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": 0.05003027134129561,
+  "epoch": 0.10006054268259122,
   "eval_steps": 500,
-  "global_step": 909,
+  "global_step": 1818,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6370,6 +6370,6369 @@
       "learning_rate": 9.987707779910499e-06,
       "loss": 0.9263,
       "step": 909
+    },
+    {
+      "epoch": 0.05008531014365127,
+      "grad_norm": 0.9423969388008118,
+      "learning_rate": 9.987677384947402e-06,
+      "loss": 0.9667,
+      "step": 910
+    },
+    {
+      "epoch": 0.050140348946006934,
+      "grad_norm": 1.5940319299697876,
+      "learning_rate": 9.987646952498256e-06,
+      "loss": 0.9223,
+      "step": 911
+    },
+    {
+      "epoch": 0.050195387748362595,
+      "grad_norm": 0.941792368888855,
+      "learning_rate": 9.987616482563292e-06,
+      "loss": 0.895,
+      "step": 912
+    },
+    {
+      "epoch": 0.05025042655071826,
+      "grad_norm": 3.1945221424102783,
+      "learning_rate": 9.987585975142738e-06,
+      "loss": 0.837,
+      "step": 913
+    },
+    {
+      "epoch": 0.05030546535307392,
+      "grad_norm": 2.0819199085235596,
+      "learning_rate": 9.98755543023682e-06,
+      "loss": 0.918,
+      "step": 914
+    },
+    {
+      "epoch": 0.05036050415542958,
+      "grad_norm": 0.984282910823822,
+      "learning_rate": 9.987524847845773e-06,
+      "loss": 0.8589,
+      "step": 915
+    },
+    {
+      "epoch": 0.05041554295778524,
+      "grad_norm": 0.9021026492118835,
+      "learning_rate": 9.987494227969823e-06,
+      "loss": 0.9053,
+      "step": 916
+    },
+    {
+      "epoch": 0.0504705817601409,
+      "grad_norm": 2.6515462398529053,
+      "learning_rate": 9.9874635706092e-06,
+      "loss": 0.8874,
+      "step": 917
+    },
+    {
+      "epoch": 0.050525620562496563,
+      "grad_norm": 0.8893095254898071,
+      "learning_rate": 9.98743287576414e-06,
+      "loss": 0.8259,
+      "step": 918
+    },
+    {
+      "epoch": 0.05058065936485222,
+      "grad_norm": 0.9897775650024414,
+      "learning_rate": 9.987402143434868e-06,
+      "loss": 0.877,
+      "step": 919
+    },
+    {
+      "epoch": 0.05063569816720788,
+      "grad_norm": 0.9391944408416748,
+      "learning_rate": 9.987371373621614e-06,
+      "loss": 0.9363,
+      "step": 920
+    },
+    {
+      "epoch": 0.05069073696956354,
+      "grad_norm": 0.9585913419723511,
+      "learning_rate": 9.987340566324615e-06,
+      "loss": 0.8704,
+      "step": 921
+    },
+    {
+      "epoch": 0.0507457757719192,
+      "grad_norm": 0.9210980534553528,
+      "learning_rate": 9.987309721544098e-06,
+      "loss": 0.9321,
+      "step": 922
+    },
+    {
+      "epoch": 0.05080081457427486,
+      "grad_norm": 1.0713307857513428,
+      "learning_rate": 9.987278839280295e-06,
+      "loss": 0.9489,
+      "step": 923
+    },
+    {
+      "epoch": 0.050855853376630525,
+      "grad_norm": 1.0178636312484741,
+      "learning_rate": 9.98724791953344e-06,
+      "loss": 0.853,
+      "step": 924
+    },
+    {
+      "epoch": 0.050910892178986186,
+      "grad_norm": 0.9782636761665344,
+      "learning_rate": 9.987216962303766e-06,
+      "loss": 0.924,
+      "step": 925
+    },
+    {
+      "epoch": 0.05096593098134185,
+      "grad_norm": 0.9474522471427917,
+      "learning_rate": 9.987185967591503e-06,
+      "loss": 0.8619,
+      "step": 926
+    },
+    {
+      "epoch": 0.05102096978369751,
+      "grad_norm": 1.1875778436660767,
+      "learning_rate": 9.987154935396885e-06,
+      "loss": 1.012,
+      "step": 927
+    },
+    {
+      "epoch": 0.05107600858605317,
+      "grad_norm": 1.0585243701934814,
+      "learning_rate": 9.987123865720147e-06,
+      "loss": 0.9019,
+      "step": 928
+    },
+    {
+      "epoch": 0.051131047388408825,
+      "grad_norm": 0.9848800897598267,
+      "learning_rate": 9.98709275856152e-06,
+      "loss": 0.9665,
+      "step": 929
+    },
+    {
+      "epoch": 0.051186086190764486,
+      "grad_norm": 1.04201078414917,
+      "learning_rate": 9.987061613921238e-06,
+      "loss": 0.9269,
+      "step": 930
+    },
+    {
+      "epoch": 0.05124112499312015,
+      "grad_norm": 1.1748600006103516,
+      "learning_rate": 9.987030431799537e-06,
+      "loss": 0.8565,
+      "step": 931
+    },
+    {
+      "epoch": 0.05129616379547581,
+      "grad_norm": 1.879232406616211,
+      "learning_rate": 9.98699921219665e-06,
+      "loss": 0.8535,
+      "step": 932
+    },
+    {
+      "epoch": 0.05135120259783147,
+      "grad_norm": 0.9837847948074341,
+      "learning_rate": 9.986967955112812e-06,
+      "loss": 0.927,
+      "step": 933
+    },
+    {
+      "epoch": 0.05140624140018713,
+      "grad_norm": 0.8637211918830872,
+      "learning_rate": 9.986936660548257e-06,
+      "loss": 0.7903,
+      "step": 934
+    },
+    {
+      "epoch": 0.05146128020254279,
+      "grad_norm": 0.9078792929649353,
+      "learning_rate": 9.986905328503222e-06,
+      "loss": 0.9135,
+      "step": 935
+    },
+    {
+      "epoch": 0.051516319004898455,
+      "grad_norm": 0.9763005971908569,
+      "learning_rate": 9.98687395897794e-06,
+      "loss": 0.9006,
+      "step": 936
+    },
+    {
+      "epoch": 0.051571357807254116,
+      "grad_norm": 1.0174345970153809,
+      "learning_rate": 9.98684255197265e-06,
+      "loss": 0.9294,
+      "step": 937
+    },
+    {
+      "epoch": 0.05162639660960978,
+      "grad_norm": 0.8709769248962402,
+      "learning_rate": 9.986811107487584e-06,
+      "loss": 0.7986,
+      "step": 938
+    },
+    {
+      "epoch": 0.05168143541196544,
+      "grad_norm": 0.8717525601387024,
+      "learning_rate": 9.986779625522983e-06,
+      "loss": 0.8705,
+      "step": 939
+    },
+    {
+      "epoch": 0.05173647421432109,
+      "grad_norm": 0.9682945013046265,
+      "learning_rate": 9.98674810607908e-06,
+      "loss": 0.8127,
+      "step": 940
+    },
+    {
+      "epoch": 0.051791513016676755,
+      "grad_norm": 1.0248037576675415,
+      "learning_rate": 9.986716549156113e-06,
+      "loss": 0.9217,
+      "step": 941
+    },
+    {
+      "epoch": 0.051846551819032416,
+      "grad_norm": 0.9883397221565247,
+      "learning_rate": 9.98668495475432e-06,
+      "loss": 0.853,
+      "step": 942
+    },
+    {
+      "epoch": 0.05190159062138808,
+      "grad_norm": 0.9271108508110046,
+      "learning_rate": 9.986653322873937e-06,
+      "loss": 0.8807,
+      "step": 943
+    },
+    {
+      "epoch": 0.05195662942374374,
+      "grad_norm": 0.9027101397514343,
+      "learning_rate": 9.986621653515203e-06,
+      "loss": 0.88,
+      "step": 944
+    },
+    {
+      "epoch": 0.0520116682260994,
+      "grad_norm": 0.9807021617889404,
+      "learning_rate": 9.986589946678354e-06,
+      "loss": 0.8922,
+      "step": 945
+    },
+    {
+      "epoch": 0.05206670702845506,
+      "grad_norm": 0.8779157400131226,
+      "learning_rate": 9.98655820236363e-06,
+      "loss": 0.8988,
+      "step": 946
+    },
+    {
+      "epoch": 0.05212174583081072,
+      "grad_norm": 0.8182910680770874,
+      "learning_rate": 9.986526420571272e-06,
+      "loss": 0.7534,
+      "step": 947
+    },
+    {
+      "epoch": 0.052176784633166384,
+      "grad_norm": 0.9205981492996216,
+      "learning_rate": 9.986494601301513e-06,
+      "loss": 0.7516,
+      "step": 948
+    },
+    {
+      "epoch": 0.052231823435522046,
+      "grad_norm": 0.9829681515693665,
+      "learning_rate": 9.986462744554598e-06,
+      "loss": 0.9358,
+      "step": 949
+    },
+    {
+      "epoch": 0.05228686223787771,
+      "grad_norm": 0.8869890570640564,
+      "learning_rate": 9.986430850330762e-06,
+      "loss": 0.7933,
+      "step": 950
+    },
+    {
+      "epoch": 0.05234190104023336,
+      "grad_norm": 1.0226716995239258,
+      "learning_rate": 9.986398918630248e-06,
+      "loss": 0.9523,
+      "step": 951
+    },
+    {
+      "epoch": 0.05239693984258902,
+      "grad_norm": 0.9549778699874878,
+      "learning_rate": 9.986366949453293e-06,
+      "loss": 0.9368,
+      "step": 952
+    },
+    {
+      "epoch": 0.052451978644944684,
+      "grad_norm": 0.860454797744751,
+      "learning_rate": 9.98633494280014e-06,
+      "loss": 0.7618,
+      "step": 953
+    },
+    {
+      "epoch": 0.052507017447300346,
+      "grad_norm": 0.9623841643333435,
+      "learning_rate": 9.986302898671027e-06,
+      "loss": 0.8356,
+      "step": 954
+    },
+    {
+      "epoch": 0.05256205624965601,
+      "grad_norm": 0.9236606359481812,
+      "learning_rate": 9.986270817066196e-06,
+      "loss": 0.921,
+      "step": 955
+    },
+    {
+      "epoch": 0.05261709505201167,
+      "grad_norm": 1.0599812269210815,
+      "learning_rate": 9.98623869798589e-06,
+      "loss": 0.8082,
+      "step": 956
+    },
+    {
+      "epoch": 0.05267213385436733,
+      "grad_norm": 1.0321687459945679,
+      "learning_rate": 9.986206541430347e-06,
+      "loss": 0.9001,
+      "step": 957
+    },
+    {
+      "epoch": 0.05272717265672299,
+      "grad_norm": 0.8884543776512146,
+      "learning_rate": 9.986174347399813e-06,
+      "loss": 0.8317,
+      "step": 958
+    },
+    {
+      "epoch": 0.05278221145907865,
+      "grad_norm": 0.9592668414115906,
+      "learning_rate": 9.986142115894526e-06,
+      "loss": 0.9955,
+      "step": 959
+    },
+    {
+      "epoch": 0.052837250261434314,
+      "grad_norm": 0.9604032039642334,
+      "learning_rate": 9.986109846914729e-06,
+      "loss": 0.876,
+      "step": 960
+    },
+    {
+      "epoch": 0.052892289063789975,
+      "grad_norm": 0.9837536811828613,
+      "learning_rate": 9.986077540460664e-06,
+      "loss": 0.8247,
+      "step": 961
+    },
+    {
+      "epoch": 0.05294732786614563,
+      "grad_norm": 0.8570861220359802,
+      "learning_rate": 9.986045196532576e-06,
+      "loss": 0.879,
+      "step": 962
+    },
+    {
+      "epoch": 0.05300236666850129,
+      "grad_norm": 0.8441471457481384,
+      "learning_rate": 9.986012815130708e-06,
+      "loss": 0.8979,
+      "step": 963
+    },
+    {
+      "epoch": 0.05305740547085695,
+      "grad_norm": 0.8976197838783264,
+      "learning_rate": 9.985980396255302e-06,
+      "loss": 0.9382,
+      "step": 964
+    },
+    {
+      "epoch": 0.053112444273212614,
+      "grad_norm": 0.9685307741165161,
+      "learning_rate": 9.985947939906599e-06,
+      "loss": 0.8627,
+      "step": 965
+    },
+    {
+      "epoch": 0.053167483075568275,
+      "grad_norm": 0.8939018249511719,
+      "learning_rate": 9.98591544608485e-06,
+      "loss": 0.9221,
+      "step": 966
+    },
+    {
+      "epoch": 0.05322252187792394,
+      "grad_norm": 0.9218310713768005,
+      "learning_rate": 9.985882914790292e-06,
+      "loss": 0.8356,
+      "step": 967
+    },
+    {
+      "epoch": 0.0532775606802796,
+      "grad_norm": 0.9342261552810669,
+      "learning_rate": 9.985850346023174e-06,
+      "loss": 0.971,
+      "step": 968
+    },
+    {
+      "epoch": 0.05333259948263526,
+      "grad_norm": 1.0860705375671387,
+      "learning_rate": 9.985817739783741e-06,
+      "loss": 0.906,
+      "step": 969
+    },
+    {
+      "epoch": 0.05338763828499092,
+      "grad_norm": 0.8675006031990051,
+      "learning_rate": 9.985785096072234e-06,
+      "loss": 0.906,
+      "step": 970
+    },
+    {
+      "epoch": 0.05344267708734658,
+      "grad_norm": 0.8170626163482666,
+      "learning_rate": 9.985752414888903e-06,
+      "loss": 0.8109,
+      "step": 971
+    },
+    {
+      "epoch": 0.05349771588970224,
+      "grad_norm": 0.936434805393219,
+      "learning_rate": 9.98571969623399e-06,
+      "loss": 0.9219,
+      "step": 972
+    },
+    {
+      "epoch": 0.0535527546920579,
+      "grad_norm": 0.9316715002059937,
+      "learning_rate": 9.985686940107741e-06,
+      "loss": 0.8569,
+      "step": 973
+    },
+    {
+      "epoch": 0.05360779349441356,
+      "grad_norm": 1.183008074760437,
+      "learning_rate": 9.985654146510405e-06,
+      "loss": 0.837,
+      "step": 974
+    },
+    {
+      "epoch": 0.05366283229676922,
+      "grad_norm": 1.0788745880126953,
+      "learning_rate": 9.98562131544223e-06,
+      "loss": 0.8822,
+      "step": 975
+    },
+    {
+      "epoch": 0.05371787109912488,
+      "grad_norm": 0.9285461902618408,
+      "learning_rate": 9.985588446903455e-06,
+      "loss": 0.9279,
+      "step": 976
+    },
+    {
+      "epoch": 0.053772909901480544,
+      "grad_norm": 0.9389022588729858,
+      "learning_rate": 9.985555540894334e-06,
+      "loss": 0.9839,
+      "step": 977
+    },
+    {
+      "epoch": 0.053827948703836205,
+      "grad_norm": 0.8920616507530212,
+      "learning_rate": 9.985522597415112e-06,
+      "loss": 0.9205,
+      "step": 978
+    },
+    {
+      "epoch": 0.053882987506191866,
+      "grad_norm": 0.9755093455314636,
+      "learning_rate": 9.985489616466035e-06,
+      "loss": 0.8987,
+      "step": 979
+    },
+    {
+      "epoch": 0.05393802630854753,
+      "grad_norm": 0.96027010679245,
+      "learning_rate": 9.985456598047356e-06,
+      "loss": 0.8543,
+      "step": 980
+    },
+    {
+      "epoch": 0.05399306511090319,
+      "grad_norm": 1.0489718914031982,
+      "learning_rate": 9.985423542159317e-06,
+      "loss": 0.9179,
+      "step": 981
+    },
+    {
+      "epoch": 0.05404810391325885,
+      "grad_norm": 0.8665526509284973,
+      "learning_rate": 9.985390448802171e-06,
+      "loss": 0.9047,
+      "step": 982
+    },
+    {
+      "epoch": 0.054103142715614505,
+      "grad_norm": 0.8849464654922485,
+      "learning_rate": 9.985357317976163e-06,
+      "loss": 0.8892,
+      "step": 983
+    },
+    {
+      "epoch": 0.054158181517970166,
+      "grad_norm": 1.0083115100860596,
+      "learning_rate": 9.985324149681545e-06,
+      "loss": 0.7713,
+      "step": 984
+    },
+    {
+      "epoch": 0.05421322032032583,
+      "grad_norm": 0.8233863711357117,
+      "learning_rate": 9.985290943918565e-06,
+      "loss": 0.7967,
+      "step": 985
+    },
+    {
+      "epoch": 0.05426825912268149,
+      "grad_norm": 0.9615303874015808,
+      "learning_rate": 9.985257700687472e-06,
+      "loss": 0.8576,
+      "step": 986
+    },
+    {
+      "epoch": 0.05432329792503715,
+      "grad_norm": 0.8856416344642639,
+      "learning_rate": 9.985224419988517e-06,
+      "loss": 0.8614,
+      "step": 987
+    },
+    {
+      "epoch": 0.05437833672739281,
+      "grad_norm": 0.968325674533844,
+      "learning_rate": 9.98519110182195e-06,
+      "loss": 0.8247,
+      "step": 988
+    },
+    {
+      "epoch": 0.05443337552974847,
+      "grad_norm": 0.878402054309845,
+      "learning_rate": 9.985157746188021e-06,
+      "loss": 0.8661,
+      "step": 989
+    },
+    {
+      "epoch": 0.054488414332104135,
+      "grad_norm": 0.8376438021659851,
+      "learning_rate": 9.985124353086981e-06,
+      "loss": 0.8554,
+      "step": 990
+    },
+    {
+      "epoch": 0.054543453134459796,
+      "grad_norm": 1.0293036699295044,
+      "learning_rate": 9.98509092251908e-06,
+      "loss": 0.8049,
+      "step": 991
+    },
+    {
+      "epoch": 0.05459849193681546,
+      "grad_norm": 1.2345234155654907,
+      "learning_rate": 9.98505745448457e-06,
+      "loss": 1.0358,
+      "step": 992
+    },
+    {
+      "epoch": 0.05465353073917112,
+      "grad_norm": 0.9974482655525208,
+      "learning_rate": 9.985023948983703e-06,
+      "loss": 0.9329,
+      "step": 993
+    },
+    {
+      "epoch": 0.05470856954152677,
+      "grad_norm": 1.383955478668213,
+      "learning_rate": 9.984990406016732e-06,
+      "loss": 0.8688,
+      "step": 994
+    },
+    {
+      "epoch": 0.054763608343882435,
+      "grad_norm": 0.9369306564331055,
+      "learning_rate": 9.984956825583906e-06,
+      "loss": 0.8308,
+      "step": 995
+    },
+    {
+      "epoch": 0.054818647146238096,
+      "grad_norm": 0.8676120042800903,
+      "learning_rate": 9.984923207685478e-06,
+      "loss": 0.8283,
+      "step": 996
+    },
+    {
+      "epoch": 0.05487368594859376,
+      "grad_norm": 0.9218453168869019,
+      "learning_rate": 9.984889552321704e-06,
+      "loss": 0.7247,
+      "step": 997
+    },
+    {
+      "epoch": 0.05492872475094942,
+      "grad_norm": 0.8575478196144104,
+      "learning_rate": 9.984855859492833e-06,
+      "loss": 0.8462,
+      "step": 998
+    },
+    {
+      "epoch": 0.05498376355330508,
+      "grad_norm": 1.0042616128921509,
+      "learning_rate": 9.98482212919912e-06,
+      "loss": 0.9383,
+      "step": 999
+    },
+    {
+      "epoch": 0.05503880235566074,
+      "grad_norm": 0.8642181158065796,
+      "learning_rate": 9.984788361440817e-06,
+      "loss": 0.8805,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0550938411580164,
+      "grad_norm": 0.8413823843002319,
+      "learning_rate": 9.984754556218178e-06,
+      "loss": 0.8161,
+      "step": 1001
+    },
+    {
+      "epoch": 0.055148879960372064,
+      "grad_norm": 0.9473856091499329,
+      "learning_rate": 9.984720713531462e-06,
+      "loss": 0.8425,
+      "step": 1002
+    },
+    {
+      "epoch": 0.055203918762727726,
+      "grad_norm": 0.7854379415512085,
+      "learning_rate": 9.984686833380917e-06,
+      "loss": 0.7506,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05525895756508339,
+      "grad_norm": 0.9481745958328247,
+      "learning_rate": 9.984652915766801e-06,
+      "loss": 0.954,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05531399636743904,
+      "grad_norm": 0.767803966999054,
+      "learning_rate": 9.984618960689366e-06,
+      "loss": 0.8113,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0553690351697947,
+      "grad_norm": 0.8957781195640564,
+      "learning_rate": 9.984584968148871e-06,
+      "loss": 0.9042,
+      "step": 1006
+    },
+    {
+      "epoch": 0.055424073972150364,
+      "grad_norm": 1.116646409034729,
+      "learning_rate": 9.98455093814557e-06,
+      "loss": 0.8648,
+      "step": 1007
+    },
+    {
+      "epoch": 0.055479112774506026,
+      "grad_norm": 0.9567018151283264,
+      "learning_rate": 9.98451687067972e-06,
+      "loss": 0.9446,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05553415157686169,
+      "grad_norm": 0.8470665812492371,
+      "learning_rate": 9.98448276575157e-06,
+      "loss": 0.8186,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05558919037921735,
+      "grad_norm": 0.9595193862915039,
+      "learning_rate": 9.984448623361387e-06,
+      "loss": 0.8406,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05564422918157301,
+      "grad_norm": 1.0579735040664673,
+      "learning_rate": 9.98441444350942e-06,
+      "loss": 0.9676,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05569926798392867,
+      "grad_norm": 0.8693701028823853,
+      "learning_rate": 9.98438022619593e-06,
+      "loss": 0.9451,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05575430678628433,
+      "grad_norm": 0.9251859784126282,
+      "learning_rate": 9.98434597142117e-06,
+      "loss": 0.7858,
+      "step": 1013
+    },
+    {
+      "epoch": 0.055809345588639994,
+      "grad_norm": 0.8584280014038086,
+      "learning_rate": 9.984311679185402e-06,
+      "loss": 0.8481,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05586438439099565,
+      "grad_norm": 0.8903968334197998,
+      "learning_rate": 9.98427734948888e-06,
+      "loss": 0.7832,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05591942319335131,
+      "grad_norm": 0.905581533908844,
+      "learning_rate": 9.984242982331864e-06,
+      "loss": 0.8088,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05597446199570697,
+      "grad_norm": 0.9866476655006409,
+      "learning_rate": 9.984208577714612e-06,
+      "loss": 0.8366,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05602950079806263,
+      "grad_norm": 0.8843809962272644,
+      "learning_rate": 9.984174135637384e-06,
+      "loss": 0.8961,
+      "step": 1018
+    },
+    {
+      "epoch": 0.056084539600418294,
+      "grad_norm": 0.9071753621101379,
+      "learning_rate": 9.984139656100435e-06,
+      "loss": 0.8671,
+      "step": 1019
+    },
+    {
+      "epoch": 0.056139578402773956,
+      "grad_norm": 0.9894018173217773,
+      "learning_rate": 9.984105139104028e-06,
+      "loss": 0.9099,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05619461720512962,
+      "grad_norm": 0.8432741165161133,
+      "learning_rate": 9.98407058464842e-06,
+      "loss": 0.7817,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05624965600748528,
