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": 0.7699711260827719,
+  "epoch": 1.1549566891241578,
   "eval_steps": 500,
-  "global_step": 20000,
+  "global_step": 30000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -28007,6 +28007,14006 @@
       "learning_rate": 0.00013542003106948943,
       "loss": 1.1504,
       "step": 20000
+    },
+    {
+      "epoch": 0.7701636188642926,
+      "grad_norm": 1.804743766784668,
+      "learning_rate": 0.00013539175029099405,
+      "loss": 1.1263,
+      "step": 20005
+    },
+    {
+      "epoch": 0.7703561116458133,
+      "grad_norm": 1.006145715713501,
+      "learning_rate": 0.00013536346627615755,
+      "loss": 1.2072,
+      "step": 20010
+    },
+    {
+      "epoch": 0.770548604427334,
+      "grad_norm": 1.6577104330062866,
+      "learning_rate": 0.0001353351790275663,
+      "loss": 0.9571,
+      "step": 20015
+    },
+    {
+      "epoch": 0.7707410972088546,
+      "grad_norm": 1.1842409372329712,
+      "learning_rate": 0.00013530688854780705,
+      "loss": 1.0826,
+      "step": 20020
+    },
+    {
+      "epoch": 0.7709335899903753,
+      "grad_norm": 2.0414535999298096,
+      "learning_rate": 0.00013527859483946668,
+      "loss": 0.9676,
+      "step": 20025
+    },
+    {
+      "epoch": 0.7711260827718961,
+      "grad_norm": 1.6052886247634888,
+      "learning_rate": 0.00013525029790513254,
+      "loss": 1.211,
+      "step": 20030
+    },
+    {
+      "epoch": 0.7713185755534168,
+      "grad_norm": 1.241951823234558,
+      "learning_rate": 0.00013522199774739218,
+      "loss": 1.0928,
+      "step": 20035
+    },
+    {
+      "epoch": 0.7715110683349374,
+      "grad_norm": 1.220537543296814,
+      "learning_rate": 0.0001351936943688334,
+      "loss": 1.0877,
+      "step": 20040
+    },
+    {
+      "epoch": 0.7717035611164581,
+      "grad_norm": 1.6344817876815796,
+      "learning_rate": 0.00013516538777204445,
+      "loss": 0.9949,
+      "step": 20045
+    },
+    {
+      "epoch": 0.7718960538979788,
+      "grad_norm": 1.798012137413025,
+      "learning_rate": 0.00013513707795961374,
+      "loss": 1.1155,
+      "step": 20050
+    },
+    {
+      "epoch": 0.7720885466794996,
+      "grad_norm": 1.1736531257629395,
+      "learning_rate": 0.00013510876493413,
+      "loss": 1.1333,
+      "step": 20055
+    },
+    {
+      "epoch": 0.7722810394610202,
+      "grad_norm": 1.3480515480041504,
+      "learning_rate": 0.00013508044869818225,
+      "loss": 1.316,
+      "step": 20060
+    },
+    {
+      "epoch": 0.7724735322425409,
+      "grad_norm": 1.5204951763153076,
+      "learning_rate": 0.00013505212925435994,
+      "loss": 1.157,
+      "step": 20065
+    },
+    {
+      "epoch": 0.7726660250240616,
+      "grad_norm": 1.116921067237854,
+      "learning_rate": 0.00013502380660525255,
+      "loss": 0.9601,
+      "step": 20070
+    },
+    {
+      "epoch": 0.7728585178055823,
+      "grad_norm": 2.0013625621795654,
+      "learning_rate": 0.00013499548075345002,
+      "loss": 0.9889,
+      "step": 20075
+    },
+    {
+      "epoch": 0.773051010587103,
+      "grad_norm": 2.779212474822998,
+      "learning_rate": 0.00013496715170154266,
+      "loss": 1.1228,
+      "step": 20080
+    },
+    {
+      "epoch": 0.7732435033686237,
+      "grad_norm": 1.983289361000061,
+      "learning_rate": 0.00013493881945212088,
+      "loss": 1.1228,
+      "step": 20085
+    },
+    {
+      "epoch": 0.7734359961501444,
+      "grad_norm": 1.5862370729446411,
+      "learning_rate": 0.0001349104840077755,
+      "loss": 1.2704,
