+      "learning_rate": 9.606142712575707e-06,
+      "loss": 0.8807,
+      "step": 4709
+    },
+    {
+      "epoch": 0.2592327590951621,
+      "grad_norm": 0.7155291438102722,
+      "learning_rate": 9.605974067296601e-06,
+      "loss": 0.7852,
+      "step": 4710
+    },
+    {
+      "epoch": 0.25928779789751777,
+      "grad_norm": 0.7222152352333069,
+      "learning_rate": 9.605805387400228e-06,
+      "loss": 0.7362,
+      "step": 4711
+    },
+    {
+      "epoch": 0.2593428366998734,
+      "grad_norm": 0.8350114226341248,
+      "learning_rate": 9.605636672887854e-06,
+      "loss": 0.7201,
+      "step": 4712
+    },
+    {
+      "epoch": 0.2593978755022291,
+      "grad_norm": 0.6805943250656128,
+      "learning_rate": 9.605467923760747e-06,
+      "loss": 0.6936,
+      "step": 4713
+    },
+    {
+      "epoch": 0.2594529143045847,
+      "grad_norm": 0.7863980531692505,
+      "learning_rate": 9.605299140020177e-06,
+      "loss": 0.9079,
+      "step": 4714
+    },
+    {
+      "epoch": 0.2595079531069404,
+      "grad_norm": 0.838843584060669,
+      "learning_rate": 9.60513032166741e-06,
+      "loss": 0.839,
+      "step": 4715
+    },
+    {
+      "epoch": 0.25956299190929605,
+      "grad_norm": 0.7872797250747681,
+      "learning_rate": 9.60496146870372e-06,
+      "loss": 0.9164,
+      "step": 4716
+    },
+    {
+      "epoch": 0.25961803071165174,
+      "grad_norm": 0.7300794720649719,
+      "learning_rate": 9.604792581130369e-06,
+      "loss": 0.8227,
+      "step": 4717
+    },
+    {
+      "epoch": 0.25967306951400737,
+      "grad_norm": 0.8420879244804382,
+      "learning_rate": 9.60462365894863e-06,
+      "loss": 0.7865,
+      "step": 4718
+    },
+    {
+      "epoch": 0.25972810831636306,
+      "grad_norm": 0.807697057723999,
+      "learning_rate": 9.604454702159771e-06,
+      "loss": 0.9081,
+      "step": 4719
+    },
+    {
+      "epoch": 0.2597831471187187,
+      "grad_norm": 0.9041245579719543,
+      "learning_rate": 9.604285710765064e-06,
+      "loss": 0.8102,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2598381859210744,
+      "grad_norm": 0.7061690092086792,
+      "learning_rate": 9.604116684765779e-06,
+      "loss": 0.762,
+      "step": 4721
+    },
+    {
+      "epoch": 0.25989322472343,
+      "grad_norm": 0.7790346741676331,
+      "learning_rate": 9.603947624163186e-06,
+      "loss": 0.8038,
+      "step": 4722
+    },
+    {
+      "epoch": 0.2599482635257857,
+      "grad_norm": 0.8109704256057739,
+      "learning_rate": 9.603778528958553e-06,
+      "loss": 0.9105,
+      "step": 4723
+    },
+    {
+      "epoch": 0.26000330232814134,
+      "grad_norm": 0.7396997213363647,
+      "learning_rate": 9.603609399153153e-06,
+      "loss": 0.8384,
+      "step": 4724
+    },
+    {
+      "epoch": 0.260058341130497,
+      "grad_norm": 0.8594317436218262,
+      "learning_rate": 9.603440234748257e-06,
+      "loss": 0.8301,
+      "step": 4725
+    },
+    {
+      "epoch": 0.26011337993285266,
+      "grad_norm": 0.7087241411209106,
+      "learning_rate": 9.603271035745138e-06,
+      "loss": 0.6652,
+      "step": 4726
+    },
+    {
+      "epoch": 0.2601684187352083,
+      "grad_norm": 0.7405440211296082,
+      "learning_rate": 9.603101802145065e-06,
+      "loss": 0.7804,
+      "step": 4727
+    },
+    {
+      "epoch": 0.260223457537564,
+      "grad_norm": 0.8637508749961853,
+      "learning_rate": 9.602932533949312e-06,
+      "loss": 0.8509,
+      "step": 4728
+    },
+    {
+      "epoch": 0.2602784963399196,
+      "grad_norm": 0.7040451765060425,
+      "learning_rate": 9.60276323115915e-06,
+      "loss": 0.7842,
+      "step": 4729
+    },
+    {
+      "epoch": 0.2603335351422753,
+      "grad_norm": 0.7743955254554749,
+      "learning_rate": 9.602593893775852e-06,
+      "loss": 0.8492,
+      "step": 4730
+    },
+    {
+      "epoch": 0.26038857394463094,
+      "grad_norm": 0.7110480070114136,
+      "learning_rate": 9.602424521800688e-06,
+      "loss": 0.7227,
+      "step": 4731
+    },
+    {
+      "epoch": 0.26044361274698663,
+      "grad_norm": 1.0066583156585693,
+      "learning_rate": 9.602255115234936e-06,
+      "loss": 0.8825,
+      "step": 4732
+    },
+    {
+      "epoch": 0.26049865154934226,
+      "grad_norm": 0.7746492624282837,
+      "learning_rate": 9.602085674079864e-06,
+      "loss": 0.8316,
+      "step": 4733
+    },
+    {
+      "epoch": 0.26055369035169795,
+      "grad_norm": 0.7394356727600098,
+      "learning_rate": 9.60191619833675e-06,
+      "loss": 0.746,
+      "step": 4734
+    },
+    {
+      "epoch": 0.2606087291540536,
+      "grad_norm": 0.7140582203865051,
+      "learning_rate": 9.601746688006866e-06,
+      "loss": 0.7204,
+      "step": 4735
+    },
+    {
+      "epoch": 0.2606637679564093,
+      "grad_norm": 0.753399133682251,
+      "learning_rate": 9.601577143091483e-06,
+      "loss": 0.8157,
+      "step": 4736
+    },
+    {
+      "epoch": 0.2607188067587649,
+      "grad_norm": 0.674320638179779,
+      "learning_rate": 9.601407563591881e-06,
+      "loss": 0.7279,
+      "step": 4737
+    },
+    {
+      "epoch": 0.2607738455611206,
+      "grad_norm": 0.855944037437439,
+      "learning_rate": 9.60123794950933e-06,
+      "loss": 0.804,
+      "step": 4738
+    },
+    {
+      "epoch": 0.26082888436347623,
+      "grad_norm": 0.6833948493003845,
+      "learning_rate": 9.601068300845106e-06,
+      "loss": 0.701,
+      "step": 4739
+    },
+    {
+      "epoch": 0.2608839231658319,
+      "grad_norm": 0.8085536360740662,
+      "learning_rate": 9.600898617600485e-06,
+      "loss": 0.8435,
+      "step": 4740
+    },
+    {
+      "epoch": 0.26093896196818755,
+      "grad_norm": 0.752849817276001,
+      "learning_rate": 9.600728899776741e-06,
+      "loss": 0.7205,
+      "step": 4741
+    },
+    {
+      "epoch": 0.26099400077054324,
+      "grad_norm": 0.7320554852485657,
+      "learning_rate": 9.600559147375151e-06,
+      "loss": 0.7556,
+      "step": 4742
+    },
+    {
+      "epoch": 0.2610490395728989,
+      "grad_norm": 0.7789202928543091,
+      "learning_rate": 9.600389360396988e-06,
+      "loss": 0.8467,
+      "step": 4743
+    },
+    {
+      "epoch": 0.26110407837525457,
+      "grad_norm": 0.8480898141860962,
+      "learning_rate": 9.600219538843532e-06,
+      "loss": 0.7762,
+      "step": 4744
+    },
+    {
+      "epoch": 0.2611591171776102,
+      "grad_norm": 0.8382542133331299,
+      "learning_rate": 9.600049682716055e-06,
+      "loss": 0.9051,
+      "step": 4745
+    },
+    {
+      "epoch": 0.2612141559799659,
+      "grad_norm": 0.8319274187088013,
+      "learning_rate": 9.599879792015838e-06,
+      "loss": 0.8221,
+      "step": 4746
+    },
+    {
+      "epoch": 0.2612691947823215,
+      "grad_norm": 0.7325875163078308,
+      "learning_rate": 9.599709866744156e-06,
+      "loss": 0.7968,
+      "step": 4747
+    },
+    {
+      "epoch": 0.2613242335846772,
+      "grad_norm": 0.7053360342979431,
+      "learning_rate": 9.599539906902285e-06,
+      "loss": 0.7073,
+      "step": 4748
+    },
+    {
+      "epoch": 0.26137927238703285,
+      "grad_norm": 0.763017475605011,
+      "learning_rate": 9.599369912491503e-06,
+      "loss": 0.7031,
+      "step": 4749
+    },
+    {
+      "epoch": 0.26143431118938854,
+      "grad_norm": 0.6816151738166809,
+      "learning_rate": 9.599199883513088e-06,
+      "loss": 0.7295,
+      "step": 4750
+    },
+    {
+      "epoch": 0.26148934999174417,
+      "grad_norm": 0.8143941164016724,
+      "learning_rate": 9.599029819968319e-06,
+      "loss": 0.8449,
+      "step": 4751
+    },
+    {
+      "epoch": 0.26154438879409986,
+      "grad_norm": 0.8093858361244202,
+      "learning_rate": 9.598859721858471e-06,
+      "loss": 0.8397,
+      "step": 4752
+    },
+    {
+      "epoch": 0.2615994275964555,
+      "grad_norm": 0.7431835532188416,
+      "learning_rate": 9.598689589184827e-06,
+      "loss": 0.7299,
+      "step": 4753
+    },
+    {
+      "epoch": 0.2616544663988112,
+      "grad_norm": 0.9871510863304138,
+      "learning_rate": 9.59851942194866e-06,
+      "loss": 0.7992,
+      "step": 4754
+    },
+    {
+      "epoch": 0.2617095052011668,
+      "grad_norm": 0.9304273724555969,
+      "learning_rate": 9.598349220151254e-06,
+      "loss": 0.7519,
+      "step": 4755
+    },
+    {
+      "epoch": 0.2617645440035225,
+      "grad_norm": 0.9361812472343445,
+      "learning_rate": 9.598178983793886e-06,
+      "loss": 0.8131,
+      "step": 4756
+    },
+    {
+      "epoch": 0.26181958280587814,
+      "grad_norm": 0.7783429622650146,
+      "learning_rate": 9.598008712877835e-06,
+      "loss": 0.7351,
+      "step": 4757
+    },
+    {
+      "epoch": 0.2618746216082338,
+      "grad_norm": 0.8739376068115234,
+      "learning_rate": 9.597838407404381e-06,
+      "loss": 0.9458,
+      "step": 4758
+    },
+    {
+      "epoch": 0.26192966041058946,
+      "grad_norm": 0.7076277732849121,
+      "learning_rate": 9.597668067374805e-06,
+      "loss": 0.7632,
+      "step": 4759
+    },
+    {
+      "epoch": 0.26198469921294515,
+      "grad_norm": 0.7652345299720764,
+      "learning_rate": 9.597497692790386e-06,
+      "loss": 0.8018,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2620397380153008,
+      "grad_norm": 0.7332149147987366,
+      "learning_rate": 9.597327283652405e-06,
+      "loss": 0.8223,
+      "step": 4761
+    },
+    {
+      "epoch": 0.26209477681765647,
+      "grad_norm": 0.8361638784408569,
+      "learning_rate": 9.597156839962145e-06,
+      "loss": 0.8784,
+      "step": 4762
+    },
+    {
+      "epoch": 0.2621498156200121,
+      "grad_norm": 1.183772325515747,
+      "learning_rate": 9.596986361720882e-06,
+      "loss": 0.8768,
+      "step": 4763
+    },
+    {
+      "epoch": 0.2622048544223678,
+      "grad_norm": 0.9895418882369995,
+      "learning_rate": 9.596815848929902e-06,
+      "loss": 0.714,
+      "step": 4764
+    },
+    {
+      "epoch": 0.26225989322472343,
+      "grad_norm": 0.8210558295249939,
+      "learning_rate": 9.59664530159048e-06,
+      "loss": 0.7246,
+      "step": 4765
+    },
+    {
+      "epoch": 0.2623149320270791,
+      "grad_norm": 0.8003455996513367,
+      "learning_rate": 9.596474719703908e-06,
+      "loss": 0.8385,
+      "step": 4766
+    },
+    {
+      "epoch": 0.26236997082943475,
+      "grad_norm": 0.7555826306343079,
+      "learning_rate": 9.59630410327146e-06,
+      "loss": 0.7243,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2624250096317904,
+      "grad_norm": 0.7746273279190063,
+      "learning_rate": 9.596133452294421e-06,
+      "loss": 0.8763,
+      "step": 4768
+    },
+    {
+      "epoch": 0.2624800484341461,
+      "grad_norm": 0.7238507866859436,
+      "learning_rate": 9.595962766774074e-06,
+      "loss": 0.8302,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2625350872365017,
+      "grad_norm": 0.7874132394790649,
+      "learning_rate": 9.595792046711699e-06,
+      "loss": 0.7979,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2625901260388574,
+      "grad_norm": 0.8792033791542053,
+      "learning_rate": 9.595621292108583e-06,
+      "loss": 0.8555,
+      "step": 4771
+    },
+    {
+      "epoch": 0.26264516484121303,
+      "grad_norm": 0.7026945948600769,
+      "learning_rate": 9.595450502966006e-06,
+      "loss": 0.718,
+      "step": 4772
+    },
+    {
+      "epoch": 0.2627002036435687,
+      "grad_norm": 0.7747959494590759,
+      "learning_rate": 9.595279679285254e-06,
+      "loss": 0.8329,
+      "step": 4773
+    },
+    {
+      "epoch": 0.26275524244592435,
+      "grad_norm": 0.697979748249054,
+      "learning_rate": 9.59510882106761e-06,
+      "loss": 0.7456,
+      "step": 4774
+    },
+    {
+      "epoch": 0.26281028124828004,
+      "grad_norm": 0.7600447535514832,
+      "learning_rate": 9.594937928314359e-06,
+      "loss": 0.875,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2628653200506357,
+      "grad_norm": 0.7591384649276733,
+      "learning_rate": 9.594767001026783e-06,
+      "loss": 0.7607,
+      "step": 4776
+    },
+    {
+      "epoch": 0.26292035885299136,
+      "grad_norm": 0.9267380833625793,
+      "learning_rate": 9.59459603920617e-06,
+      "loss": 0.8926,
+      "step": 4777
+    },
+    {
+      "epoch": 0.262975397655347,
+      "grad_norm": 0.7751328349113464,
+      "learning_rate": 9.594425042853802e-06,
+      "loss": 0.7449,
+      "step": 4778
+    },
+    {
+      "epoch": 0.2630304364577027,
+      "grad_norm": 0.7066012620925903,
+      "learning_rate": 9.594254011970966e-06,
+      "loss": 0.8374,
+      "step": 4779
+    },
+    {
+      "epoch": 0.2630854752600583,
+      "grad_norm": 0.7564317584037781,
+      "learning_rate": 9.594082946558945e-06,
+      "loss": 0.735,
+      "step": 4780
+    },
+    {
+      "epoch": 0.263140514062414,
+      "grad_norm": 0.8151416182518005,
+      "learning_rate": 9.593911846619027e-06,
+      "loss": 0.8575,
+      "step": 4781
+    },
+    {
+      "epoch": 0.26319555286476964,
+      "grad_norm": 0.719261646270752,
+      "learning_rate": 9.593740712152497e-06,
+      "loss": 0.7981,
+      "step": 4782
+    },
+    {
+      "epoch": 0.26325059166712533,
+      "grad_norm": 0.8627344369888306,
+      "learning_rate": 9.593569543160642e-06,
+      "loss": 0.895,
+      "step": 4783
+    },
+    {
+      "epoch": 0.26330563046948097,
+      "grad_norm": 1.293272614479065,
+      "learning_rate": 9.593398339644748e-06,
+      "loss": 0.7531,
+      "step": 4784
+    },
+    {
+      "epoch": 0.26336066927183666,
+      "grad_norm": 0.8475207686424255,
+      "learning_rate": 9.593227101606102e-06,
+      "loss": 0.9091,
+      "step": 4785
+    },
+    {
+      "epoch": 0.2634157080741923,
+      "grad_norm": 0.78054279088974,
+      "learning_rate": 9.593055829045989e-06,
+      "loss": 0.7692,
+      "step": 4786
+    },
+    {
+      "epoch": 0.263470746876548,
+      "grad_norm": 0.7677399516105652,
+      "learning_rate": 9.592884521965699e-06,
+      "loss": 0.6232,
+      "step": 4787
+    },
+    {
+      "epoch": 0.2635257856789036,
+      "grad_norm": 0.7232677340507507,
+      "learning_rate": 9.59271318036652e-06,
+      "loss": 0.8087,
+      "step": 4788
+    },
+    {
+      "epoch": 0.2635808244812593,
+      "grad_norm": 0.8728463649749756,
+      "learning_rate": 9.592541804249735e-06,
+      "loss": 0.7824,
+      "step": 4789
+    },
+    {
+      "epoch": 0.26363586328361494,
+      "grad_norm": 0.7569910883903503,
+      "learning_rate": 9.592370393616637e-06,
+      "loss": 0.7418,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2636909020859706,
+      "grad_norm": 0.7631934285163879,
+      "learning_rate": 9.592198948468511e-06,
+      "loss": 0.7929,
+      "step": 4791
+    },
+    {
+      "epoch": 0.26374594088832626,
+      "grad_norm": 0.8021631240844727,
+      "learning_rate": 9.592027468806649e-06,
+      "loss": 0.8111,
+      "step": 4792
+    },
+    {
+      "epoch": 0.26380097969068195,
+      "grad_norm": 0.9454651474952698,
+      "learning_rate": 9.591855954632336e-06,
+      "loss": 0.8239,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2638560184930376,
+      "grad_norm": 0.672924280166626,
+      "learning_rate": 9.591684405946863e-06,
+      "loss": 0.6877,
+      "step": 4794
+    },
+    {
+      "epoch": 0.26391105729539327,
+      "grad_norm": 0.7942802906036377,
+      "learning_rate": 9.59151282275152e-06,
+      "loss": 0.9002,
+      "step": 4795
+    },
+    {
+      "epoch": 0.2639660960977489,
+      "grad_norm": 0.7131155133247375,
+      "learning_rate": 9.591341205047596e-06,
+      "loss": 0.7692,
+      "step": 4796
+    },
+    {
+      "epoch": 0.2640211349001046,
+      "grad_norm": 1.0395869016647339,
+      "learning_rate": 9.59116955283638e-06,
+      "loss": 0.8352,
+      "step": 4797
+    },
+    {
+      "epoch": 0.2640761737024602,
+      "grad_norm": 0.9503256678581238,
+      "learning_rate": 9.590997866119163e-06,
+      "loss": 1.0287,
+      "step": 4798
+    },
+    {
+      "epoch": 0.2641312125048159,
+      "grad_norm": 0.7539612054824829,
+      "learning_rate": 9.590826144897235e-06,
+      "loss": 0.872,
+      "step": 4799
+    },
+    {
+      "epoch": 0.26418625130717155,
+      "grad_norm": 0.7067893743515015,
+      "learning_rate": 9.590654389171885e-06,
+      "loss": 0.7636,
+      "step": 4800
+    },
+    {
+      "epoch": 0.26424129010952724,
+      "grad_norm": 0.7355281710624695,
+      "learning_rate": 9.590482598944407e-06,
+      "loss": 0.7715,
+      "step": 4801
+    },
+    {
+      "epoch": 0.26429632891188287,
+      "grad_norm": 0.7589674592018127,
+      "learning_rate": 9.590310774216089e-06,
+      "loss": 0.7451,