+      "grad_norm": 0.9538390040397644,
+      "learning_rate": 9.984035992733873e-06,
+      "loss": 0.8689,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05630469480984094,
+      "grad_norm": 0.9263421297073364,
+      "learning_rate": 9.984001363360645e-06,
+      "loss": 0.9066,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0563597336121966,
+      "grad_norm": 0.8921047449111938,
+      "learning_rate": 9.983966696528996e-06,
+      "loss": 0.8304,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05641477241455226,
+      "grad_norm": 0.8379812240600586,
+      "learning_rate": 9.983931992239188e-06,
+      "loss": 0.866,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05646981121690792,
+      "grad_norm": 0.9444219470024109,
+      "learning_rate": 9.983897250491481e-06,
+      "loss": 0.9456,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05652485001926358,
+      "grad_norm": 1.0268759727478027,
+      "learning_rate": 9.983862471286137e-06,
+      "loss": 0.8277,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05657988882161924,
+      "grad_norm": 1.3949217796325684,
+      "learning_rate": 9.983827654623418e-06,
+      "loss": 0.9721,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0566349276239749,
+      "grad_norm": 0.8899377584457397,
+      "learning_rate": 9.983792800503582e-06,
+      "loss": 0.8794,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05668996642633056,
+      "grad_norm": 0.989072322845459,
+      "learning_rate": 9.983757908926895e-06,
+      "loss": 0.8852,
+      "step": 1030
+    },
+    {
+      "epoch": 0.056745005228686224,
+      "grad_norm": 0.9797759056091309,
+      "learning_rate": 9.983722979893615e-06,
+      "loss": 1.0405,
+      "step": 1031
+    },
+    {
+      "epoch": 0.056800044031041885,
+      "grad_norm": 0.9044767618179321,
+      "learning_rate": 9.98368801340401e-06,
+      "loss": 0.7243,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05685508283339755,
+      "grad_norm": 1.116324782371521,
+      "learning_rate": 9.983653009458338e-06,
+      "loss": 0.9183,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05691012163575321,
+      "grad_norm": 0.9373337030410767,
+      "learning_rate": 9.983617968056866e-06,
+      "loss": 0.9417,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05696516043810887,
+      "grad_norm": 1.0587197542190552,
+      "learning_rate": 9.983582889199855e-06,
+      "loss": 0.896,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05702019924046453,
+      "grad_norm": 1.0080119371414185,
+      "learning_rate": 9.983547772887568e-06,
+      "loss": 0.924,
+      "step": 1036
+    },
+    {
+      "epoch": 0.057075238042820185,
+      "grad_norm": 0.847091019153595,
+      "learning_rate": 9.98351261912027e-06,
+      "loss": 0.7443,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05713027684517585,
+      "grad_norm": 0.9876272082328796,
+      "learning_rate": 9.983477427898225e-06,
+      "loss": 0.9365,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05718531564753151,
+      "grad_norm": 0.9188169240951538,
+      "learning_rate": 9.983442199221698e-06,
+      "loss": 0.9213,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05724035444988717,
+      "grad_norm": 0.932399332523346,
+      "learning_rate": 9.983406933090954e-06,
+      "loss": 0.958,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05729539325224283,
+      "grad_norm": 0.9126465320587158,
+      "learning_rate": 9.983371629506258e-06,
+      "loss": 0.8913,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05735043205459849,
+      "grad_norm": 0.80904620885849,
+      "learning_rate": 9.983336288467873e-06,
+      "loss": 0.7719,
+      "step": 1042
+    },
+    {
+      "epoch": 0.057405470856954154,
+      "grad_norm": 0.873833417892456,
+      "learning_rate": 9.983300909976067e-06,
+      "loss": 0.9201,
+      "step": 1043
+    },
+    {
+      "epoch": 0.057460509659309815,
+      "grad_norm": 0.8331829309463501,
+      "learning_rate": 9.983265494031107e-06,
+      "loss": 0.8605,
+      "step": 1044
+    },
+    {
+      "epoch": 0.057515548461665476,
+      "grad_norm": 0.8364768624305725,
+      "learning_rate": 9.983230040633255e-06,
+      "loss": 0.8627,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05757058726402114,
+      "grad_norm": 0.9226736426353455,
+      "learning_rate": 9.98319454978278e-06,
+      "loss": 0.9759,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05762562606637679,
+      "grad_norm": 0.8174427151679993,
+      "learning_rate": 9.98315902147995e-06,
+      "loss": 0.8066,
+      "step": 1047
+    },
+    {
+      "epoch": 0.057680664868732454,
+      "grad_norm": 0.9154924750328064,
+      "learning_rate": 9.98312345572503e-06,
+      "loss": 0.9112,
+      "step": 1048
+    },
+    {
+      "epoch": 0.057735703671088115,
+      "grad_norm": 0.8884655237197876,
+      "learning_rate": 9.983087852518289e-06,
+      "loss": 0.8699,
+      "step": 1049
+    },
+    {
+      "epoch": 0.057790742473443776,
+      "grad_norm": 0.8849230408668518,
+      "learning_rate": 9.983052211859992e-06,
+      "loss": 0.8999,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05784578127579944,
+      "grad_norm": 1.025843858718872,
+      "learning_rate": 9.98301653375041e-06,
+      "loss": 0.7764,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0579008200781551,
+      "grad_norm": 0.900505006313324,
+      "learning_rate": 9.98298081818981e-06,
+      "loss": 0.9196,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05795585888051076,
+      "grad_norm": 0.9506704211235046,
+      "learning_rate": 9.982945065178457e-06,
+      "loss": 0.8319,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05801089768286642,
+      "grad_norm": 0.9439849853515625,
+      "learning_rate": 9.982909274716626e-06,
+      "loss": 0.8561,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05806593648522208,
+      "grad_norm": 0.8761240243911743,
+      "learning_rate": 9.982873446804579e-06,
+      "loss": 0.9681,
+      "step": 1055
+    },
+    {
+      "epoch": 0.058120975287577745,
+      "grad_norm": 0.8756145238876343,
+      "learning_rate": 9.982837581442592e-06,
+      "loss": 0.8452,
+      "step": 1056
+    },
+    {
+      "epoch": 0.058176014089933406,
+      "grad_norm": 0.8732383847236633,
+      "learning_rate": 9.982801678630932e-06,
+      "loss": 0.9018,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05823105289228906,
+      "grad_norm": 0.8338272571563721,
+      "learning_rate": 9.982765738369867e-06,
+      "loss": 0.9308,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05828609169464472,
+      "grad_norm": 0.843163013458252,
+      "learning_rate": 9.982729760659669e-06,
+      "loss": 0.7802,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05834113049700038,
+      "grad_norm": 1.2007580995559692,
+      "learning_rate": 9.982693745500606e-06,
+      "loss": 0.8406,
+      "step": 1060
+    },
+    {
+      "epoch": 0.058396169299356045,
+      "grad_norm": 0.8760073781013489,
+      "learning_rate": 9.982657692892954e-06,
+      "loss": 0.8528,
+      "step": 1061
+    },
+    {
+      "epoch": 0.058451208101711706,
+      "grad_norm": 0.925309419631958,
+      "learning_rate": 9.982621602836978e-06,
+      "loss": 0.9601,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05850624690406737,
+      "grad_norm": 0.9277135133743286,
+      "learning_rate": 9.982585475332952e-06,
+      "loss": 0.8405,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05856128570642303,
+      "grad_norm": 0.928044319152832,
+      "learning_rate": 9.98254931038115e-06,
+      "loss": 0.8259,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05861632450877869,
+      "grad_norm": 0.8363838195800781,
+      "learning_rate": 9.982513107981837e-06,
+      "loss": 0.8655,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05867136331113435,
+      "grad_norm": 0.9800984859466553,
+      "learning_rate": 9.982476868135292e-06,
+      "loss": 0.9285,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05872640211349001,
+      "grad_norm": 0.8062636256217957,
+      "learning_rate": 9.982440590841785e-06,
+      "loss": 0.754,
+      "step": 1067
+    },
+    {
+      "epoch": 0.058781440915845674,
+      "grad_norm": 1.2010705471038818,
+      "learning_rate": 9.982404276101586e-06,
+      "loss": 0.9872,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05883647971820133,
+      "grad_norm": 1.0036406517028809,
+      "learning_rate": 9.982367923914971e-06,
+      "loss": 0.8724,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05889151852055699,
+      "grad_norm": 0.8768866658210754,
+      "learning_rate": 9.982331534282212e-06,
+      "loss": 0.838,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05894655732291265,
+      "grad_norm": 0.7892739176750183,
+      "learning_rate": 9.982295107203584e-06,
+      "loss": 0.6974,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05900159612526831,
+      "grad_norm": 0.863315999507904,
+      "learning_rate": 9.982258642679358e-06,
+      "loss": 0.9282,
+      "step": 1072
+    },
+    {
+      "epoch": 0.059056634927623974,
+      "grad_norm": 0.8645132780075073,
+      "learning_rate": 9.982222140709812e-06,
+      "loss": 0.8504,
+      "step": 1073
+    },
+    {
+      "epoch": 0.059111673729979636,
+      "grad_norm": 1.0003199577331543,
+      "learning_rate": 9.982185601295216e-06,
+      "loss": 1.0293,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0591667125323353,
+      "grad_norm": 0.8391831517219543,
+      "learning_rate": 9.982149024435848e-06,
+      "loss": 0.8609,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05922175133469096,
+      "grad_norm": 0.9940230846405029,
+      "learning_rate": 9.982112410131981e-06,
+      "loss": 0.9623,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05927679013704662,
+      "grad_norm": 1.0670262575149536,
+      "learning_rate": 9.98207575838389e-06,
+      "loss": 0.9952,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05933182893940228,
+      "grad_norm": 0.8506165742874146,
+      "learning_rate": 9.982039069191853e-06,
+      "loss": 0.8401,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05938686774175794,
+      "grad_norm": 0.8956409096717834,
+      "learning_rate": 9.982002342556144e-06,
+      "loss": 0.8779,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0594419065441136,
+      "grad_norm": 0.8955749273300171,
+      "learning_rate": 9.981965578477038e-06,
+      "loss": 0.8946,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05949694534646926,
+      "grad_norm": 0.9035234451293945,
+      "learning_rate": 9.981928776954811e-06,
+      "loss": 0.9352,
+      "step": 1081
+    },
+    {
+      "epoch": 0.05955198414882492,
+      "grad_norm": 0.8748759627342224,
+      "learning_rate": 9.981891937989743e-06,
+      "loss": 0.8803,
+      "step": 1082
+    },
+    {
+      "epoch": 0.05960702295118058,
+      "grad_norm": 0.9966281056404114,
+      "learning_rate": 9.981855061582108e-06,
+      "loss": 0.9304,
+      "step": 1083
+    },
+    {
+      "epoch": 0.05966206175353624,
+      "grad_norm": 0.8696668148040771,
+      "learning_rate": 9.981818147732183e-06,
+      "loss": 0.8706,
+      "step": 1084
+    },
+    {
+      "epoch": 0.059717100555891904,
+      "grad_norm": 0.9823188185691833,
+      "learning_rate": 9.981781196440249e-06,
+      "loss": 0.9431,
+      "step": 1085
+    },
+    {
+      "epoch": 0.059772139358247565,
+      "grad_norm": 0.8401583433151245,
+      "learning_rate": 9.981744207706577e-06,
+      "loss": 0.8369,
+      "step": 1086
+    },
+    {
+      "epoch": 0.05982717816060323,
+      "grad_norm": 0.8775757551193237,
+      "learning_rate": 9.981707181531452e-06,
+      "loss": 0.9516,
+      "step": 1087
+    },
+    {
+      "epoch": 0.05988221696295889,
+      "grad_norm": 0.9153465628623962,
+      "learning_rate": 9.981670117915148e-06,
+      "loss": 0.8997,
+      "step": 1088
+    },
+    {
+      "epoch": 0.05993725576531455,
+      "grad_norm": 0.9053078889846802,
+      "learning_rate": 9.981633016857946e-06,
+      "loss": 0.9452,
+      "step": 1089
+    },
+    {
+      "epoch": 0.059992294567670204,
+      "grad_norm": 0.9154480695724487,
+      "learning_rate": 9.981595878360123e-06,
+      "loss": 0.8293,
+      "step": 1090
+    },
+    {
+      "epoch": 0.060047333370025865,
+      "grad_norm": 0.85718834400177,
+      "learning_rate": 9.981558702421958e-06,
+      "loss": 0.876,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06010237217238153,
+      "grad_norm": 0.9437130689620972,
+      "learning_rate": 9.981521489043734e-06,
+      "loss": 0.9731,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06015741097473719,
+      "grad_norm": 0.9014891386032104,
+      "learning_rate": 9.981484238225725e-06,
+      "loss": 0.811,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06021244977709285,
+      "grad_norm": 0.8942846655845642,
+      "learning_rate": 9.981446949968216e-06,
+      "loss": 0.808,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06026748857944851,
+      "grad_norm": 0.855297863483429,
+      "learning_rate": 9.981409624271483e-06,
+      "loss": 0.8319,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06032252738180417,
+      "grad_norm": 0.9310913681983948,
+      "learning_rate": 9.981372261135811e-06,
+      "loss": 0.899,
+      "step": 1096
+    },
+    {
+      "epoch": 0.060377566184159834,
+      "grad_norm": 0.8472979664802551,
+      "learning_rate": 9.981334860561478e-06,
+      "loss": 0.8818,
+      "step": 1097
+    },
+    {
+      "epoch": 0.060432604986515495,
+      "grad_norm": 0.896617591381073,
+      "learning_rate": 9.981297422548764e-06,
+      "loss": 0.8991,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06048764378887116,
+      "grad_norm": 0.8543037176132202,
+      "learning_rate": 9.981259947097954e-06,
+      "loss": 0.8595,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06054268259122682,
+      "grad_norm": 0.8794904947280884,
+      "learning_rate": 9.981222434209327e-06,
+      "loss": 0.8561,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06059772139358247,
+      "grad_norm": 0.8882116675376892,
+      "learning_rate": 9.981184883883165e-06,
+      "loss": 0.8099,
+      "step": 1101
+    },
+    {
+      "epoch": 0.060652760195938134,
+      "grad_norm": 1.0068262815475464,
+      "learning_rate": 9.98114729611975e-06,
+      "loss": 0.8104,
+      "step": 1102
+    },
+    {
+      "epoch": 0.060707798998293795,
+      "grad_norm": 1.072316288948059,
+      "learning_rate": 9.981109670919366e-06,
+      "loss": 0.9877,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06076283780064946,
+      "grad_norm": 0.9959045052528381,
+      "learning_rate": 9.981072008282298e-06,
+      "loss": 0.906,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06081787660300512,
+      "grad_norm": 0.8712790608406067,
+      "learning_rate": 9.981034308208823e-06,
+      "loss": 0.8725,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06087291540536078,
+      "grad_norm": 0.9114679098129272,
+      "learning_rate": 9.980996570699228e-06,
+      "loss": 0.8385,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06092795420771644,
+      "grad_norm": 1.0024466514587402,
+      "learning_rate": 9.980958795753796e-06,
+      "loss": 0.8661,
+      "step": 1107
+    },
+    {
+      "epoch": 0.0609829930100721,
+      "grad_norm": 0.9578461050987244,
+      "learning_rate": 9.98092098337281e-06,
+      "loss": 0.9358,
+      "step": 1108
+    },
+    {
+      "epoch": 0.061038031812427763,
+      "grad_norm": 0.8677787780761719,
+      "learning_rate": 9.980883133556557e-06,
+      "loss": 0.8146,
+      "step": 1109
+    },
+    {
+      "epoch": 0.061093070614783425,
+      "grad_norm": 0.9072276949882507,
+      "learning_rate": 9.98084524630532e-06,
+      "loss": 0.91,
+      "step": 1110
+    },
+    {
+      "epoch": 0.061148109417139086,
+      "grad_norm": 0.8827292919158936,
+      "learning_rate": 9.980807321619381e-06,
+      "loss": 0.8854,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06120314821949474,
+      "grad_norm": 1.0012744665145874,
+      "learning_rate": 9.98076935949903e-06,
+      "loss": 0.8242,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0612581870218504,
+      "grad_norm": 0.9152620434761047,
+      "learning_rate": 9.980731359944548e-06,
+      "loss": 0.8832,
+      "step": 1113
+    },
+    {
+      "epoch": 0.061313225824206063,
+      "grad_norm": 0.8986824750900269,
+      "learning_rate": 9.980693322956222e-06,
+      "loss": 0.7975,
+      "step": 1114
+    },
+    {
+      "epoch": 0.061368264626561725,
+      "grad_norm": 0.9373019933700562,
+      "learning_rate": 9.98065524853434e-06,
+      "loss": 0.9541,
+      "step": 1115
+    },
+    {
+      "epoch": 0.061423303428917386,
+      "grad_norm": 0.9875593781471252,
+      "learning_rate": 9.980617136679185e-06,
+      "loss": 1.0052,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06147834223127305,
+      "grad_norm": 1.0664819478988647,
+      "learning_rate": 9.980578987391045e-06,
+      "loss": 0.8584,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06153338103362871,
+      "grad_norm": 0.9149377942085266,
+      "learning_rate": 9.980540800670207e-06,
+      "loss": 0.8467,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06158841983598437,
+      "grad_norm": 0.9303194284439087,
+      "learning_rate": 9.980502576516959e-06,
+      "loss": 0.8219,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06164345863834003,
+      "grad_norm": 0.9059457778930664,
+      "learning_rate": 9.980464314931583e-06,
+      "loss": 0.8459,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06169849744069569,
+      "grad_norm": 0.9368849396705627,
+      "learning_rate": 9.980426015914375e-06,
+      "loss": 0.8933,
+      "step": 1121
+    },
+    {
+      "epoch": 0.061753536243051355,
+      "grad_norm": 0.8188626766204834,
+      "learning_rate": 9.980387679465615e-06,
+      "loss": 0.807,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06180857504540701,
+      "grad_norm": 1.027171015739441,
+      "learning_rate": 9.980349305585595e-06,
+      "loss": 0.8919,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06186361384776267,
+      "grad_norm": 0.831649899482727,
+      "learning_rate": 9.980310894274603e-06,
+      "loss": 0.8109,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06191865265011833,
+      "grad_norm": 1.0170252323150635,
+      "learning_rate": 9.980272445532928e-06,
+      "loss": 0.9537,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06197369145247399,
+      "grad_norm": 0.97837233543396,
+      "learning_rate": 9.980233959360858e-06,
+      "loss": 0.9104,
+      "step": 1126
+    },
+    {
+      "epoch": 0.062028730254829655,
+      "grad_norm": 0.9548324942588806,
+      "learning_rate": 9.980195435758681e-06,
+      "loss": 0.9473,
+      "step": 1127
+    },
+    {
+      "epoch": 0.062083769057185316,
+      "grad_norm": 0.8675842881202698,
+      "learning_rate": 9.980156874726692e-06,
+      "loss": 0.8313,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06213880785954098,
+      "grad_norm": 0.8948968052864075,
+      "learning_rate": 9.980118276265173e-06,
+      "loss": 0.8008,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06219384666189664,
+      "grad_norm": 0.8914239406585693,
+      "learning_rate": 9.98007964037442e-06,
+      "loss": 0.7642,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0622488854642523,
+      "grad_norm": 0.9499951004981995,
+      "learning_rate": 9.980040967054723e-06,
+      "loss": 0.8669,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06230392426660796,
+      "grad_norm": 0.8959251642227173,
+      "learning_rate": 9.980002256306369e-06,
+      "loss": 0.9177,
+      "step": 1132
+    },
+    {
+      "epoch": 0.062358963068963616,
+      "grad_norm": 0.8634380102157593,
+      "learning_rate": 9.97996350812965e-06,
+      "loss": 0.8252,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06241400187131928,
+      "grad_norm": 0.9380598068237305,
+      "learning_rate": 9.97992472252486e-06,
+      "loss": 0.9335,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06246904067367494,
+      "grad_norm": 0.8373183608055115,
+      "learning_rate": 9.97988589949229e-06,