+      "step": 20090
+    },
+    {
+      "epoch": 0.773628488931665,
+      "grad_norm": 1.318428874015808,
+      "learning_rate": 0.00013488214537109766,
+      "loss": 1.0183,
+      "step": 20095
+    },
+    {
+      "epoch": 0.7738209817131858,
+      "grad_norm": 1.2089256048202515,
+      "learning_rate": 0.00013485380354467868,
+      "loss": 1.2204,
+      "step": 20100
+    },
+    {
+      "epoch": 0.7740134744947065,
+      "grad_norm": 1.587889552116394,
+      "learning_rate": 0.00013482545853111023,
+      "loss": 1.1076,
+      "step": 20105
+    },
+    {
+      "epoch": 0.7742059672762271,
+      "grad_norm": 1.168034553527832,
+      "learning_rate": 0.00013479711033298432,
+      "loss": 1.1339,
+      "step": 20110
+    },
+    {
+      "epoch": 0.7743984600577478,
+      "grad_norm": 1.2924379110336304,
+      "learning_rate": 0.0001347687589528932,
+      "loss": 1.1975,
+      "step": 20115
+    },
+    {
+      "epoch": 0.7745909528392685,
+      "grad_norm": 1.0745187997817993,
+      "learning_rate": 0.00013474040439342933,
+      "loss": 1.185,
+      "step": 20120
+    },
+    {
+      "epoch": 0.7747834456207893,
+      "grad_norm": 1.319265365600586,
+      "learning_rate": 0.00013471204665718568,
+      "loss": 0.9424,
+      "step": 20125
+    },
+    {
+      "epoch": 0.7749759384023099,
+      "grad_norm": 1.1227604150772095,
+      "learning_rate": 0.00013468368574675528,
+      "loss": 1.1125,
+      "step": 20130
+    },
+    {
+      "epoch": 0.7751684311838306,
+      "grad_norm": 0.9614366888999939,
+      "learning_rate": 0.00013465532166473157,
+      "loss": 1.2149,
+      "step": 20135
+    },
+    {
+      "epoch": 0.7753609239653513,
+      "grad_norm": 1.4985253810882568,
+      "learning_rate": 0.00013462695441370827,
+      "loss": 1.0558,
+      "step": 20140
+    },
+    {
+      "epoch": 0.775553416746872,
+      "grad_norm": 2.495373487472534,
+      "learning_rate": 0.00013459858399627938,
+      "loss": 1.1317,
+      "step": 20145
+    },
+    {
+      "epoch": 0.7757459095283927,
+      "grad_norm": 1.7242252826690674,
+      "learning_rate": 0.00013457021041503916,
+      "loss": 1.1204,
+      "step": 20150
+    },
+    {
+      "epoch": 0.7759384023099134,
+      "grad_norm": 1.0122432708740234,
+      "learning_rate": 0.00013454183367258223,
+      "loss": 1.0719,
+      "step": 20155
+    },
+    {
+      "epoch": 0.7761308950914341,
+      "grad_norm": 1.057358741760254,
+      "learning_rate": 0.00013451345377150342,
+      "loss": 1.0651,
+      "step": 20160
+    },
+    {
+      "epoch": 0.7763233878729547,
+      "grad_norm": 0.9141941666603088,
+      "learning_rate": 0.00013448507071439788,
+      "loss": 1.0958,
+      "step": 20165
+    },
+    {
+      "epoch": 0.7765158806544754,
+      "grad_norm": 1.0692884922027588,
+      "learning_rate": 0.00013445668450386106,
+      "loss": 0.9905,
+      "step": 20170
+    },
+    {
+      "epoch": 0.7767083734359962,
+      "grad_norm": 1.3968685865402222,
+      "learning_rate": 0.00013442829514248873,
+      "loss": 1.0801,
+      "step": 20175
+    },
+    {
+      "epoch": 0.7769008662175169,
+      "grad_norm": 1.427119255065918,
+      "learning_rate": 0.00013439990263287686,
+      "loss": 1.2859,
+      "step": 20180
+    },
+    {
+      "epoch": 0.7770933589990375,
+      "grad_norm": 1.9501906633377075,
+      "learning_rate": 0.00013437150697762177,
+      "loss": 1.1413,
+      "step": 20185
+    },
+    {
+      "epoch": 0.7772858517805582,
+      "grad_norm": 0.9638822674751282,