+      "step": 4802
+    },
+    {
+      "epoch": 0.26435136771423856,
+      "grad_norm": 0.701386034488678,
+      "learning_rate": 9.590138914988226e-06,
+      "loss": 0.7317,
+      "step": 4803
+    },
+    {
+      "epoch": 0.2644064065165942,
+      "grad_norm": 0.7663118243217468,
+      "learning_rate": 9.589967021262105e-06,
+      "loss": 0.8227,
+      "step": 4804
+    },
+    {
+      "epoch": 0.2644614453189499,
+      "grad_norm": 0.7059655785560608,
+      "learning_rate": 9.589795093039023e-06,
+      "loss": 0.7829,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2645164841213055,
+      "grad_norm": 0.7377020120620728,
+      "learning_rate": 9.58962313032027e-06,
+      "loss": 0.8308,
+      "step": 4806
+    },
+    {
+      "epoch": 0.2645715229236612,
+      "grad_norm": 0.8635388612747192,
+      "learning_rate": 9.589451133107134e-06,
+      "loss": 0.7882,
+      "step": 4807
+    },
+    {
+      "epoch": 0.26462656172601684,
+      "grad_norm": 0.8282824754714966,
+      "learning_rate": 9.589279101400915e-06,
+      "loss": 0.8055,
+      "step": 4808
+    },
+    {
+      "epoch": 0.26468160052837253,
+      "grad_norm": 0.7026814818382263,
+      "learning_rate": 9.589107035202903e-06,
+      "loss": 0.7567,
+      "step": 4809
+    },
+    {
+      "epoch": 0.26473663933072816,
+      "grad_norm": 0.7575708031654358,
+      "learning_rate": 9.588934934514392e-06,
+      "loss": 0.7456,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2647916781330838,
+      "grad_norm": 0.9732069969177246,
+      "learning_rate": 9.588762799336671e-06,
+      "loss": 0.8217,
+      "step": 4811
+    },
+    {
+      "epoch": 0.2648467169354395,
+      "grad_norm": 0.786803126335144,
+      "learning_rate": 9.58859062967104e-06,
+      "loss": 0.729,
+      "step": 4812
+    },
+    {
+      "epoch": 0.2649017557377951,
+      "grad_norm": 0.8068973422050476,
+      "learning_rate": 9.588418425518789e-06,
+      "loss": 0.8204,
+      "step": 4813
+    },
+    {
+      "epoch": 0.2649567945401508,
+      "grad_norm": 0.8222702145576477,
+      "learning_rate": 9.588246186881213e-06,
+      "loss": 0.8349,
+      "step": 4814
+    },
+    {
+      "epoch": 0.26501183334250644,
+      "grad_norm": 0.7560802698135376,
+      "learning_rate": 9.588073913759608e-06,
+      "loss": 0.7601,
+      "step": 4815
+    },
+    {
+      "epoch": 0.26506687214486213,
+      "grad_norm": 0.9221365451812744,
+      "learning_rate": 9.587901606155266e-06,
+      "loss": 0.7725,
+      "step": 4816
+    },
+    {
+      "epoch": 0.26512191094721776,
+      "grad_norm": 0.8092262744903564,
+      "learning_rate": 9.587729264069485e-06,
+      "loss": 0.9074,
+      "step": 4817
+    },
+    {
+      "epoch": 0.26517694974957345,
+      "grad_norm": 0.8183920979499817,
+      "learning_rate": 9.587556887503557e-06,
+      "loss": 0.8321,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2652319885519291,
+      "grad_norm": 0.7023420929908752,
+      "learning_rate": 9.587384476458781e-06,
+      "loss": 0.7842,
+      "step": 4819
+    },
+    {
+      "epoch": 0.2652870273542848,
+      "grad_norm": 1.2864880561828613,
+      "learning_rate": 9.58721203093645e-06,
+      "loss": 0.7519,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2653420661566404,
+      "grad_norm": 0.8133784532546997,
+      "learning_rate": 9.587039550937864e-06,
+      "loss": 0.8208,
+      "step": 4821
+    },
+    {
+      "epoch": 0.2653971049589961,
+      "grad_norm": 0.739732027053833,
+      "learning_rate": 9.586867036464314e-06,
+      "loss": 0.8553,
+      "step": 4822
+    },
+    {
+      "epoch": 0.26545214376135173,
+      "grad_norm": 0.7539162635803223,
+      "learning_rate": 9.5866944875171e-06,
+      "loss": 0.7385,
+      "step": 4823
+    },
+    {
+      "epoch": 0.2655071825637074,
+      "grad_norm": 0.8012336492538452,
+      "learning_rate": 9.58652190409752e-06,
+      "loss": 0.8343,
+      "step": 4824
+    },
+    {
+      "epoch": 0.26556222136606306,
+      "grad_norm": 0.7972521185874939,
+      "learning_rate": 9.586349286206865e-06,
+      "loss": 0.8481,
+      "step": 4825
+    },
+    {
+      "epoch": 0.26561726016841875,
+      "grad_norm": 0.7772900462150574,
+      "learning_rate": 9.58617663384644e-06,
+      "loss": 0.7655,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2656722989707744,
+      "grad_norm": 0.677916944026947,
+      "learning_rate": 9.586003947017537e-06,
+      "loss": 0.696,
+      "step": 4827
+    },
+    {
+      "epoch": 0.26572733777313007,
+      "grad_norm": 0.8254117369651794,
+      "learning_rate": 9.585831225721455e-06,
+      "loss": 0.7841,
+      "step": 4828
+    },
+    {
+      "epoch": 0.2657823765754857,
+      "grad_norm": 0.7256904244422913,
+      "learning_rate": 9.585658469959496e-06,
+      "loss": 0.8057,
+      "step": 4829
+    },
+    {
+      "epoch": 0.2658374153778414,
+      "grad_norm": 0.7651757001876831,
+      "learning_rate": 9.585485679732953e-06,
+      "loss": 0.7918,
+      "step": 4830
+    },
+    {
+      "epoch": 0.265892454180197,
+      "grad_norm": 0.7581052184104919,
+      "learning_rate": 9.58531285504313e-06,
+      "loss": 0.759,
+      "step": 4831
+    },
+    {
+      "epoch": 0.2659474929825527,
+      "grad_norm": 0.7190486192703247,
+      "learning_rate": 9.58513999589132e-06,
+      "loss": 0.7403,
+      "step": 4832
+    },
+    {
+      "epoch": 0.26600253178490835,
+      "grad_norm": 0.8603141903877258,
+      "learning_rate": 9.584967102278825e-06,
+      "loss": 0.8944,
+      "step": 4833
+    },
+    {
+      "epoch": 0.26605757058726404,
+      "grad_norm": 0.806297779083252,
+      "learning_rate": 9.584794174206947e-06,
+      "loss": 0.7039,
+      "step": 4834
+    },
+    {
+      "epoch": 0.26611260938961967,
+      "grad_norm": 0.7604451775550842,
+      "learning_rate": 9.584621211676981e-06,
+      "loss": 0.8076,
+      "step": 4835
+    },
+    {
+      "epoch": 0.26616764819197536,
+      "grad_norm": 0.7276773452758789,
+      "learning_rate": 9.584448214690232e-06,
+      "loss": 0.786,
+      "step": 4836
+    },
+    {
+      "epoch": 0.266222686994331,
+      "grad_norm": 0.8737080693244934,
+      "learning_rate": 9.584275183247994e-06,
+      "loss": 0.8071,
+      "step": 4837
+    },
+    {
+      "epoch": 0.2662777257966867,
+      "grad_norm": 0.8447219133377075,
+      "learning_rate": 9.584102117351574e-06,
+      "loss": 0.7682,
+      "step": 4838
+    },
+    {
+      "epoch": 0.2663327645990423,
+      "grad_norm": 0.7001703381538391,
+      "learning_rate": 9.583929017002268e-06,
+      "loss": 0.7077,
+      "step": 4839
+    },
+    {
+      "epoch": 0.266387803401398,
+      "grad_norm": 0.7935730218887329,
+      "learning_rate": 9.583755882201377e-06,
+      "loss": 0.8122,
+      "step": 4840
+    },
+    {
+      "epoch": 0.26644284220375364,
+      "grad_norm": 0.8763312697410583,
+      "learning_rate": 9.583582712950207e-06,
+      "loss": 0.8241,
+      "step": 4841
+    },
+    {
+      "epoch": 0.2664978810061093,
+      "grad_norm": 0.7910245656967163,
+      "learning_rate": 9.583409509250055e-06,
+      "loss": 0.7717,
+      "step": 4842
+    },
+    {
+      "epoch": 0.26655291980846496,
+      "grad_norm": 0.7975226640701294,
+      "learning_rate": 9.583236271102222e-06,
+      "loss": 0.7165,
+      "step": 4843
+    },
+    {
+      "epoch": 0.26660795861082065,
+      "grad_norm": 0.8060342073440552,
+      "learning_rate": 9.583062998508014e-06,
+      "loss": 0.7659,
+      "step": 4844
+    },
+    {
+      "epoch": 0.2666629974131763,
+      "grad_norm": 0.8779375553131104,
+      "learning_rate": 9.582889691468732e-06,
+      "loss": 0.8207,
+      "step": 4845
+    },
+    {
+      "epoch": 0.266718036215532,
+      "grad_norm": 0.7409310936927795,
+      "learning_rate": 9.582716349985677e-06,
+      "loss": 0.8439,
+      "step": 4846
+    },
+    {
+      "epoch": 0.2667730750178876,
+      "grad_norm": 0.8871899843215942,
+      "learning_rate": 9.582542974060152e-06,
+      "loss": 0.8305,
+      "step": 4847
+    },
+    {
+      "epoch": 0.2668281138202433,
+      "grad_norm": 0.9003115296363831,
+      "learning_rate": 9.58236956369346e-06,
+      "loss": 0.8334,
+      "step": 4848
+    },
+    {
+      "epoch": 0.26688315262259893,
+      "grad_norm": 1.0149577856063843,
+      "learning_rate": 9.582196118886909e-06,
+      "loss": 0.7962,
+      "step": 4849
+    },
+    {
+      "epoch": 0.2669381914249546,
+      "grad_norm": 0.785214900970459,
+      "learning_rate": 9.582022639641795e-06,
+      "loss": 0.7806,
+      "step": 4850
+    },
+    {
+      "epoch": 0.26699323022731025,
+      "grad_norm": 0.9833952188491821,
+      "learning_rate": 9.581849125959426e-06,
+      "loss": 0.7607,
+      "step": 4851
+    },
+    {
+      "epoch": 0.26704826902966594,
+      "grad_norm": 1.404751181602478,
+      "learning_rate": 9.581675577841104e-06,
+      "loss": 0.9046,
+      "step": 4852
+    },
+    {
+      "epoch": 0.2671033078320216,
+      "grad_norm": 0.791159451007843,
+      "learning_rate": 9.581501995288137e-06,
+      "loss": 0.6582,
+      "step": 4853
+    },
+    {
+      "epoch": 0.2671583466343772,
+      "grad_norm": 0.8507272005081177,
+      "learning_rate": 9.581328378301827e-06,
+      "loss": 0.8946,
+      "step": 4854
+    },
+    {
+      "epoch": 0.2672133854367329,
+      "grad_norm": 0.7372786998748779,
+      "learning_rate": 9.58115472688348e-06,
+      "loss": 0.7865,
+      "step": 4855
+    },
+    {
+      "epoch": 0.26726842423908853,
+      "grad_norm": 0.8293853998184204,
+      "learning_rate": 9.580981041034398e-06,
+      "loss": 0.9113,
+      "step": 4856
+    },
+    {
+      "epoch": 0.2673234630414442,
+      "grad_norm": 0.7212402820587158,
+      "learning_rate": 9.580807320755889e-06,
+      "loss": 0.7149,
+      "step": 4857
+    },
+    {
+      "epoch": 0.26737850184379985,
+      "grad_norm": 0.7885197401046753,
+      "learning_rate": 9.58063356604926e-06,
+      "loss": 0.8651,
+      "step": 4858
+    },
+    {
+      "epoch": 0.26743354064615554,
+      "grad_norm": 0.8444308042526245,
+      "learning_rate": 9.580459776915814e-06,
+      "loss": 0.7968,
+      "step": 4859
+    },
+    {
+      "epoch": 0.2674885794485112,
+      "grad_norm": 0.7974254488945007,
+      "learning_rate": 9.58028595335686e-06,
+      "loss": 0.8499,
+      "step": 4860
+    },
+    {
+      "epoch": 0.26754361825086687,
+      "grad_norm": 0.7491242289543152,
+      "learning_rate": 9.580112095373702e-06,
+      "loss": 0.8278,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2675986570532225,
+      "grad_norm": 0.6856499314308167,
+      "learning_rate": 9.579938202967646e-06,
+      "loss": 0.7466,
+      "step": 4862
+    },
+    {
+      "epoch": 0.2676536958555782,
+      "grad_norm": 0.7347447872161865,
+      "learning_rate": 9.579764276140002e-06,
+      "loss": 0.8046,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2677087346579338,
+      "grad_norm": 0.6797083020210266,
+      "learning_rate": 9.579590314892077e-06,
+      "loss": 0.7012,
+      "step": 4864
+    },
+    {
+      "epoch": 0.2677637734602895,
+      "grad_norm": 0.8219562768936157,
+      "learning_rate": 9.579416319225175e-06,
+      "loss": 0.7592,
+      "step": 4865
+    },
+    {
+      "epoch": 0.26781881226264515,
+      "grad_norm": 0.7388357520103455,
+      "learning_rate": 9.579242289140607e-06,
+      "loss": 0.8179,
+      "step": 4866
+    },
+    {
+      "epoch": 0.26787385106500083,
+      "grad_norm": 0.7394490838050842,
+      "learning_rate": 9.579068224639679e-06,
+      "loss": 0.694,
+      "step": 4867
+    },
+    {
+      "epoch": 0.26792888986735647,
+      "grad_norm": 0.7309017181396484,
+      "learning_rate": 9.578894125723699e-06,
+      "loss": 0.7882,
+      "step": 4868
+    },
+    {
+      "epoch": 0.26798392866971216,
+      "grad_norm": 0.7785035967826843,
+      "learning_rate": 9.578719992393978e-06,
+      "loss": 0.8142,
+      "step": 4869
+    },
+    {
+      "epoch": 0.2680389674720678,
+      "grad_norm": 0.8983079195022583,
+      "learning_rate": 9.57854582465182e-06,
+      "loss": 0.7809,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2680940062744235,
+      "grad_norm": 0.7433765530586243,
+      "learning_rate": 9.578371622498542e-06,
+      "loss": 0.8937,
+      "step": 4871
+    },
+    {
+      "epoch": 0.2681490450767791,
+      "grad_norm": 0.8808667659759521,
+      "learning_rate": 9.578197385935446e-06,
+      "loss": 0.7821,
+      "step": 4872
+    },
+    {
+      "epoch": 0.2682040838791348,
+      "grad_norm": 0.825794517993927,
+      "learning_rate": 9.578023114963843e-06,
+      "loss": 0.8228,
+      "step": 4873
+    },
+    {
+      "epoch": 0.26825912268149044,
+      "grad_norm": 1.0165129899978638,
+      "learning_rate": 9.577848809585046e-06,
+      "loss": 0.7964,
+      "step": 4874
+    },
+    {
+      "epoch": 0.2683141614838461,
+      "grad_norm": 0.742028534412384,
+      "learning_rate": 9.577674469800362e-06,
+      "loss": 0.9126,
+      "step": 4875
+    },
+    {
+      "epoch": 0.26836920028620176,
+      "grad_norm": 0.7571890354156494,
+      "learning_rate": 9.577500095611101e-06,
+      "loss": 0.879,
+      "step": 4876
+    },
+    {
+      "epoch": 0.26842423908855745,
+      "grad_norm": 0.7577160596847534,
+      "learning_rate": 9.577325687018575e-06,
+      "loss": 0.8048,
+      "step": 4877
+    },
+    {
+      "epoch": 0.2684792778909131,
+      "grad_norm": 0.7704411745071411,
+      "learning_rate": 9.577151244024095e-06,
+      "loss": 0.7451,
+      "step": 4878
+    },
+    {
+      "epoch": 0.26853431669326877,
+      "grad_norm": 0.8323166966438293,
+      "learning_rate": 9.57697676662897e-06,
+      "loss": 0.7591,
+      "step": 4879
+    },
+    {
+      "epoch": 0.2685893554956244,
+      "grad_norm": 0.7257028222084045,
+      "learning_rate": 9.576802254834516e-06,
+      "loss": 0.7941,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2686443942979801,
+      "grad_norm": 0.8170442581176758,
+      "learning_rate": 9.57662770864204e-06,
+      "loss": 0.8617,
+      "step": 4881
+    },
+    {
+      "epoch": 0.2686994331003357,
+      "grad_norm": 0.7435339689254761,
+      "learning_rate": 9.576453128052852e-06,
+      "loss": 0.7683,
+      "step": 4882
+    },
+    {
+      "epoch": 0.2687544719026914,
+      "grad_norm": 0.7932955026626587,
+      "learning_rate": 9.576278513068271e-06,
+      "loss": 0.7103,
+      "step": 4883
+    },
+    {
+      "epoch": 0.26880951070504705,
+      "grad_norm": 0.8008469939231873,
+      "learning_rate": 9.576103863689604e-06,
+      "loss": 0.8144,
+      "step": 4884
+    },
+    {
+      "epoch": 0.26886454950740274,
+      "grad_norm": 0.8573774695396423,
+      "learning_rate": 9.575929179918167e-06,
+      "loss": 0.8992,
+      "step": 4885
+    },
+    {
+      "epoch": 0.2689195883097584,
+      "grad_norm": 0.7326993942260742,
+      "learning_rate": 9.57575446175527e-06,
+      "loss": 0.699,
+      "step": 4886
+    },
+    {
+      "epoch": 0.26897462711211406,
+      "grad_norm": 0.8249791264533997,
+      "learning_rate": 9.575579709202228e-06,
+      "loss": 0.7445,
+      "step": 4887
+    },
+    {
+      "epoch": 0.2690296659144697,
+      "grad_norm": 0.7136644124984741,
+      "learning_rate": 9.575404922260351e-06,
+      "loss": 0.779,
+      "step": 4888
+    },
+    {
+      "epoch": 0.2690847047168254,
+      "grad_norm": 1.0130438804626465,
+      "learning_rate": 9.575230100930958e-06,
+      "loss": 0.8535,
+      "step": 4889
+    },
+    {
+      "epoch": 0.269139743519181,
+      "grad_norm": 0.6784926652908325,
+      "learning_rate": 9.575055245215358e-06,
+      "loss": 0.6745,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2691947823215367,
+      "grad_norm": 0.7492508888244629,
+      "learning_rate": 9.57488035511487e-06,
+      "loss": 0.6748,
+      "step": 4891
+    },
+    {
+      "epoch": 0.26924982112389234,
+      "grad_norm": 0.7951217889785767,
+      "learning_rate": 9.574705430630807e-06,
+      "loss": 0.8119,
+      "step": 4892
+    },
+    {
+      "epoch": 0.26930485992624803,
+      "grad_norm": 0.9756677746772766,
+      "learning_rate": 9.574530471764478e-06,
+      "loss": 0.855,
+      "step": 4893
+    },
+    {
+      "epoch": 0.26935989872860366,
+      "grad_norm": 0.7806811928749084,
+      "learning_rate": 9.574355478517206e-06,
+      "loss": 0.8432,
+      "step": 4894
+    },
+    {
+      "epoch": 0.26941493753095935,
+      "grad_norm": 0.7814774513244629,
+      "learning_rate": 9.574180450890301e-06,
+      "loss": 0.8226,
+      "step": 4895
+    },
+    {
+      "epoch": 0.269469976333315,