+      "loss": 0.848,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0625240794760306,
+      "grad_norm": 0.9649023413658142,
+      "learning_rate": 9.97984703903223e-06,
+      "loss": 0.9648,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06257911827838626,
+      "grad_norm": 0.9972373843193054,
+      "learning_rate": 9.979808141144972e-06,
+      "loss": 0.9104,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06263415708074192,
+      "grad_norm": 0.8230985403060913,
+      "learning_rate": 9.97976920583081e-06,
+      "loss": 0.8393,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06268919588309758,
+      "grad_norm": 0.9775324463844299,
+      "learning_rate": 9.979730233090034e-06,
+      "loss": 0.8385,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06274423468545325,
+      "grad_norm": 0.8288110494613647,
+      "learning_rate": 9.97969122292294e-06,
+      "loss": 0.7308,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06279927348780891,
+      "grad_norm": 0.8980758786201477,
+      "learning_rate": 9.979652175329819e-06,
+      "loss": 0.863,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06285431229016457,
+      "grad_norm": 7.43889045715332,
+      "learning_rate": 9.979613090310965e-06,
+      "loss": 0.9412,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06290935109252023,
+      "grad_norm": 0.9758191704750061,
+      "learning_rate": 9.97957396786667e-06,
+      "loss": 0.8896,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06296438989487589,
+      "grad_norm": 0.8211693167686462,
+      "learning_rate": 9.979534807997234e-06,
+      "loss": 0.7352,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06301942869723155,
+      "grad_norm": 0.8643441796302795,
+      "learning_rate": 9.979495610702945e-06,
+      "loss": 0.8701,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06307446749958721,
+      "grad_norm": 1.0199437141418457,
+      "learning_rate": 9.9794563759841e-06,
+      "loss": 0.9025,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06312950630194288,
+      "grad_norm": 0.8367893695831299,
+      "learning_rate": 9.979417103840994e-06,
+      "loss": 0.8491,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06318454510429854,
+      "grad_norm": 0.9411819577217102,
+      "learning_rate": 9.979377794273923e-06,
+      "loss": 0.8501,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06323958390665418,
+      "grad_norm": 1.1497365236282349,
+      "learning_rate": 9.97933844728318e-06,
+      "loss": 1.0227,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06329462270900985,
+      "grad_norm": 0.9892984628677368,
+      "learning_rate": 9.979299062869064e-06,
+      "loss": 0.8942,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06334966151136551,
+      "grad_norm": 0.947952926158905,
+      "learning_rate": 9.979259641031867e-06,
+      "loss": 1.0149,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06340470031372117,
+      "grad_norm": 0.9060251712799072,
+      "learning_rate": 9.979220181771889e-06,
+      "loss": 0.8607,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06345973911607683,
+      "grad_norm": 0.8331984281539917,
+      "learning_rate": 9.979180685089424e-06,
+      "loss": 0.8777,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06351477791843249,
+      "grad_norm": 0.9133188724517822,
+      "learning_rate": 9.97914115098477e-06,
+      "loss": 0.7409,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06356981672078815,
+      "grad_norm": 0.9095513820648193,
+      "learning_rate": 9.979101579458224e-06,
+      "loss": 0.8938,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06362485552314381,
+      "grad_norm": 0.9584553241729736,
+      "learning_rate": 9.979061970510082e-06,
+      "loss": 0.8765,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06367989432549948,
+      "grad_norm": 0.8742124438285828,
+      "learning_rate": 9.979022324140644e-06,
+      "loss": 0.8564,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06373493312785514,
+      "grad_norm": 0.8776904344558716,
+      "learning_rate": 9.978982640350208e-06,
+      "loss": 0.8713,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0637899719302108,
+      "grad_norm": 0.8667464852333069,
+      "learning_rate": 9.97894291913907e-06,
+      "loss": 0.8705,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06384501073256646,
+      "grad_norm": 0.9028087854385376,
+      "learning_rate": 9.978903160507531e-06,
+      "loss": 0.8297,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06390004953492212,
+      "grad_norm": 0.900812029838562,
+      "learning_rate": 9.978863364455887e-06,
+      "loss": 0.8456,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06395508833727778,
+      "grad_norm": 0.9667207598686218,
+      "learning_rate": 9.97882353098444e-06,
+      "loss": 0.8081,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06401012713963344,
+      "grad_norm": 0.8959711194038391,
+      "learning_rate": 9.978783660093488e-06,
+      "loss": 0.8455,
+      "step": 1163
+    },
+    {
+      "epoch": 0.0640651659419891,
+      "grad_norm": 0.8519117832183838,
+      "learning_rate": 9.97874375178333e-06,
+      "loss": 0.849,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06412020474434477,
+      "grad_norm": 1.0532654523849487,
+      "learning_rate": 9.978703806054267e-06,
+      "loss": 0.7356,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06417524354670043,
+      "grad_norm": 1.0313252210617065,
+      "learning_rate": 9.9786638229066e-06,
+      "loss": 1.024,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06423028234905609,
+      "grad_norm": 1.0567537546157837,
+      "learning_rate": 9.978623802340627e-06,
+      "loss": 0.9423,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06428532115141175,
+      "grad_norm": 0.8198097348213196,
+      "learning_rate": 9.97858374435665e-06,
+      "loss": 0.829,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06434035995376741,
+      "grad_norm": 0.8718193173408508,
+      "learning_rate": 9.97854364895497e-06,
+      "loss": 0.7184,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06439539875612307,
+      "grad_norm": 0.8037594556808472,
+      "learning_rate": 9.978503516135892e-06,
+      "loss": 0.7961,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06445043755847872,
+      "grad_norm": 0.9052229523658752,
+      "learning_rate": 9.978463345899709e-06,
+      "loss": 0.8016,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06450547636083438,
+      "grad_norm": 1.0194638967514038,
+      "learning_rate": 9.978423138246731e-06,
+      "loss": 0.9045,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06456051516319004,
+      "grad_norm": 0.953078031539917,
+      "learning_rate": 9.978382893177259e-06,
+      "loss": 0.9661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.0646155539655457,
+      "grad_norm": 0.8777341842651367,
+      "learning_rate": 9.978342610691592e-06,
+      "loss": 0.8685,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06467059276790137,
+      "grad_norm": 1.0178394317626953,
+      "learning_rate": 9.978302290790034e-06,
+      "loss": 0.9075,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06472563157025703,
+      "grad_norm": 0.935694694519043,
+      "learning_rate": 9.978261933472889e-06,
+      "loss": 0.8438,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06478067037261269,
+      "grad_norm": 1.0022411346435547,
+      "learning_rate": 9.97822153874046e-06,
+      "loss": 0.8701,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06483570917496835,
+      "grad_norm": 1.0371203422546387,
+      "learning_rate": 9.97818110659305e-06,
+      "loss": 0.9111,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06489074797732401,
+      "grad_norm": 0.7972478866577148,
+      "learning_rate": 9.978140637030963e-06,
+      "loss": 0.8602,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06494578677967967,
+      "grad_norm": 0.8556679487228394,
+      "learning_rate": 9.978100130054505e-06,
+      "loss": 0.9149,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06500082558203533,
+      "grad_norm": 0.92474365234375,
+      "learning_rate": 9.978059585663979e-06,
+      "loss": 0.8608,
+      "step": 1181
+    },
+    {
+      "epoch": 0.065055864384391,
+      "grad_norm": 1.0170830488204956,
+      "learning_rate": 9.978019003859687e-06,
+      "loss": 0.9986,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06511090318674666,
+      "grad_norm": 0.9405049681663513,
+      "learning_rate": 9.97797838464194e-06,
+      "loss": 0.9023,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06516594198910232,
+      "grad_norm": 0.9351203441619873,
+      "learning_rate": 9.977937728011038e-06,
+      "loss": 0.8698,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06522098079145798,
+      "grad_norm": 0.8620241284370422,
+      "learning_rate": 9.97789703396729e-06,
+      "loss": 0.9393,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06527601959381364,
+      "grad_norm": 0.9440441131591797,
+      "learning_rate": 9.977856302511e-06,
+      "loss": 0.8249,
+      "step": 1186
+    },
+    {
+      "epoch": 0.0653310583961693,
+      "grad_norm": 0.8311079144477844,
+      "learning_rate": 9.977815533642474e-06,
+      "loss": 0.8614,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06538609719852496,
+      "grad_norm": 0.8911672830581665,
+      "learning_rate": 9.977774727362018e-06,
+      "loss": 0.7909,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06544113600088063,
+      "grad_norm": 0.9237088561058044,
+      "learning_rate": 9.97773388366994e-06,
+      "loss": 0.7116,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06549617480323629,
+      "grad_norm": 1.1155747175216675,
+      "learning_rate": 9.977693002566549e-06,
+      "loss": 0.9248,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06555121360559195,
+      "grad_norm": 0.9386736750602722,
+      "learning_rate": 9.977652084052148e-06,
+      "loss": 0.8307,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0656062524079476,
+      "grad_norm": 1.1666040420532227,
+      "learning_rate": 9.977611128127044e-06,
+      "loss": 0.9723,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06566129121030326,
+      "grad_norm": 1.2366368770599365,
+      "learning_rate": 9.977570134791552e-06,
+      "loss": 0.8253,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06571633001265892,
+      "grad_norm": 0.823443591594696,
+      "learning_rate": 9.977529104045971e-06,
+      "loss": 0.7472,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06577136881501458,
+      "grad_norm": 0.9481683969497681,
+      "learning_rate": 9.977488035890617e-06,
+      "loss": 0.887,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06582640761737024,
+      "grad_norm": 0.9120422005653381,
+      "learning_rate": 9.977446930325794e-06,
+      "loss": 0.867,
+      "step": 1196
+    },
+    {
+      "epoch": 0.0658814464197259,
+      "grad_norm": 0.8595587015151978,
+      "learning_rate": 9.977405787351811e-06,
+      "loss": 0.8532,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06593648522208156,
+      "grad_norm": 0.8590419888496399,
+      "learning_rate": 9.97736460696898e-06,
+      "loss": 0.8998,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06599152402443723,
+      "grad_norm": 0.9670939445495605,
+      "learning_rate": 9.977323389177609e-06,
+      "loss": 0.8964,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06604656282679289,
+      "grad_norm": 0.8870261907577515,
+      "learning_rate": 9.977282133978006e-06,
+      "loss": 0.9542,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06610160162914855,
+      "grad_norm": 0.942294180393219,
+      "learning_rate": 9.977240841370484e-06,
+      "loss": 0.8681,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06615664043150421,
+      "grad_norm": 0.9632517099380493,
+      "learning_rate": 9.977199511355353e-06,
+      "loss": 0.7327,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06621167923385987,
+      "grad_norm": 4.8085479736328125,
+      "learning_rate": 9.97715814393292e-06,
+      "loss": 0.8528,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06626671803621553,
+      "grad_norm": 0.9084093570709229,
+      "learning_rate": 9.977116739103503e-06,
+      "loss": 0.7836,
+      "step": 1204
+    },
+    {
+      "epoch": 0.0663217568385712,
+      "grad_norm": 0.8961902260780334,
+      "learning_rate": 9.977075296867406e-06,
+      "loss": 0.854,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06637679564092686,
+      "grad_norm": 0.8727987408638,
+      "learning_rate": 9.977033817224945e-06,
+      "loss": 0.7931,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06643183444328252,
+      "grad_norm": 0.8263267874717712,
+      "learning_rate": 9.976992300176428e-06,
+      "loss": 0.852,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06648687324563818,
+      "grad_norm": 1.0499639511108398,
+      "learning_rate": 9.97695074572217e-06,
+      "loss": 1.0427,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06654191204799384,
+      "grad_norm": 0.9337313771247864,
+      "learning_rate": 9.976909153862482e-06,
+      "loss": 0.8035,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0665969508503495,
+      "grad_norm": 0.8795992732048035,
+      "learning_rate": 9.976867524597678e-06,
+      "loss": 0.9022,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06665198965270516,
+      "grad_norm": 0.9787294268608093,
+      "learning_rate": 9.976825857928069e-06,
+      "loss": 0.8259,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06670702845506082,
+      "grad_norm": 0.8570082187652588,
+      "learning_rate": 9.976784153853969e-06,
+      "loss": 0.8567,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06676206725741649,
+      "grad_norm": 1.0620380640029907,
+      "learning_rate": 9.976742412375694e-06,
+      "loss": 0.851,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06681710605977213,
+      "grad_norm": 0.8545439839363098,
+      "learning_rate": 9.976700633493551e-06,
+      "loss": 0.8827,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0668721448621278,
+      "grad_norm": 0.8543682098388672,
+      "learning_rate": 9.97665881720786e-06,
+      "loss": 0.8524,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06692718366448346,
+      "grad_norm": 0.7748527526855469,
+      "learning_rate": 9.976616963518935e-06,
+      "loss": 0.7459,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06698222246683912,
+      "grad_norm": 0.9876659512519836,
+      "learning_rate": 9.976575072427087e-06,
+      "loss": 0.8426,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06703726126919478,
+      "grad_norm": 0.8763901591300964,
+      "learning_rate": 9.976533143932635e-06,
+      "loss": 0.8561,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06709230007155044,
+      "grad_norm": 0.7816654443740845,
+      "learning_rate": 9.97649117803589e-06,
+      "loss": 0.8361,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0671473388739061,
+      "grad_norm": 0.8659802675247192,
+      "learning_rate": 9.97644917473717e-06,
+      "loss": 0.897,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06720237767626176,
+      "grad_norm": 0.9180877208709717,
+      "learning_rate": 9.97640713403679e-06,
+      "loss": 0.9516,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06725741647861742,
+      "grad_norm": 0.9624410271644592,
+      "learning_rate": 9.976365055935067e-06,
+      "loss": 0.9119,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06731245528097309,
+      "grad_norm": 0.8291105031967163,
+      "learning_rate": 9.976322940432314e-06,
+      "loss": 0.788,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06736749408332875,
+      "grad_norm": 0.9858983755111694,
+      "learning_rate": 9.976280787528854e-06,
+      "loss": 0.8794,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06742253288568441,
+      "grad_norm": 0.8283948302268982,
+      "learning_rate": 9.976238597224996e-06,
+      "loss": 0.8571,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06747757168804007,
+      "grad_norm": 0.8585363626480103,
+      "learning_rate": 9.976196369521063e-06,
+      "loss": 0.9005,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06753261049039573,
+      "grad_norm": 0.847882091999054,
+      "learning_rate": 9.976154104417369e-06,
+      "loss": 0.8058,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06758764929275139,
+      "grad_norm": 0.9045611023902893,
+      "learning_rate": 9.976111801914232e-06,
+      "loss": 0.7864,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06764268809510705,
+      "grad_norm": 0.805932879447937,
+      "learning_rate": 9.976069462011972e-06,
+      "loss": 0.8436,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06769772689746271,
+      "grad_norm": 0.8809003233909607,
+      "learning_rate": 9.976027084710906e-06,
+      "loss": 0.7876,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06775276569981838,
+      "grad_norm": 0.8681740760803223,
+      "learning_rate": 9.975984670011352e-06,
+      "loss": 0.877,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06780780450217404,
+      "grad_norm": 0.9909854531288147,
+      "learning_rate": 9.975942217913627e-06,
+      "loss": 0.8957,
+      "step": 1232
+    },
+    {
+      "epoch": 0.0678628433045297,
+      "grad_norm": 0.9213934540748596,
+      "learning_rate": 9.975899728418056e-06,
+      "loss": 0.8344,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06791788210688536,
+      "grad_norm": 0.8289967179298401,
+      "learning_rate": 9.975857201524952e-06,
+      "loss": 0.876,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06797292090924101,
+      "grad_norm": 0.891812264919281,
+      "learning_rate": 9.97581463723464e-06,
+      "loss": 0.8611,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06802795971159667,
+      "grad_norm": 1.0301382541656494,
+      "learning_rate": 9.975772035547435e-06,
+      "loss": 0.8177,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06808299851395233,
+      "grad_norm": 0.8380662798881531,
+      "learning_rate": 9.975729396463659e-06,
+      "loss": 0.8631,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06813803731630799,
+      "grad_norm": 0.9226046204566956,
+      "learning_rate": 9.975686719983633e-06,
+      "loss": 0.8927,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06819307611866365,
+      "grad_norm": 0.8917136192321777,
+      "learning_rate": 9.975644006107679e-06,
+      "loss": 0.9048,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06824811492101931,
+      "grad_norm": 0.8559191226959229,
+      "learning_rate": 9.975601254836114e-06,
+      "loss": 0.8169,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06830315372337498,
+      "grad_norm": 0.9345341920852661,
+      "learning_rate": 9.975558466169263e-06,
+      "loss": 0.7929,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06835819252573064,
+      "grad_norm": 0.9155850410461426,
+      "learning_rate": 9.975515640107447e-06,
+      "loss": 0.8825,
+      "step": 1242
+    },
+    {
+      "epoch": 0.0684132313280863,
+      "grad_norm": 0.899712860584259,
+      "learning_rate": 9.975472776650987e-06,
+      "loss": 0.825,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06846827013044196,
+      "grad_norm": 0.8280880451202393,
+      "learning_rate": 9.975429875800206e-06,
+      "loss": 0.8539,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06852330893279762,
+      "grad_norm": 0.9589636325836182,
+      "learning_rate": 9.975386937555426e-06,
+      "loss": 0.9465,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06857834773515328,
+      "grad_norm": 1.1027253866195679,
+      "learning_rate": 9.97534396191697e-06,
+      "loss": 0.87,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06863338653750894,
+      "grad_norm": 1.0510318279266357,
+      "learning_rate": 9.975300948885158e-06,
+      "loss": 0.8569,
+      "step": 1247
+    },
+    {
+      "epoch": 0.0686884253398646,
+      "grad_norm": 0.8897958397865295,
+      "learning_rate": 9.975257898460317e-06,
+      "loss": 0.8431,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06874346414222027,
+      "grad_norm": 0.8827036619186401,
+      "learning_rate": 9.975214810642771e-06,
+      "loss": 0.922,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06879850294457593,