+      "learning_rate": 0.00013434310817932006,
+      "loss": 1.0041,
+      "step": 20190
+    },
+    {
+      "epoch": 0.7774783445620789,
+      "grad_norm": 1.9460463523864746,
+      "learning_rate": 0.00013431470624056862,
+      "loss": 1.1464,
+      "step": 20195
+    },
+    {
+      "epoch": 0.7776708373435997,
+      "grad_norm": 1.4077537059783936,
+      "learning_rate": 0.00013428630116396457,
+      "loss": 1.086,
+      "step": 20200
+    },
+    {
+      "epoch": 0.7778633301251203,
+      "grad_norm": 1.545357346534729,
+      "learning_rate": 0.00013425789295210545,
+      "loss": 1.0626,
+      "step": 20205
+    },
+    {
+      "epoch": 0.778055822906641,
+      "grad_norm": 1.826687216758728,
+      "learning_rate": 0.00013422948160758894,
+      "loss": 1.1287,
+      "step": 20210
+    },
+    {
+      "epoch": 0.7782483156881617,
+      "grad_norm": 1.4467684030532837,
+      "learning_rate": 0.00013420106713301307,
+      "loss": 1.0774,
+      "step": 20215
+    },
+    {
+      "epoch": 0.7784408084696823,
+      "grad_norm": 1.0557719469070435,
+      "learning_rate": 0.00013417264953097618,
+      "loss": 0.968,
+      "step": 20220
+    },
+    {
+      "epoch": 0.7786333012512031,
+      "grad_norm": 1.5922484397888184,
+      "learning_rate": 0.00013414422880407686,
+      "loss": 1.1911,
+      "step": 20225
+    },
+    {
+      "epoch": 0.7788257940327238,
+      "grad_norm": 1.2360528707504272,
+      "learning_rate": 0.000134115804954914,
+      "loss": 1.123,
+      "step": 20230
+    },
+    {
+      "epoch": 0.7790182868142445,
+      "grad_norm": 1.3299078941345215,
+      "learning_rate": 0.0001340873779860868,
+      "loss": 1.1586,
+      "step": 20235
+    },
+    {
+      "epoch": 0.7792107795957651,
+      "grad_norm": 1.2143315076828003,
+      "learning_rate": 0.00013405894790019467,
+      "loss": 1.0069,
+      "step": 20240
+    },
+    {
+      "epoch": 0.7794032723772859,
+      "grad_norm": 0.9542683959007263,
+      "learning_rate": 0.00013403051469983737,
+      "loss": 1.2013,
+      "step": 20245
+    },
+    {
+      "epoch": 0.7795957651588066,
+      "grad_norm": 1.3608813285827637,
+      "learning_rate": 0.00013400207838761496,
+      "loss": 1.1684,
+      "step": 20250
+    },
+    {
+      "epoch": 0.7797882579403272,
+      "grad_norm": 0.4999949038028717,
+      "learning_rate": 0.0001339736389661277,
+      "loss": 1.1161,
+      "step": 20255
+    },
+    {
+      "epoch": 0.7799807507218479,
+      "grad_norm": 1.8794244527816772,
+      "learning_rate": 0.00013394519643797622,
+      "loss": 0.8838,
+      "step": 20260
+    },
+    {
+      "epoch": 0.7801732435033686,
+      "grad_norm": 0.8917430639266968,
+      "learning_rate": 0.00013391675080576138,
+      "loss": 0.9322,
+      "step": 20265
+    },
+    {
+      "epoch": 0.7803657362848894,
+      "grad_norm": 0.8630626797676086,
+      "learning_rate": 0.0001338883020720844,
+      "loss": 1.0805,
+      "step": 20270
+    },
+    {
+      "epoch": 0.78055822906641,
+      "grad_norm": 1.7655322551727295,
+      "learning_rate": 0.00013385985023954664,
+      "loss": 1.0011,
+      "step": 20275
+    },
+    {
+      "epoch": 0.7807507218479307,
+      "grad_norm": 2.459364652633667,
+      "learning_rate": 0.00013383139531074987,
+      "loss": 1.2634,
+      "step": 20280
+    },
+    {
+      "epoch": 0.7809432146294514,
+      "grad_norm": 0.9355961084365845,
+      "learning_rate": 0.00013380293728829613,
+      "loss": 1.1082,
+      "step": 20285
+    },
+    {
+      "epoch": 0.781135707410972,