+      "grad_norm": 0.7745325565338135,
+      "learning_rate": 9.574005388885081e-06,
+      "loss": 0.7722,
+      "step": 4896
+    },
+    {
+      "epoch": 0.2695250151356706,
+      "grad_norm": 0.7805666327476501,
+      "learning_rate": 9.573830292502862e-06,
+      "loss": 0.8357,
+      "step": 4897
+    },
+    {
+      "epoch": 0.2695800539380263,
+      "grad_norm": 0.8428031802177429,
+      "learning_rate": 9.573655161744958e-06,
+      "loss": 0.8056,
+      "step": 4898
+    },
+    {
+      "epoch": 0.26963509274038194,
+      "grad_norm": 0.7896600961685181,
+      "learning_rate": 9.573479996612684e-06,
+      "loss": 0.7984,
+      "step": 4899
+    },
+    {
+      "epoch": 0.26969013154273763,
+      "grad_norm": 0.7718683481216431,
+      "learning_rate": 9.57330479710736e-06,
+      "loss": 0.7527,
+      "step": 4900
+    },
+    {
+      "epoch": 0.26974517034509327,
+      "grad_norm": 0.7868129014968872,
+      "learning_rate": 9.573129563230302e-06,
+      "loss": 0.7876,
+      "step": 4901
+    },
+    {
+      "epoch": 0.26980020914744895,
+      "grad_norm": 0.8493777513504028,
+      "learning_rate": 9.572954294982826e-06,
+      "loss": 0.864,
+      "step": 4902
+    },
+    {
+      "epoch": 0.2698552479498046,
+      "grad_norm": 0.7492502331733704,
+      "learning_rate": 9.57277899236625e-06,
+      "loss": 0.8236,
+      "step": 4903
+    },
+    {
+      "epoch": 0.2699102867521603,
+      "grad_norm": 1.0534250736236572,
+      "learning_rate": 9.57260365538189e-06,
+      "loss": 0.8012,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2699653255545159,
+      "grad_norm": 0.7557470202445984,
+      "learning_rate": 9.572428284031065e-06,
+      "loss": 0.9084,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2700203643568716,
+      "grad_norm": 0.8055123686790466,
+      "learning_rate": 9.572252878315094e-06,
+      "loss": 0.7468,
+      "step": 4906
+    },
+    {
+      "epoch": 0.27007540315922723,
+      "grad_norm": 0.8399039506912231,
+      "learning_rate": 9.572077438235294e-06,
+      "loss": 0.9293,
+      "step": 4907
+    },
+    {
+      "epoch": 0.2701304419615829,
+      "grad_norm": 0.9800041317939758,
+      "learning_rate": 9.571901963792983e-06,
+      "loss": 0.8664,
+      "step": 4908
+    },
+    {
+      "epoch": 0.27018548076393856,
+      "grad_norm": 0.7732129096984863,
+      "learning_rate": 9.571726454989482e-06,
+      "loss": 0.7227,
+      "step": 4909
+    },
+    {
+      "epoch": 0.27024051956629425,
+      "grad_norm": 0.730754017829895,
+      "learning_rate": 9.571550911826109e-06,
+      "loss": 0.6467,
+      "step": 4910
+    },
+    {
+      "epoch": 0.2702955583686499,
+      "grad_norm": 0.8245325684547424,
+      "learning_rate": 9.57137533430418e-06,
+      "loss": 0.7847,
+      "step": 4911
+    },
+    {
+      "epoch": 0.27035059717100557,
+      "grad_norm": 0.8606786131858826,
+      "learning_rate": 9.57119972242502e-06,
+      "loss": 0.9556,
+      "step": 4912
+    },
+    {
+      "epoch": 0.2704056359733612,
+      "grad_norm": 0.7480195164680481,
+      "learning_rate": 9.571024076189947e-06,
+      "loss": 0.8504,
+      "step": 4913
+    },
+    {
+      "epoch": 0.2704606747757169,
+      "grad_norm": 0.718913197517395,
+      "learning_rate": 9.57084839560028e-06,
+      "loss": 0.7869,
+      "step": 4914
+    },
+    {
+      "epoch": 0.2705157135780725,
+      "grad_norm": 0.9778180122375488,
+      "learning_rate": 9.57067268065734e-06,
+      "loss": 0.8514,
+      "step": 4915
+    },
+    {
+      "epoch": 0.2705707523804282,
+      "grad_norm": 0.7394844889640808,
+      "learning_rate": 9.570496931362448e-06,
+      "loss": 0.7906,
+      "step": 4916
+    },
+    {
+      "epoch": 0.27062579118278385,
+      "grad_norm": 0.7648600339889526,
+      "learning_rate": 9.570321147716923e-06,
+      "loss": 0.8194,
+      "step": 4917
+    },
+    {
+      "epoch": 0.27068082998513954,
+      "grad_norm": 0.8002632260322571,
+      "learning_rate": 9.57014532972209e-06,
+      "loss": 0.8079,
+      "step": 4918
+    },
+    {
+      "epoch": 0.27073586878749517,
+      "grad_norm": 0.8668341040611267,
+      "learning_rate": 9.569969477379267e-06,
+      "loss": 0.8954,
+      "step": 4919
+    },
+    {
+      "epoch": 0.27079090758985086,
+      "grad_norm": 0.7403327226638794,
+      "learning_rate": 9.569793590689775e-06,
+      "loss": 0.7755,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2708459463922065,
+      "grad_norm": 0.7399682998657227,
+      "learning_rate": 9.569617669654938e-06,
+      "loss": 0.8203,
+      "step": 4921
+    },
+    {
+      "epoch": 0.2709009851945622,
+      "grad_norm": 0.788600504398346,
+      "learning_rate": 9.56944171427608e-06,
+      "loss": 0.7565,
+      "step": 4922
+    },
+    {
+      "epoch": 0.2709560239969178,
+      "grad_norm": 0.7044861912727356,
+      "learning_rate": 9.56926572455452e-06,
+      "loss": 0.7073,
+      "step": 4923
+    },
+    {
+      "epoch": 0.2710110627992735,
+      "grad_norm": 0.8195114135742188,
+      "learning_rate": 9.569089700491581e-06,
+      "loss": 0.8658,
+      "step": 4924
+    },
+    {
+      "epoch": 0.27106610160162914,
+      "grad_norm": 0.7792258858680725,
+      "learning_rate": 9.568913642088589e-06,
+      "loss": 0.8628,
+      "step": 4925
+    },
+    {
+      "epoch": 0.27112114040398483,
+      "grad_norm": 0.764930248260498,
+      "learning_rate": 9.568737549346862e-06,
+      "loss": 0.7761,
+      "step": 4926
+    },
+    {
+      "epoch": 0.27117617920634046,
+      "grad_norm": 0.7226328253746033,
+      "learning_rate": 9.56856142226773e-06,
+      "loss": 0.7208,
+      "step": 4927
+    },
+    {
+      "epoch": 0.27123121800869615,
+      "grad_norm": 0.8726598620414734,
+      "learning_rate": 9.568385260852512e-06,
+      "loss": 0.8599,
+      "step": 4928
+    },
+    {
+      "epoch": 0.2712862568110518,
+      "grad_norm": 1.0126571655273438,
+      "learning_rate": 9.568209065102533e-06,
+      "loss": 0.8145,
+      "step": 4929
+    },
+    {
+      "epoch": 0.2713412956134075,
+      "grad_norm": 0.7764692306518555,
+      "learning_rate": 9.568032835019116e-06,
+      "loss": 0.6758,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2713963344157631,
+      "grad_norm": 0.6955474019050598,
+      "learning_rate": 9.567856570603589e-06,
+      "loss": 0.7461,
+      "step": 4931
+    },
+    {
+      "epoch": 0.2714513732181188,
+      "grad_norm": 0.7136832475662231,
+      "learning_rate": 9.567680271857274e-06,
+      "loss": 0.7692,
+      "step": 4932
+    },
+    {
+      "epoch": 0.27150641202047443,
+      "grad_norm": 1.2288198471069336,
+      "learning_rate": 9.567503938781497e-06,
+      "loss": 0.7815,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2715614508228301,
+      "grad_norm": 0.9182234406471252,
+      "learning_rate": 9.567327571377584e-06,
+      "loss": 0.8822,
+      "step": 4934
+    },
+    {
+      "epoch": 0.27161648962518575,
+      "grad_norm": 0.7684763669967651,
+      "learning_rate": 9.567151169646859e-06,
+      "loss": 0.7618,
+      "step": 4935
+    },
+    {
+      "epoch": 0.27167152842754144,
+      "grad_norm": 0.872360348701477,
+      "learning_rate": 9.566974733590647e-06,
+      "loss": 0.7975,
+      "step": 4936
+    },
+    {
+      "epoch": 0.2717265672298971,
+      "grad_norm": 0.9010463356971741,
+      "learning_rate": 9.566798263210277e-06,
+      "loss": 0.7159,
+      "step": 4937
+    },
+    {
+      "epoch": 0.27178160603225276,
+      "grad_norm": 0.7254281044006348,
+      "learning_rate": 9.566621758507072e-06,
+      "loss": 0.6724,
+      "step": 4938
+    },
+    {
+      "epoch": 0.2718366448346084,
+      "grad_norm": 0.8478212356567383,
+      "learning_rate": 9.566445219482363e-06,
+      "loss": 0.659,
+      "step": 4939
+    },
+    {
+      "epoch": 0.27189168363696403,
+      "grad_norm": 0.9038714170455933,
+      "learning_rate": 9.56626864613747e-06,
+      "loss": 0.8766,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2719467224393197,
+      "grad_norm": 0.9704582691192627,
+      "learning_rate": 9.566092038473728e-06,
+      "loss": 0.8972,
+      "step": 4941
+    },
+    {
+      "epoch": 0.27200176124167535,
+      "grad_norm": 0.7069430947303772,
+      "learning_rate": 9.565915396492459e-06,
+      "loss": 0.8116,
+      "step": 4942
+    },
+    {
+      "epoch": 0.27205680004403104,
+      "grad_norm": 0.7432642579078674,
+      "learning_rate": 9.565738720194993e-06,
+      "loss": 0.847,
+      "step": 4943
+    },
+    {
+      "epoch": 0.2721118388463867,
+      "grad_norm": 0.6813814043998718,
+      "learning_rate": 9.565562009582655e-06,
+      "loss": 0.7146,
+      "step": 4944
+    },
+    {
+      "epoch": 0.27216687764874237,
+      "grad_norm": 0.7447707056999207,
+      "learning_rate": 9.565385264656776e-06,
+      "loss": 0.7696,
+      "step": 4945
+    },
+    {
+      "epoch": 0.272221916451098,
+      "grad_norm": 0.875073254108429,
+      "learning_rate": 9.565208485418685e-06,
+      "loss": 0.8714,
+      "step": 4946
+    },
+    {
+      "epoch": 0.2722769552534537,
+      "grad_norm": 0.7753880620002747,
+      "learning_rate": 9.565031671869707e-06,
+      "loss": 0.739,
+      "step": 4947
+    },
+    {
+      "epoch": 0.2723319940558093,
+      "grad_norm": 0.749264121055603,
+      "learning_rate": 9.564854824011172e-06,
+      "loss": 0.7957,
+      "step": 4948
+    },
+    {
+      "epoch": 0.272387032858165,
+      "grad_norm": 0.6733991503715515,
+      "learning_rate": 9.564677941844412e-06,
+      "loss": 0.7402,
+      "step": 4949
+    },
+    {
+      "epoch": 0.27244207166052065,
+      "grad_norm": 0.7426447868347168,
+      "learning_rate": 9.564501025370753e-06,
+      "loss": 0.7977,
+      "step": 4950
+    },
+    {
+      "epoch": 0.27249711046287634,
+      "grad_norm": 0.7930514812469482,
+      "learning_rate": 9.564324074591529e-06,
+      "loss": 0.8485,
+      "step": 4951
+    },
+    {
+      "epoch": 0.27255214926523197,
+      "grad_norm": 0.8087072968482971,
+      "learning_rate": 9.564147089508064e-06,
+      "loss": 0.9215,
+      "step": 4952
+    },
+    {
+      "epoch": 0.27260718806758766,
+      "grad_norm": 0.7560327053070068,
+      "learning_rate": 9.563970070121694e-06,
+      "loss": 0.7966,
+      "step": 4953
+    },
+    {
+      "epoch": 0.2726622268699433,
+      "grad_norm": 0.735573947429657,
+      "learning_rate": 9.563793016433744e-06,
+      "loss": 0.7737,
+      "step": 4954
+    },
+    {
+      "epoch": 0.272717265672299,
+      "grad_norm": 0.7603545784950256,
+      "learning_rate": 9.563615928445548e-06,
+      "loss": 0.7717,
+      "step": 4955
+    },
+    {
+      "epoch": 0.2727723044746546,
+      "grad_norm": 0.7185375094413757,
+      "learning_rate": 9.563438806158437e-06,
+      "loss": 0.8057,
+      "step": 4956
+    },
+    {
+      "epoch": 0.2728273432770103,
+      "grad_norm": 0.7619272470474243,
+      "learning_rate": 9.56326164957374e-06,
+      "loss": 0.8173,
+      "step": 4957
+    },
+    {
+      "epoch": 0.27288238207936594,
+      "grad_norm": 0.7868000864982605,
+      "learning_rate": 9.563084458692793e-06,
+      "loss": 0.6855,
+      "step": 4958
+    },
+    {
+      "epoch": 0.2729374208817216,
+      "grad_norm": 0.7949535846710205,
+      "learning_rate": 9.562907233516923e-06,
+      "loss": 0.7754,
+      "step": 4959
+    },
+    {
+      "epoch": 0.27299245968407726,
+      "grad_norm": 0.7037919163703918,
+      "learning_rate": 9.562729974047462e-06,
+      "loss": 0.7419,
+      "step": 4960
+    },
+    {
+      "epoch": 0.27304749848643295,
+      "grad_norm": 0.7236568927764893,
+      "learning_rate": 9.562552680285746e-06,
+      "loss": 0.7135,
+      "step": 4961
+    },
+    {
+      "epoch": 0.2731025372887886,
+      "grad_norm": 0.8410467505455017,
+      "learning_rate": 9.562375352233105e-06,
+      "loss": 0.8507,
+      "step": 4962
+    },
+    {
+      "epoch": 0.27315757609114427,
+      "grad_norm": 0.8043560981750488,
+      "learning_rate": 9.562197989890871e-06,
+      "loss": 0.8484,
+      "step": 4963
+    },
+    {
+      "epoch": 0.2732126148934999,
+      "grad_norm": 0.6926127672195435,
+      "learning_rate": 9.56202059326038e-06,
+      "loss": 0.8087,
+      "step": 4964
+    },
+    {
+      "epoch": 0.2732676536958556,
+      "grad_norm": 0.7149024605751038,
+      "learning_rate": 9.561843162342961e-06,
+      "loss": 0.7349,
+      "step": 4965
+    },
+    {
+      "epoch": 0.27332269249821123,
+      "grad_norm": 0.7165781855583191,
+      "learning_rate": 9.561665697139952e-06,
+      "loss": 0.8139,
+      "step": 4966
+    },
+    {
+      "epoch": 0.2733777313005669,
+      "grad_norm": 0.7481133341789246,
+      "learning_rate": 9.561488197652684e-06,
+      "loss": 0.7712,
+      "step": 4967
+    },
+    {
+      "epoch": 0.27343277010292255,
+      "grad_norm": 0.6928209066390991,
+      "learning_rate": 9.561310663882491e-06,
+      "loss": 0.7524,
+      "step": 4968
+    },
+    {
+      "epoch": 0.27348780890527824,
+      "grad_norm": 0.7397856116294861,
+      "learning_rate": 9.561133095830708e-06,
+      "loss": 0.718,
+      "step": 4969
+    },
+    {
+      "epoch": 0.2735428477076339,
+      "grad_norm": 0.7712383270263672,
+      "learning_rate": 9.560955493498672e-06,
+      "loss": 0.8201,
+      "step": 4970
+    },
+    {
+      "epoch": 0.27359788650998956,
+      "grad_norm": 0.96076899766922,
+      "learning_rate": 9.560777856887714e-06,
+      "loss": 0.8555,
+      "step": 4971
+    },
+    {
+      "epoch": 0.2736529253123452,
+      "grad_norm": 0.7331019639968872,
+      "learning_rate": 9.56060018599917e-06,
+      "loss": 0.8315,
+      "step": 4972
+    },
+    {
+      "epoch": 0.2737079641147009,
+      "grad_norm": 0.7157140970230103,
+      "learning_rate": 9.560422480834374e-06,
+      "loss": 0.7177,
+      "step": 4973
+    },
+    {
+      "epoch": 0.2737630029170565,
+      "grad_norm": 0.807614266872406,
+      "learning_rate": 9.560244741394666e-06,
+      "loss": 0.8413,
+      "step": 4974
+    },
+    {
+      "epoch": 0.2738180417194122,
+      "grad_norm": 0.7618574500083923,
+      "learning_rate": 9.560066967681378e-06,
+      "loss": 0.8248,
+      "step": 4975
+    },
+    {
+      "epoch": 0.27387308052176784,
+      "grad_norm": 0.7886885404586792,
+      "learning_rate": 9.559889159695848e-06,
+      "loss": 0.8793,
+      "step": 4976
+    },
+    {
+      "epoch": 0.27392811932412353,
+      "grad_norm": 1.0090755224227905,
+      "learning_rate": 9.559711317439411e-06,
+      "loss": 0.9255,
+      "step": 4977
+    },
+    {
+      "epoch": 0.27398315812647916,
+      "grad_norm": 0.7855443358421326,
+      "learning_rate": 9.559533440913405e-06,
+      "loss": 0.8001,
+      "step": 4978
+    },
+    {
+      "epoch": 0.27403819692883485,
+      "grad_norm": 0.768741250038147,
+      "learning_rate": 9.559355530119165e-06,
+      "loss": 0.8109,
+      "step": 4979
+    },
+    {
+      "epoch": 0.2740932357311905,
+      "grad_norm": 0.759589672088623,
+      "learning_rate": 9.55917758505803e-06,
+      "loss": 0.8001,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2741482745335462,
+      "grad_norm": 0.7937445640563965,
+      "learning_rate": 9.558999605731338e-06,
+      "loss": 0.8924,
+      "step": 4981
+    },
+    {
+      "epoch": 0.2742033133359018,
+      "grad_norm": 0.9041592478752136,
+      "learning_rate": 9.558821592140423e-06,
+      "loss": 0.9167,
+      "step": 4982
+    },
+    {
+      "epoch": 0.27425835213825744,
+      "grad_norm": 0.6971380710601807,
+      "learning_rate": 9.558643544286627e-06,
+      "loss": 0.7589,
+      "step": 4983
+    },
+    {
+      "epoch": 0.27431339094061313,
+      "grad_norm": 0.9292929172515869,
+      "learning_rate": 9.558465462171287e-06,
+      "loss": 0.9566,
+      "step": 4984
+    },
+    {
+      "epoch": 0.27436842974296877,
+      "grad_norm": 0.8320629000663757,
+      "learning_rate": 9.558287345795738e-06,
+      "loss": 0.8854,
+      "step": 4985
+    },
+    {
+      "epoch": 0.27442346854532446,
+      "grad_norm": 0.797272801399231,
+      "learning_rate": 9.558109195161325e-06,
+      "loss": 0.7838,
+      "step": 4986
+    },
+    {
+      "epoch": 0.2744785073476801,
+      "grad_norm": 0.9702700972557068,
+      "learning_rate": 9.557931010269382e-06,
+      "loss": 0.8593,
+      "step": 4987
+    },
+    {
+      "epoch": 0.2745335461500358,
+      "grad_norm": 0.8309103846549988,
+      "learning_rate": 9.557752791121248e-06,
+      "loss": 0.8902,
+      "step": 4988
+    },
+    {
+      "epoch": 0.2745885849523914,
+      "grad_norm": 0.706667959690094,