+      "grad_norm": 0.8798324465751648,
+      "learning_rate": 9.97517168543284e-06,
+      "loss": 0.7837,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06885354174693159,
+      "grad_norm": 0.9053803086280823,
+      "learning_rate": 9.975128522830853e-06,
+      "loss": 0.82,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06890858054928725,
+      "grad_norm": 0.8362607359886169,
+      "learning_rate": 9.975085322837129e-06,
+      "loss": 0.7684,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06896361935164291,
+      "grad_norm": 0.8898602724075317,
+      "learning_rate": 9.975042085451997e-06,
+      "loss": 0.8205,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06901865815399857,
+      "grad_norm": 0.9210274815559387,
+      "learning_rate": 9.97499881067578e-06,
+      "loss": 0.8364,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06907369695635424,
+      "grad_norm": 1.0881952047348022,
+      "learning_rate": 9.974955498508804e-06,
+      "loss": 0.8234,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0691287357587099,
+      "grad_norm": 0.8875024914741516,
+      "learning_rate": 9.974912148951394e-06,
+      "loss": 0.7974,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06918377456106554,
+      "grad_norm": 0.9065666794776917,
+      "learning_rate": 9.974868762003876e-06,
+      "loss": 0.7721,
+      "step": 1257
+    },
+    {
+      "epoch": 0.0692388133634212,
+      "grad_norm": 0.8904553651809692,
+      "learning_rate": 9.974825337666576e-06,
+      "loss": 0.8551,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06929385216577687,
+      "grad_norm": 0.8586102724075317,
+      "learning_rate": 9.974781875939821e-06,
+      "loss": 0.8666,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06934889096813253,
+      "grad_norm": 0.9103402495384216,
+      "learning_rate": 9.974738376823935e-06,
+      "loss": 0.8361,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06940392977048819,
+      "grad_norm": 0.8657701015472412,
+      "learning_rate": 9.974694840319249e-06,
+      "loss": 0.8217,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06945896857284385,
+      "grad_norm": 0.865703821182251,
+      "learning_rate": 9.974651266426088e-06,
+      "loss": 0.8751,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06951400737519951,
+      "grad_norm": 0.8932577967643738,
+      "learning_rate": 9.974607655144779e-06,
+      "loss": 0.8709,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06956904617755517,
+      "grad_norm": 0.8417405486106873,
+      "learning_rate": 9.97456400647565e-06,
+      "loss": 0.8104,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06962408497991084,
+      "grad_norm": 0.8578035235404968,
+      "learning_rate": 9.974520320419032e-06,
+      "loss": 0.9173,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0696791237822665,
+      "grad_norm": 0.957539439201355,
+      "learning_rate": 9.974476596975249e-06,
+      "loss": 0.8955,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06973416258462216,
+      "grad_norm": 0.851222038269043,
+      "learning_rate": 9.974432836144632e-06,
+      "loss": 0.8696,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06978920138697782,
+      "grad_norm": 0.8178789615631104,
+      "learning_rate": 9.974389037927508e-06,
+      "loss": 0.7921,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06984424018933348,
+      "grad_norm": 0.954091489315033,
+      "learning_rate": 9.97434520232421e-06,
+      "loss": 0.9362,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06989927899168914,
+      "grad_norm": 0.8525053858757019,
+      "learning_rate": 9.974301329335063e-06,
+      "loss": 0.7996,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0699543177940448,
+      "grad_norm": 0.9340476393699646,
+      "learning_rate": 9.9742574189604e-06,
+      "loss": 0.9091,
+      "step": 1271
+    },
+    {
+      "epoch": 0.07000935659640047,
+      "grad_norm": 0.7946187257766724,
+      "learning_rate": 9.974213471200548e-06,
+      "loss": 0.874,
+      "step": 1272
+    },
+    {
+      "epoch": 0.07006439539875613,
+      "grad_norm": 0.8048381209373474,
+      "learning_rate": 9.97416948605584e-06,
+      "loss": 0.8557,
+      "step": 1273
+    },
+    {
+      "epoch": 0.07011943420111179,
+      "grad_norm": 0.9849064946174622,
+      "learning_rate": 9.974125463526607e-06,
+      "loss": 0.8154,
+      "step": 1274
+    },
+    {
+      "epoch": 0.07017447300346745,
+      "grad_norm": 0.9030239582061768,
+      "learning_rate": 9.974081403613178e-06,
+      "loss": 0.9411,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07022951180582311,
+      "grad_norm": 0.8869300484657288,
+      "learning_rate": 9.974037306315882e-06,
+      "loss": 0.8978,
+      "step": 1276
+    },
+    {
+      "epoch": 0.07028455060817877,
+      "grad_norm": 0.8558536767959595,
+      "learning_rate": 9.973993171635057e-06,
+      "loss": 0.8937,
+      "step": 1277
+    },
+    {
+      "epoch": 0.07033958941053442,
+      "grad_norm": 0.9005453586578369,
+      "learning_rate": 9.973948999571029e-06,
+      "loss": 0.9336,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07039462821289008,
+      "grad_norm": 0.8489978909492493,
+      "learning_rate": 9.973904790124131e-06,
+      "loss": 0.8267,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07044966701524574,
+      "grad_norm": 0.8295948505401611,
+      "learning_rate": 9.973860543294696e-06,
+      "loss": 0.8478,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0705047058176014,
+      "grad_norm": 0.8111379742622375,
+      "learning_rate": 9.973816259083058e-06,
+      "loss": 0.8333,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07055974461995707,
+      "grad_norm": 0.9380189776420593,
+      "learning_rate": 9.973771937489547e-06,
+      "loss": 0.9718,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07061478342231273,
+      "grad_norm": 1.251194953918457,
+      "learning_rate": 9.973727578514499e-06,
+      "loss": 0.9531,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07066982222466839,
+      "grad_norm": 0.9897224307060242,
+      "learning_rate": 9.973683182158243e-06,
+      "loss": 0.7853,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07072486102702405,
+      "grad_norm": 0.8409335017204285,
+      "learning_rate": 9.973638748421119e-06,
+      "loss": 0.7692,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07077989982937971,
+      "grad_norm": 0.9019681215286255,
+      "learning_rate": 9.973594277303456e-06,
+      "loss": 0.8135,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07083493863173537,
+      "grad_norm": 0.9236096739768982,
+      "learning_rate": 9.973549768805588e-06,
+      "loss": 0.9304,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07088997743409103,
+      "grad_norm": 0.9244743585586548,
+      "learning_rate": 9.973505222927854e-06,
+      "loss": 0.9056,
+      "step": 1288
+    },
+    {
+      "epoch": 0.0709450162364467,
+      "grad_norm": 1.3418753147125244,
+      "learning_rate": 9.973460639670585e-06,
+      "loss": 0.8419,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07100005503880236,
+      "grad_norm": 0.8715767860412598,
+      "learning_rate": 9.973416019034117e-06,
+      "loss": 0.9704,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07105509384115802,
+      "grad_norm": 0.9609012007713318,
+      "learning_rate": 9.973371361018787e-06,
+      "loss": 0.8807,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07111013264351368,
+      "grad_norm": 0.8085873126983643,
+      "learning_rate": 9.973326665624927e-06,
+      "loss": 0.7947,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07116517144586934,
+      "grad_norm": 0.919280469417572,
+      "learning_rate": 9.973281932852877e-06,
+      "loss": 0.9743,
+      "step": 1293
+    },
+    {
+      "epoch": 0.071220210248225,
+      "grad_norm": 1.0651074647903442,
+      "learning_rate": 9.973237162702968e-06,
+      "loss": 0.7164,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07127524905058066,
+      "grad_norm": 0.987251341342926,
+      "learning_rate": 9.973192355175542e-06,
+      "loss": 0.9286,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07133028785293632,
+      "grad_norm": 1.5507274866104126,
+      "learning_rate": 9.973147510270935e-06,
+      "loss": 0.9733,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07138532665529199,
+      "grad_norm": 0.8439416885375977,
+      "learning_rate": 9.97310262798948e-06,
+      "loss": 0.7462,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07144036545764765,
+      "grad_norm": 0.9604889750480652,
+      "learning_rate": 9.973057708331519e-06,
+      "loss": 1.0006,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07149540426000331,
+      "grad_norm": 0.8568960428237915,
+      "learning_rate": 9.973012751297386e-06,
+      "loss": 0.878,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07155044306235896,
+      "grad_norm": 0.8169522285461426,
+      "learning_rate": 9.972967756887419e-06,
+      "loss": 0.8241,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07160548186471462,
+      "grad_norm": 0.875738799571991,
+      "learning_rate": 9.97292272510196e-06,
+      "loss": 0.854,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07166052066707028,
+      "grad_norm": 0.7877739071846008,
+      "learning_rate": 9.972877655941345e-06,
+      "loss": 0.779,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07171555946942594,
+      "grad_norm": 0.8148574829101562,
+      "learning_rate": 9.972832549405912e-06,
+      "loss": 0.6965,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0717705982717816,
+      "grad_norm": 0.936720609664917,
+      "learning_rate": 9.972787405495998e-06,
+      "loss": 0.798,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07182563707413726,
+      "grad_norm": 0.8932886123657227,
+      "learning_rate": 9.972742224211949e-06,
+      "loss": 0.9196,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07188067587649292,
+      "grad_norm": 0.899246871471405,
+      "learning_rate": 9.972697005554099e-06,
+      "loss": 0.8081,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07193571467884859,
+      "grad_norm": 0.8789899349212646,
+      "learning_rate": 9.972651749522788e-06,
+      "loss": 0.89,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07199075348120425,
+      "grad_norm": 1.2412173748016357,
+      "learning_rate": 9.97260645611836e-06,
+      "loss": 0.9866,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07204579228355991,
+      "grad_norm": 0.8655833005905151,
+      "learning_rate": 9.972561125341152e-06,
+      "loss": 0.8144,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07210083108591557,
+      "grad_norm": 0.8705299496650696,
+      "learning_rate": 9.972515757191506e-06,
+      "loss": 0.8431,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07215586988827123,
+      "grad_norm": 0.8813188672065735,
+      "learning_rate": 9.972470351669761e-06,
+      "loss": 0.859,
+      "step": 1311
+    },
+    {
+      "epoch": 0.0722109086906269,
+      "grad_norm": 2.043627977371216,
+      "learning_rate": 9.972424908776262e-06,
+      "loss": 0.9886,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07226594749298255,
+      "grad_norm": 0.9167500734329224,
+      "learning_rate": 9.972379428511348e-06,
+      "loss": 0.7203,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07232098629533822,
+      "grad_norm": 1.3145136833190918,
+      "learning_rate": 9.972333910875358e-06,
+      "loss": 0.9325,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07237602509769388,
+      "grad_norm": 0.834710419178009,
+      "learning_rate": 9.972288355868641e-06,
+      "loss": 0.9361,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07243106390004954,
+      "grad_norm": 0.9039230942726135,
+      "learning_rate": 9.972242763491535e-06,
+      "loss": 0.8027,
+      "step": 1316
+    },
+    {
+      "epoch": 0.0724861027024052,
+      "grad_norm": 0.8911495208740234,
+      "learning_rate": 9.972197133744384e-06,
+      "loss": 0.951,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07254114150476086,
+      "grad_norm": 1.0752439498901367,
+      "learning_rate": 9.972151466627529e-06,
+      "loss": 0.8421,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07259618030711652,
+      "grad_norm": 0.926135778427124,
+      "learning_rate": 9.972105762141314e-06,
+      "loss": 0.8901,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07265121910947218,
+      "grad_norm": 0.8166295289993286,
+      "learning_rate": 9.972060020286085e-06,
+      "loss": 0.7845,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07270625791182783,
+      "grad_norm": 1.0000934600830078,
+      "learning_rate": 9.972014241062182e-06,
+      "loss": 0.8383,
+      "step": 1321
+    },
+    {
+      "epoch": 0.0727612967141835,
+      "grad_norm": 1.2617899179458618,
+      "learning_rate": 9.971968424469951e-06,
+      "loss": 0.9826,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07281633551653915,
+      "grad_norm": 0.8451040983200073,
+      "learning_rate": 9.971922570509738e-06,
+      "loss": 0.8262,
+      "step": 1323
+    },
+    {
+      "epoch": 0.07287137431889482,
+      "grad_norm": 0.8101939558982849,
+      "learning_rate": 9.971876679181884e-06,
+      "loss": 0.6904,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07292641312125048,
+      "grad_norm": 0.8805514574050903,
+      "learning_rate": 9.971830750486736e-06,
+      "loss": 0.8491,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07298145192360614,
+      "grad_norm": 0.8236901164054871,
+      "learning_rate": 9.97178478442464e-06,
+      "loss": 0.8462,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0730364907259618,
+      "grad_norm": 0.9183042645454407,
+      "learning_rate": 9.971738780995938e-06,
+      "loss": 0.7577,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07309152952831746,
+      "grad_norm": 0.8425934314727783,
+      "learning_rate": 9.971692740200982e-06,
+      "loss": 0.8462,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07314656833067312,
+      "grad_norm": 0.9114993214607239,
+      "learning_rate": 9.971646662040112e-06,
+      "loss": 0.9132,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07320160713302878,
+      "grad_norm": 0.8516649603843689,
+      "learning_rate": 9.971600546513675e-06,
+      "loss": 0.8819,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07325664593538445,
+      "grad_norm": 1.0859558582305908,
+      "learning_rate": 9.971554393622023e-06,
+      "loss": 0.9929,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07331168473774011,
+      "grad_norm": 0.8906900882720947,
+      "learning_rate": 9.971508203365497e-06,
+      "loss": 0.9166,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07336672354009577,
+      "grad_norm": 0.8931803703308105,
+      "learning_rate": 9.971461975744445e-06,
+      "loss": 0.864,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07342176234245143,
+      "grad_norm": 0.8404982686042786,
+      "learning_rate": 9.971415710759216e-06,
+      "loss": 0.8609,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07347680114480709,
+      "grad_norm": 0.8016490340232849,
+      "learning_rate": 9.971369408410157e-06,
+      "loss": 0.7694,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07353183994716275,
+      "grad_norm": 0.7700600028038025,
+      "learning_rate": 9.971323068697618e-06,
+      "loss": 0.7875,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07358687874951841,
+      "grad_norm": 0.8679799437522888,
+      "learning_rate": 9.971276691621946e-06,
+      "loss": 0.8409,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07364191755187408,
+      "grad_norm": 0.8329173922538757,
+      "learning_rate": 9.971230277183486e-06,
+      "loss": 0.8707,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07369695635422974,
+      "grad_norm": 0.8790140151977539,
+      "learning_rate": 9.97118382538259e-06,
+      "loss": 0.7631,
+      "step": 1339
+    },
+    {
+      "epoch": 0.0737519951565854,
+      "grad_norm": 1.1895341873168945,
+      "learning_rate": 9.97113733621961e-06,
+      "loss": 0.8555,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07380703395894106,
+      "grad_norm": 0.8531593680381775,
+      "learning_rate": 9.97109080969489e-06,
+      "loss": 0.7192,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07386207276129672,
+      "grad_norm": 1.0388946533203125,
+      "learning_rate": 9.971044245808784e-06,
+      "loss": 0.8182,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07391711156365237,
+      "grad_norm": 0.8858556747436523,
+      "learning_rate": 9.970997644561639e-06,
+      "loss": 0.7981,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07397215036600803,
+      "grad_norm": 0.8710204362869263,
+      "learning_rate": 9.970951005953807e-06,
+      "loss": 0.7667,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07402718916836369,
+      "grad_norm": 0.9788708090782166,
+      "learning_rate": 9.970904329985638e-06,
+      "loss": 0.9693,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07408222797071935,
+      "grad_norm": 0.7805914878845215,
+      "learning_rate": 9.970857616657482e-06,
+      "loss": 0.6683,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07413726677307501,
+      "grad_norm": 0.9977933168411255,
+      "learning_rate": 9.97081086596969e-06,
+      "loss": 0.8288,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07419230557543068,
+      "grad_norm": 0.829115629196167,
+      "learning_rate": 9.970764077922617e-06,
+      "loss": 0.8361,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07424734437778634,
+      "grad_norm": 1.226120114326477,
+      "learning_rate": 9.97071725251661e-06,
+      "loss": 1.0008,
+      "step": 1349
+    },
+    {
+      "epoch": 0.074302383180142,
+      "grad_norm": 0.8997750878334045,
+      "learning_rate": 9.970670389752021e-06,
+      "loss": 0.8048,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07435742198249766,
+      "grad_norm": 1.0885238647460938,
+      "learning_rate": 9.970623489629205e-06,
+      "loss": 0.9202,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07441246078485332,
+      "grad_norm": 0.8736100792884827,
+      "learning_rate": 9.970576552148515e-06,
+      "loss": 0.8515,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07446749958720898,
+      "grad_norm": 0.9211294651031494,
+      "learning_rate": 9.970529577310301e-06,
+      "loss": 0.9389,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07452253838956464,
+      "grad_norm": 0.9334765672683716,
+      "learning_rate": 9.970482565114917e-06,
+      "loss": 0.8165,
+      "step": 1354
+    },
+    {
+      "epoch": 0.0745775771919203,
+      "grad_norm": 0.8307162523269653,
+      "learning_rate": 9.970435515562717e-06,
+      "loss": 0.7829,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07463261599427597,
+      "grad_norm": 0.987634003162384,
+      "learning_rate": 9.970388428654055e-06,
+      "loss": 0.848,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07468765479663163,
+      "grad_norm": 1.094752311706543,
+      "learning_rate": 9.970341304389281e-06,
+      "loss": 1.003,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07474269359898729,
+      "grad_norm": 0.9865909814834595,
+      "learning_rate": 9.970294142768755e-06,
+      "loss": 0.9116,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07479773240134295,
+      "grad_norm": 0.8404149413108826,
+      "learning_rate": 9.970246943792828e-06,
+      "loss": 0.8699,
+      "step": 1359
+    },
+    {
+      "epoch": 0.07485277120369861,
+      "grad_norm": 0.9602416753768921,
+      "learning_rate": 9.970199707461855e-06,
+      "loss": 0.8166,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07490781000605427,
+      "grad_norm": 0.9748693704605103,
+      "learning_rate": 9.970152433776193e-06,
+      "loss": 0.8767,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07496284880840993,
+      "grad_norm": 0.8721657991409302,
+      "learning_rate": 9.970105122736194e-06,
+      "loss": 0.8825,
+      "step": 1362
+    },
+    {
+      "epoch": 0.0750178876107656,
+      "grad_norm": 0.8683610558509827,
+      "learning_rate": 9.970057774342215e-06,
+      "loss": 0.7873,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07507292641312124,