+      "grad_norm": 1.9479714632034302,
+      "learning_rate": 0.00013377447617478772,
+      "loss": 1.2329,
+      "step": 20290
+    },
+    {
+      "epoch": 0.7813282001924928,
+      "grad_norm": 1.408583164215088,
+      "learning_rate": 0.00013374601197282715,
+      "loss": 1.0494,
+      "step": 20295
+    },
+    {
+      "epoch": 0.7815206929740135,
+      "grad_norm": 1.0006494522094727,
+      "learning_rate": 0.00013371754468501735,
+      "loss": 1.0935,
+      "step": 20300
+    },
+    {
+      "epoch": 0.7817131857555342,
+      "grad_norm": 1.8096888065338135,
+      "learning_rate": 0.00013368907431396147,
+      "loss": 1.14,
+      "step": 20305
+    },
+    {
+      "epoch": 0.7819056785370548,
+      "grad_norm": 1.373366117477417,
+      "learning_rate": 0.00013366060086226288,
+      "loss": 1.0523,
+      "step": 20310
+    },
+    {
+      "epoch": 0.7820981713185755,
+      "grad_norm": 1.7965810298919678,
+      "learning_rate": 0.00013363212433252538,
+      "loss": 1.1531,
+      "step": 20315
+    },
+    {
+      "epoch": 0.7822906641000963,
+      "grad_norm": 1.3974156379699707,
+      "learning_rate": 0.00013360364472735285,
+      "loss": 1.0295,
+      "step": 20320
+    },
+    {
+      "epoch": 0.782483156881617,
+      "grad_norm": 2.2229206562042236,
+      "learning_rate": 0.00013357516204934961,
+      "loss": 1.2168,
+      "step": 20325
+    },
+    {
+      "epoch": 0.7826756496631376,
+      "grad_norm": 2.140164852142334,
+      "learning_rate": 0.00013354667630112026,
+      "loss": 1.1943,
+      "step": 20330
+    },
+    {
+      "epoch": 0.7828681424446583,
+      "grad_norm": 1.12675940990448,
+      "learning_rate": 0.00013351818748526953,
+      "loss": 1.1229,
+      "step": 20335
+    },
+    {
+      "epoch": 0.783060635226179,
+      "grad_norm": 0.9767451286315918,
+      "learning_rate": 0.00013348969560440262,
+      "loss": 1.0961,
+      "step": 20340
+    },
+    {
+      "epoch": 0.7832531280076998,
+      "grad_norm": 2.367725372314453,
+      "learning_rate": 0.00013346120066112492,
+      "loss": 0.9969,
+      "step": 20345
+    },
+    {
+      "epoch": 0.7834456207892204,
+      "grad_norm": 1.4590439796447754,
+      "learning_rate": 0.00013343270265804205,
+      "loss": 1.2416,
+      "step": 20350
+    },
+    {
+      "epoch": 0.7836381135707411,
+      "grad_norm": 1.0241880416870117,
+      "learning_rate": 0.00013340420159776,
+      "loss": 1.1929,
+      "step": 20355
+    },
+    {
+      "epoch": 0.7838306063522618,
+      "grad_norm": 1.1249408721923828,
+      "learning_rate": 0.000133375697482885,
+      "loss": 1.1346,
+      "step": 20360
+    },
+    {
+      "epoch": 0.7840230991337824,
+      "grad_norm": 0.3698437809944153,
+      "learning_rate": 0.00013334719031602357,
+      "loss": 0.9347,
+      "step": 20365
+    },
+    {
+      "epoch": 0.7842155919153032,
+      "grad_norm": 0.9476606845855713,
+      "learning_rate": 0.00013331868009978248,
+      "loss": 1.0438,
+      "step": 20370
+    },
+    {
+      "epoch": 0.7844080846968239,
+      "grad_norm": 1.432447910308838,
+      "learning_rate": 0.0001332901668367688,
+      "loss": 1.0981,
+      "step": 20375
+    },
+    {
+      "epoch": 0.7846005774783446,
+      "grad_norm": 0.9563504457473755,
+      "learning_rate": 0.0001332616505295899,
+      "loss": 1.0899,
+      "step": 20380
+    },
+    {
+      "epoch": 0.7847930702598652,
+      "grad_norm": 1.3649405241012573,
+      "learning_rate": 0.00013323313118085341,
+      "loss": 1.2495,
+      "step": 20385