+      "learning_rate": 9.557574537718265e-06,
+      "loss": 0.7259,
+      "step": 4989
+    },
+    {
+      "epoch": 0.2746436237547471,
+      "grad_norm": 0.770239531993866,
+      "learning_rate": 9.557396250061771e-06,
+      "loss": 0.8644,
+      "step": 4990
+    },
+    {
+      "epoch": 0.27469866255710274,
+      "grad_norm": 0.8695803880691528,
+      "learning_rate": 9.557217928153108e-06,
+      "loss": 0.895,
+      "step": 4991
+    },
+    {
+      "epoch": 0.2747537013594584,
+      "grad_norm": 0.7525948286056519,
+      "learning_rate": 9.557039571993614e-06,
+      "loss": 0.7029,
+      "step": 4992
+    },
+    {
+      "epoch": 0.27480874016181406,
+      "grad_norm": 0.7616680264472961,
+      "learning_rate": 9.556861181584631e-06,
+      "loss": 0.8025,
+      "step": 4993
+    },
+    {
+      "epoch": 0.27486377896416975,
+      "grad_norm": 0.7216167449951172,
+      "learning_rate": 9.5566827569275e-06,
+      "loss": 0.8314,
+      "step": 4994
+    },
+    {
+      "epoch": 0.2749188177665254,
+      "grad_norm": 0.7412614226341248,
+      "learning_rate": 9.55650429802356e-06,
+      "loss": 0.7877,
+      "step": 4995
+    },
+    {
+      "epoch": 0.27497385656888107,
+      "grad_norm": 0.7176525592803955,
+      "learning_rate": 9.556325804874154e-06,
+      "loss": 0.7615,
+      "step": 4996
+    },
+    {
+      "epoch": 0.2750288953712367,
+      "grad_norm": 0.7544515132904053,
+      "learning_rate": 9.556147277480623e-06,
+      "loss": 0.8352,
+      "step": 4997
+    },
+    {
+      "epoch": 0.2750839341735924,
+      "grad_norm": 0.7318205833435059,
+      "learning_rate": 9.555968715844309e-06,
+      "loss": 0.7403,
+      "step": 4998
+    },
+    {
+      "epoch": 0.275138972975948,
+      "grad_norm": 0.7495027780532837,
+      "learning_rate": 9.555790119966552e-06,
+      "loss": 0.7611,
+      "step": 4999
+    },
+    {
+      "epoch": 0.2751940117783037,
+      "grad_norm": 0.7544401288032532,
+      "learning_rate": 9.555611489848697e-06,
+      "loss": 0.8594,
+      "step": 5000
+    },
+    {
+      "epoch": 0.27524905058065935,
+      "grad_norm": 0.7698250412940979,
+      "learning_rate": 9.555432825492084e-06,
+      "loss": 0.8438,
+      "step": 5001
+    },
+    {
+      "epoch": 0.27530408938301504,
+      "grad_norm": 0.7668892741203308,
+      "learning_rate": 9.555254126898059e-06,
+      "loss": 0.8082,
+      "step": 5002
+    },
+    {
+      "epoch": 0.27535912818537067,
+      "grad_norm": 0.9170669317245483,
+      "learning_rate": 9.555075394067963e-06,
+      "loss": 0.7443,
+      "step": 5003
+    },
+    {
+      "epoch": 0.27541416698772636,
+      "grad_norm": 0.7890255451202393,
+      "learning_rate": 9.55489662700314e-06,
+      "loss": 0.8269,
+      "step": 5004
+    },
+    {
+      "epoch": 0.275469205790082,
+      "grad_norm": 0.6740512847900391,
+      "learning_rate": 9.554717825704932e-06,
+      "loss": 0.6906,
+      "step": 5005
+    },
+    {
+      "epoch": 0.2755242445924377,
+      "grad_norm": 0.8032376170158386,
+      "learning_rate": 9.554538990174685e-06,
+      "loss": 0.812,
+      "step": 5006
+    },
+    {
+      "epoch": 0.2755792833947933,
+      "grad_norm": 0.6932135224342346,
+      "learning_rate": 9.554360120413741e-06,
+      "loss": 0.7823,
+      "step": 5007
+    },
+    {
+      "epoch": 0.275634322197149,
+      "grad_norm": 0.7447643876075745,
+      "learning_rate": 9.554181216423447e-06,
+      "loss": 0.8753,
+      "step": 5008
+    },
+    {
+      "epoch": 0.27568936099950464,
+      "grad_norm": 0.8035081624984741,
+      "learning_rate": 9.554002278205145e-06,
+      "loss": 0.7135,
+      "step": 5009
+    },
+    {
+      "epoch": 0.27574439980186033,
+      "grad_norm": 0.7544171214103699,
+      "learning_rate": 9.553823305760182e-06,
+      "loss": 0.7574,
+      "step": 5010
+    },
+    {
+      "epoch": 0.27579943860421596,
+      "grad_norm": 0.6648419499397278,
+      "learning_rate": 9.553644299089902e-06,
+      "loss": 0.7566,
+      "step": 5011
+    },
+    {
+      "epoch": 0.27585447740657165,
+      "grad_norm": 0.7481752038002014,
+      "learning_rate": 9.55346525819565e-06,
+      "loss": 0.7862,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2759095162089273,
+      "grad_norm": 0.7000668048858643,
+      "learning_rate": 9.55328618307877e-06,
+      "loss": 0.7767,
+      "step": 5013
+    },
+    {
+      "epoch": 0.275964555011283,
+      "grad_norm": 0.7435166239738464,
+      "learning_rate": 9.553107073740612e-06,
+      "loss": 0.6888,
+      "step": 5014
+    },
+    {
+      "epoch": 0.2760195938136386,
+      "grad_norm": 0.7593170404434204,
+      "learning_rate": 9.552927930182521e-06,
+      "loss": 0.7272,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2760746326159943,
+      "grad_norm": 0.870079755783081,
+      "learning_rate": 9.55274875240584e-06,
+      "loss": 0.8692,
+      "step": 5016
+    },
+    {
+      "epoch": 0.27612967141834993,
+      "grad_norm": 0.8550307750701904,
+      "learning_rate": 9.55256954041192e-06,
+      "loss": 0.8729,
+      "step": 5017
+    },
+    {
+      "epoch": 0.2761847102207056,
+      "grad_norm": 0.888830304145813,
+      "learning_rate": 9.552390294202105e-06,
+      "loss": 0.8607,
+      "step": 5018
+    },
+    {
+      "epoch": 0.27623974902306125,
+      "grad_norm": 0.8295729160308838,
+      "learning_rate": 9.552211013777743e-06,
+      "loss": 0.8722,
+      "step": 5019
+    },
+    {
+      "epoch": 0.27629478782541694,
+      "grad_norm": 0.7732356190681458,
+      "learning_rate": 9.552031699140182e-06,
+      "loss": 0.8332,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2763498266277726,
+      "grad_norm": 0.9132987856864929,
+      "learning_rate": 9.55185235029077e-06,
+      "loss": 0.769,
+      "step": 5021
+    },
+    {
+      "epoch": 0.27640486543012827,
+      "grad_norm": 0.7221076488494873,
+      "learning_rate": 9.551672967230851e-06,
+      "loss": 0.8505,
+      "step": 5022
+    },
+    {
+      "epoch": 0.2764599042324839,
+      "grad_norm": 0.8526949882507324,
+      "learning_rate": 9.551493549961778e-06,
+      "loss": 0.8002,
+      "step": 5023
+    },
+    {
+      "epoch": 0.2765149430348396,
+      "grad_norm": 0.9513188004493713,
+      "learning_rate": 9.551314098484901e-06,
+      "loss": 0.8558,
+      "step": 5024
+    },
+    {
+      "epoch": 0.2765699818371952,
+      "grad_norm": 0.7543003559112549,
+      "learning_rate": 9.551134612801563e-06,
+      "loss": 0.8292,
+      "step": 5025
+    },
+    {
+      "epoch": 0.27662502063955086,
+      "grad_norm": 0.7531017065048218,
+      "learning_rate": 9.550955092913115e-06,
+      "loss": 0.7837,
+      "step": 5026
+    },
+    {
+      "epoch": 0.27668005944190655,
+      "grad_norm": 0.8725717663764954,
+      "learning_rate": 9.550775538820907e-06,
+      "loss": 0.8362,
+      "step": 5027
+    },
+    {
+      "epoch": 0.2767350982442622,
+      "grad_norm": 0.8122721910476685,
+      "learning_rate": 9.550595950526288e-06,
+      "loss": 0.8539,
+      "step": 5028
+    },
+    {
+      "epoch": 0.27679013704661787,
+      "grad_norm": 0.7756829261779785,
+      "learning_rate": 9.550416328030608e-06,
+      "loss": 0.787,
+      "step": 5029
+    },
+    {
+      "epoch": 0.2768451758489735,
+      "grad_norm": 0.9086001515388489,
+      "learning_rate": 9.550236671335218e-06,
+      "loss": 0.7972,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2769002146513292,
+      "grad_norm": 0.7857060432434082,
+      "learning_rate": 9.550056980441466e-06,
+      "loss": 0.7577,
+      "step": 5031
+    },
+    {
+      "epoch": 0.2769552534536848,
+      "grad_norm": 0.8190392851829529,
+      "learning_rate": 9.549877255350703e-06,
+      "loss": 0.81,
+      "step": 5032
+    },
+    {
+      "epoch": 0.2770102922560405,
+      "grad_norm": 0.7714588642120361,
+      "learning_rate": 9.549697496064283e-06,
+      "loss": 0.7916,
+      "step": 5033
+    },
+    {
+      "epoch": 0.27706533105839615,
+      "grad_norm": 0.7178533673286438,
+      "learning_rate": 9.549517702583552e-06,
+      "loss": 0.8001,
+      "step": 5034
+    },
+    {
+      "epoch": 0.27712036986075184,
+      "grad_norm": 0.7552955150604248,
+      "learning_rate": 9.549337874909865e-06,
+      "loss": 0.8361,
+      "step": 5035
+    },
+    {
+      "epoch": 0.27717540866310747,
+      "grad_norm": 0.7823992371559143,
+      "learning_rate": 9.549158013044573e-06,
+      "loss": 0.7033,
+      "step": 5036
+    },
+    {
+      "epoch": 0.27723044746546316,
+      "grad_norm": 0.731504499912262,
+      "learning_rate": 9.548978116989026e-06,
+      "loss": 0.73,
+      "step": 5037
+    },
+    {
+      "epoch": 0.2772854862678188,
+      "grad_norm": 0.7455994486808777,
+      "learning_rate": 9.548798186744578e-06,
+      "loss": 0.8005,
+      "step": 5038
+    },
+    {
+      "epoch": 0.2773405250701745,
+      "grad_norm": 0.7020164728164673,
+      "learning_rate": 9.54861822231258e-06,
+      "loss": 0.6707,
+      "step": 5039
+    },
+    {
+      "epoch": 0.2773955638725301,
+      "grad_norm": 0.7526360750198364,
+      "learning_rate": 9.548438223694385e-06,
+      "loss": 0.7686,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2774506026748858,
+      "grad_norm": 0.7268579006195068,
+      "learning_rate": 9.548258190891344e-06,
+      "loss": 0.7039,
+      "step": 5041
+    },
+    {
+      "epoch": 0.27750564147724144,
+      "grad_norm": 0.9361631274223328,
+      "learning_rate": 9.548078123904815e-06,
+      "loss": 0.8023,
+      "step": 5042
+    },
+    {
+      "epoch": 0.2775606802795971,
+      "grad_norm": 0.7786710262298584,
+      "learning_rate": 9.547898022736147e-06,
+      "loss": 0.6866,
+      "step": 5043
+    },
+    {
+      "epoch": 0.27761571908195276,
+      "grad_norm": 0.7175624370574951,
+      "learning_rate": 9.547717887386695e-06,
+      "loss": 0.7554,
+      "step": 5044
+    },
+    {
+      "epoch": 0.27767075788430845,
+      "grad_norm": 0.9157657623291016,
+      "learning_rate": 9.547537717857813e-06,
+      "loss": 0.7936,
+      "step": 5045
+    },
+    {
+      "epoch": 0.2777257966866641,
+      "grad_norm": 0.7881377935409546,
+      "learning_rate": 9.547357514150854e-06,
+      "loss": 0.8198,
+      "step": 5046
+    },
+    {
+      "epoch": 0.2777808354890198,
+      "grad_norm": 1.0444039106369019,
+      "learning_rate": 9.547177276267173e-06,
+      "loss": 0.7954,
+      "step": 5047
+    },
+    {
+      "epoch": 0.2778358742913754,
+      "grad_norm": 0.7889506220817566,
+      "learning_rate": 9.546997004208124e-06,
+      "loss": 0.7697,
+      "step": 5048
+    },
+    {
+      "epoch": 0.2778909130937311,
+      "grad_norm": 0.7304134368896484,
+      "learning_rate": 9.546816697975066e-06,
+      "loss": 0.7034,
+      "step": 5049
+    },
+    {
+      "epoch": 0.27794595189608673,
+      "grad_norm": 0.7783082723617554,
+      "learning_rate": 9.546636357569347e-06,
+      "loss": 0.8185,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2780009906984424,
+      "grad_norm": 0.750712513923645,
+      "learning_rate": 9.54645598299233e-06,
+      "loss": 0.7336,
+      "step": 5051
+    },
+    {
+      "epoch": 0.27805602950079805,
+      "grad_norm": 0.7849590182304382,
+      "learning_rate": 9.546275574245364e-06,
+      "loss": 0.8088,
+      "step": 5052
+    },
+    {
+      "epoch": 0.27811106830315374,
+      "grad_norm": 0.8490208983421326,
+      "learning_rate": 9.546095131329809e-06,
+      "loss": 0.8507,
+      "step": 5053
+    },
+    {
+      "epoch": 0.2781661071055094,
+      "grad_norm": 0.8107250928878784,
+      "learning_rate": 9.54591465424702e-06,
+      "loss": 0.7787,
+      "step": 5054
+    },
+    {
+      "epoch": 0.27822114590786506,
+      "grad_norm": 0.8278594613075256,
+      "learning_rate": 9.54573414299835e-06,
+      "loss": 0.7836,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2782761847102207,
+      "grad_norm": 0.7982015013694763,
+      "learning_rate": 9.545553597585163e-06,
+      "loss": 0.7672,
+      "step": 5056
+    },
+    {
+      "epoch": 0.2783312235125764,
+      "grad_norm": 0.7311522364616394,
+      "learning_rate": 9.54537301800881e-06,
+      "loss": 0.7571,
+      "step": 5057
+    },
+    {
+      "epoch": 0.278386262314932,
+      "grad_norm": 0.8039999604225159,
+      "learning_rate": 9.545192404270651e-06,
+      "loss": 0.764,
+      "step": 5058
+    },
+    {
+      "epoch": 0.2784413011172877,
+      "grad_norm": 0.7810946702957153,
+      "learning_rate": 9.545011756372042e-06,
+      "loss": 0.9217,
+      "step": 5059
+    },
+    {
+      "epoch": 0.27849633991964334,
+      "grad_norm": 0.7092248797416687,
+      "learning_rate": 9.544831074314343e-06,
+      "loss": 0.7599,
+      "step": 5060
+    },
+    {
+      "epoch": 0.27855137872199903,
+      "grad_norm": 0.831550657749176,
+      "learning_rate": 9.544650358098908e-06,
+      "loss": 0.7278,
+      "step": 5061
+    },
+    {
+      "epoch": 0.27860641752435467,
+      "grad_norm": 0.7645474076271057,
+      "learning_rate": 9.544469607727098e-06,
+      "loss": 0.7945,
+      "step": 5062
+    },
+    {
+      "epoch": 0.27866145632671036,
+      "grad_norm": 0.6956788301467896,
+      "learning_rate": 9.544288823200273e-06,
+      "loss": 0.749,
+      "step": 5063
+    },
+    {
+      "epoch": 0.278716495129066,
+      "grad_norm": 0.7262974381446838,
+      "learning_rate": 9.544108004519786e-06,
+      "loss": 0.8074,
+      "step": 5064
+    },
+    {
+      "epoch": 0.2787715339314217,
+      "grad_norm": 0.7439202666282654,
+      "learning_rate": 9.543927151687001e-06,
+      "loss": 0.9403,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2788265727337773,
+      "grad_norm": 0.8468778133392334,
+      "learning_rate": 9.543746264703277e-06,
+      "loss": 0.8182,
+      "step": 5066
+    },
+    {
+      "epoch": 0.278881611536133,
+      "grad_norm": 0.8396204113960266,
+      "learning_rate": 9.54356534356997e-06,
+      "loss": 0.8067,
+      "step": 5067
+    },
+    {
+      "epoch": 0.27893665033848863,
+      "grad_norm": 0.718758225440979,
+      "learning_rate": 9.543384388288445e-06,
+      "loss": 0.8172,
+      "step": 5068
+    },
+    {
+      "epoch": 0.27899168914084427,
+      "grad_norm": 0.7562685012817383,
+      "learning_rate": 9.543203398860056e-06,
+      "loss": 0.9053,
+      "step": 5069
+    },
+    {
+      "epoch": 0.27904672794319996,
+      "grad_norm": 0.9592792987823486,
+      "learning_rate": 9.543022375286169e-06,
+      "loss": 0.9375,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2791017667455556,
+      "grad_norm": 0.7162739634513855,
+      "learning_rate": 9.54284131756814e-06,
+      "loss": 0.7297,
+      "step": 5071
+    },
+    {
+      "epoch": 0.2791568055479113,
+      "grad_norm": 0.7703517079353333,
+      "learning_rate": 9.542660225707335e-06,
+      "loss": 0.8863,
+      "step": 5072
+    },
+    {
+      "epoch": 0.2792118443502669,
+      "grad_norm": 0.7860418558120728,
+      "learning_rate": 9.542479099705109e-06,
+      "loss": 0.8335,
+      "step": 5073
+    },
+    {
+      "epoch": 0.2792668831526226,
+      "grad_norm": 0.8880825042724609,
+      "learning_rate": 9.542297939562825e-06,
+      "loss": 0.8344,
+      "step": 5074
+    },
+    {
+      "epoch": 0.27932192195497824,
+      "grad_norm": 0.7900505661964417,
+      "learning_rate": 9.542116745281849e-06,
+      "loss": 0.7613,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2793769607573339,
+      "grad_norm": 0.7446081042289734,
+      "learning_rate": 9.541935516863536e-06,
+      "loss": 0.6615,
+      "step": 5076
+    },
+    {
+      "epoch": 0.27943199955968956,
+      "grad_norm": 0.7831308245658875,
+      "learning_rate": 9.541754254309254e-06,
+      "loss": 0.779,
+      "step": 5077
+    },
+    {
+      "epoch": 0.27948703836204525,
+      "grad_norm": 0.9007606506347656,
+      "learning_rate": 9.541572957620361e-06,
+      "loss": 0.8883,
+      "step": 5078
+    },
+    {
+      "epoch": 0.2795420771644009,
+      "grad_norm": 0.8033407330513,
+      "learning_rate": 9.541391626798222e-06,
+      "loss": 0.7354,
+      "step": 5079
+    },
+    {
+      "epoch": 0.27959711596675657,
+      "grad_norm": 0.9259470105171204,
+      "learning_rate": 9.5412102618442e-06,
+      "loss": 0.7602,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2796521547691122,
+      "grad_norm": 0.786523163318634,
+      "learning_rate": 9.541028862759656e-06,
+      "loss": 0.7402,
+      "step": 5081
+    },
+    {
+      "epoch": 0.2797071935714679,
+      "grad_norm": 0.8053372502326965,
+      "learning_rate": 9.540847429545954e-06,
+      "loss": 0.825,