+      "grad_norm": 0.856396496295929,
+      "learning_rate": 9.970010388594613e-06,
+      "loss": 0.8505,
+      "step": 1364
+    },
+    {
+      "epoch": 0.0751279652154769,
+      "grad_norm": 1.0709880590438843,
+      "learning_rate": 9.969962965493744e-06,
+      "loss": 0.9519,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07518300401783257,
+      "grad_norm": 0.8839450478553772,
+      "learning_rate": 9.969915505039963e-06,
+      "loss": 0.8041,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07523804282018823,
+      "grad_norm": 0.89545738697052,
+      "learning_rate": 9.969868007233627e-06,
+      "loss": 0.8713,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07529308162254389,
+      "grad_norm": 0.9870849251747131,
+      "learning_rate": 9.969820472075094e-06,
+      "loss": 0.8655,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07534812042489955,
+      "grad_norm": 1.3123797178268433,
+      "learning_rate": 9.96977289956472e-06,
+      "loss": 1.0425,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07540315922725521,
+      "grad_norm": 0.8538400530815125,
+      "learning_rate": 9.969725289702865e-06,
+      "loss": 0.7052,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07545819802961087,
+      "grad_norm": 0.933397114276886,
+      "learning_rate": 9.969677642489884e-06,
+      "loss": 0.9819,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07551323683196653,
+      "grad_norm": 0.8428112268447876,
+      "learning_rate": 9.969629957926134e-06,
+      "loss": 0.7313,
+      "step": 1372
+    },
+    {
+      "epoch": 0.0755682756343222,
+      "grad_norm": 0.9023239612579346,
+      "learning_rate": 9.96958223601198e-06,
+      "loss": 0.8297,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07562331443667786,
+      "grad_norm": 0.8971324563026428,
+      "learning_rate": 9.969534476747771e-06,
+      "loss": 0.8832,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07567835323903352,
+      "grad_norm": 0.8709388375282288,
+      "learning_rate": 9.969486680133874e-06,
+      "loss": 0.743,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07573339204138918,
+      "grad_norm": 0.9094591736793518,
+      "learning_rate": 9.969438846170644e-06,
+      "loss": 0.8294,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07578843084374484,
+      "grad_norm": 1.0753988027572632,
+      "learning_rate": 9.969390974858444e-06,
+      "loss": 0.7479,
+      "step": 1377
+    },
+    {
+      "epoch": 0.0758434696461005,
+      "grad_norm": 0.933775007724762,
+      "learning_rate": 9.96934306619763e-06,
+      "loss": 0.8235,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07589850844845616,
+      "grad_norm": 0.8419735431671143,
+      "learning_rate": 9.969295120188565e-06,
+      "loss": 0.8103,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07595354725081183,
+      "grad_norm": 0.8912790417671204,
+      "learning_rate": 9.969247136831606e-06,
+      "loss": 0.911,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07600858605316749,
+      "grad_norm": 0.8780983090400696,
+      "learning_rate": 9.969199116127118e-06,
+      "loss": 0.8619,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07606362485552315,
+      "grad_norm": 0.8503809571266174,
+      "learning_rate": 9.969151058075459e-06,
+      "loss": 0.8093,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07611866365787881,
+      "grad_norm": 0.8633087277412415,
+      "learning_rate": 9.96910296267699e-06,
+      "loss": 0.7524,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07617370246023447,
+      "grad_norm": 1.1203595399856567,
+      "learning_rate": 9.969054829932074e-06,
+      "loss": 0.945,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07622874126259013,
+      "grad_norm": 0.8766878843307495,
+      "learning_rate": 9.969006659841072e-06,
+      "loss": 0.7537,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07628378006494578,
+      "grad_norm": 0.9795958399772644,
+      "learning_rate": 9.968958452404345e-06,
+      "loss": 0.7963,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07633881886730144,
+      "grad_norm": 0.9117506146430969,
+      "learning_rate": 9.968910207622257e-06,
+      "loss": 0.9469,
+      "step": 1387
+    },
+    {
+      "epoch": 0.0763938576696571,
+      "grad_norm": 0.9731466770172119,
+      "learning_rate": 9.96886192549517e-06,
+      "loss": 0.9536,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07644889647201276,
+      "grad_norm": 0.8923571109771729,
+      "learning_rate": 9.968813606023446e-06,
+      "loss": 0.8362,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07650393527436843,
+      "grad_norm": 0.8819600343704224,
+      "learning_rate": 9.96876524920745e-06,
+      "loss": 0.6938,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07655897407672409,
+      "grad_norm": 0.9629887342453003,
+      "learning_rate": 9.968716855047545e-06,
+      "loss": 0.9104,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07661401287907975,
+      "grad_norm": 0.992770254611969,
+      "learning_rate": 9.968668423544093e-06,
+      "loss": 0.944,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07666905168143541,
+      "grad_norm": 0.8578491806983948,
+      "learning_rate": 9.96861995469746e-06,
+      "loss": 0.898,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07672409048379107,
+      "grad_norm": 1.1169229745864868,
+      "learning_rate": 9.968571448508008e-06,
+      "loss": 0.8324,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07677912928614673,
+      "grad_norm": 0.9600160121917725,
+      "learning_rate": 9.968522904976106e-06,
+      "loss": 0.9519,
+      "step": 1395
+    },
+    {
+      "epoch": 0.0768341680885024,
+      "grad_norm": 0.8271373510360718,
+      "learning_rate": 9.968474324102112e-06,
+      "loss": 0.8576,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07688920689085806,
+      "grad_norm": 0.9437325596809387,
+      "learning_rate": 9.968425705886397e-06,
+      "loss": 0.9201,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07694424569321372,
+      "grad_norm": 0.8679039478302002,
+      "learning_rate": 9.968377050329325e-06,
+      "loss": 0.8893,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07699928449556938,
+      "grad_norm": 1.0178717374801636,
+      "learning_rate": 9.96832835743126e-06,
+      "loss": 0.9718,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07705432329792504,
+      "grad_norm": 0.8354432582855225,
+      "learning_rate": 9.96827962719257e-06,
+      "loss": 0.83,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0771093621002807,
+      "grad_norm": 1.2244631052017212,
+      "learning_rate": 9.968230859613619e-06,
+      "loss": 0.907,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07716440090263636,
+      "grad_norm": 0.9099625945091248,
+      "learning_rate": 9.968182054694775e-06,
+      "loss": 0.809,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07721943970499202,
+      "grad_norm": 0.8591424226760864,
+      "learning_rate": 9.968133212436404e-06,
+      "loss": 0.8869,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07727447850734769,
+      "grad_norm": 1.068003535270691,
+      "learning_rate": 9.968084332838876e-06,
+      "loss": 0.8747,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07732951730970335,
+      "grad_norm": 0.8503691554069519,
+      "learning_rate": 9.968035415902555e-06,
+      "loss": 0.7478,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07738455611205901,
+      "grad_norm": 0.9209537506103516,
+      "learning_rate": 9.967986461627808e-06,
+      "loss": 0.9052,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07743959491441466,
+      "grad_norm": 0.8447962999343872,
+      "learning_rate": 9.967937470015006e-06,
+      "loss": 0.7897,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07749463371677032,
+      "grad_norm": 0.8731846809387207,
+      "learning_rate": 9.967888441064515e-06,
+      "loss": 0.837,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07754967251912598,
+      "grad_norm": 0.9810444712638855,
+      "learning_rate": 9.967839374776705e-06,
+      "loss": 0.8236,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07760471132148164,
+      "grad_norm": 0.8283190131187439,
+      "learning_rate": 9.967790271151944e-06,
+      "loss": 0.8443,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0776597501238373,
+      "grad_norm": 0.7999932765960693,
+      "learning_rate": 9.9677411301906e-06,
+      "loss": 0.7945,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07771478892619296,
+      "grad_norm": 0.9435983300209045,
+      "learning_rate": 9.967691951893044e-06,
+      "loss": 0.9745,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07776982772854862,
+      "grad_norm": 0.8885984420776367,
+      "learning_rate": 9.967642736259646e-06,
+      "loss": 0.9163,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07782486653090429,
+      "grad_norm": 0.993928074836731,
+      "learning_rate": 9.967593483290776e-06,
+      "loss": 0.7797,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07787990533325995,
+      "grad_norm": 1.058830976486206,
+      "learning_rate": 9.9675441929868e-06,
+      "loss": 0.8671,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07793494413561561,
+      "grad_norm": 1.0469766855239868,
+      "learning_rate": 9.967494865348093e-06,
+      "loss": 0.8671,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07798998293797127,
+      "grad_norm": 0.902729868888855,
+      "learning_rate": 9.967445500375025e-06,
+      "loss": 0.8748,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07804502174032693,
+      "grad_norm": 0.90755295753479,
+      "learning_rate": 9.967396098067965e-06,
+      "loss": 0.8279,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07810006054268259,
+      "grad_norm": 0.8822374939918518,
+      "learning_rate": 9.967346658427287e-06,
+      "loss": 0.9386,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07815509934503825,
+      "grad_norm": 0.9201469421386719,
+      "learning_rate": 9.96729718145336e-06,
+      "loss": 0.8684,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07821013814739392,
+      "grad_norm": 0.9451109766960144,
+      "learning_rate": 9.967247667146558e-06,
+      "loss": 0.7854,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07826517694974958,
+      "grad_norm": 0.9146197438240051,
+      "learning_rate": 9.96719811550725e-06,
+      "loss": 0.8496,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07832021575210524,
+      "grad_norm": 0.9771224856376648,
+      "learning_rate": 9.967148526535813e-06,
+      "loss": 0.9657,
+      "step": 1423
+    },
+    {
+      "epoch": 0.0783752545544609,
+      "grad_norm": 0.8437683582305908,
+      "learning_rate": 9.967098900232616e-06,
+      "loss": 0.8336,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07843029335681656,
+      "grad_norm": 0.8232185244560242,
+      "learning_rate": 9.967049236598034e-06,
+      "loss": 0.8878,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07848533215917222,
+      "grad_norm": 1.0200369358062744,
+      "learning_rate": 9.96699953563244e-06,
+      "loss": 0.8135,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07854037096152788,
+      "grad_norm": 0.8779187202453613,
+      "learning_rate": 9.966949797336208e-06,
+      "loss": 0.9124,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07859540976388354,
+      "grad_norm": 0.9557466506958008,
+      "learning_rate": 9.966900021709708e-06,
+      "loss": 0.9118,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07865044856623919,
+      "grad_norm": 0.8431050777435303,
+      "learning_rate": 9.966850208753317e-06,
+      "loss": 0.8361,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07870548736859485,
+      "grad_norm": 0.9269648194313049,
+      "learning_rate": 9.966800358467412e-06,
+      "loss": 0.9194,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07876052617095052,
+      "grad_norm": 0.818681538105011,
+      "learning_rate": 9.966750470852363e-06,
+      "loss": 0.7483,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07881556497330618,
+      "grad_norm": 0.8788284659385681,
+      "learning_rate": 9.966700545908547e-06,
+      "loss": 0.858,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07887060377566184,
+      "grad_norm": 0.7734160423278809,
+      "learning_rate": 9.966650583636342e-06,
+      "loss": 0.694,
+      "step": 1433
+    },
+    {
+      "epoch": 0.0789256425780175,
+      "grad_norm": 0.8846608996391296,
+      "learning_rate": 9.966600584036117e-06,
+      "loss": 0.8144,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07898068138037316,
+      "grad_norm": 0.9740058183670044,
+      "learning_rate": 9.966550547108254e-06,
+      "loss": 0.9314,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07903572018272882,
+      "grad_norm": 0.8731759786605835,
+      "learning_rate": 9.966500472853124e-06,
+      "loss": 0.8475,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07909075898508448,
+      "grad_norm": 0.8984843492507935,
+      "learning_rate": 9.966450361271109e-06,
+      "loss": 0.7803,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07914579778744014,
+      "grad_norm": 0.8897966742515564,
+      "learning_rate": 9.96640021236258e-06,
+      "loss": 0.8879,
+      "step": 1438
+    },
+    {
+      "epoch": 0.0792008365897958,
+      "grad_norm": 0.80704265832901,
+      "learning_rate": 9.966350026127917e-06,
+      "loss": 0.7585,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07925587539215147,
+      "grad_norm": 1.0807467699050903,
+      "learning_rate": 9.966299802567499e-06,
+      "loss": 1.078,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07931091419450713,
+      "grad_norm": 0.7994028925895691,
+      "learning_rate": 9.966249541681697e-06,
+      "loss": 0.8074,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07936595299686279,
+      "grad_norm": 0.877592921257019,
+      "learning_rate": 9.966199243470895e-06,
+      "loss": 0.8084,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07942099179921845,
+      "grad_norm": 0.7704572081565857,
+      "learning_rate": 9.966148907935469e-06,
+      "loss": 0.7206,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07947603060157411,
+      "grad_norm": 0.8222140669822693,
+      "learning_rate": 9.966098535075797e-06,
+      "loss": 0.7768,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07953106940392977,
+      "grad_norm": 1.389320731163025,
+      "learning_rate": 9.966048124892257e-06,
+      "loss": 1.0356,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07958610820628544,
+      "grad_norm": 0.9082457423210144,
+      "learning_rate": 9.965997677385229e-06,
+      "loss": 0.7379,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0796411470086411,
+      "grad_norm": 0.8029153943061829,
+      "learning_rate": 9.965947192555093e-06,
+      "loss": 0.7826,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07969618581099676,
+      "grad_norm": 0.8752758502960205,
+      "learning_rate": 9.965896670402227e-06,
+      "loss": 0.8526,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07975122461335242,
+      "grad_norm": 1.0665404796600342,
+      "learning_rate": 9.965846110927009e-06,
+      "loss": 0.858,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07980626341570807,
+      "grad_norm": 0.9468502402305603,
+      "learning_rate": 9.96579551412982e-06,
+      "loss": 0.9658,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07986130221806373,
+      "grad_norm": 1.0239403247833252,
+      "learning_rate": 9.965744880011046e-06,
+      "loss": 0.7995,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07991634102041939,
+      "grad_norm": 0.9808099865913391,
+      "learning_rate": 9.965694208571059e-06,
+      "loss": 1.0173,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07997137982277505,
+      "grad_norm": 0.9338780641555786,
+      "learning_rate": 9.965643499810245e-06,
+      "loss": 0.7917,
+      "step": 1453
+    },
+    {
+      "epoch": 0.08002641862513071,
+      "grad_norm": 0.9294295310974121,
+      "learning_rate": 9.965592753728981e-06,
+      "loss": 0.88,
+      "step": 1454
+    },
+    {
+      "epoch": 0.08008145742748637,
+      "grad_norm": 1.0261508226394653,
+      "learning_rate": 9.965541970327654e-06,
+      "loss": 0.8825,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08013649622984204,
+      "grad_norm": 0.8964946269989014,
+      "learning_rate": 9.965491149606642e-06,
+      "loss": 0.81,
+      "step": 1456
+    },
+    {
+      "epoch": 0.0801915350321977,
+      "grad_norm": 0.9468267560005188,
+      "learning_rate": 9.965440291566329e-06,
+      "loss": 0.9453,
+      "step": 1457
+    },
+    {
+      "epoch": 0.08024657383455336,
+      "grad_norm": 0.8289040327072144,
+      "learning_rate": 9.965389396207092e-06,
+      "loss": 0.7373,
+      "step": 1458
+    },
+    {
+      "epoch": 0.08030161263690902,
+      "grad_norm": 0.8782384991645813,
+      "learning_rate": 9.965338463529322e-06,
+      "loss": 0.9199,
+      "step": 1459
+    },
+    {
+      "epoch": 0.08035665143926468,
+      "grad_norm": 0.8613787293434143,
+      "learning_rate": 9.965287493533395e-06,
+      "loss": 0.8719,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08041169024162034,
+      "grad_norm": 0.8474903106689453,
+      "learning_rate": 9.965236486219696e-06,
+      "loss": 0.8033,
+      "step": 1461
+    },
+    {
+      "epoch": 0.080466729043976,
+      "grad_norm": 1.1442681550979614,
+      "learning_rate": 9.965185441588609e-06,
+      "loss": 0.8996,
+      "step": 1462
+    },
+    {
+      "epoch": 0.08052176784633167,
+      "grad_norm": 1.564138412475586,
+      "learning_rate": 9.965134359640518e-06,
+      "loss": 0.7451,
+      "step": 1463
+    },
+    {
+      "epoch": 0.08057680664868733,
+      "grad_norm": 0.9211083054542542,
+      "learning_rate": 9.965083240375806e-06,
+      "loss": 0.8939,
+      "step": 1464
+    },
+    {
+      "epoch": 0.08063184545104299,
+      "grad_norm": 0.9503418207168579,
+      "learning_rate": 9.965032083794856e-06,
+      "loss": 0.8544,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08068688425339865,
+      "grad_norm": 0.9304021596908569,
+      "learning_rate": 9.964980889898055e-06,
+      "loss": 0.9192,
+      "step": 1466
+    },
+    {
+      "epoch": 0.08074192305575431,
+      "grad_norm": 0.8430425524711609,
+      "learning_rate": 9.964929658685787e-06,
+      "loss": 0.8586,
+      "step": 1467
+    },
+    {
+      "epoch": 0.08079696185810997,
+      "grad_norm": 0.8671759366989136,
+      "learning_rate": 9.964878390158437e-06,
+      "loss": 0.8807,
+      "step": 1468
+    },
+    {
+      "epoch": 0.08085200066046563,
+      "grad_norm": 0.9548830986022949,
+      "learning_rate": 9.964827084316389e-06,
+      "loss": 0.9033,
+      "step": 1469
+    },
+    {
+      "epoch": 0.0809070394628213,
+      "grad_norm": 0.8736767768859863,
+      "learning_rate": 9.964775741160029e-06,
+      "loss": 0.8509,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08096207826517696,
+      "grad_norm": 0.8827025890350342,
+      "learning_rate": 9.964724360689745e-06,
+      "loss": 0.897,
+      "step": 1471
+    },
+    {
+      "epoch": 0.0810171170675326,
+      "grad_norm": 1.02822744846344,
+      "learning_rate": 9.964672942905921e-06,
+      "loss": 1.0371,
+      "step": 1472
+    },
+    {
+      "epoch": 0.08107215586988827,
+      "grad_norm": 0.8619557619094849,
+      "learning_rate": 9.964621487808946e-06,
+      "loss": 0.7654,
+      "step": 1473
+    },
+    {
+      "epoch": 0.08112719467224393,
+      "grad_norm": 0.7855951189994812,
+      "learning_rate": 9.9645699953992e-06,
+      "loss": 0.7767,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08118223347459959,
+      "grad_norm": 0.8139809370040894,
+      "learning_rate": 9.96451846567708e-06,
+      "loss": 0.7535,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08123727227695525,
+      "grad_norm": 0.8491657376289368,
+      "learning_rate": 9.964466898642966e-06,
+      "loss": 0.854,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08129231107931091,
+      "grad_norm": 0.8968605399131775,
+      "learning_rate": 9.964415294297247e-06,
+      "loss": 0.8914,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08134734988166657,