+    },
+    {
+      "epoch": 0.784985563041386,
+      "grad_norm": 1.109165906906128,
+      "learning_rate": 0.00013320460879316719,
+      "loss": 1.122,
+      "step": 20390
+    },
+    {
+      "epoch": 0.7851780558229067,
+      "grad_norm": 1.2185308933258057,
+      "learning_rate": 0.0001331760833691395,
+      "loss": 1.0217,
+      "step": 20395
+    },
+    {
+      "epoch": 0.7853705486044273,
+      "grad_norm": 1.0969536304473877,
+      "learning_rate": 0.00013314755491137872,
+      "loss": 1.0269,
+      "step": 20400
+    },
+    {
+      "epoch": 0.785563041385948,
+      "grad_norm": 1.2872892618179321,
+      "learning_rate": 0.00013311902342249364,
+      "loss": 1.2322,
+      "step": 20405
+    },
+    {
+      "epoch": 0.7857555341674687,
+      "grad_norm": 1.7610993385314941,
+      "learning_rate": 0.0001330904889050933,
+      "loss": 0.9873,
+      "step": 20410
+    },
+    {
+      "epoch": 0.7859480269489895,
+      "grad_norm": 2.3899195194244385,
+      "learning_rate": 0.00013306195136178687,
+      "loss": 1.4362,
+      "step": 20415
+    },
+    {
+      "epoch": 0.7861405197305101,
+      "grad_norm": 0.9143069982528687,
+      "learning_rate": 0.00013303341079518404,
+      "loss": 1.0994,
+      "step": 20420
+    },
+    {
+      "epoch": 0.7863330125120308,
+      "grad_norm": 0.9000052809715271,
+      "learning_rate": 0.00013300486720789465,
+      "loss": 1.0955,
+      "step": 20425
+    },
+    {
+      "epoch": 0.7865255052935515,
+      "grad_norm": 1.6691547632217407,
+      "learning_rate": 0.00013297632060252875,
+      "loss": 1.1152,
+      "step": 20430
+    },
+    {
+      "epoch": 0.7867179980750721,
+      "grad_norm": 1.8236645460128784,
+      "learning_rate": 0.0001329477709816968,
+      "loss": 0.9309,
+      "step": 20435
+    },
+    {
+      "epoch": 0.7869104908565929,
+      "grad_norm": 1.151060700416565,
+      "learning_rate": 0.00013291921834800947,
+      "loss": 1.0847,
+      "step": 20440
+    },
+    {
+      "epoch": 0.7871029836381136,
+      "grad_norm": 0.9802745580673218,
+      "learning_rate": 0.00013289066270407766,
+      "loss": 1.18,
+      "step": 20445
+    },
+    {
+      "epoch": 0.7872954764196343,
+      "grad_norm": 1.4023500680923462,
+      "learning_rate": 0.00013286210405251262,
+      "loss": 1.1138,
+      "step": 20450
+    },
+    {
+      "epoch": 0.7874879692011549,
+      "grad_norm": 2.108933925628662,
+      "learning_rate": 0.0001328335423959259,
+      "loss": 1.245,
+      "step": 20455
+    },
+    {
+      "epoch": 0.7876804619826756,
+      "grad_norm": 1.586288332939148,
+      "learning_rate": 0.0001328049777369292,
+      "loss": 1.1811,
+      "step": 20460
+    },
+    {
+      "epoch": 0.7878729547641964,
+      "grad_norm": 1.2598310708999634,
+      "learning_rate": 0.00013277641007813457,
+      "loss": 0.9954,
+      "step": 20465
+    },
+    {
+      "epoch": 0.7880654475457171,
+      "grad_norm": 0.8444984555244446,
+      "learning_rate": 0.00013274783942215443,
+      "loss": 0.9792,
+      "step": 20470
+    },
+    {
+      "epoch": 0.7882579403272377,
+      "grad_norm": 1.5442594289779663,
+      "learning_rate": 0.00013271926577160126,
+      "loss": 1.1249,
+      "step": 20475
+    },
+    {
+      "epoch": 0.7884504331087584,
+      "grad_norm": 1.1979268789291382,
+      "learning_rate": 0.000132690689129088,
+      "loss": 1.1101,
+      "step": 20480
+    },
+    {
+      "epoch": 0.7886429258902791,
+      "grad_norm": 1.0047119855880737,
+      "learning_rate": 0.00013266210949722777,