+      "step": 5082
+    },
+    {
+      "epoch": 0.2797622323738235,
+      "grad_norm": 0.8578022122383118,
+      "learning_rate": 9.54066596220446e-06,
+      "loss": 0.7866,
+      "step": 5083
+    },
+    {
+      "epoch": 0.2798172711761792,
+      "grad_norm": 0.916161835193634,
+      "learning_rate": 9.540484460736535e-06,
+      "loss": 0.5961,
+      "step": 5084
+    },
+    {
+      "epoch": 0.27987230997853485,
+      "grad_norm": 0.7843562960624695,
+      "learning_rate": 9.540302925143545e-06,
+      "loss": 0.764,
+      "step": 5085
+    },
+    {
+      "epoch": 0.27992734878089054,
+      "grad_norm": 0.7392510771751404,
+      "learning_rate": 9.540121355426852e-06,
+      "loss": 0.8038,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2799823875832462,
+      "grad_norm": 0.7406296133995056,
+      "learning_rate": 9.539939751587825e-06,
+      "loss": 0.8202,
+      "step": 5087
+    },
+    {
+      "epoch": 0.28003742638560186,
+      "grad_norm": 0.7274924516677856,
+      "learning_rate": 9.539758113627823e-06,
+      "loss": 0.7691,
+      "step": 5088
+    },
+    {
+      "epoch": 0.2800924651879575,
+      "grad_norm": 0.8563184142112732,
+      "learning_rate": 9.539576441548218e-06,
+      "loss": 0.8341,
+      "step": 5089
+    },
+    {
+      "epoch": 0.2801475039903132,
+      "grad_norm": 0.7708351016044617,
+      "learning_rate": 9.539394735350366e-06,
+      "loss": 0.7126,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2802025427926688,
+      "grad_norm": 0.7314836382865906,
+      "learning_rate": 9.539212995035642e-06,
+      "loss": 0.7465,
+      "step": 5091
+    },
+    {
+      "epoch": 0.2802575815950245,
+      "grad_norm": 0.7594754695892334,
+      "learning_rate": 9.539031220605409e-06,
+      "loss": 0.7563,
+      "step": 5092
+    },
+    {
+      "epoch": 0.28031262039738014,
+      "grad_norm": 0.699414074420929,
+      "learning_rate": 9.53884941206103e-06,
+      "loss": 0.7847,
+      "step": 5093
+    },
+    {
+      "epoch": 0.28036765919973583,
+      "grad_norm": 0.8013063073158264,
+      "learning_rate": 9.538667569403877e-06,
+      "loss": 0.7769,
+      "step": 5094
+    },
+    {
+      "epoch": 0.28042269800209146,
+      "grad_norm": 0.7778805494308472,
+      "learning_rate": 9.538485692635312e-06,
+      "loss": 0.7646,
+      "step": 5095
+    },
+    {
+      "epoch": 0.28047773680444715,
+      "grad_norm": 0.785649299621582,
+      "learning_rate": 9.538303781756702e-06,
+      "loss": 0.8162,
+      "step": 5096
+    },
+    {
+      "epoch": 0.2805327756068028,
+      "grad_norm": 0.7073212265968323,
+      "learning_rate": 9.538121836769417e-06,
+      "loss": 0.7208,
+      "step": 5097
+    },
+    {
+      "epoch": 0.2805878144091585,
+      "grad_norm": 0.7545642852783203,
+      "learning_rate": 9.53793985767482e-06,
+      "loss": 0.8673,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2806428532115141,
+      "grad_norm": 0.6818416118621826,
+      "learning_rate": 9.537757844474285e-06,
+      "loss": 0.7576,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2806978920138698,
+      "grad_norm": 0.6718038320541382,
+      "learning_rate": 9.537575797169176e-06,
+      "loss": 0.6683,
+      "step": 5100
+    },
+    {
+      "epoch": 0.28075293081622543,
+      "grad_norm": 0.7851004600524902,
+      "learning_rate": 9.53739371576086e-06,
+      "loss": 0.8871,
+      "step": 5101
+    },
+    {
+      "epoch": 0.2808079696185811,
+      "grad_norm": 0.7565650343894958,
+      "learning_rate": 9.53721160025071e-06,
+      "loss": 0.8799,
+      "step": 5102
+    },
+    {
+      "epoch": 0.28086300842093676,
+      "grad_norm": 0.7522932887077332,
+      "learning_rate": 9.537029450640091e-06,
+      "loss": 0.838,
+      "step": 5103
+    },
+    {
+      "epoch": 0.28091804722329244,
+      "grad_norm": 0.929634690284729,
+      "learning_rate": 9.536847266930375e-06,
+      "loss": 0.7997,
+      "step": 5104
+    },
+    {
+      "epoch": 0.2809730860256481,
+      "grad_norm": 0.8050084710121155,
+      "learning_rate": 9.536665049122928e-06,
+      "loss": 0.7652,
+      "step": 5105
+    },
+    {
+      "epoch": 0.28102812482800377,
+      "grad_norm": 0.7401233315467834,
+      "learning_rate": 9.53648279721912e-06,
+      "loss": 0.7904,
+      "step": 5106
+    },
+    {
+      "epoch": 0.2810831636303594,
+      "grad_norm": 0.7125453948974609,
+      "learning_rate": 9.536300511220322e-06,
+      "loss": 0.7349,
+      "step": 5107
+    },
+    {
+      "epoch": 0.2811382024327151,
+      "grad_norm": 0.7165758609771729,
+      "learning_rate": 9.536118191127905e-06,
+      "loss": 0.7314,
+      "step": 5108
+    },
+    {
+      "epoch": 0.2811932412350707,
+      "grad_norm": 0.7507439851760864,
+      "learning_rate": 9.535935836943237e-06,
+      "loss": 0.7603,
+      "step": 5109
+    },
+    {
+      "epoch": 0.2812482800374264,
+      "grad_norm": 0.7832109332084656,
+      "learning_rate": 9.535753448667688e-06,
+      "loss": 0.7279,
+      "step": 5110
+    },
+    {
+      "epoch": 0.28130331883978205,
+      "grad_norm": 0.7346609234809875,
+      "learning_rate": 9.535571026302633e-06,
+      "loss": 0.6882,
+      "step": 5111
+    },
+    {
+      "epoch": 0.2813583576421377,
+      "grad_norm": 0.7569608688354492,
+      "learning_rate": 9.535388569849437e-06,
+      "loss": 0.8451,
+      "step": 5112
+    },
+    {
+      "epoch": 0.28141339644449337,
+      "grad_norm": 0.7319865822792053,
+      "learning_rate": 9.535206079309478e-06,
+      "loss": 0.8161,
+      "step": 5113
+    },
+    {
+      "epoch": 0.281468435246849,
+      "grad_norm": 0.7744631171226501,
+      "learning_rate": 9.535023554684122e-06,
+      "loss": 0.8025,
+      "step": 5114
+    },
+    {
+      "epoch": 0.2815234740492047,
+      "grad_norm": 0.6867525577545166,
+      "learning_rate": 9.534840995974743e-06,
+      "loss": 0.7693,
+      "step": 5115
+    },
+    {
+      "epoch": 0.2815785128515603,
+      "grad_norm": 0.7625848054885864,
+      "learning_rate": 9.534658403182715e-06,
+      "loss": 0.8034,
+      "step": 5116
+    },
+    {
+      "epoch": 0.281633551653916,
+      "grad_norm": 0.7369832992553711,
+      "learning_rate": 9.534475776309406e-06,
+      "loss": 0.873,
+      "step": 5117
+    },
+    {
+      "epoch": 0.28168859045627165,
+      "grad_norm": 0.7267127633094788,
+      "learning_rate": 9.534293115356191e-06,
+      "loss": 0.7954,
+      "step": 5118
+    },
+    {
+      "epoch": 0.28174362925862734,
+      "grad_norm": 0.7244247794151306,
+      "learning_rate": 9.534110420324443e-06,
+      "loss": 0.7784,
+      "step": 5119
+    },
+    {
+      "epoch": 0.28179866806098297,
+      "grad_norm": 0.8207812905311584,
+      "learning_rate": 9.533927691215534e-06,
+      "loss": 0.8696,
+      "step": 5120
+    },
+    {
+      "epoch": 0.28185370686333866,
+      "grad_norm": 0.8669891357421875,
+      "learning_rate": 9.53374492803084e-06,
+      "loss": 0.8203,
+      "step": 5121
+    },
+    {
+      "epoch": 0.2819087456656943,
+      "grad_norm": 0.7650816440582275,
+      "learning_rate": 9.533562130771732e-06,
+      "loss": 0.77,
+      "step": 5122
+    },
+    {
+      "epoch": 0.28196378446805,
+      "grad_norm": 0.7664972543716431,
+      "learning_rate": 9.533379299439584e-06,
+      "loss": 0.7187,
+      "step": 5123
+    },
+    {
+      "epoch": 0.2820188232704056,
+      "grad_norm": 0.7921896576881409,
+      "learning_rate": 9.533196434035772e-06,
+      "loss": 0.8669,
+      "step": 5124
+    },
+    {
+      "epoch": 0.2820738620727613,
+      "grad_norm": 0.7714456915855408,
+      "learning_rate": 9.533013534561669e-06,
+      "loss": 0.8783,
+      "step": 5125
+    },
+    {
+      "epoch": 0.28212890087511694,
+      "grad_norm": 0.7222065329551697,
+      "learning_rate": 9.532830601018648e-06,
+      "loss": 0.7449,
+      "step": 5126
+    },
+    {
+      "epoch": 0.28218393967747263,
+      "grad_norm": 0.718142569065094,
+      "learning_rate": 9.532647633408085e-06,
+      "loss": 0.8226,
+      "step": 5127
+    },
+    {
+      "epoch": 0.28223897847982826,
+      "grad_norm": 0.730592668056488,
+      "learning_rate": 9.532464631731357e-06,
+      "loss": 0.7878,
+      "step": 5128
+    },
+    {
+      "epoch": 0.28229401728218395,
+      "grad_norm": 0.7841802835464478,
+      "learning_rate": 9.532281595989839e-06,
+      "loss": 0.8262,
+      "step": 5129
+    },
+    {
+      "epoch": 0.2823490560845396,
+      "grad_norm": 0.8617212772369385,
+      "learning_rate": 9.532098526184904e-06,
+      "loss": 0.8368,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2824040948868953,
+      "grad_norm": 0.6968556642532349,
+      "learning_rate": 9.53191542231793e-06,
+      "loss": 0.6848,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2824591336892509,
+      "grad_norm": 0.7872157096862793,
+      "learning_rate": 9.531732284390294e-06,
+      "loss": 0.7898,
+      "step": 5132
+    },
+    {
+      "epoch": 0.2825141724916066,
+      "grad_norm": 0.7727276086807251,
+      "learning_rate": 9.53154911240337e-06,
+      "loss": 0.8506,
+      "step": 5133
+    },
+    {
+      "epoch": 0.28256921129396223,
+      "grad_norm": 0.7279896140098572,
+      "learning_rate": 9.531365906358536e-06,
+      "loss": 0.7415,
+      "step": 5134
+    },
+    {
+      "epoch": 0.2826242500963179,
+      "grad_norm": 0.7457457780838013,
+      "learning_rate": 9.53118266625717e-06,
+      "loss": 0.7652,
+      "step": 5135
+    },
+    {
+      "epoch": 0.28267928889867355,
+      "grad_norm": 0.8989270329475403,
+      "learning_rate": 9.530999392100646e-06,
+      "loss": 0.9085,
+      "step": 5136
+    },
+    {
+      "epoch": 0.28273432770102924,
+      "grad_norm": 0.9622626304626465,
+      "learning_rate": 9.530816083890347e-06,
+      "loss": 0.8726,
+      "step": 5137
+    },
+    {
+      "epoch": 0.2827893665033849,
+      "grad_norm": 0.7712846994400024,
+      "learning_rate": 9.530632741627643e-06,
+      "loss": 0.765,
+      "step": 5138
+    },
+    {
+      "epoch": 0.28284440530574056,
+      "grad_norm": 0.8320727348327637,
+      "learning_rate": 9.530449365313918e-06,
+      "loss": 0.7828,
+      "step": 5139
+    },
+    {
+      "epoch": 0.2828994441080962,
+      "grad_norm": 0.9310963153839111,
+      "learning_rate": 9.530265954950549e-06,
+      "loss": 0.8482,
+      "step": 5140
+    },
+    {
+      "epoch": 0.2829544829104519,
+      "grad_norm": 0.9984502792358398,
+      "learning_rate": 9.530082510538914e-06,
+      "loss": 0.8673,
+      "step": 5141
+    },
+    {
+      "epoch": 0.2830095217128075,
+      "grad_norm": 0.8300992250442505,
+      "learning_rate": 9.52989903208039e-06,
+      "loss": 0.8232,
+      "step": 5142
+    },
+    {
+      "epoch": 0.2830645605151632,
+      "grad_norm": 0.930052638053894,
+      "learning_rate": 9.529715519576356e-06,
+      "loss": 0.7766,
+      "step": 5143
+    },
+    {
+      "epoch": 0.28311959931751884,
+      "grad_norm": 0.8038359880447388,
+      "learning_rate": 9.529531973028194e-06,
+      "loss": 0.712,
+      "step": 5144
+    },
+    {
+      "epoch": 0.28317463811987453,
+      "grad_norm": 0.856250524520874,
+      "learning_rate": 9.529348392437283e-06,
+      "loss": 0.8578,
+      "step": 5145
+    },
+    {
+      "epoch": 0.28322967692223017,
+      "grad_norm": 0.7602483630180359,
+      "learning_rate": 9.529164777805002e-06,
+      "loss": 0.749,
+      "step": 5146
+    },
+    {
+      "epoch": 0.28328471572458586,
+      "grad_norm": 0.8946549892425537,
+      "learning_rate": 9.52898112913273e-06,
+      "loss": 0.8101,
+      "step": 5147
+    },
+    {
+      "epoch": 0.2833397545269415,
+      "grad_norm": 0.8015615344047546,
+      "learning_rate": 9.52879744642185e-06,
+      "loss": 0.8203,
+      "step": 5148
+    },
+    {
+      "epoch": 0.2833947933292972,
+      "grad_norm": 0.7767183780670166,
+      "learning_rate": 9.528613729673738e-06,
+      "loss": 0.8409,
+      "step": 5149
+    },
+    {
+      "epoch": 0.2834498321316528,
+      "grad_norm": 0.7604000568389893,
+      "learning_rate": 9.52842997888978e-06,
+      "loss": 0.8853,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2835048709340085,
+      "grad_norm": 0.7079401016235352,
+      "learning_rate": 9.528246194071353e-06,
+      "loss": 0.6855,
+      "step": 5151
+    },
+    {
+      "epoch": 0.28355990973636414,
+      "grad_norm": 0.7616782188415527,
+      "learning_rate": 9.52806237521984e-06,
+      "loss": 0.785,
+      "step": 5152
+    },
+    {
+      "epoch": 0.2836149485387198,
+      "grad_norm": 0.7408583760261536,
+      "learning_rate": 9.527878522336622e-06,
+      "loss": 0.7105,
+      "step": 5153
+    },
+    {
+      "epoch": 0.28366998734107546,
+      "grad_norm": 0.694821834564209,
+      "learning_rate": 9.52769463542308e-06,
+      "loss": 0.6552,
+      "step": 5154
+    },
+    {
+      "epoch": 0.2837250261434311,
+      "grad_norm": 0.796925961971283,
+      "learning_rate": 9.5275107144806e-06,
+      "loss": 0.7122,
+      "step": 5155
+    },
+    {
+      "epoch": 0.2837800649457868,
+      "grad_norm": 0.8001971244812012,
+      "learning_rate": 9.527326759510558e-06,
+      "loss": 0.8528,
+      "step": 5156
+    },
+    {
+      "epoch": 0.2838351037481424,
+      "grad_norm": 0.8605831265449524,
+      "learning_rate": 9.527142770514341e-06,
+      "loss": 0.7948,
+      "step": 5157
+    },
+    {
+      "epoch": 0.2838901425504981,
+      "grad_norm": 0.8380078077316284,
+      "learning_rate": 9.526958747493334e-06,
+      "loss": 0.8184,
+      "step": 5158
+    },
+    {
+      "epoch": 0.28394518135285374,
+      "grad_norm": 0.8758485317230225,
+      "learning_rate": 9.526774690448913e-06,
+      "loss": 0.7625,
+      "step": 5159
+    },
+    {
+      "epoch": 0.2840002201552094,
+      "grad_norm": 0.7078989744186401,
+      "learning_rate": 9.526590599382466e-06,
+      "loss": 0.8179,
+      "step": 5160
+    },
+    {
+      "epoch": 0.28405525895756506,
+      "grad_norm": 0.6668990850448608,
+      "learning_rate": 9.526406474295376e-06,
+      "loss": 0.7169,
+      "step": 5161
+    },
+    {
+      "epoch": 0.28411029775992075,
+      "grad_norm": 0.7666084170341492,
+      "learning_rate": 9.526222315189026e-06,
+      "loss": 0.8511,
+      "step": 5162
+    },
+    {
+      "epoch": 0.2841653365622764,
+      "grad_norm": 0.7390545606613159,
+      "learning_rate": 9.526038122064802e-06,
+      "loss": 0.7926,
+      "step": 5163
+    },
+    {
+      "epoch": 0.28422037536463207,
+      "grad_norm": 0.7972092032432556,
+      "learning_rate": 9.525853894924086e-06,
+      "loss": 0.9166,
+      "step": 5164
+    },
+    {
+      "epoch": 0.2842754141669877,
+      "grad_norm": 0.8988455533981323,
+      "learning_rate": 9.525669633768265e-06,
+      "loss": 0.9497,
+      "step": 5165
+    },
+    {
+      "epoch": 0.2843304529693434,
+      "grad_norm": 0.7092710137367249,
+      "learning_rate": 9.525485338598722e-06,
+      "loss": 0.7241,
+      "step": 5166
+    },
+    {
+      "epoch": 0.28438549177169903,
+      "grad_norm": 0.8630063533782959,
+      "learning_rate": 9.525301009416843e-06,
+      "loss": 0.8318,
+      "step": 5167
+    },
+    {
+      "epoch": 0.2844405305740547,
+      "grad_norm": 0.7336890697479248,
+      "learning_rate": 9.52511664622401e-06,
+      "loss": 0.7077,
+      "step": 5168
+    },
+    {
+      "epoch": 0.28449556937641035,
+      "grad_norm": 0.8156722784042358,
+      "learning_rate": 9.524932249021615e-06,
+      "loss": 0.8573,
+      "step": 5169
+    },
+    {
+      "epoch": 0.28455060817876604,
+      "grad_norm": 0.7061388492584229,
+      "learning_rate": 9.524747817811038e-06,
+      "loss": 0.7432,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2846056469811217,
+      "grad_norm": 0.7948413491249084,
+      "learning_rate": 9.52456335259367e-06,
+      "loss": 0.8082,
+      "step": 5171
+    },
+    {
+      "epoch": 0.28466068578347736,
+      "grad_norm": 0.7208091020584106,
+      "learning_rate": 9.524378853370893e-06,
+      "loss": 0.7027,
+      "step": 5172
+    },
+    {
+      "epoch": 0.284715724585833,
+      "grad_norm": 0.8377540111541748,
+      "learning_rate": 9.524194320144096e-06,
+      "loss": 0.7093,
+      "step": 5173
+    },
+    {
+      "epoch": 0.2847707633881887,
+      "grad_norm": 0.8734563589096069,
+      "learning_rate": 9.524009752914666e-06,
+      "loss": 0.8422,
+      "step": 5174
+    },
+    {
+      "epoch": 0.2848258021905443,
+      "grad_norm": 0.7303940653800964,
+      "learning_rate": 9.523825151683989e-06,
+      "loss": 0.811,
+      "step": 5175
+    },
+    {
+      "epoch": 0.2848808409929,