+      "grad_norm": 0.8692505359649658,
+      "learning_rate": 9.964363652640313e-06,
+      "loss": 0.9245,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08140238868402223,
+      "grad_norm": 0.8916530013084412,
+      "learning_rate": 9.964311973672549e-06,
+      "loss": 0.7662,
+      "step": 1479
+    },
+    {
+      "epoch": 0.0814574274863779,
+      "grad_norm": 0.8239215612411499,
+      "learning_rate": 9.964260257394347e-06,
+      "loss": 0.9191,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08151246628873356,
+      "grad_norm": 0.8672100901603699,
+      "learning_rate": 9.964208503806092e-06,
+      "loss": 0.7656,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08156750509108922,
+      "grad_norm": 0.9195712208747864,
+      "learning_rate": 9.964156712908177e-06,
+      "loss": 0.8656,
+      "step": 1482
+    },
+    {
+      "epoch": 0.08162254389344488,
+      "grad_norm": 0.8282535672187805,
+      "learning_rate": 9.964104884700986e-06,
+      "loss": 0.8264,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08167758269580054,
+      "grad_norm": 0.8492032289505005,
+      "learning_rate": 9.964053019184913e-06,
+      "loss": 0.7816,
+      "step": 1484
+    },
+    {
+      "epoch": 0.0817326214981562,
+      "grad_norm": 0.8491117358207703,
+      "learning_rate": 9.964001116360347e-06,
+      "loss": 0.7885,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08178766030051186,
+      "grad_norm": 0.9415153861045837,
+      "learning_rate": 9.963949176227677e-06,
+      "loss": 0.8165,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08184269910286752,
+      "grad_norm": 0.8462526202201843,
+      "learning_rate": 9.963897198787294e-06,
+      "loss": 0.8498,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08189773790522319,
+      "grad_norm": 0.8591959476470947,
+      "learning_rate": 9.963845184039586e-06,
+      "loss": 0.8906,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08195277670757885,
+      "grad_norm": 0.840761661529541,
+      "learning_rate": 9.963793131984949e-06,
+      "loss": 0.7831,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08200781550993451,
+      "grad_norm": 0.931404173374176,
+      "learning_rate": 9.96374104262377e-06,
+      "loss": 0.889,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08206285431229017,
+      "grad_norm": 0.9048783779144287,
+      "learning_rate": 9.963688915956443e-06,
+      "loss": 0.8321,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08211789311464583,
+      "grad_norm": 0.9145931601524353,
+      "learning_rate": 9.96363675198336e-06,
+      "loss": 0.9918,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08217293191700148,
+      "grad_norm": 0.9256643652915955,
+      "learning_rate": 9.963584550704908e-06,
+      "loss": 0.8731,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08222797071935714,
+      "grad_norm": 1.0212007761001587,
+      "learning_rate": 9.963532312121486e-06,
+      "loss": 0.9077,
+      "step": 1494
+    },
+    {
+      "epoch": 0.0822830095217128,
+      "grad_norm": 0.9206242561340332,
+      "learning_rate": 9.963480036233483e-06,
+      "loss": 0.9076,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08233804832406846,
+      "grad_norm": 0.8846865296363831,
+      "learning_rate": 9.963427723041294e-06,
+      "loss": 0.6826,
+      "step": 1496
+    },
+    {
+      "epoch": 0.08239308712642412,
+      "grad_norm": 0.8745351433753967,
+      "learning_rate": 9.963375372545309e-06,
+      "loss": 0.7935,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08244812592877979,
+      "grad_norm": 0.9019666314125061,
+      "learning_rate": 9.963322984745924e-06,
+      "loss": 0.8435,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08250316473113545,
+      "grad_norm": 0.8586859703063965,
+      "learning_rate": 9.963270559643531e-06,
+      "loss": 0.8118,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08255820353349111,
+      "grad_norm": 0.9192817807197571,
+      "learning_rate": 9.963218097238528e-06,
+      "loss": 0.824,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08261324233584677,
+      "grad_norm": 0.8972243070602417,
+      "learning_rate": 9.963165597531304e-06,
+      "loss": 0.8404,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08266828113820243,
+      "grad_norm": 0.8953961133956909,
+      "learning_rate": 9.963113060522256e-06,
+      "loss": 0.9031,
+      "step": 1502
+    },
+    {
+      "epoch": 0.0827233199405581,
+      "grad_norm": 0.9551270604133606,
+      "learning_rate": 9.963060486211779e-06,
+      "loss": 0.9177,
+      "step": 1503
+    },
+    {
+      "epoch": 0.08277835874291375,
+      "grad_norm": 0.8524616956710815,
+      "learning_rate": 9.963007874600268e-06,
+      "loss": 0.8582,
+      "step": 1504
+    },
+    {
+      "epoch": 0.08283339754526942,
+      "grad_norm": 0.8148764371871948,
+      "learning_rate": 9.962955225688118e-06,
+      "loss": 0.6859,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08288843634762508,
+      "grad_norm": 0.9110590219497681,
+      "learning_rate": 9.962902539475728e-06,
+      "loss": 0.7189,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08294347514998074,
+      "grad_norm": 0.8700116872787476,
+      "learning_rate": 9.962849815963487e-06,
+      "loss": 0.9462,
+      "step": 1507
+    },
+    {
+      "epoch": 0.0829985139523364,
+      "grad_norm": 0.877109706401825,
+      "learning_rate": 9.962797055151797e-06,
+      "loss": 0.8138,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08305355275469206,
+      "grad_norm": 0.7818365097045898,
+      "learning_rate": 9.962744257041053e-06,
+      "loss": 0.8474,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08310859155704772,
+      "grad_norm": 0.88360196352005,
+      "learning_rate": 9.96269142163165e-06,
+      "loss": 0.8724,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08316363035940338,
+      "grad_norm": 0.8982682228088379,
+      "learning_rate": 9.962638548923988e-06,
+      "loss": 0.9687,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08321866916175905,
+      "grad_norm": 0.7362002730369568,
+      "learning_rate": 9.962585638918462e-06,
+      "loss": 0.7666,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08327370796411471,
+      "grad_norm": 1.0993375778198242,
+      "learning_rate": 9.962532691615472e-06,
+      "loss": 0.8869,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08332874676647037,
+      "grad_norm": 0.8684842586517334,
+      "learning_rate": 9.962479707015415e-06,
+      "loss": 0.872,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08338378556882602,
+      "grad_norm": 1.0598478317260742,
+      "learning_rate": 9.962426685118689e-06,
+      "loss": 0.9102,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08343882437118168,
+      "grad_norm": 0.8492125272750854,
+      "learning_rate": 9.96237362592569e-06,
+      "loss": 0.7554,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08349386317353734,
+      "grad_norm": 0.8489052653312683,
+      "learning_rate": 9.962320529436821e-06,
+      "loss": 0.9139,
+      "step": 1517
+    },
+    {
+      "epoch": 0.083548901975893,
+      "grad_norm": 0.8650774359703064,
+      "learning_rate": 9.962267395652479e-06,
+      "loss": 0.8717,
+      "step": 1518
+    },
+    {
+      "epoch": 0.08360394077824866,
+      "grad_norm": 0.8393206596374512,
+      "learning_rate": 9.962214224573064e-06,
+      "loss": 0.8256,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08365897958060432,
+      "grad_norm": 0.8304896354675293,
+      "learning_rate": 9.962161016198974e-06,
+      "loss": 0.8232,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08371401838295998,
+      "grad_norm": 0.8718386292457581,
+      "learning_rate": 9.962107770530612e-06,
+      "loss": 0.8206,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08376905718531565,
+      "grad_norm": 0.9109341502189636,
+      "learning_rate": 9.962054487568373e-06,
+      "loss": 0.9576,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08382409598767131,
+      "grad_norm": 0.9543303847312927,
+      "learning_rate": 9.962001167312663e-06,
+      "loss": 0.8816,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08387913479002697,
+      "grad_norm": 0.9992844462394714,
+      "learning_rate": 9.961947809763881e-06,
+      "loss": 0.8682,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08393417359238263,
+      "grad_norm": 0.8092770576477051,
+      "learning_rate": 9.961894414922425e-06,
+      "loss": 0.6352,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08398921239473829,
+      "grad_norm": 0.9888653755187988,
+      "learning_rate": 9.961840982788703e-06,
+      "loss": 0.8721,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08404425119709395,
+      "grad_norm": 1.0092703104019165,
+      "learning_rate": 9.961787513363108e-06,
+      "loss": 0.7776,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08409928999944961,
+      "grad_norm": 0.8654646277427673,
+      "learning_rate": 9.961734006646049e-06,
+      "loss": 0.8835,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08415432880180528,
+      "grad_norm": 0.7630153298377991,
+      "learning_rate": 9.961680462637924e-06,
+      "loss": 0.7501,
+      "step": 1529
+    },
+    {
+      "epoch": 0.08420936760416094,
+      "grad_norm": 1.1883158683776855,
+      "learning_rate": 9.961626881339138e-06,
+      "loss": 0.9476,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0842644064065166,
+      "grad_norm": 0.8710927963256836,
+      "learning_rate": 9.96157326275009e-06,
+      "loss": 0.749,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08431944520887226,
+      "grad_norm": 0.9500633478164673,
+      "learning_rate": 9.961519606871188e-06,
+      "loss": 0.8994,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08437448401122792,
+      "grad_norm": 0.873257577419281,
+      "learning_rate": 9.961465913702833e-06,
+      "loss": 0.816,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08442952281358358,
+      "grad_norm": 0.8007022142410278,
+      "learning_rate": 9.961412183245426e-06,
+      "loss": 0.787,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08448456161593924,
+      "grad_norm": 0.8998435139656067,
+      "learning_rate": 9.961358415499374e-06,
+      "loss": 0.8741,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08453960041829489,
+      "grad_norm": 0.9152502417564392,
+      "learning_rate": 9.961304610465081e-06,
+      "loss": 0.9749,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08459463922065055,
+      "grad_norm": 0.8961958289146423,
+      "learning_rate": 9.961250768142949e-06,
+      "loss": 0.8683,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08464967802300621,
+      "grad_norm": 0.8683995008468628,
+      "learning_rate": 9.961196888533387e-06,
+      "loss": 0.8347,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08470471682536188,
+      "grad_norm": 0.835221529006958,
+      "learning_rate": 9.961142971636795e-06,
+      "loss": 0.8936,
+      "step": 1539
+    },
+    {
+      "epoch": 0.08475975562771754,
+      "grad_norm": 0.8666725158691406,
+      "learning_rate": 9.96108901745358e-06,
+      "loss": 0.7344,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0848147944300732,
+      "grad_norm": 0.9509082436561584,
+      "learning_rate": 9.96103502598415e-06,
+      "loss": 0.8965,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08486983323242886,
+      "grad_norm": 0.8134233951568604,
+      "learning_rate": 9.960980997228908e-06,
+      "loss": 0.797,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08492487203478452,
+      "grad_norm": 1.0432242155075073,
+      "learning_rate": 9.96092693118826e-06,
+      "loss": 0.8754,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08497991083714018,
+      "grad_norm": 0.9560218453407288,
+      "learning_rate": 9.960872827862613e-06,
+      "loss": 0.9238,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08503494963949584,
+      "grad_norm": 0.8471649885177612,
+      "learning_rate": 9.960818687252374e-06,
+      "loss": 0.8622,
+      "step": 1545
+    },
+    {
+      "epoch": 0.0850899884418515,
+      "grad_norm": 1.2584747076034546,
+      "learning_rate": 9.960764509357951e-06,
+      "loss": 0.8007,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08514502724420717,
+      "grad_norm": 0.8730618953704834,
+      "learning_rate": 9.960710294179748e-06,
+      "loss": 0.7412,
+      "step": 1547
+    },
+    {
+      "epoch": 0.08520006604656283,
+      "grad_norm": 0.8361592292785645,
+      "learning_rate": 9.960656041718176e-06,
+      "loss": 0.7018,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08525510484891849,
+      "grad_norm": 0.8351722359657288,
+      "learning_rate": 9.96060175197364e-06,
+      "loss": 0.843,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08531014365127415,
+      "grad_norm": 0.8665090203285217,
+      "learning_rate": 9.960547424946549e-06,
+      "loss": 0.8235,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08536518245362981,
+      "grad_norm": 0.9254478812217712,
+      "learning_rate": 9.960493060637313e-06,
+      "loss": 0.8122,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08542022125598547,
+      "grad_norm": 0.8712261319160461,
+      "learning_rate": 9.960438659046337e-06,
+      "loss": 0.823,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08547526005834113,
+      "grad_norm": 0.9027207493782043,
+      "learning_rate": 9.960384220174033e-06,
+      "loss": 0.7964,
+      "step": 1553
+    },
+    {
+      "epoch": 0.0855302988606968,
+      "grad_norm": 0.854626476764679,
+      "learning_rate": 9.960329744020808e-06,
+      "loss": 0.755,
+      "step": 1554
+    },
+    {
+      "epoch": 0.08558533766305246,
+      "grad_norm": 0.9398048520088196,
+      "learning_rate": 9.960275230587073e-06,
+      "loss": 0.8607,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08564037646540812,
+      "grad_norm": 1.008002758026123,
+      "learning_rate": 9.960220679873238e-06,
+      "loss": 0.9711,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08569541526776378,
+      "grad_norm": 0.8999453783035278,
+      "learning_rate": 9.96016609187971e-06,
+      "loss": 0.8233,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08575045407011943,
+      "grad_norm": 0.8912106156349182,
+      "learning_rate": 9.960111466606903e-06,
+      "loss": 0.8271,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08580549287247509,
+      "grad_norm": 0.9269998073577881,
+      "learning_rate": 9.960056804055227e-06,
+      "loss": 0.7959,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08586053167483075,
+      "grad_norm": 1.083815336227417,
+      "learning_rate": 9.96000210422509e-06,
+      "loss": 0.9436,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08591557047718641,
+      "grad_norm": 0.8906280398368835,
+      "learning_rate": 9.959947367116905e-06,
+      "loss": 0.9317,
+      "step": 1561
+    },
+    {
+      "epoch": 0.08597060927954207,
+      "grad_norm": 1.211696743965149,
+      "learning_rate": 9.959892592731084e-06,
+      "loss": 0.9076,
+      "step": 1562
+    },
+    {
+      "epoch": 0.08602564808189773,
+      "grad_norm": 0.9050534963607788,
+      "learning_rate": 9.959837781068038e-06,
+      "loss": 0.8728,
+      "step": 1563
+    },
+    {
+      "epoch": 0.0860806868842534,
+      "grad_norm": 0.9384796619415283,
+      "learning_rate": 9.959782932128178e-06,
+      "loss": 0.9277,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08613572568660906,
+      "grad_norm": 0.795844316482544,
+      "learning_rate": 9.959728045911915e-06,
+      "loss": 0.7666,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08619076448896472,
+      "grad_norm": 0.925956666469574,
+      "learning_rate": 9.959673122419668e-06,
+      "loss": 0.815,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08624580329132038,
+      "grad_norm": 0.898047924041748,
+      "learning_rate": 9.959618161651843e-06,
+      "loss": 0.8131,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08630084209367604,
+      "grad_norm": 0.8656220436096191,
+      "learning_rate": 9.959563163608856e-06,
+      "loss": 0.9336,
+      "step": 1568
+    },
+    {
+      "epoch": 0.0863558808960317,
+      "grad_norm": 0.9184645414352417,
+      "learning_rate": 9.95950812829112e-06,
+      "loss": 0.9557,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08641091969838736,
+      "grad_norm": 0.8607667684555054,
+      "learning_rate": 9.959453055699048e-06,
+      "loss": 0.8272,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08646595850074303,
+      "grad_norm": 0.9561272263526917,
+      "learning_rate": 9.959397945833056e-06,
+      "loss": 0.8876,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08652099730309869,
+      "grad_norm": 0.8562412261962891,
+      "learning_rate": 9.959342798693556e-06,
+      "loss": 0.8404,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08657603610545435,
+      "grad_norm": 0.8924610614776611,
+      "learning_rate": 9.95928761428096e-06,
+      "loss": 0.8779,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08663107490781001,
+      "grad_norm": 0.8343208432197571,
+      "learning_rate": 9.95923239259569e-06,
+      "loss": 0.8992,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08668611371016567,
+      "grad_norm": 0.8835015296936035,
+      "learning_rate": 9.959177133638155e-06,
+      "loss": 1.0026,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08674115251252133,
+      "grad_norm": 0.9540221095085144,
+      "learning_rate": 9.959121837408771e-06,
+      "loss": 0.8507,
+      "step": 1576
+    },
+    {
+      "epoch": 0.086796191314877,
+      "grad_norm": 1.087817668914795,
+      "learning_rate": 9.959066503907957e-06,
+      "loss": 0.8607,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08685123011723266,
+      "grad_norm": 0.8072447180747986,
+      "learning_rate": 9.959011133136124e-06,
+      "loss": 0.882,
+      "step": 1578
+    },
+    {
+      "epoch": 0.0869062689195883,
+      "grad_norm": 0.7646876573562622,
+      "learning_rate": 9.958955725093694e-06,
+      "loss": 0.7653,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08696130772194396,
+      "grad_norm": 0.8979537487030029,
+      "learning_rate": 9.958900279781078e-06,
+      "loss": 0.9033,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08701634652429963,
+      "grad_norm": 0.9445611834526062,
+      "learning_rate": 9.958844797198696e-06,
+      "loss": 0.9423,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08707138532665529,
+      "grad_norm": 0.8836671113967896,
+      "learning_rate": 9.958789277346963e-06,
+      "loss": 0.839,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08712642412901095,
+      "grad_norm": 1.0333542823791504,
+      "learning_rate": 9.958733720226296e-06,
+      "loss": 0.9211,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08718146293136661,
+      "grad_norm": 0.8084085583686829,
+      "learning_rate": 9.958678125837117e-06,
+      "loss": 0.8387,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08723650173372227,
+      "grad_norm": 0.7769419550895691,
+      "learning_rate": 9.958622494179838e-06,
+      "loss": 0.8307,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08729154053607793,
+      "grad_norm": 0.8387578129768372,
+      "learning_rate": 9.95856682525488e-06,
+      "loss": 0.8001,
+      "step": 1586
+    },
+    {
+      "epoch": 0.0873465793384336,
+      "grad_norm": 0.8989812731742859,
+      "learning_rate": 9.95851111906266e-06,
+      "loss": 0.7752,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08740161814078926,
+      "grad_norm": 0.8558734655380249,
+      "learning_rate": 9.958455375603602e-06,
+      "loss": 0.8149,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08745665694314492,
+      "grad_norm": 0.8890896439552307,
+      "learning_rate": 9.958399594878117e-06,
+      "loss": 0.8232,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08751169574550058,
+      "grad_norm": 0.875912070274353,
+      "learning_rate": 9.95834377688663e-06,
+      "loss": 0.7458,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08756673454785624,
+      "grad_norm": 0.808355987071991,
+      "learning_rate": 9.958287921629557e-06,
+      "loss": 0.8296,
+      "step": 1591
+    },
+    {
+      "epoch": 0.0876217733502119,