+      "loss": 1.2296,
+      "step": 20485
+    },
+    {
+      "epoch": 0.7888354186717998,
+      "grad_norm": 0.9286414980888367,
+      "learning_rate": 0.000132633526878634,
+      "loss": 1.0575,
+      "step": 20490
+    },
+    {
+      "epoch": 0.7890279114533205,
+      "grad_norm": 1.0984097719192505,
+      "learning_rate": 0.00013260494127592036,
+      "loss": 1.2199,
+      "step": 20495
+    },
+    {
+      "epoch": 0.7892204042348412,
+      "grad_norm": 2.0137503147125244,
+      "learning_rate": 0.00013257635269170082,
+      "loss": 1.1294,
+      "step": 20500
+    },
+    {
+      "epoch": 0.7894128970163619,
+      "grad_norm": 1.5907988548278809,
+      "learning_rate": 0.00013254776112858966,
+      "loss": 1.0603,
+      "step": 20505
+    },
+    {
+      "epoch": 0.7896053897978825,
+      "grad_norm": 1.6451369524002075,
+      "learning_rate": 0.00013251916658920133,
+      "loss": 1.2061,
+      "step": 20510
+    },
+    {
+      "epoch": 0.7897978825794033,
+      "grad_norm": 1.7516169548034668,
+      "learning_rate": 0.00013249056907615065,
+      "loss": 1.0927,
+      "step": 20515
+    },
+    {
+      "epoch": 0.789990375360924,
+      "grad_norm": 2.2118046283721924,
+      "learning_rate": 0.00013246196859205265,
+      "loss": 1.2632,
+      "step": 20520
+    },
+    {
+      "epoch": 0.7901828681424446,
+      "grad_norm": 1.0801787376403809,
+      "learning_rate": 0.00013243336513952265,
+      "loss": 1.104,
+      "step": 20525
+    },
+    {
+      "epoch": 0.7903753609239653,
+      "grad_norm": 1.737065076828003,
+      "learning_rate": 0.00013240475872117625,
+      "loss": 1.1356,
+      "step": 20530
+    },
+    {
+      "epoch": 0.790567853705486,
+      "grad_norm": 2.103365659713745,
+      "learning_rate": 0.0001323761493396294,
+      "loss": 1.1386,
+      "step": 20535
+    },
+    {
+      "epoch": 0.7907603464870068,
+      "grad_norm": 1.2037795782089233,
+      "learning_rate": 0.0001323475369974981,
+      "loss": 1.1696,
+      "step": 20540
+    },
+    {
+      "epoch": 0.7909528392685274,
+      "grad_norm": 2.0184690952301025,
+      "learning_rate": 0.0001323189216973988,
+      "loss": 1.2378,
+      "step": 20545
+    },
+    {
+      "epoch": 0.7911453320500481,
+      "grad_norm": 1.1056318283081055,
+      "learning_rate": 0.0001322903034419483,
+      "loss": 1.0001,
+      "step": 20550
+    },
+    {
+      "epoch": 0.7913378248315688,
+      "grad_norm": 1.8512266874313354,
+      "learning_rate": 0.00013226168223376335,
+      "loss": 1.2393,
+      "step": 20555
+    },
+    {
+      "epoch": 0.7915303176130896,
+      "grad_norm": 1.0075215101242065,
+      "learning_rate": 0.00013223305807546134,
+      "loss": 1.0399,
+      "step": 20560
+    },
+    {
+      "epoch": 0.7917228103946102,
+      "grad_norm": 1.206949234008789,
+      "learning_rate": 0.00013220443096965967,
+      "loss": 1.131,
+      "step": 20565
+    },
+    {
+      "epoch": 0.7919153031761309,
+      "grad_norm": 1.0401040315628052,
+      "learning_rate": 0.00013217580091897613,
+      "loss": 1.0425,
+      "step": 20570
+    },
+    {
+      "epoch": 0.7921077959576516,
+      "grad_norm": 1.1868935823440552,
+      "learning_rate": 0.00013214716792602873,
+      "loss": 1.2813,
+      "step": 20575
+    },
+    {
+      "epoch": 0.7923002887391722,
+      "grad_norm": 1.0238397121429443,
+      "learning_rate": 0.00013211853199343577,
+      "loss": 1.1308,
+      "step": 20580
+    },
+    {
+      "epoch": 0.792492781520693,
+      "grad_norm": 1.0374101400375366,