+      "grad_norm": 0.7653842568397522,
+      "learning_rate": 9.523640516453455e-06,
+      "loss": 0.8595,
+      "step": 5176
+    },
+    {
+      "epoch": 0.28493587979525564,
+      "grad_norm": 0.7366930246353149,
+      "learning_rate": 9.523455847224448e-06,
+      "loss": 0.7832,
+      "step": 5177
+    },
+    {
+      "epoch": 0.28499091859761133,
+      "grad_norm": 0.7908505797386169,
+      "learning_rate": 9.523271143998357e-06,
+      "loss": 0.8115,
+      "step": 5178
+    },
+    {
+      "epoch": 0.28504595739996696,
+      "grad_norm": 0.8176048398017883,
+      "learning_rate": 9.523086406776572e-06,
+      "loss": 0.8377,
+      "step": 5179
+    },
+    {
+      "epoch": 0.28510099620232265,
+      "grad_norm": 0.724086344242096,
+      "learning_rate": 9.52290163556048e-06,
+      "loss": 0.7804,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2851560350046783,
+      "grad_norm": 0.6461299657821655,
+      "learning_rate": 9.52271683035147e-06,
+      "loss": 0.5727,
+      "step": 5181
+    },
+    {
+      "epoch": 0.285211073807034,
+      "grad_norm": 0.7275353074073792,
+      "learning_rate": 9.522531991150932e-06,
+      "loss": 0.8345,
+      "step": 5182
+    },
+    {
+      "epoch": 0.2852661126093896,
+      "grad_norm": 0.7321951985359192,
+      "learning_rate": 9.522347117960253e-06,
+      "loss": 0.8832,
+      "step": 5183
+    },
+    {
+      "epoch": 0.2853211514117453,
+      "grad_norm": 0.7526552677154541,
+      "learning_rate": 9.522162210780825e-06,
+      "loss": 0.831,
+      "step": 5184
+    },
+    {
+      "epoch": 0.28537619021410093,
+      "grad_norm": 0.7592381238937378,
+      "learning_rate": 9.521977269614036e-06,
+      "loss": 0.7293,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2854312290164566,
+      "grad_norm": 0.8060448169708252,
+      "learning_rate": 9.521792294461274e-06,
+      "loss": 0.819,
+      "step": 5186
+    },
+    {
+      "epoch": 0.28548626781881226,
+      "grad_norm": 0.7178553342819214,
+      "learning_rate": 9.521607285323932e-06,
+      "loss": 0.7526,
+      "step": 5187
+    },
+    {
+      "epoch": 0.28554130662116795,
+      "grad_norm": 0.8186969757080078,
+      "learning_rate": 9.521422242203401e-06,
+      "loss": 0.8526,
+      "step": 5188
+    },
+    {
+      "epoch": 0.2855963454235236,
+      "grad_norm": 0.8480883240699768,
+      "learning_rate": 9.521237165101071e-06,
+      "loss": 0.8088,
+      "step": 5189
+    },
+    {
+      "epoch": 0.28565138422587927,
+      "grad_norm": 0.8053719401359558,
+      "learning_rate": 9.521052054018333e-06,
+      "loss": 0.928,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2857064230282349,
+      "grad_norm": 0.6937163472175598,
+      "learning_rate": 9.52086690895658e-06,
+      "loss": 0.7418,
+      "step": 5191
+    },
+    {
+      "epoch": 0.2857614618305906,
+      "grad_norm": 1.0616179704666138,
+      "learning_rate": 9.520681729917196e-06,
+      "loss": 0.8726,
+      "step": 5192
+    },
+    {
+      "epoch": 0.2858165006329462,
+      "grad_norm": 0.7504106163978577,
+      "learning_rate": 9.520496516901582e-06,
+      "loss": 0.844,
+      "step": 5193
+    },
+    {
+      "epoch": 0.2858715394353019,
+      "grad_norm": 0.7634509205818176,
+      "learning_rate": 9.520311269911127e-06,
+      "loss": 0.7595,
+      "step": 5194
+    },
+    {
+      "epoch": 0.28592657823765755,
+      "grad_norm": 0.7069799900054932,
+      "learning_rate": 9.52012598894722e-06,
+      "loss": 0.7566,
+      "step": 5195
+    },
+    {
+      "epoch": 0.28598161704001324,
+      "grad_norm": 0.695737361907959,
+      "learning_rate": 9.519940674011256e-06,
+      "loss": 0.7534,
+      "step": 5196
+    },
+    {
+      "epoch": 0.28603665584236887,
+      "grad_norm": 0.7212124466896057,
+      "learning_rate": 9.51975532510463e-06,
+      "loss": 0.8237,
+      "step": 5197
+    },
+    {
+      "epoch": 0.2860916946447245,
+      "grad_norm": 0.7274062633514404,
+      "learning_rate": 9.519569942228732e-06,
+      "loss": 0.756,
+      "step": 5198
+    },
+    {
+      "epoch": 0.2861467334470802,
+      "grad_norm": 0.7038697600364685,
+      "learning_rate": 9.519384525384956e-06,
+      "loss": 0.7308,
+      "step": 5199
+    },
+    {
+      "epoch": 0.2862017722494358,
+      "grad_norm": 0.6897109150886536,
+      "learning_rate": 9.519199074574694e-06,
+      "loss": 0.7858,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2862568110517915,
+      "grad_norm": 0.8471527099609375,
+      "learning_rate": 9.519013589799343e-06,
+      "loss": 0.8198,
+      "step": 5201
+    },
+    {
+      "epoch": 0.28631184985414715,
+      "grad_norm": 0.6828129291534424,
+      "learning_rate": 9.518828071060295e-06,
+      "loss": 0.7734,
+      "step": 5202
+    },
+    {
+      "epoch": 0.28636688865650284,
+      "grad_norm": 0.7437755465507507,
+      "learning_rate": 9.518642518358946e-06,
+      "loss": 0.7669,
+      "step": 5203
+    },
+    {
+      "epoch": 0.28642192745885847,
+      "grad_norm": 0.8841923475265503,
+      "learning_rate": 9.518456931696689e-06,
+      "loss": 0.8201,
+      "step": 5204
+    },
+    {
+      "epoch": 0.28647696626121416,
+      "grad_norm": 0.9514154195785522,
+      "learning_rate": 9.518271311074917e-06,
+      "loss": 0.7864,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2865320050635698,
+      "grad_norm": 0.830795407295227,
+      "learning_rate": 9.51808565649503e-06,
+      "loss": 0.8024,
+      "step": 5206
+    },
+    {
+      "epoch": 0.2865870438659255,
+      "grad_norm": 0.7274934649467468,
+      "learning_rate": 9.51789996795842e-06,
+      "loss": 0.7631,
+      "step": 5207
+    },
+    {
+      "epoch": 0.2866420826682811,
+      "grad_norm": 0.7004290223121643,
+      "learning_rate": 9.517714245466482e-06,
+      "loss": 0.7344,
+      "step": 5208
+    },
+    {
+      "epoch": 0.2866971214706368,
+      "grad_norm": 0.8559010624885559,
+      "learning_rate": 9.517528489020614e-06,
+      "loss": 0.7502,
+      "step": 5209
+    },
+    {
+      "epoch": 0.28675216027299244,
+      "grad_norm": 0.8913494348526001,
+      "learning_rate": 9.517342698622212e-06,
+      "loss": 0.8908,
+      "step": 5210
+    },
+    {
+      "epoch": 0.28680719907534813,
+      "grad_norm": 0.8375207781791687,
+      "learning_rate": 9.51715687427267e-06,
+      "loss": 0.7701,
+      "step": 5211
+    },
+    {
+      "epoch": 0.28686223787770376,
+      "grad_norm": 1.1804776191711426,
+      "learning_rate": 9.516971015973386e-06,
+      "loss": 0.8449,
+      "step": 5212
+    },
+    {
+      "epoch": 0.28691727668005945,
+      "grad_norm": 0.7260473370552063,
+      "learning_rate": 9.516785123725758e-06,
+      "loss": 0.7978,
+      "step": 5213
+    },
+    {
+      "epoch": 0.2869723154824151,
+      "grad_norm": 0.8159041404724121,
+      "learning_rate": 9.516599197531182e-06,
+      "loss": 0.7454,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2870273542847708,
+      "grad_norm": 0.7850227952003479,
+      "learning_rate": 9.516413237391056e-06,
+      "loss": 0.8082,
+      "step": 5215
+    },
+    {
+      "epoch": 0.2870823930871264,
+      "grad_norm": 0.7596960067749023,
+      "learning_rate": 9.516227243306774e-06,
+      "loss": 0.7286,
+      "step": 5216
+    },
+    {
+      "epoch": 0.2871374318894821,
+      "grad_norm": 0.8763321042060852,
+      "learning_rate": 9.516041215279741e-06,
+      "loss": 0.8685,
+      "step": 5217
+    },
+    {
+      "epoch": 0.28719247069183773,
+      "grad_norm": 1.2130110263824463,
+      "learning_rate": 9.515855153311349e-06,
+      "loss": 0.8374,
+      "step": 5218
+    },
+    {
+      "epoch": 0.2872475094941934,
+      "grad_norm": 0.7578628063201904,
+      "learning_rate": 9.515669057402999e-06,
+      "loss": 0.793,
+      "step": 5219
+    },
+    {
+      "epoch": 0.28730254829654905,
+      "grad_norm": 0.9085225462913513,
+      "learning_rate": 9.515482927556088e-06,
+      "loss": 0.8366,
+      "step": 5220
+    },
+    {
+      "epoch": 0.28735758709890474,
+      "grad_norm": 0.7107900977134705,
+      "learning_rate": 9.515296763772017e-06,
+      "loss": 0.6571,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2874126259012604,
+      "grad_norm": 0.7742018699645996,
+      "learning_rate": 9.515110566052183e-06,
+      "loss": 0.8387,
+      "step": 5222
+    },
+    {
+      "epoch": 0.28746766470361607,
+      "grad_norm": 0.8934319615364075,
+      "learning_rate": 9.514924334397987e-06,
+      "loss": 0.8546,
+      "step": 5223
+    },
+    {
+      "epoch": 0.2875227035059717,
+      "grad_norm": 0.720245897769928,
+      "learning_rate": 9.51473806881083e-06,
+      "loss": 0.7459,
+      "step": 5224
+    },
+    {
+      "epoch": 0.2875777423083274,
+      "grad_norm": 0.7074370384216309,
+      "learning_rate": 9.514551769292109e-06,
+      "loss": 0.8598,
+      "step": 5225
+    },
+    {
+      "epoch": 0.287632781110683,
+      "grad_norm": 0.7608621120452881,
+      "learning_rate": 9.514365435843226e-06,
+      "loss": 0.7263,
+      "step": 5226
+    },
+    {
+      "epoch": 0.2876878199130387,
+      "grad_norm": 0.7581011652946472,
+      "learning_rate": 9.51417906846558e-06,
+      "loss": 0.7498,
+      "step": 5227
+    },
+    {
+      "epoch": 0.28774285871539435,
+      "grad_norm": 0.8184412121772766,
+      "learning_rate": 9.513992667160572e-06,
+      "loss": 0.6889,
+      "step": 5228
+    },
+    {
+      "epoch": 0.28779789751775003,
+      "grad_norm": 0.6835145354270935,
+      "learning_rate": 9.513806231929605e-06,
+      "loss": 0.7399,
+      "step": 5229
+    },
+    {
+      "epoch": 0.28785293632010567,
+      "grad_norm": 0.7601536512374878,
+      "learning_rate": 9.513619762774077e-06,
+      "loss": 0.846,
+      "step": 5230
+    },
+    {
+      "epoch": 0.28790797512246136,
+      "grad_norm": 0.781491219997406,
+      "learning_rate": 9.513433259695392e-06,
+      "loss": 0.8326,
+      "step": 5231
+    },
+    {
+      "epoch": 0.287963013924817,
+      "grad_norm": 0.7978106141090393,
+      "learning_rate": 9.513246722694951e-06,
+      "loss": 0.7917,
+      "step": 5232
+    },
+    {
+      "epoch": 0.2880180527271727,
+      "grad_norm": 0.8071381449699402,
+      "learning_rate": 9.513060151774156e-06,
+      "loss": 0.8054,
+      "step": 5233
+    },
+    {
+      "epoch": 0.2880730915295283,
+      "grad_norm": 0.815567135810852,
+      "learning_rate": 9.512873546934406e-06,
+      "loss": 0.8647,
+      "step": 5234
+    },
+    {
+      "epoch": 0.288128130331884,
+      "grad_norm": 0.8255048990249634,
+      "learning_rate": 9.512686908177111e-06,
+      "loss": 0.9011,
+      "step": 5235
+    },
+    {
+      "epoch": 0.28818316913423964,
+      "grad_norm": 0.8392062187194824,
+      "learning_rate": 9.512500235503666e-06,
+      "loss": 0.8778,
+      "step": 5236
+    },
+    {
+      "epoch": 0.2882382079365953,
+      "grad_norm": 0.7256191372871399,
+      "learning_rate": 9.512313528915478e-06,
+      "loss": 0.7231,
+      "step": 5237
+    },
+    {
+      "epoch": 0.28829324673895096,
+      "grad_norm": 0.9041032195091248,
+      "learning_rate": 9.51212678841395e-06,
+      "loss": 0.8469,
+      "step": 5238
+    },
+    {
+      "epoch": 0.28834828554130665,
+      "grad_norm": 0.7857525944709778,
+      "learning_rate": 9.511940014000485e-06,
+      "loss": 0.7447,
+      "step": 5239
+    },
+    {
+      "epoch": 0.2884033243436623,
+      "grad_norm": 0.6925225257873535,
+      "learning_rate": 9.511753205676485e-06,
+      "loss": 0.8302,
+      "step": 5240
+    },
+    {
+      "epoch": 0.2884583631460179,
+      "grad_norm": 0.7253623008728027,
+      "learning_rate": 9.511566363443356e-06,
+      "loss": 0.8373,
+      "step": 5241
+    },
+    {
+      "epoch": 0.2885134019483736,
+      "grad_norm": 0.7198607921600342,
+      "learning_rate": 9.511379487302504e-06,
+      "loss": 0.79,
+      "step": 5242
+    },
+    {
+      "epoch": 0.28856844075072924,
+      "grad_norm": 0.7966421246528625,
+      "learning_rate": 9.511192577255328e-06,
+      "loss": 0.7933,
+      "step": 5243
+    },
+    {
+      "epoch": 0.2886234795530849,
+      "grad_norm": 0.9159359931945801,
+      "learning_rate": 9.511005633303239e-06,
+      "loss": 0.7254,
+      "step": 5244
+    },
+    {
+      "epoch": 0.28867851835544056,
+      "grad_norm": 0.9514481425285339,
+      "learning_rate": 9.510818655447638e-06,
+      "loss": 0.8916,
+      "step": 5245
+    },
+    {
+      "epoch": 0.28873355715779625,
+      "grad_norm": 0.7505099773406982,
+      "learning_rate": 9.510631643689932e-06,
+      "loss": 0.765,
+      "step": 5246
+    },
+    {
+      "epoch": 0.2887885959601519,
+      "grad_norm": 0.7824658751487732,
+      "learning_rate": 9.510444598031526e-06,
+      "loss": 0.6972,
+      "step": 5247
+    },
+    {
+      "epoch": 0.2888436347625076,
+      "grad_norm": 0.7778681516647339,
+      "learning_rate": 9.510257518473824e-06,
+      "loss": 0.8705,
+      "step": 5248
+    },
+    {
+      "epoch": 0.2888986735648632,
+      "grad_norm": 0.6785199642181396,
+      "learning_rate": 9.510070405018235e-06,
+      "loss": 0.6889,
+      "step": 5249
+    },
+    {
+      "epoch": 0.2889537123672189,
+      "grad_norm": 0.7045316100120544,
+      "learning_rate": 9.509883257666164e-06,
+      "loss": 0.7979,
+      "step": 5250
+    },
+    {
+      "epoch": 0.28900875116957453,
+      "grad_norm": 1.3174562454223633,
+      "learning_rate": 9.509696076419018e-06,
+      "loss": 0.8802,
+      "step": 5251
+    },
+    {
+      "epoch": 0.2890637899719302,
+      "grad_norm": 1.1800767183303833,
+      "learning_rate": 9.509508861278205e-06,
+      "loss": 0.9246,
+      "step": 5252
+    },
+    {
+      "epoch": 0.28911882877428585,
+      "grad_norm": 0.7057580947875977,
+      "learning_rate": 9.509321612245128e-06,
+      "loss": 0.7565,
+      "step": 5253
+    },
+    {
+      "epoch": 0.28917386757664154,
+      "grad_norm": 0.7681905031204224,
+      "learning_rate": 9.509134329321197e-06,
+      "loss": 0.8678,
+      "step": 5254
+    },
+    {
+      "epoch": 0.2892289063789972,
+      "grad_norm": 0.96025550365448,
+      "learning_rate": 9.50894701250782e-06,
+      "loss": 0.9108,
+      "step": 5255
+    },
+    {
+      "epoch": 0.28928394518135286,
+      "grad_norm": 0.7786841988563538,
+      "learning_rate": 9.508759661806405e-06,
+      "loss": 0.7747,
+      "step": 5256
+    },
+    {
+      "epoch": 0.2893389839837085,
+      "grad_norm": 0.7073540091514587,
+      "learning_rate": 9.508572277218358e-06,
+      "loss": 0.7573,
+      "step": 5257
+    },
+    {
+      "epoch": 0.2893940227860642,
+      "grad_norm": 0.6648856401443481,
+      "learning_rate": 9.50838485874509e-06,
+      "loss": 0.7294,
+      "step": 5258
+    },
+    {
+      "epoch": 0.2894490615884198,
+      "grad_norm": 0.6794270873069763,
+      "learning_rate": 9.508197406388007e-06,
+      "loss": 0.7001,
+      "step": 5259
+    },
+    {
+      "epoch": 0.2895041003907755,
+      "grad_norm": 0.6819350123405457,
+      "learning_rate": 9.50800992014852e-06,
+      "loss": 0.7114,
+      "step": 5260
+    },
+    {
+      "epoch": 0.28955913919313114,
+      "grad_norm": 0.6616997122764587,
+      "learning_rate": 9.507822400028036e-06,
+      "loss": 0.7108,
+      "step": 5261
+    },
+    {
+      "epoch": 0.28961417799548683,
+      "grad_norm": 0.7447230219841003,
+      "learning_rate": 9.507634846027966e-06,
+      "loss": 0.7865,
+      "step": 5262
+    },
+    {
+      "epoch": 0.28966921679784247,
+      "grad_norm": 0.7826278209686279,
+      "learning_rate": 9.50744725814972e-06,
+      "loss": 0.7922,
+      "step": 5263
+    },
+    {
+      "epoch": 0.28972425560019816,
+      "grad_norm": 0.8054459095001221,
+      "learning_rate": 9.507259636394706e-06,
+      "loss": 0.795,
+      "step": 5264
+    },
+    {
+      "epoch": 0.2897792944025538,
+      "grad_norm": 0.9539191722869873,
+      "learning_rate": 9.507071980764335e-06,
+      "loss": 0.9495,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2898343332049095,
+      "grad_norm": 0.8877993226051331,
+      "learning_rate": 9.506884291260017e-06,
+      "loss": 0.8418,
+      "step": 5266
+    },
+    {
+      "epoch": 0.2898893720072651,
+      "grad_norm": 0.6620327234268188,
+      "learning_rate": 9.506696567883164e-06,
+      "loss": 0.6285,
+      "step": 5267
+    },
+    {
+      "epoch": 0.2899444108096208,
+      "grad_norm": 0.7604434490203857,
+      "learning_rate": 9.506508810635187e-06,
+      "loss": 0.8562,
+      "step": 5268
+    },
+    {
+      "epoch": 0.28999944961197643,
+      "grad_norm": 0.8181812763214111,