+      "grad_norm": 0.9637090563774109,
+      "learning_rate": 9.958232029107318e-06,
+      "loss": 0.8769,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08767681215256756,
+      "grad_norm": 0.8980715870857239,
+      "learning_rate": 9.958176099320336e-06,
+      "loss": 0.7995,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08773185095492322,
+      "grad_norm": 0.9369860291481018,
+      "learning_rate": 9.95812013226903e-06,
+      "loss": 0.8545,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08778688975727889,
+      "grad_norm": 0.8589349389076233,
+      "learning_rate": 9.958064127953819e-06,
+      "loss": 0.8693,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08784192855963455,
+      "grad_norm": 0.929207444190979,
+      "learning_rate": 9.958008086375126e-06,
+      "loss": 0.811,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08789696736199021,
+      "grad_norm": 1.0825661420822144,
+      "learning_rate": 9.957952007533371e-06,
+      "loss": 1.0145,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08795200616434587,
+      "grad_norm": 0.8818382024765015,
+      "learning_rate": 9.957895891428978e-06,
+      "loss": 0.7771,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08800704496670153,
+      "grad_norm": 0.882780909538269,
+      "learning_rate": 9.957839738062363e-06,
+      "loss": 0.8857,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08806208376905718,
+      "grad_norm": 0.9136924743652344,
+      "learning_rate": 9.957783547433955e-06,
+      "loss": 0.8873,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08811712257141284,
+      "grad_norm": 0.8896858096122742,
+      "learning_rate": 9.95772731954417e-06,
+      "loss": 0.8463,
+      "step": 1601
+    },
+    {
+      "epoch": 0.0881721613737685,
+      "grad_norm": 0.8671631813049316,
+      "learning_rate": 9.957671054393436e-06,
+      "loss": 0.8333,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08822720017612416,
+      "grad_norm": 0.9442896246910095,
+      "learning_rate": 9.957614751982172e-06,
+      "loss": 0.9676,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08828223897847982,
+      "grad_norm": 0.8249240517616272,
+      "learning_rate": 9.957558412310803e-06,
+      "loss": 0.7746,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08833727778083549,
+      "grad_norm": 0.8125253319740295,
+      "learning_rate": 9.957502035379751e-06,
+      "loss": 0.7816,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08839231658319115,
+      "grad_norm": 0.8467233777046204,
+      "learning_rate": 9.957445621189442e-06,
+      "loss": 0.7697,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08844735538554681,
+      "grad_norm": 0.8322175145149231,
+      "learning_rate": 9.957389169740299e-06,
+      "loss": 0.7561,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08850239418790247,
+      "grad_norm": 0.869163453578949,
+      "learning_rate": 9.957332681032746e-06,
+      "loss": 0.8984,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08855743299025813,
+      "grad_norm": 0.8755944967269897,
+      "learning_rate": 9.957276155067206e-06,
+      "loss": 0.8016,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08861247179261379,
+      "grad_norm": 0.8152669668197632,
+      "learning_rate": 9.957219591844108e-06,
+      "loss": 0.7763,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08866751059496945,
+      "grad_norm": 0.979752779006958,
+      "learning_rate": 9.957162991363871e-06,
+      "loss": 0.7755,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08872254939732512,
+      "grad_norm": 1.0481054782867432,
+      "learning_rate": 9.957106353626926e-06,
+      "loss": 0.9395,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08877758819968078,
+      "grad_norm": 0.7773686647415161,
+      "learning_rate": 9.957049678633697e-06,
+      "loss": 0.7713,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08883262700203644,
+      "grad_norm": 0.838979959487915,
+      "learning_rate": 9.956992966384609e-06,
+      "loss": 0.7909,
+      "step": 1614
+    },
+    {
+      "epoch": 0.0888876658043921,
+      "grad_norm": 0.9527049660682678,
+      "learning_rate": 9.956936216880089e-06,
+      "loss": 0.7944,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08894270460674776,
+      "grad_norm": 0.7967305183410645,
+      "learning_rate": 9.956879430120561e-06,
+      "loss": 0.7703,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08899774340910342,
+      "grad_norm": 0.9065802097320557,
+      "learning_rate": 9.956822606106456e-06,
+      "loss": 0.8188,
+      "step": 1617
+    },
+    {
+      "epoch": 0.08905278221145908,
+      "grad_norm": 0.7329322099685669,
+      "learning_rate": 9.956765744838199e-06,
+      "loss": 0.8043,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08910782101381474,
+      "grad_norm": 0.864973247051239,
+      "learning_rate": 9.95670884631622e-06,
+      "loss": 0.8334,
+      "step": 1619
+    },
+    {
+      "epoch": 0.0891628598161704,
+      "grad_norm": 1.073559045791626,
+      "learning_rate": 9.95665191054094e-06,
+      "loss": 0.7755,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08921789861852607,
+      "grad_norm": 0.7347918748855591,
+      "learning_rate": 9.956594937512794e-06,
+      "loss": 0.7556,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08927293742088172,
+      "grad_norm": 0.8756610751152039,
+      "learning_rate": 9.956537927232205e-06,
+      "loss": 0.8129,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08932797622323738,
+      "grad_norm": 0.9132435917854309,
+      "learning_rate": 9.956480879699605e-06,
+      "loss": 0.8221,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08938301502559304,
+      "grad_norm": 1.1978256702423096,
+      "learning_rate": 9.956423794915421e-06,
+      "loss": 0.8651,
+      "step": 1624
+    },
+    {
+      "epoch": 0.0894380538279487,
+      "grad_norm": 0.8493894934654236,
+      "learning_rate": 9.956366672880082e-06,
+      "loss": 0.7267,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08949309263030436,
+      "grad_norm": 1.0971951484680176,
+      "learning_rate": 9.956309513594019e-06,
+      "loss": 0.7852,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08954813143266002,
+      "grad_norm": 0.899974524974823,
+      "learning_rate": 9.95625231705766e-06,
+      "loss": 0.8868,
+      "step": 1627
+    },
+    {
+      "epoch": 0.08960317023501568,
+      "grad_norm": 0.8995566368103027,
+      "learning_rate": 9.956195083271436e-06,
+      "loss": 0.87,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08965820903737134,
+      "grad_norm": 0.8924218416213989,
+      "learning_rate": 9.956137812235776e-06,
+      "loss": 0.7885,
+      "step": 1629
+    },
+    {
+      "epoch": 0.089713247839727,
+      "grad_norm": 0.9232820868492126,
+      "learning_rate": 9.956080503951108e-06,
+      "loss": 0.7923,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08976828664208267,
+      "grad_norm": 0.9298982620239258,
+      "learning_rate": 9.956023158417869e-06,
+      "loss": 0.8625,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08982332544443833,
+      "grad_norm": 0.86515212059021,
+      "learning_rate": 9.955965775636488e-06,
+      "loss": 0.7683,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08987836424679399,
+      "grad_norm": 0.8016952276229858,
+      "learning_rate": 9.955908355607392e-06,
+      "loss": 0.8122,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08993340304914965,
+      "grad_norm": 0.842703640460968,
+      "learning_rate": 9.955850898331015e-06,
+      "loss": 0.8487,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08998844185150531,
+      "grad_norm": 0.8239083886146545,
+      "learning_rate": 9.95579340380779e-06,
+      "loss": 0.8701,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09004348065386097,
+      "grad_norm": 0.8575418591499329,
+      "learning_rate": 9.955735872038149e-06,
+      "loss": 0.8263,
+      "step": 1636
+    },
+    {
+      "epoch": 0.09009851945621664,
+      "grad_norm": 0.8884586095809937,
+      "learning_rate": 9.955678303022522e-06,
+      "loss": 0.8112,
+      "step": 1637
+    },
+    {
+      "epoch": 0.0901535582585723,
+      "grad_norm": 0.9024681448936462,
+      "learning_rate": 9.955620696761345e-06,
+      "loss": 0.9174,
+      "step": 1638
+    },
+    {
+      "epoch": 0.09020859706092796,
+      "grad_norm": 0.8151944875717163,
+      "learning_rate": 9.955563053255049e-06,
+      "loss": 0.806,
+      "step": 1639
+    },
+    {
+      "epoch": 0.09026363586328362,
+      "grad_norm": 0.8292184472084045,
+      "learning_rate": 9.955505372504069e-06,
+      "loss": 0.8007,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09031867466563928,
+      "grad_norm": 0.9445936679840088,
+      "learning_rate": 9.955447654508835e-06,
+      "loss": 0.7089,
+      "step": 1641
+    },
+    {
+      "epoch": 0.09037371346799494,
+      "grad_norm": 0.781579315662384,
+      "learning_rate": 9.955389899269782e-06,
+      "loss": 0.8224,
+      "step": 1642
+    },
+    {
+      "epoch": 0.09042875227035059,
+      "grad_norm": 0.9028880596160889,
+      "learning_rate": 9.955332106787348e-06,
+      "loss": 0.7976,
+      "step": 1643
+    },
+    {
+      "epoch": 0.09048379107270625,
+      "grad_norm": 1.0336887836456299,
+      "learning_rate": 9.955274277061963e-06,
+      "loss": 0.9296,
+      "step": 1644
+    },
+    {
+      "epoch": 0.09053882987506191,
+      "grad_norm": 0.8894197940826416,
+      "learning_rate": 9.955216410094062e-06,
+      "loss": 0.815,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09059386867741757,
+      "grad_norm": 0.8955528140068054,
+      "learning_rate": 9.955158505884083e-06,
+      "loss": 0.8707,
+      "step": 1646
+    },
+    {
+      "epoch": 0.09064890747977324,
+      "grad_norm": 0.8012683987617493,
+      "learning_rate": 9.955100564432458e-06,
+      "loss": 0.7467,
+      "step": 1647
+    },
+    {
+      "epoch": 0.0907039462821289,
+      "grad_norm": 0.917969286441803,
+      "learning_rate": 9.955042585739623e-06,
+      "loss": 0.8835,
+      "step": 1648
+    },
+    {
+      "epoch": 0.09075898508448456,
+      "grad_norm": 0.8066666722297668,
+      "learning_rate": 9.954984569806014e-06,
+      "loss": 0.8338,
+      "step": 1649
+    },
+    {
+      "epoch": 0.09081402388684022,
+      "grad_norm": 1.1324070692062378,
+      "learning_rate": 9.954926516632069e-06,
+      "loss": 0.8245,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09086906268919588,
+      "grad_norm": 0.8196014761924744,
+      "learning_rate": 9.954868426218222e-06,
+      "loss": 0.7897,
+      "step": 1651
+    },
+    {
+      "epoch": 0.09092410149155154,
+      "grad_norm": 0.8713478446006775,
+      "learning_rate": 9.95481029856491e-06,
+      "loss": 0.891,
+      "step": 1652
+    },
+    {
+      "epoch": 0.0909791402939072,
+      "grad_norm": 0.8489059805870056,
+      "learning_rate": 9.954752133672569e-06,
+      "loss": 0.7748,
+      "step": 1653
+    },
+    {
+      "epoch": 0.09103417909626287,
+      "grad_norm": 0.8914602994918823,
+      "learning_rate": 9.954693931541638e-06,
+      "loss": 0.8657,
+      "step": 1654
+    },
+    {
+      "epoch": 0.09108921789861853,
+      "grad_norm": 0.9031614661216736,
+      "learning_rate": 9.954635692172555e-06,
+      "loss": 0.7409,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09114425670097419,
+      "grad_norm": 0.8680000305175781,
+      "learning_rate": 9.954577415565756e-06,
+      "loss": 0.8535,
+      "step": 1656
+    },
+    {
+      "epoch": 0.09119929550332985,
+      "grad_norm": 0.830596923828125,
+      "learning_rate": 9.954519101721679e-06,
+      "loss": 0.8601,
+      "step": 1657
+    },
+    {
+      "epoch": 0.09125433430568551,
+      "grad_norm": 0.9041332602500916,
+      "learning_rate": 9.954460750640762e-06,
+      "loss": 0.9104,
+      "step": 1658
+    },
+    {
+      "epoch": 0.09130937310804117,
+      "grad_norm": 0.7786296606063843,
+      "learning_rate": 9.954402362323445e-06,
+      "loss": 0.7671,
+      "step": 1659
+    },
+    {
+      "epoch": 0.09136441191039683,
+      "grad_norm": 1.0363564491271973,
+      "learning_rate": 9.954343936770165e-06,
+      "loss": 0.9339,
+      "step": 1660
+    },
+    {
+      "epoch": 0.0914194507127525,
+      "grad_norm": 0.8049986958503723,
+      "learning_rate": 9.954285473981363e-06,
+      "loss": 0.8125,
+      "step": 1661
+    },
+    {
+      "epoch": 0.09147448951510816,
+      "grad_norm": 0.7842011451721191,
+      "learning_rate": 9.954226973957477e-06,
+      "loss": 0.7153,
+      "step": 1662
+    },
+    {
+      "epoch": 0.09152952831746382,
+      "grad_norm": 0.8929729461669922,
+      "learning_rate": 9.954168436698948e-06,
+      "loss": 0.9563,
+      "step": 1663
+    },
+    {
+      "epoch": 0.09158456711981948,
+      "grad_norm": 0.8850226402282715,
+      "learning_rate": 9.954109862206216e-06,
+      "loss": 0.8257,
+      "step": 1664
+    },
+    {
+      "epoch": 0.09163960592217513,
+      "grad_norm": 0.8673348426818848,
+      "learning_rate": 9.954051250479719e-06,
+      "loss": 0.9489,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09169464472453079,
+      "grad_norm": 0.8726119995117188,
+      "learning_rate": 9.9539926015199e-06,
+      "loss": 0.8222,
+      "step": 1666
+    },
+    {
+      "epoch": 0.09174968352688645,
+      "grad_norm": 0.7609312534332275,
+      "learning_rate": 9.953933915327196e-06,
+      "loss": 0.7749,
+      "step": 1667
+    },
+    {
+      "epoch": 0.09180472232924211,
+      "grad_norm": 0.857404887676239,
+      "learning_rate": 9.953875191902055e-06,
+      "loss": 0.8496,
+      "step": 1668
+    },
+    {
+      "epoch": 0.09185976113159777,
+      "grad_norm": 0.7835526466369629,
+      "learning_rate": 9.953816431244909e-06,
+      "loss": 0.7258,
+      "step": 1669
+    },
+    {
+      "epoch": 0.09191479993395343,
+      "grad_norm": 0.944984495639801,
+      "learning_rate": 9.95375763335621e-06,
+      "loss": 0.902,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0919698387363091,
+      "grad_norm": 0.9038936495780945,
+      "learning_rate": 9.953698798236391e-06,
+      "loss": 0.7559,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09202487753866476,
+      "grad_norm": 0.8450848460197449,
+      "learning_rate": 9.953639925885898e-06,
+      "loss": 0.8338,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09207991634102042,
+      "grad_norm": 0.827419102191925,
+      "learning_rate": 9.953581016305175e-06,
+      "loss": 0.8167,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09213495514337608,
+      "grad_norm": 0.8517075777053833,
+      "learning_rate": 9.953522069494663e-06,
+      "loss": 0.8681,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09218999394573174,
+      "grad_norm": 0.9504323601722717,
+      "learning_rate": 9.953463085454804e-06,
+      "loss": 0.8688,
+      "step": 1675
+    },
+    {
+      "epoch": 0.0922450327480874,
+      "grad_norm": 0.8905719518661499,
+      "learning_rate": 9.953404064186044e-06,
+      "loss": 0.8818,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09230007155044306,
+      "grad_norm": 0.9223340153694153,
+      "learning_rate": 9.953345005688822e-06,
+      "loss": 0.8752,
+      "step": 1677
+    },
+    {
+      "epoch": 0.09235511035279872,
+      "grad_norm": 1.0500547885894775,
+      "learning_rate": 9.953285909963588e-06,
+      "loss": 0.7816,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09241014915515439,
+      "grad_norm": 0.8407441973686218,
+      "learning_rate": 9.953226777010781e-06,
+      "loss": 0.745,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09246518795751005,
+      "grad_norm": 0.7997288107872009,
+      "learning_rate": 9.953167606830847e-06,
+      "loss": 0.8171,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09252022675986571,
+      "grad_norm": 0.9752318859100342,
+      "learning_rate": 9.953108399424234e-06,
+      "loss": 0.8719,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09257526556222137,
+      "grad_norm": 0.8524298667907715,
+      "learning_rate": 9.953049154791382e-06,
+      "loss": 0.8257,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09263030436457703,
+      "grad_norm": 0.9460529088973999,
+      "learning_rate": 9.952989872932739e-06,
+      "loss": 0.7278,
+      "step": 1683
+    },
+    {
+      "epoch": 0.0926853431669327,
+      "grad_norm": 0.8959575891494751,
+      "learning_rate": 9.95293055384875e-06,
+      "loss": 0.903,
+      "step": 1684
+    },
+    {
+      "epoch": 0.09274038196928835,
+      "grad_norm": 0.8764386177062988,
+      "learning_rate": 9.95287119753986e-06,
+      "loss": 0.7958,
+      "step": 1685
+    },
+    {
+      "epoch": 0.092795420771644,
+      "grad_norm": 0.9611337184906006,
+      "learning_rate": 9.952811804006517e-06,
+      "loss": 0.8726,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09285045957399966,
+      "grad_norm": 0.8155574202537537,
+      "learning_rate": 9.952752373249165e-06,
+      "loss": 0.7882,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09290549837635532,
+      "grad_norm": 0.8789697289466858,
+      "learning_rate": 9.952692905268253e-06,
+      "loss": 0.8642,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09296053717871099,
+      "grad_norm": 0.7910027503967285,
+      "learning_rate": 9.952633400064227e-06,
+      "loss": 0.7852,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09301557598106665,
+      "grad_norm": 0.815819501876831,
+      "learning_rate": 9.952573857637533e-06,
+      "loss": 0.8606,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09307061478342231,
+      "grad_norm": 0.9840701818466187,
+      "learning_rate": 9.95251427798862e-06,
+      "loss": 0.9349,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09312565358577797,
+      "grad_norm": 0.8715788722038269,
+      "learning_rate": 9.952454661117936e-06,
+      "loss": 0.813,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09318069238813363,
+      "grad_norm": 0.8287779092788696,
+      "learning_rate": 9.952395007025926e-06,
+      "loss": 0.8346,
+      "step": 1693
+    },
+    {
+      "epoch": 0.0932357311904893,
+      "grad_norm": 0.9375059008598328,
+      "learning_rate": 9.952335315713044e-06,
+      "loss": 0.8868,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09329076999284495,
+      "grad_norm": 0.9063667058944702,
+      "learning_rate": 9.952275587179734e-06,
+      "loss": 0.9562,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09334580879520062,
+      "grad_norm": 0.816643476486206,
+      "learning_rate": 9.952215821426447e-06,
+      "loss": 0.7456,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09340084759755628,
+      "grad_norm": 0.9004347324371338,
+      "learning_rate": 9.95215601845363e-06,
+      "loss": 0.8545,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09345588639991194,
+      "grad_norm": 0.919195830821991,
+      "learning_rate": 9.952096178261736e-06,
+      "loss": 0.9347,
+      "step": 1698
+    },
+    {
+      "epoch": 0.0935109252022676,
+      "grad_norm": 0.8313261866569519,
+      "learning_rate": 9.952036300851211e-06,
+      "loss": 0.9169,
+      "step": 1699
+    },
+    {
+      "epoch": 0.09356596400462326,
+      "grad_norm": 0.8674910664558411,
+      "learning_rate": 9.951976386222507e-06,
+      "loss": 0.7621,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09362100280697892,
+      "grad_norm": 0.8931052684783936,
+      "learning_rate": 9.951916434376074e-06,
+      "loss": 0.8702,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09367604160933458,
+      "grad_norm": 0.8748393058776855,
+      "learning_rate": 9.951856445312364e-06,
+      "loss": 0.7446,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09373108041169025,
+      "grad_norm": 1.005459189414978,
+      "learning_rate": 9.951796419031825e-06,
+      "loss": 0.9843,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09378611921404591,
+      "grad_norm": 1.0155184268951416,
+      "learning_rate": 9.95173635553491e-06,
+      "loss": 0.8868,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09384115801640157,
+      "grad_norm": 2.1387271881103516,
+      "learning_rate": 9.951676254822072e-06,
+      "loss": 0.8691,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09389619681875723,