+      "learning_rate": 0.00013208989312381585,
+      "loss": 0.9725,
+      "step": 20585
+    },
+    {
+      "epoch": 0.7926852743022137,
+      "grad_norm": 1.1706335544586182,
+      "learning_rate": 0.00013206125131978778,
+      "loss": 1.2237,
+      "step": 20590
+    },
+    {
+      "epoch": 0.7928777670837344,
+      "grad_norm": 1.3591363430023193,
+      "learning_rate": 0.00013203260658397066,
+      "loss": 1.217,
+      "step": 20595
+    },
+    {
+      "epoch": 0.793070259865255,
+      "grad_norm": 1.0355279445648193,
+      "learning_rate": 0.00013200395891898388,
+      "loss": 1.1943,
+      "step": 20600
+    },
+    {
+      "epoch": 0.7932627526467757,
+      "grad_norm": 1.4872701168060303,
+      "learning_rate": 0.000131975308327447,
+      "loss": 1.0404,
+      "step": 20605
+    },
+    {
+      "epoch": 0.7934552454282965,
+      "grad_norm": 1.1777255535125732,
+      "learning_rate": 0.00013194665481198006,
+      "loss": 1.1498,
+      "step": 20610
+    },
+    {
+      "epoch": 0.7936477382098172,
+      "grad_norm": 1.446821689605713,
+      "learning_rate": 0.00013191799837520312,
+      "loss": 1.1557,
+      "step": 20615
+    },
+    {
+      "epoch": 0.7938402309913378,
+      "grad_norm": 1.1746444702148438,
+      "learning_rate": 0.00013188933901973668,
+      "loss": 1.0335,
+      "step": 20620
+    },
+    {
+      "epoch": 0.7940327237728585,
+      "grad_norm": 1.7785364389419556,
+      "learning_rate": 0.00013186067674820145,
+      "loss": 1.2201,
+      "step": 20625
+    },
+    {
+      "epoch": 0.7942252165543792,
+      "grad_norm": 2.1019060611724854,
+      "learning_rate": 0.0001318320115632184,
+      "loss": 1.3937,
+      "step": 20630
+    },
+    {
+      "epoch": 0.7944177093359,
+      "grad_norm": 0.9918488264083862,
+      "learning_rate": 0.0001318033434674087,
+      "loss": 0.899,
+      "step": 20635
+    },
+    {
+      "epoch": 0.7946102021174206,
+      "grad_norm": 1.4679549932479858,
+      "learning_rate": 0.00013177467246339395,
+      "loss": 1.1425,
+      "step": 20640
+    },
+    {
+      "epoch": 0.7948026948989413,
+      "grad_norm": 1.2556777000427246,
+      "learning_rate": 0.0001317459985537959,
+      "loss": 1.0685,
+      "step": 20645
+    },
+    {
+      "epoch": 0.794995187680462,
+      "grad_norm": 1.1215819120407104,
+      "learning_rate": 0.00013171732174123657,
+      "loss": 1.1001,
+      "step": 20650
+    },
+    {
+      "epoch": 0.7951876804619826,
+      "grad_norm": 1.128730297088623,
+      "learning_rate": 0.0001316886420283383,
+      "loss": 1.1104,
+      "step": 20655
+    },
+    {
+      "epoch": 0.7953801732435034,
+      "grad_norm": 0.9813867807388306,
+      "learning_rate": 0.0001316599594177236,
+      "loss": 1.1989,
+      "step": 20660
+    },
+    {
+      "epoch": 0.7955726660250241,
+      "grad_norm": 1.1886361837387085,
+      "learning_rate": 0.00013163127391201536,
+      "loss": 1.1553,
+      "step": 20665
+    },
+    {
+      "epoch": 0.7957651588065447,
+      "grad_norm": 1.1876790523529053,
+      "learning_rate": 0.00013160258551383664,
+      "loss": 0.9751,
+      "step": 20670
+    },
+    {
+      "epoch": 0.7959576515880654,
+      "grad_norm": 0.8354418873786926,
+      "learning_rate": 0.00013157389422581087,
+      "loss": 1.1553,
+      "step": 20675
+    },
+    {
+      "epoch": 0.7961501443695861,
+      "grad_norm": 1.0195808410644531,
+      "learning_rate": 0.00013154520005056157,
+      "loss": 1.173,
+      "step": 20680
+    },
+    {