+      "learning_rate": 9.506321019517494e-06,
+      "loss": 0.905,
+      "step": 5269
+    },
+    {
+      "epoch": 0.2900544884143321,
+      "grad_norm": 0.7776391506195068,
+      "learning_rate": 9.5061331945315e-06,
+      "loss": 0.8871,
+      "step": 5270
+    },
+    {
+      "epoch": 0.29010952721668776,
+      "grad_norm": 0.8125039339065552,
+      "learning_rate": 9.505945335678613e-06,
+      "loss": 0.7254,
+      "step": 5271
+    },
+    {
+      "epoch": 0.29016456601904345,
+      "grad_norm": 0.7229846715927124,
+      "learning_rate": 9.50575744296025e-06,
+      "loss": 0.8192,
+      "step": 5272
+    },
+    {
+      "epoch": 0.2902196048213991,
+      "grad_norm": 0.72443026304245,
+      "learning_rate": 9.505569516377817e-06,
+      "loss": 0.7813,
+      "step": 5273
+    },
+    {
+      "epoch": 0.29027464362375477,
+      "grad_norm": 0.6798073053359985,
+      "learning_rate": 9.505381555932731e-06,
+      "loss": 0.7655,
+      "step": 5274
+    },
+    {
+      "epoch": 0.2903296824261104,
+      "grad_norm": 1.0805624723434448,
+      "learning_rate": 9.505193561626404e-06,
+      "loss": 0.9035,
+      "step": 5275
+    },
+    {
+      "epoch": 0.2903847212284661,
+      "grad_norm": 0.7579694986343384,
+      "learning_rate": 9.505005533460247e-06,
+      "loss": 0.8612,
+      "step": 5276
+    },
+    {
+      "epoch": 0.2904397600308217,
+      "grad_norm": 1.2496099472045898,
+      "learning_rate": 9.504817471435676e-06,
+      "loss": 0.813,
+      "step": 5277
+    },
+    {
+      "epoch": 0.2904947988331774,
+      "grad_norm": 0.6915673017501831,
+      "learning_rate": 9.504629375554102e-06,
+      "loss": 0.6891,
+      "step": 5278
+    },
+    {
+      "epoch": 0.29054983763553305,
+      "grad_norm": 0.8581767082214355,
+      "learning_rate": 9.504441245816937e-06,
+      "loss": 0.7137,
+      "step": 5279
+    },
+    {
+      "epoch": 0.29060487643788874,
+      "grad_norm": 0.7469545006752014,
+      "learning_rate": 9.504253082225601e-06,
+      "loss": 0.7621,
+      "step": 5280
+    },
+    {
+      "epoch": 0.29065991524024437,
+      "grad_norm": 0.7725615501403809,
+      "learning_rate": 9.504064884781503e-06,
+      "loss": 0.7988,
+      "step": 5281
+    },
+    {
+      "epoch": 0.29071495404260006,
+      "grad_norm": 1.0187722444534302,
+      "learning_rate": 9.503876653486058e-06,
+      "loss": 0.7772,
+      "step": 5282
+    },
+    {
+      "epoch": 0.2907699928449557,
+      "grad_norm": 0.675574779510498,
+      "learning_rate": 9.503688388340683e-06,
+      "loss": 0.7096,
+      "step": 5283
+    },
+    {
+      "epoch": 0.2908250316473113,
+      "grad_norm": 0.7980207800865173,
+      "learning_rate": 9.503500089346792e-06,
+      "loss": 0.8291,
+      "step": 5284
+    },
+    {
+      "epoch": 0.290880070449667,
+      "grad_norm": 0.6891655325889587,
+      "learning_rate": 9.503311756505797e-06,
+      "loss": 0.7186,
+      "step": 5285
+    },
+    {
+      "epoch": 0.29093510925202265,
+      "grad_norm": 0.7273408770561218,
+      "learning_rate": 9.50312338981912e-06,
+      "loss": 0.7483,
+      "step": 5286
+    },
+    {
+      "epoch": 0.29099014805437834,
+      "grad_norm": 0.7346869111061096,
+      "learning_rate": 9.50293498928817e-06,
+      "loss": 0.766,
+      "step": 5287
+    },
+    {
+      "epoch": 0.291045186856734,
+      "grad_norm": 0.7627394795417786,
+      "learning_rate": 9.502746554914368e-06,
+      "loss": 0.867,
+      "step": 5288
+    },
+    {
+      "epoch": 0.29110022565908966,
+      "grad_norm": 0.8477200865745544,
+      "learning_rate": 9.502558086699128e-06,
+      "loss": 0.8317,
+      "step": 5289
+    },
+    {
+      "epoch": 0.2911552644614453,
+      "grad_norm": 0.7696006894111633,
+      "learning_rate": 9.502369584643867e-06,
+      "loss": 0.7814,
+      "step": 5290
+    },
+    {
+      "epoch": 0.291210303263801,
+      "grad_norm": 0.7614455819129944,
+      "learning_rate": 9.502181048749999e-06,
+      "loss": 0.7398,
+      "step": 5291
+    },
+    {
+      "epoch": 0.2912653420661566,
+      "grad_norm": 0.7877628207206726,
+      "learning_rate": 9.501992479018946e-06,
+      "loss": 0.8731,
+      "step": 5292
+    },
+    {
+      "epoch": 0.2913203808685123,
+      "grad_norm": 0.7455846667289734,
+      "learning_rate": 9.50180387545212e-06,
+      "loss": 0.7059,
+      "step": 5293
+    },
+    {
+      "epoch": 0.29137541967086794,
+      "grad_norm": 1.145520567893982,
+      "learning_rate": 9.501615238050944e-06,
+      "loss": 0.6968,
+      "step": 5294
+    },
+    {
+      "epoch": 0.29143045847322363,
+      "grad_norm": 0.8100234866142273,
+      "learning_rate": 9.501426566816831e-06,
+      "loss": 0.8122,
+      "step": 5295
+    },
+    {
+      "epoch": 0.29148549727557926,
+      "grad_norm": 0.6813066005706787,
+      "learning_rate": 9.501237861751203e-06,
+      "loss": 0.6718,
+      "step": 5296
+    },
+    {
+      "epoch": 0.29154053607793495,
+      "grad_norm": 0.7400195002555847,
+      "learning_rate": 9.501049122855473e-06,
+      "loss": 0.802,
+      "step": 5297
+    },
+    {
+      "epoch": 0.2915955748802906,
+      "grad_norm": 0.7948681712150574,
+      "learning_rate": 9.500860350131065e-06,
+      "loss": 0.8237,
+      "step": 5298
+    },
+    {
+      "epoch": 0.2916506136826463,
+      "grad_norm": 0.772093653678894,
+      "learning_rate": 9.500671543579394e-06,
+      "loss": 0.7687,
+      "step": 5299
+    },
+    {
+      "epoch": 0.2917056524850019,
+      "grad_norm": 0.7468486428260803,
+      "learning_rate": 9.500482703201881e-06,
+      "loss": 0.7827,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2917606912873576,
+      "grad_norm": 0.7284440398216248,
+      "learning_rate": 9.500293828999945e-06,
+      "loss": 0.8086,
+      "step": 5301
+    },
+    {
+      "epoch": 0.29181573008971323,
+      "grad_norm": 0.8014211654663086,
+      "learning_rate": 9.500104920975005e-06,
+      "loss": 0.8409,
+      "step": 5302
+    },
+    {
+      "epoch": 0.2918707688920689,
+      "grad_norm": 0.7588346004486084,
+      "learning_rate": 9.49991597912848e-06,
+      "loss": 0.7149,
+      "step": 5303
+    },
+    {
+      "epoch": 0.29192580769442456,
+      "grad_norm": 0.8098518252372742,
+      "learning_rate": 9.499727003461794e-06,
+      "loss": 0.8375,
+      "step": 5304
+    },
+    {
+      "epoch": 0.29198084649678024,
+      "grad_norm": 0.8502426743507385,
+      "learning_rate": 9.499537993976363e-06,
+      "loss": 0.8177,
+      "step": 5305
+    },
+    {
+      "epoch": 0.2920358852991359,
+      "grad_norm": 0.8010903596878052,
+      "learning_rate": 9.499348950673607e-06,
+      "loss": 0.8457,
+      "step": 5306
+    },
+    {
+      "epoch": 0.29209092410149157,
+      "grad_norm": 0.6628156304359436,
+      "learning_rate": 9.49915987355495e-06,
+      "loss": 0.7327,
+      "step": 5307
+    },
+    {
+      "epoch": 0.2921459629038472,
+      "grad_norm": 0.7414939999580383,
+      "learning_rate": 9.49897076262181e-06,
+      "loss": 0.8271,
+      "step": 5308
+    },
+    {
+      "epoch": 0.2922010017062029,
+      "grad_norm": 0.7490847706794739,
+      "learning_rate": 9.498781617875613e-06,
+      "loss": 0.7689,
+      "step": 5309
+    },
+    {
+      "epoch": 0.2922560405085585,
+      "grad_norm": 0.7913424968719482,
+      "learning_rate": 9.498592439317777e-06,
+      "loss": 0.8571,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2923110793109142,
+      "grad_norm": 0.6903867125511169,
+      "learning_rate": 9.498403226949724e-06,
+      "loss": 0.7325,
+      "step": 5311
+    },
+    {
+      "epoch": 0.29236611811326985,
+      "grad_norm": 0.8087130188941956,
+      "learning_rate": 9.498213980772875e-06,
+      "loss": 0.8167,
+      "step": 5312
+    },
+    {
+      "epoch": 0.29242115691562554,
+      "grad_norm": 1.1316752433776855,
+      "learning_rate": 9.498024700788655e-06,
+      "loss": 0.912,
+      "step": 5313
+    },
+    {
+      "epoch": 0.29247619571798117,
+      "grad_norm": 0.8701719045639038,
+      "learning_rate": 9.497835386998486e-06,
+      "loss": 0.8728,
+      "step": 5314
+    },
+    {
+      "epoch": 0.29253123452033686,
+      "grad_norm": 0.6688953638076782,
+      "learning_rate": 9.49764603940379e-06,
+      "loss": 0.6561,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2925862733226925,
+      "grad_norm": 0.8067505359649658,
+      "learning_rate": 9.49745665800599e-06,
+      "loss": 0.8419,
+      "step": 5316
+    },
+    {
+      "epoch": 0.2926413121250482,
+      "grad_norm": 0.7157390117645264,
+      "learning_rate": 9.49726724280651e-06,
+      "loss": 0.7964,
+      "step": 5317
+    },
+    {
+      "epoch": 0.2926963509274038,
+      "grad_norm": 0.7038627862930298,
+      "learning_rate": 9.497077793806772e-06,
+      "loss": 0.7343,
+      "step": 5318
+    },
+    {
+      "epoch": 0.2927513897297595,
+      "grad_norm": 0.7674478888511658,
+      "learning_rate": 9.4968883110082e-06,
+      "loss": 0.7624,
+      "step": 5319
+    },
+    {
+      "epoch": 0.29280642853211514,
+      "grad_norm": 0.6708847284317017,
+      "learning_rate": 9.496698794412223e-06,
+      "loss": 0.6554,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2928614673344708,
+      "grad_norm": 0.8332329392433167,
+      "learning_rate": 9.49650924402026e-06,
+      "loss": 0.9357,
+      "step": 5321
+    },
+    {
+      "epoch": 0.29291650613682646,
+      "grad_norm": 0.7601341605186462,
+      "learning_rate": 9.496319659833737e-06,
+      "loss": 0.8208,
+      "step": 5322
+    },
+    {
+      "epoch": 0.29297154493918215,
+      "grad_norm": 0.8320396542549133,
+      "learning_rate": 9.496130041854077e-06,
+      "loss": 0.8423,
+      "step": 5323
+    },
+    {
+      "epoch": 0.2930265837415378,
+      "grad_norm": 0.8242839574813843,
+      "learning_rate": 9.49594039008271e-06,
+      "loss": 0.9101,
+      "step": 5324
+    },
+    {
+      "epoch": 0.29308162254389347,
+      "grad_norm": 0.8906320333480835,
+      "learning_rate": 9.495750704521058e-06,
+      "loss": 0.7343,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2931366613462491,
+      "grad_norm": 0.7964318990707397,
+      "learning_rate": 9.495560985170546e-06,
+      "loss": 0.7789,
+      "step": 5326
+    },
+    {
+      "epoch": 0.29319170014860474,
+      "grad_norm": 0.8267771601676941,
+      "learning_rate": 9.495371232032602e-06,
+      "loss": 0.7447,
+      "step": 5327
+    },
+    {
+      "epoch": 0.29324673895096043,
+      "grad_norm": 0.8120046257972717,
+      "learning_rate": 9.49518144510865e-06,
+      "loss": 0.7803,
+      "step": 5328
+    },
+    {
+      "epoch": 0.29330177775331606,
+      "grad_norm": 0.7314801812171936,
+      "learning_rate": 9.494991624400119e-06,
+      "loss": 0.6758,
+      "step": 5329
+    },
+    {
+      "epoch": 0.29335681655567175,
+      "grad_norm": 0.6989930272102356,
+      "learning_rate": 9.494801769908433e-06,
+      "loss": 0.7945,
+      "step": 5330
+    },
+    {
+      "epoch": 0.2934118553580274,
+      "grad_norm": 0.7804785966873169,
+      "learning_rate": 9.494611881635021e-06,
+      "loss": 0.7977,
+      "step": 5331
+    },
+    {
+      "epoch": 0.2934668941603831,
+      "grad_norm": 0.8377045392990112,
+      "learning_rate": 9.494421959581308e-06,
+      "loss": 0.8077,
+      "step": 5332
+    },
+    {
+      "epoch": 0.2935219329627387,
+      "grad_norm": 0.7463418245315552,
+      "learning_rate": 9.494232003748724e-06,
+      "loss": 0.783,
+      "step": 5333
+    },
+    {
+      "epoch": 0.2935769717650944,
+      "grad_norm": 0.7598912715911865,
+      "learning_rate": 9.494042014138695e-06,
+      "loss": 0.7869,
+      "step": 5334
+    },
+    {
+      "epoch": 0.29363201056745003,
+      "grad_norm": 0.7634113430976868,
+      "learning_rate": 9.493851990752648e-06,
+      "loss": 0.8108,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2936870493698057,
+      "grad_norm": 0.8056474328041077,
+      "learning_rate": 9.493661933592013e-06,
+      "loss": 0.7921,
+      "step": 5336
+    },
+    {
+      "epoch": 0.29374208817216135,
+      "grad_norm": 0.8699371218681335,
+      "learning_rate": 9.493471842658219e-06,
+      "loss": 0.8833,
+      "step": 5337
+    },
+    {
+      "epoch": 0.29379712697451704,
+      "grad_norm": 0.8803261518478394,
+      "learning_rate": 9.493281717952691e-06,
+      "loss": 0.7848,
+      "step": 5338
+    },
+    {
+      "epoch": 0.2938521657768727,
+      "grad_norm": 0.7678453922271729,
+      "learning_rate": 9.493091559476864e-06,
+      "loss": 0.836,
+      "step": 5339
+    },
+    {
+      "epoch": 0.29390720457922836,
+      "grad_norm": 0.7653701305389404,
+      "learning_rate": 9.49290136723216e-06,
+      "loss": 0.8215,
+      "step": 5340
+    },
+    {
+      "epoch": 0.293962243381584,
+      "grad_norm": 0.768120527267456,
+      "learning_rate": 9.492711141220013e-06,
+      "loss": 0.7498,
+      "step": 5341
+    },
+    {
+      "epoch": 0.2940172821839397,
+      "grad_norm": 0.7665749788284302,
+      "learning_rate": 9.492520881441854e-06,
+      "loss": 0.7883,
+      "step": 5342
+    },
+    {
+      "epoch": 0.2940723209862953,
+      "grad_norm": 0.7405015230178833,
+      "learning_rate": 9.492330587899108e-06,
+      "loss": 0.8112,
+      "step": 5343
+    },
+    {
+      "epoch": 0.294127359788651,
+      "grad_norm": 0.7183459997177124,
+      "learning_rate": 9.492140260593208e-06,
+      "loss": 0.8227,
+      "step": 5344
+    },
+    {
+      "epoch": 0.29418239859100664,
+      "grad_norm": 0.7453572154045105,
+      "learning_rate": 9.491949899525585e-06,
+      "loss": 0.8148,
+      "step": 5345
+    },
+    {
+      "epoch": 0.29423743739336233,
+      "grad_norm": 0.8963750600814819,
+      "learning_rate": 9.491759504697669e-06,
+      "loss": 0.9261,
+      "step": 5346
+    },
+    {
+      "epoch": 0.29429247619571797,
+      "grad_norm": 0.7631667256355286,
+      "learning_rate": 9.49156907611089e-06,
+      "loss": 0.7708,
+      "step": 5347
+    },
+    {
+      "epoch": 0.29434751499807366,
+      "grad_norm": 0.6324381232261658,
+      "learning_rate": 9.49137861376668e-06,
+      "loss": 0.6688,
+      "step": 5348
+    },
+    {
+      "epoch": 0.2944025538004293,
+      "grad_norm": 0.6969807147979736,
+      "learning_rate": 9.491188117666472e-06,
+      "loss": 0.7516,
+      "step": 5349
+    },
+    {
+      "epoch": 0.294457592602785,
+      "grad_norm": 1.633340835571289,
+      "learning_rate": 9.490997587811697e-06,
+      "loss": 0.8111,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2945126314051406,
+      "grad_norm": 0.7084371447563171,
+      "learning_rate": 9.490807024203785e-06,
+      "loss": 0.8375,
+      "step": 5351
+    },
+    {
+      "epoch": 0.2945676702074963,
+      "grad_norm": 0.7335958480834961,
+      "learning_rate": 9.490616426844169e-06,
+      "loss": 0.7884,
+      "step": 5352
+    },
+    {
+      "epoch": 0.29462270900985194,
+      "grad_norm": 0.7560276985168457,
+      "learning_rate": 9.490425795734282e-06,
+      "loss": 0.8918,
+      "step": 5353
+    },
+    {
+      "epoch": 0.2946777478122076,
+      "grad_norm": 0.9185894727706909,
+      "learning_rate": 9.490235130875557e-06,
+      "loss": 0.7976,
+      "step": 5354
+    },
+    {
+      "epoch": 0.29473278661456326,
+      "grad_norm": 0.7871553897857666,
+      "learning_rate": 9.490044432269427e-06,
+      "loss": 0.8564,
+      "step": 5355
+    },
+    {
+      "epoch": 0.29478782541691895,
+      "grad_norm": 0.8736812472343445,
+      "learning_rate": 9.489853699917326e-06,
+      "loss": 0.8114,
+      "step": 5356
+    },
+    {
+      "epoch": 0.2948428642192746,
+      "grad_norm": 0.8068968653678894,
+      "learning_rate": 9.489662933820684e-06,
+      "loss": 0.9198,
+      "step": 5357
+    },
+    {
+      "epoch": 0.29489790302163027,
+      "grad_norm": 0.7816325426101685,
+      "learning_rate": 9.489472133980939e-06,
+      "loss": 0.8012,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2949529418239859,
+      "grad_norm": 0.7248200178146362,
+      "learning_rate": 9.489281300399522e-06,
+      "loss": 0.8099,
+      "step": 5359
+    },
+    {
+      "epoch": 0.2950079806263416,
+      "grad_norm": 0.7887724041938782,
+      "learning_rate": 9.48909043307787e-06,
+      "loss": 0.884,
+      "step": 5360
+    },
+    {
+      "epoch": 0.2950630194286972,
+      "grad_norm": 0.765163004398346,
+      "learning_rate": 9.488899532017415e-06,
+      "loss": 0.8563,
+      "step": 5361
+    },
+    {
+      "epoch": 0.2951180582310529,
+      "grad_norm": 0.7658557295799255,
+      "learning_rate": 9.488708597219592e-06,
+      "loss": 0.8897,