+      "grad_norm": 0.9768403768539429,
+      "learning_rate": 9.951616116893757e-06,
+      "loss": 0.8409,
+      "step": 1706
+    },
+    {
+      "epoch": 0.09395123562111289,
+      "grad_norm": 0.7994607090950012,
+      "learning_rate": 9.951555941750424e-06,
+      "loss": 0.7836,
+      "step": 1707
+    },
+    {
+      "epoch": 0.09400627442346854,
+      "grad_norm": 0.8460201025009155,
+      "learning_rate": 9.95149572939252e-06,
+      "loss": 0.8216,
+      "step": 1708
+    },
+    {
+      "epoch": 0.0940613132258242,
+      "grad_norm": 0.8904135227203369,
+      "learning_rate": 9.951435479820499e-06,
+      "loss": 0.9053,
+      "step": 1709
+    },
+    {
+      "epoch": 0.09411635202817986,
+      "grad_norm": 0.9084494113922119,
+      "learning_rate": 9.951375193034815e-06,
+      "loss": 0.9308,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09417139083053552,
+      "grad_norm": 1.0826482772827148,
+      "learning_rate": 9.951314869035921e-06,
+      "loss": 0.8468,
+      "step": 1711
+    },
+    {
+      "epoch": 0.09422642963289118,
+      "grad_norm": 0.8068915009498596,
+      "learning_rate": 9.95125450782427e-06,
+      "loss": 0.8253,
+      "step": 1712
+    },
+    {
+      "epoch": 0.09428146843524685,
+      "grad_norm": 0.8445400595664978,
+      "learning_rate": 9.951194109400316e-06,
+      "loss": 0.8386,
+      "step": 1713
+    },
+    {
+      "epoch": 0.09433650723760251,
+      "grad_norm": 0.8180645704269409,
+      "learning_rate": 9.951133673764513e-06,
+      "loss": 0.7907,
+      "step": 1714
+    },
+    {
+      "epoch": 0.09439154603995817,
+      "grad_norm": 0.8111036419868469,
+      "learning_rate": 9.951073200917311e-06,
+      "loss": 0.7918,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09444658484231383,
+      "grad_norm": 0.862042248249054,
+      "learning_rate": 9.951012690859172e-06,
+      "loss": 0.783,
+      "step": 1716
+    },
+    {
+      "epoch": 0.09450162364466949,
+      "grad_norm": 0.8189615607261658,
+      "learning_rate": 9.950952143590544e-06,
+      "loss": 0.8192,
+      "step": 1717
+    },
+    {
+      "epoch": 0.09455666244702515,
+      "grad_norm": 0.9714062809944153,
+      "learning_rate": 9.950891559111887e-06,
+      "loss": 0.774,
+      "step": 1718
+    },
+    {
+      "epoch": 0.09461170124938081,
+      "grad_norm": 0.9691846370697021,
+      "learning_rate": 9.950830937423655e-06,
+      "loss": 0.8347,
+      "step": 1719
+    },
+    {
+      "epoch": 0.09466674005173648,
+      "grad_norm": 0.8488250970840454,
+      "learning_rate": 9.950770278526301e-06,
+      "loss": 0.8228,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09472177885409214,
+      "grad_norm": 0.8638359904289246,
+      "learning_rate": 9.950709582420282e-06,
+      "loss": 0.8973,
+      "step": 1721
+    },
+    {
+      "epoch": 0.0947768176564478,
+      "grad_norm": 1.0148643255233765,
+      "learning_rate": 9.950648849106058e-06,
+      "loss": 0.9638,
+      "step": 1722
+    },
+    {
+      "epoch": 0.09483185645880346,
+      "grad_norm": 0.8870131969451904,
+      "learning_rate": 9.95058807858408e-06,
+      "loss": 0.8259,
+      "step": 1723
+    },
+    {
+      "epoch": 0.09488689526115912,
+      "grad_norm": 0.9134769439697266,
+      "learning_rate": 9.950527270854807e-06,
+      "loss": 0.865,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09494193406351478,
+      "grad_norm": 0.7221654653549194,
+      "learning_rate": 9.950466425918697e-06,
+      "loss": 0.7593,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09499697286587044,
+      "grad_norm": 0.9386674165725708,
+      "learning_rate": 9.950405543776207e-06,
+      "loss": 0.9508,
+      "step": 1726
+    },
+    {
+      "epoch": 0.0950520116682261,
+      "grad_norm": 0.7850627899169922,
+      "learning_rate": 9.950344624427795e-06,
+      "loss": 0.7999,
+      "step": 1727
+    },
+    {
+      "epoch": 0.09510705047058177,
+      "grad_norm": 0.921198308467865,
+      "learning_rate": 9.950283667873916e-06,
+      "loss": 0.8249,
+      "step": 1728
+    },
+    {
+      "epoch": 0.09516208927293741,
+      "grad_norm": 0.9503389000892639,
+      "learning_rate": 9.95022267411503e-06,
+      "loss": 0.901,
+      "step": 1729
+    },
+    {
+      "epoch": 0.09521712807529308,
+      "grad_norm": 0.7977343201637268,
+      "learning_rate": 9.950161643151597e-06,
+      "loss": 0.838,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09527216687764874,
+      "grad_norm": 0.9056238532066345,
+      "learning_rate": 9.950100574984072e-06,
+      "loss": 0.9756,
+      "step": 1731
+    },
+    {
+      "epoch": 0.0953272056800044,
+      "grad_norm": 0.8092935681343079,
+      "learning_rate": 9.950039469612918e-06,
+      "loss": 0.8812,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09538224448236006,
+      "grad_norm": 0.823693573474884,
+      "learning_rate": 9.949978327038592e-06,
+      "loss": 0.7914,
+      "step": 1733
+    },
+    {
+      "epoch": 0.09543728328471572,
+      "grad_norm": 0.9114876389503479,
+      "learning_rate": 9.949917147261554e-06,
+      "loss": 0.7944,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09549232208707138,
+      "grad_norm": 1.0084123611450195,
+      "learning_rate": 9.949855930282262e-06,
+      "loss": 0.8544,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09554736088942704,
+      "grad_norm": 0.842462956905365,
+      "learning_rate": 9.949794676101181e-06,
+      "loss": 0.7056,
+      "step": 1736
+    },
+    {
+      "epoch": 0.0956023996917827,
+      "grad_norm": 1.00497305393219,
+      "learning_rate": 9.949733384718766e-06,
+      "loss": 0.8372,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09565743849413837,
+      "grad_norm": 1.0166410207748413,
+      "learning_rate": 9.94967205613548e-06,
+      "loss": 0.9316,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09571247729649403,
+      "grad_norm": 0.8520192503929138,
+      "learning_rate": 9.949610690351784e-06,
+      "loss": 0.786,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09576751609884969,
+      "grad_norm": 0.8003227114677429,
+      "learning_rate": 9.949549287368139e-06,
+      "loss": 0.8003,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09582255490120535,
+      "grad_norm": 0.8657151460647583,
+      "learning_rate": 9.949487847185006e-06,
+      "loss": 0.8407,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09587759370356101,
+      "grad_norm": 1.1119858026504517,
+      "learning_rate": 9.949426369802848e-06,
+      "loss": 0.8594,
+      "step": 1742
+    },
+    {
+      "epoch": 0.09593263250591667,
+      "grad_norm": 0.8968474864959717,
+      "learning_rate": 9.949364855222126e-06,
+      "loss": 0.8254,
+      "step": 1743
+    },
+    {
+      "epoch": 0.09598767130827233,
+      "grad_norm": 0.8740531206130981,
+      "learning_rate": 9.949303303443304e-06,
+      "loss": 0.8748,
+      "step": 1744
+    },
+    {
+      "epoch": 0.096042710110628,
+      "grad_norm": 0.8833459615707397,
+      "learning_rate": 9.94924171446684e-06,
+      "loss": 0.838,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09609774891298366,
+      "grad_norm": 0.8783486485481262,
+      "learning_rate": 9.949180088293201e-06,
+      "loss": 0.7972,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09615278771533932,
+      "grad_norm": 0.9197877049446106,
+      "learning_rate": 9.949118424922852e-06,
+      "loss": 0.8669,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09620782651769498,
+      "grad_norm": 0.9771283864974976,
+      "learning_rate": 9.949056724356251e-06,
+      "loss": 0.8461,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09626286532005064,
+      "grad_norm": 0.8325022459030151,
+      "learning_rate": 9.948994986593864e-06,
+      "loss": 0.8482,
+      "step": 1749
+    },
+    {
+      "epoch": 0.0963179041224063,
+      "grad_norm": 0.9732363224029541,
+      "learning_rate": 9.948933211636158e-06,
+      "loss": 0.8825,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09637294292476195,
+      "grad_norm": 0.8229798078536987,
+      "learning_rate": 9.948871399483592e-06,
+      "loss": 0.8079,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09642798172711761,
+      "grad_norm": 0.8861554265022278,
+      "learning_rate": 9.948809550136635e-06,
+      "loss": 0.8323,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09648302052947327,
+      "grad_norm": 1.0618904829025269,
+      "learning_rate": 9.94874766359575e-06,
+      "loss": 0.8519,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09653805933182893,
+      "grad_norm": 0.8494864702224731,
+      "learning_rate": 9.948685739861403e-06,
+      "loss": 0.961,
+      "step": 1754
+    },
+    {
+      "epoch": 0.0965930981341846,
+      "grad_norm": 0.8872213959693909,
+      "learning_rate": 9.948623778934058e-06,
+      "loss": 0.9367,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09664813693654026,
+      "grad_norm": 0.8441230058670044,
+      "learning_rate": 9.948561780814181e-06,
+      "loss": 0.7654,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09670317573889592,
+      "grad_norm": 0.8072223663330078,
+      "learning_rate": 9.948499745502239e-06,
+      "loss": 0.7894,
+      "step": 1757
+    },
+    {
+      "epoch": 0.09675821454125158,
+      "grad_norm": 0.8285261392593384,
+      "learning_rate": 9.948437672998696e-06,
+      "loss": 0.8351,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09681325334360724,
+      "grad_norm": 0.9272124767303467,
+      "learning_rate": 9.94837556330402e-06,
+      "loss": 0.8708,
+      "step": 1759
+    },
+    {
+      "epoch": 0.0968682921459629,
+      "grad_norm": 0.8689375519752502,
+      "learning_rate": 9.94831341641868e-06,
+      "loss": 0.8478,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09692333094831856,
+      "grad_norm": 1.040784239768982,
+      "learning_rate": 9.94825123234314e-06,
+      "loss": 0.8915,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09697836975067423,
+      "grad_norm": 0.7819718718528748,
+      "learning_rate": 9.948189011077867e-06,
+      "loss": 0.7728,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09703340855302989,
+      "grad_norm": 0.7959379553794861,
+      "learning_rate": 9.948126752623331e-06,
+      "loss": 0.8248,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09708844735538555,
+      "grad_norm": 0.8844753503799438,
+      "learning_rate": 9.94806445698e-06,
+      "loss": 0.7742,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09714348615774121,
+      "grad_norm": 0.9168505668640137,
+      "learning_rate": 9.948002124148339e-06,
+      "loss": 0.9145,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09719852496009687,
+      "grad_norm": 0.7199662923812866,
+      "learning_rate": 9.947939754128819e-06,
+      "loss": 0.6652,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09725356376245253,
+      "grad_norm": 0.866470992565155,
+      "learning_rate": 9.947877346921909e-06,
+      "loss": 0.8293,
+      "step": 1767
+    },
+    {
+      "epoch": 0.0973086025648082,
+      "grad_norm": 0.9124754667282104,
+      "learning_rate": 9.947814902528078e-06,
+      "loss": 0.8599,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09736364136716386,
+      "grad_norm": 0.9169870615005493,
+      "learning_rate": 9.947752420947792e-06,
+      "loss": 0.8382,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09741868016951952,
+      "grad_norm": 1.0147640705108643,
+      "learning_rate": 9.947689902181526e-06,
+      "loss": 0.8425,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09747371897187518,
+      "grad_norm": 0.778575599193573,
+      "learning_rate": 9.947627346229745e-06,
+      "loss": 0.6979,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09752875777423083,
+      "grad_norm": 0.815101146697998,
+      "learning_rate": 9.947564753092922e-06,
+      "loss": 0.8617,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09758379657658649,
+      "grad_norm": 0.9556358456611633,
+      "learning_rate": 9.947502122771527e-06,
+      "loss": 0.9009,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09763883537894215,
+      "grad_norm": 0.8603761196136475,
+      "learning_rate": 9.94743945526603e-06,
+      "loss": 0.9443,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09769387418129781,
+      "grad_norm": 0.8621761798858643,
+      "learning_rate": 9.947376750576903e-06,
+      "loss": 0.7537,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09774891298365347,
+      "grad_norm": 0.7399948835372925,
+      "learning_rate": 9.947314008704616e-06,
+      "loss": 0.7477,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09780395178600913,
+      "grad_norm": 0.8855582475662231,
+      "learning_rate": 9.947251229649641e-06,
+      "loss": 0.8745,
+      "step": 1777
+    },
+    {
+      "epoch": 0.0978589905883648,
+      "grad_norm": 0.8718472719192505,
+      "learning_rate": 9.947188413412452e-06,
+      "loss": 0.9672,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09791402939072046,
+      "grad_norm": 0.8598514795303345,
+      "learning_rate": 9.947125559993517e-06,
+      "loss": 0.8278,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09796906819307612,
+      "grad_norm": 1.0373798608779907,
+      "learning_rate": 9.947062669393312e-06,
+      "loss": 0.8123,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09802410699543178,
+      "grad_norm": 1.0198705196380615,
+      "learning_rate": 9.946999741612306e-06,
+      "loss": 0.9039,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09807914579778744,
+      "grad_norm": 0.8770025968551636,
+      "learning_rate": 9.946936776650977e-06,
+      "loss": 0.8326,
+      "step": 1782
+    },
+    {
+      "epoch": 0.0981341846001431,
+      "grad_norm": 0.7970215678215027,
+      "learning_rate": 9.946873774509794e-06,
+      "loss": 0.848,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09818922340249876,
+      "grad_norm": 0.90342777967453,
+      "learning_rate": 9.946810735189231e-06,
+      "loss": 0.7993,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09824426220485442,
+      "grad_norm": 1.2095681428909302,
+      "learning_rate": 9.946747658689763e-06,
+      "loss": 0.8544,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09829930100721009,
+      "grad_norm": 0.8500953316688538,
+      "learning_rate": 9.946684545011866e-06,
+      "loss": 0.8398,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09835433980956575,
+      "grad_norm": 0.8570724725723267,
+      "learning_rate": 9.946621394156011e-06,
+      "loss": 0.9255,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09840937861192141,
+      "grad_norm": 0.8314846158027649,
+      "learning_rate": 9.946558206122672e-06,
+      "loss": 0.8398,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09846441741427707,
+      "grad_norm": 0.8894716501235962,
+      "learning_rate": 9.946494980912326e-06,
+      "loss": 0.8612,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09851945621663273,
+      "grad_norm": 0.9555756449699402,
+      "learning_rate": 9.94643171852545e-06,
+      "loss": 0.9551,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09857449501898839,
+      "grad_norm": 0.9556692838668823,
+      "learning_rate": 9.946368418962515e-06,
+      "loss": 0.8175,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09862953382134405,
+      "grad_norm": 0.7288535833358765,
+      "learning_rate": 9.946305082224e-06,
+      "loss": 0.6162,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09868457262369972,
+      "grad_norm": 0.95478355884552,
+      "learning_rate": 9.94624170831038e-06,
+      "loss": 0.9089,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09873961142605536,
+      "grad_norm": 0.9080137610435486,
+      "learning_rate": 9.946178297222133e-06,
+      "loss": 0.9443,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09879465022841102,
+      "grad_norm": 0.8060124516487122,
+      "learning_rate": 9.946114848959732e-06,
+      "loss": 0.7412,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09884968903076669,
+      "grad_norm": 0.8487932085990906,
+      "learning_rate": 9.946051363523655e-06,
+      "loss": 0.7098,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09890472783312235,
+      "grad_norm": 0.8982037901878357,
+      "learning_rate": 9.945987840914381e-06,
+      "loss": 0.8304,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09895976663547801,
+      "grad_norm": 0.8124602437019348,
+      "learning_rate": 9.945924281132386e-06,
+      "loss": 0.8441,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09901480543783367,
+      "grad_norm": 0.8081663250923157,
+      "learning_rate": 9.945860684178147e-06,
+      "loss": 0.732,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09906984424018933,
+      "grad_norm": 0.7662907242774963,
+      "learning_rate": 9.945797050052147e-06,
+      "loss": 0.7538,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09912488304254499,
+      "grad_norm": 0.8418399095535278,
+      "learning_rate": 9.945733378754856e-06,
+      "loss": 0.8488,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09917992184490065,
+      "grad_norm": 0.7298988699913025,
+      "learning_rate": 9.94566967028676e-06,
+      "loss": 0.7822,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09923496064725632,
+      "grad_norm": 0.7788695693016052,
+      "learning_rate": 9.945605924648332e-06,
+      "loss": 0.8037,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09928999944961198,
+      "grad_norm": 0.939297080039978,
+      "learning_rate": 9.945542141840054e-06,
+      "loss": 0.8654,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09934503825196764,
+      "grad_norm": 0.9274358749389648,
+      "learning_rate": 9.945478321862406e-06,
+      "loss": 0.7712,
+      "step": 1805
+    },
+    {
+      "epoch": 0.0994000770543233,
+      "grad_norm": 0.816561222076416,
+      "learning_rate": 9.945414464715866e-06,
+      "loss": 0.7676,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09945511585667896,
+      "grad_norm": 0.867915153503418,
+      "learning_rate": 9.945350570400916e-06,
+      "loss": 0.8343,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09951015465903462,
+      "grad_norm": 0.8446162939071655,
+      "learning_rate": 9.945286638918034e-06,
+      "loss": 0.8128,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09956519346139028,
+      "grad_norm": 0.8372986316680908,
+      "learning_rate": 9.945222670267703e-06,
+      "loss": 0.8611,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09962023226374594,
+      "grad_norm": 0.787836492061615,
+      "learning_rate": 9.945158664450399e-06,
+      "loss": 0.7286,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0996752710661016,
+      "grad_norm": 0.9293436408042908,
+      "learning_rate": 9.945094621466609e-06,
+      "loss": 0.8699,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09973030986845727,
+      "grad_norm": 0.8336932063102722,
+      "learning_rate": 9.94503054131681e-06,
+      "loss": 0.8222,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09978534867081293,
+      "grad_norm": 0.8310953378677368,
+      "learning_rate": 9.944966424001486e-06,
+      "loss": 0.8131,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09984038747316859,
+      "grad_norm": 0.7703443169593811,
+      "learning_rate": 9.944902269521117e-06,
+      "loss": 0.8135,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09989542627552424,
+      "grad_norm": 0.750990092754364,
+      "learning_rate": 9.944838077876186e-06,
+      "loss": 0.8137,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0999504650778799,
+      "grad_norm": 0.8502481579780579,
+      "learning_rate": 9.944773849067178e-06,
+      "loss": 0.8973,
+      "step": 1816
+    },
+    {
+      "epoch": 0.10000550388023556,
+      "grad_norm": 0.8299791812896729,
+      "learning_rate": 9.94470958309457e-06,
+      "loss": 0.8341,
+      "step": 1817
+    },
+    {
+      "epoch": 0.10006054268259122,
+      "grad_norm": 0.8519022464752197,
+      "learning_rate": 9.94464527995885e-06,
+      "loss": 0.8529,
+      "step": 1818
     }
   ],
   "logging_steps": 1,
@@ -6389,7 +12752,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.682514714121994e+18,
+  "total_flos": 5.365029428243988e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null