+      "step": 5362
+    },
+    {
+      "epoch": 0.29517309703340855,
+      "grad_norm": 0.6653227806091309,
+      "learning_rate": 9.488517628685838e-06,
+      "loss": 0.7107,
+      "step": 5363
+    },
+    {
+      "epoch": 0.29522813583576424,
+      "grad_norm": 0.787739098072052,
+      "learning_rate": 9.488326626417586e-06,
+      "loss": 0.8181,
+      "step": 5364
+    },
+    {
+      "epoch": 0.29528317463811987,
+      "grad_norm": 0.7822532057762146,
+      "learning_rate": 9.488135590416275e-06,
+      "loss": 0.8238,
+      "step": 5365
+    },
+    {
+      "epoch": 0.29533821344047556,
+      "grad_norm": 0.7797419428825378,
+      "learning_rate": 9.487944520683334e-06,
+      "loss": 0.8484,
+      "step": 5366
+    },
+    {
+      "epoch": 0.2953932522428312,
+      "grad_norm": 0.7230222225189209,
+      "learning_rate": 9.487753417220207e-06,
+      "loss": 0.8193,
+      "step": 5367
+    },
+    {
+      "epoch": 0.2954482910451869,
+      "grad_norm": 0.8256810307502747,
+      "learning_rate": 9.487562280028325e-06,
+      "loss": 0.7691,
+      "step": 5368
+    },
+    {
+      "epoch": 0.2955033298475425,
+      "grad_norm": 0.7704648375511169,
+      "learning_rate": 9.487371109109127e-06,
+      "loss": 0.8235,
+      "step": 5369
+    },
+    {
+      "epoch": 0.29555836864989815,
+      "grad_norm": 0.7580391764640808,
+      "learning_rate": 9.487179904464048e-06,
+      "loss": 0.7911,
+      "step": 5370
+    },
+    {
+      "epoch": 0.29561340745225384,
+      "grad_norm": 0.7211806774139404,
+      "learning_rate": 9.486988666094526e-06,
+      "loss": 0.7188,
+      "step": 5371
+    },
+    {
+      "epoch": 0.2956684462546095,
+      "grad_norm": 0.8375828862190247,
+      "learning_rate": 9.486797394001999e-06,
+      "loss": 0.881,
+      "step": 5372
+    },
+    {
+      "epoch": 0.29572348505696516,
+      "grad_norm": 0.8500093221664429,
+      "learning_rate": 9.486606088187903e-06,
+      "loss": 0.8632,
+      "step": 5373
+    },
+    {
+      "epoch": 0.2957785238593208,
+      "grad_norm": 0.7754727005958557,
+      "learning_rate": 9.486414748653677e-06,
+      "loss": 0.8124,
+      "step": 5374
+    },
+    {
+      "epoch": 0.2958335626616765,
+      "grad_norm": 0.9395208954811096,
+      "learning_rate": 9.486223375400759e-06,
+      "loss": 0.8046,
+      "step": 5375
+    },
+    {
+      "epoch": 0.2958886014640321,
+      "grad_norm": 0.7587517499923706,
+      "learning_rate": 9.486031968430587e-06,
+      "loss": 0.7852,
+      "step": 5376
+    },
+    {
+      "epoch": 0.2959436402663878,
+      "grad_norm": 0.6921781301498413,
+      "learning_rate": 9.485840527744599e-06,
+      "loss": 0.7392,
+      "step": 5377
+    },
+    {
+      "epoch": 0.29599867906874344,
+      "grad_norm": 0.8768522143363953,
+      "learning_rate": 9.485649053344233e-06,
+      "loss": 0.7819,
+      "step": 5378
+    },
+    {
+      "epoch": 0.29605371787109913,
+      "grad_norm": 0.7565680146217346,
+      "learning_rate": 9.485457545230932e-06,
+      "loss": 0.7489,
+      "step": 5379
+    },
+    {
+      "epoch": 0.29610875667345476,
+      "grad_norm": 0.7760992050170898,
+      "learning_rate": 9.485266003406132e-06,
+      "loss": 0.8129,
+      "step": 5380
+    },
+    {
+      "epoch": 0.29616379547581045,
+      "grad_norm": 0.7726097106933594,
+      "learning_rate": 9.485074427871272e-06,
+      "loss": 0.725,
+      "step": 5381
+    },
+    {
+      "epoch": 0.2962188342781661,
+      "grad_norm": 0.6885473728179932,
+      "learning_rate": 9.484882818627796e-06,
+      "loss": 0.685,
+      "step": 5382
+    },
+    {
+      "epoch": 0.2962738730805218,
+      "grad_norm": 0.776509702205658,
+      "learning_rate": 9.484691175677138e-06,
+      "loss": 0.8077,
+      "step": 5383
+    },
+    {
+      "epoch": 0.2963289118828774,
+      "grad_norm": 0.7436297535896301,
+      "learning_rate": 9.484499499020744e-06,
+      "loss": 0.8161,
+      "step": 5384
+    },
+    {
+      "epoch": 0.2963839506852331,
+      "grad_norm": 0.7604314088821411,
+      "learning_rate": 9.484307788660052e-06,
+      "loss": 0.825,
+      "step": 5385
+    },
+    {
+      "epoch": 0.29643898948758873,
+      "grad_norm": 0.7230789065361023,
+      "learning_rate": 9.484116044596501e-06,
+      "loss": 0.8005,
+      "step": 5386
+    },
+    {
+      "epoch": 0.2964940282899444,
+      "grad_norm": 0.820442259311676,
+      "learning_rate": 9.483924266831536e-06,
+      "loss": 0.789,
+      "step": 5387
+    },
+    {
+      "epoch": 0.29654906709230006,
+      "grad_norm": 0.7514582276344299,
+      "learning_rate": 9.483732455366596e-06,
+      "loss": 0.8531,
+      "step": 5388
+    },
+    {
+      "epoch": 0.29660410589465575,
+      "grad_norm": 0.6671503782272339,
+      "learning_rate": 9.483540610203124e-06,
+      "loss": 0.7627,
+      "step": 5389
+    },
+    {
+      "epoch": 0.2966591446970114,
+      "grad_norm": 0.6955942511558533,
+      "learning_rate": 9.483348731342559e-06,
+      "loss": 0.726,
+      "step": 5390
+    },
+    {
+      "epoch": 0.29671418349936707,
+      "grad_norm": 0.769781768321991,
+      "learning_rate": 9.483156818786347e-06,
+      "loss": 0.8064,
+      "step": 5391
+    },
+    {
+      "epoch": 0.2967692223017227,
+      "grad_norm": 1.0764707326889038,
+      "learning_rate": 9.482964872535927e-06,
+      "loss": 0.8249,
+      "step": 5392
+    },
+    {
+      "epoch": 0.2968242611040784,
+      "grad_norm": 1.0508921146392822,
+      "learning_rate": 9.482772892592744e-06,
+      "loss": 0.706,
+      "step": 5393
+    },
+    {
+      "epoch": 0.296879299906434,
+      "grad_norm": 0.6442564129829407,
+      "learning_rate": 9.482580878958239e-06,
+      "loss": 0.6025,
+      "step": 5394
+    },
+    {
+      "epoch": 0.2969343387087897,
+      "grad_norm": 0.7622735500335693,
+      "learning_rate": 9.482388831633856e-06,
+      "loss": 0.7639,
+      "step": 5395
+    },
+    {
+      "epoch": 0.29698937751114535,
+      "grad_norm": 0.8179057240486145,
+      "learning_rate": 9.482196750621038e-06,
+      "loss": 0.7641,
+      "step": 5396
+    },
+    {
+      "epoch": 0.29704441631350104,
+      "grad_norm": 0.7955192923545837,
+      "learning_rate": 9.48200463592123e-06,
+      "loss": 0.8407,
+      "step": 5397
+    },
+    {
+      "epoch": 0.29709945511585667,
+      "grad_norm": 0.7909773588180542,
+      "learning_rate": 9.481812487535875e-06,
+      "loss": 0.7833,
+      "step": 5398
+    },
+    {
+      "epoch": 0.29715449391821236,
+      "grad_norm": 0.8409042954444885,
+      "learning_rate": 9.481620305466417e-06,
+      "loss": 0.7788,
+      "step": 5399
+    },
+    {
+      "epoch": 0.297209532720568,
+      "grad_norm": 0.7521414160728455,
+      "learning_rate": 9.4814280897143e-06,
+      "loss": 0.7192,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2972645715229237,
+      "grad_norm": 0.7016280889511108,
+      "learning_rate": 9.481235840280969e-06,
+      "loss": 0.7181,
+      "step": 5401
+    },
+    {
+      "epoch": 0.2973196103252793,
+      "grad_norm": 0.7257362604141235,
+      "learning_rate": 9.48104355716787e-06,
+      "loss": 0.7845,
+      "step": 5402
+    },
+    {
+      "epoch": 0.297374649127635,
+      "grad_norm": 0.8048765659332275,
+      "learning_rate": 9.480851240376445e-06,
+      "loss": 0.7921,
+      "step": 5403
+    },
+    {
+      "epoch": 0.29742968792999064,
+      "grad_norm": 0.8715546131134033,
+      "learning_rate": 9.480658889908143e-06,
+      "loss": 0.856,
+      "step": 5404
+    },
+    {
+      "epoch": 0.2974847267323463,
+      "grad_norm": 0.7211160063743591,
+      "learning_rate": 9.480466505764408e-06,
+      "loss": 0.7687,
+      "step": 5405
+    },
+    {
+      "epoch": 0.29753976553470196,
+      "grad_norm": 0.8749645352363586,
+      "learning_rate": 9.480274087946686e-06,
+      "loss": 0.8419,
+      "step": 5406
+    },
+    {
+      "epoch": 0.29759480433705765,
+      "grad_norm": 0.7986398935317993,
+      "learning_rate": 9.480081636456424e-06,
+      "loss": 0.8309,
+      "step": 5407
+    },
+    {
+      "epoch": 0.2976498431394133,
+      "grad_norm": 0.8435508012771606,
+      "learning_rate": 9.479889151295067e-06,
+      "loss": 0.7457,
+      "step": 5408
+    },
+    {
+      "epoch": 0.297704881941769,
+      "grad_norm": 0.8725010752677917,
+      "learning_rate": 9.479696632464063e-06,
+      "loss": 0.8069,
+      "step": 5409
+    },
+    {
+      "epoch": 0.2977599207441246,
+      "grad_norm": 0.7364320158958435,
+      "learning_rate": 9.479504079964856e-06,
+      "loss": 0.8316,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2978149595464803,
+      "grad_norm": 0.7967824935913086,
+      "learning_rate": 9.479311493798898e-06,
+      "loss": 0.7689,
+      "step": 5411
+    },
+    {
+      "epoch": 0.29786999834883593,
+      "grad_norm": 0.8415414094924927,
+      "learning_rate": 9.479118873967632e-06,
+      "loss": 0.8288,
+      "step": 5412
+    },
+    {
+      "epoch": 0.29792503715119156,
+      "grad_norm": 0.9723265767097473,
+      "learning_rate": 9.478926220472508e-06,
+      "loss": 0.7422,
+      "step": 5413
+    },
+    {
+      "epoch": 0.29798007595354725,
+      "grad_norm": 0.7203155159950256,
+      "learning_rate": 9.478733533314974e-06,
+      "loss": 0.707,
+      "step": 5414
+    },
+    {
+      "epoch": 0.2980351147559029,
+      "grad_norm": 0.7643926739692688,
+      "learning_rate": 9.478540812496478e-06,
+      "loss": 0.7793,
+      "step": 5415
+    },
+    {
+      "epoch": 0.2980901535582586,
+      "grad_norm": 0.9177087545394897,
+      "learning_rate": 9.478348058018467e-06,
+      "loss": 0.865,
+      "step": 5416
+    },
+    {
+      "epoch": 0.2981451923606142,
+      "grad_norm": 0.678931713104248,
+      "learning_rate": 9.478155269882392e-06,
+      "loss": 0.7716,
+      "step": 5417
+    },
+    {
+      "epoch": 0.2982002311629699,
+      "grad_norm": 0.8440513610839844,
+      "learning_rate": 9.4779624480897e-06,
+      "loss": 0.8904,
+      "step": 5418
+    },
+    {
+      "epoch": 0.29825526996532553,
+      "grad_norm": 0.8508756756782532,
+      "learning_rate": 9.47776959264184e-06,
+      "loss": 0.7994,
+      "step": 5419
+    },
+    {
+      "epoch": 0.2983103087676812,
+      "grad_norm": 0.8736951947212219,
+      "learning_rate": 9.477576703540265e-06,
+      "loss": 0.8374,
+      "step": 5420
+    },
+    {
+      "epoch": 0.29836534757003685,
+      "grad_norm": 0.8063240051269531,
+      "learning_rate": 9.47738378078642e-06,
+      "loss": 0.7217,
+      "step": 5421
+    },
+    {
+      "epoch": 0.29842038637239254,
+      "grad_norm": 1.1495088338851929,
+      "learning_rate": 9.477190824381757e-06,
+      "loss": 0.8902,
+      "step": 5422
+    },
+    {
+      "epoch": 0.2984754251747482,
+      "grad_norm": 1.0241554975509644,
+      "learning_rate": 9.476997834327725e-06,
+      "loss": 0.9354,
+      "step": 5423
+    },
+    {
+      "epoch": 0.29853046397710387,
+      "grad_norm": 0.939950168132782,
+      "learning_rate": 9.476804810625779e-06,
+      "loss": 0.8714,
+      "step": 5424
+    },
+    {
+      "epoch": 0.2985855027794595,
+      "grad_norm": 0.7592660188674927,
+      "learning_rate": 9.476611753277364e-06,
+      "loss": 0.7513,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2986405415818152,
+      "grad_norm": 0.776153028011322,
+      "learning_rate": 9.476418662283935e-06,
+      "loss": 0.7828,
+      "step": 5426
+    },
+    {
+      "epoch": 0.2986955803841708,
+      "grad_norm": 0.9317814707756042,
+      "learning_rate": 9.47622553764694e-06,
+      "loss": 0.865,
+      "step": 5427
+    },
+    {
+      "epoch": 0.2987506191865265,
+      "grad_norm": 0.7770501971244812,
+      "learning_rate": 9.476032379367832e-06,
+      "loss": 0.7281,
+      "step": 5428
+    },
+    {
+      "epoch": 0.29880565798888215,
+      "grad_norm": 0.7815201282501221,
+      "learning_rate": 9.475839187448064e-06,
+      "loss": 0.7565,
+      "step": 5429
+    },
+    {
+      "epoch": 0.29886069679123783,
+      "grad_norm": 0.7992607951164246,
+      "learning_rate": 9.475645961889086e-06,
+      "loss": 0.8109,
+      "step": 5430
+    },
+    {
+      "epoch": 0.29891573559359347,
+      "grad_norm": 0.7780614495277405,
+      "learning_rate": 9.475452702692351e-06,
+      "loss": 0.7814,
+      "step": 5431
+    },
+    {
+      "epoch": 0.29897077439594916,
+      "grad_norm": 0.7409062385559082,
+      "learning_rate": 9.475259409859313e-06,
+      "loss": 0.7712,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2990258131983048,
+      "grad_norm": 0.7935584187507629,
+      "learning_rate": 9.47506608339142e-06,
+      "loss": 0.8301,
+      "step": 5433
+    },
+    {
+      "epoch": 0.2990808520006605,
+      "grad_norm": 0.6931030750274658,
+      "learning_rate": 9.474872723290132e-06,
+      "loss": 0.7471,
+      "step": 5434
+    },
+    {
+      "epoch": 0.2991358908030161,
+      "grad_norm": 0.7622918486595154,
+      "learning_rate": 9.474679329556894e-06,
+      "loss": 0.7727,
+      "step": 5435
+    },
+    {
+      "epoch": 0.2991909296053718,
+      "grad_norm": 0.7957701086997986,
+      "learning_rate": 9.474485902193169e-06,
+      "loss": 0.7663,
+      "step": 5436
+    },
+    {
+      "epoch": 0.29924596840772744,
+      "grad_norm": 1.0600612163543701,
+      "learning_rate": 9.474292441200404e-06,
+      "loss": 0.7861,
+      "step": 5437
+    },
+    {
+      "epoch": 0.2993010072100831,
+      "grad_norm": 0.7343600392341614,
+      "learning_rate": 9.474098946580053e-06,
+      "loss": 0.8609,
+      "step": 5438
+    },
+    {
+      "epoch": 0.29935604601243876,
+      "grad_norm": 0.7477726340293884,
+      "learning_rate": 9.473905418333573e-06,
+      "loss": 0.7683,
+      "step": 5439
+    },
+    {
+      "epoch": 0.29941108481479445,
+      "grad_norm": 0.7955546379089355,
+      "learning_rate": 9.473711856462417e-06,
+      "loss": 0.8406,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2994661236171501,
+      "grad_norm": 0.8291183114051819,
+      "learning_rate": 9.47351826096804e-06,
+      "loss": 0.6919,
+      "step": 5441
+    },
+    {
+      "epoch": 0.29952116241950577,
+      "grad_norm": 0.8899849057197571,
+      "learning_rate": 9.473324631851898e-06,
+      "loss": 0.9403,
+      "step": 5442
+    },
+    {
+      "epoch": 0.2995762012218614,
+      "grad_norm": 0.837066650390625,
+      "learning_rate": 9.473130969115445e-06,
+      "loss": 0.8676,
+      "step": 5443
+    },
+    {
+      "epoch": 0.2996312400242171,
+      "grad_norm": 0.8385708928108215,
+      "learning_rate": 9.472937272760138e-06,
+      "loss": 0.7588,
+      "step": 5444
+    },
+    {
+      "epoch": 0.2996862788265727,
+      "grad_norm": 0.6990595459938049,
+      "learning_rate": 9.472743542787431e-06,
+      "loss": 0.6769,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2997413176289284,
+      "grad_norm": 0.789165735244751,
+      "learning_rate": 9.472549779198781e-06,
+      "loss": 0.8084,
+      "step": 5446
+    },
+    {
+      "epoch": 0.29979635643128405,
+      "grad_norm": 0.8820298314094543,
+      "learning_rate": 9.472355981995643e-06,
+      "loss": 0.8262,
+      "step": 5447
+    },
+    {
+      "epoch": 0.29985139523363974,
+      "grad_norm": 0.8928382992744446,
+      "learning_rate": 9.472162151179475e-06,
+      "loss": 0.8123,
+      "step": 5448
+    },
+    {
+      "epoch": 0.2999064340359954,
+      "grad_norm": 0.7688086032867432,
+      "learning_rate": 9.471968286751735e-06,
+      "loss": 0.6846,
+      "step": 5449
+    },
+    {
+      "epoch": 0.29996147283835106,
+      "grad_norm": 0.6962918043136597,
+      "learning_rate": 9.471774388713877e-06,
+      "loss": 0.7872,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3000165116407067,
+      "grad_norm": 0.7467569708824158,
+      "learning_rate": 9.47158045706736e-06,
+      "loss": 0.8201,
+      "step": 5451
+    },
+    {
+      "epoch": 0.3000715504430624,
+      "grad_norm": 0.7651814222335815,
+      "learning_rate": 9.471386491813642e-06,
+      "loss": 0.7734,
+      "step": 5452
+    },
+    {
+      "epoch": 0.300126589245418,
+      "grad_norm": 0.8001144528388977,
+      "learning_rate": 9.47119249295418e-06,
+      "loss": 0.8266,
+      "step": 5453
+    },
+    {
+      "epoch": 0.3001816280477737,
+      "grad_norm": 0.7937704920768738,
+      "learning_rate": 9.47099846049043e-06,
+      "loss": 0.8025,
+      "step": 5454
     }
   ],
   "logging_steps": 1,
@@ -25478,7 +38204,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.0730058856487977e+19,
+  "total_flos": 1.6095088284731965e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null