+      "epoch": 0.7963426371511069,
+      "grad_norm": 1.6569643020629883,
+      "learning_rate": 0.0001315165029907127,
+      "loss": 1.0836,
+      "step": 20685
+    },
+    {
+      "epoch": 0.7965351299326275,
+      "grad_norm": 1.1490013599395752,
+      "learning_rate": 0.00013148780304888846,
+      "loss": 1.1071,
+      "step": 20690
+    },
+    {
+      "epoch": 0.7967276227141482,
+      "grad_norm": 1.1535710096359253,
+      "learning_rate": 0.0001314591002277132,
+      "loss": 0.9335,
+      "step": 20695
+    },
+    {
+      "epoch": 0.7969201154956689,
+      "grad_norm": 1.2786650657653809,
+      "learning_rate": 0.00013143039452981158,
+      "loss": 1.0348,
+      "step": 20700
+    },
+    {
+      "epoch": 0.7971126082771897,
+      "grad_norm": 1.0382118225097656,
+      "learning_rate": 0.00013140168595780863,
+      "loss": 1.0676,
+      "step": 20705
+    },
+    {
+      "epoch": 0.7973051010587103,
+      "grad_norm": 1.2393159866333008,
+      "learning_rate": 0.00013137297451432952,
+      "loss": 1.0786,
+      "step": 20710
+    },
+    {
+      "epoch": 0.797497593840231,
+      "grad_norm": 1.04021155834198,
+      "learning_rate": 0.00013134426020199968,
+      "loss": 1.1149,
+      "step": 20715
+    },
+    {
+      "epoch": 0.7976900866217517,
+      "grad_norm": 0.871838390827179,
+      "learning_rate": 0.0001313155430234449,
+      "loss": 1.1709,
+      "step": 20720
+    },
+    {
+      "epoch": 0.7978825794032723,
+      "grad_norm": 1.0589998960494995,
+      "learning_rate": 0.00013128682298129118,
+      "loss": 1.0185,
+      "step": 20725
+    },
+    {
+      "epoch": 0.7980750721847931,
+      "grad_norm": 1.4314393997192383,
+      "learning_rate": 0.0001312581000781647,
+      "loss": 1.0613,
+      "step": 20730
+    },
+    {
+      "epoch": 0.7982675649663138,
+      "grad_norm": 1.0706419944763184,
+      "learning_rate": 0.00013122937431669208,
+      "loss": 1.3543,
+      "step": 20735
+    },
+    {
+      "epoch": 0.7984600577478345,
+      "grad_norm": 1.705357313156128,
+      "learning_rate": 0.00013120064569950004,
+      "loss": 1.0586,
+      "step": 20740
+    },
+    {
+      "epoch": 0.7986525505293551,
+      "grad_norm": 1.2182518243789673,
+      "learning_rate": 0.00013117191422921566,
+      "loss": 1.1202,
+      "step": 20745
+    },
+    {
+      "epoch": 0.7988450433108758,
+      "grad_norm": 1.1450048685073853,
+      "learning_rate": 0.00013114317990846618,
+      "loss": 1.2812,
+      "step": 20750
+    },
+    {
+      "epoch": 0.7990375360923966,
+      "grad_norm": 1.8270835876464844,
+      "learning_rate": 0.00013111444273987923,
+      "loss": 1.1169,
+      "step": 20755
+    },
+    {
+      "epoch": 0.7992300288739173,
+      "grad_norm": 2.412691593170166,
+      "learning_rate": 0.0001310857027260826,
+      "loss": 1.2125,
+      "step": 20760
+    },
+    {
+      "epoch": 0.7994225216554379,
+      "grad_norm": 1.1688083410263062,
+      "learning_rate": 0.00013105695986970434,
+      "loss": 1.1396,
+      "step": 20765
+    },
+    {
+      "epoch": 0.7996150144369586,
+      "grad_norm": 1.3766731023788452,
+      "learning_rate": 0.00013102821417337288,
+      "loss": 1.0521,
+      "step": 20770
+    },
+    {
+      "epoch": 0.7998075072184793,
+      "grad_norm": 1.4227287769317627,
+      "learning_rate": 0.00013099946563971678,
+      "loss": 1.2142,
+      "step": 20775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2115936279296875,
+      "learning_rate": 0.00013097071427136485,
+      "loss": 1.1114,
+      "step": 20780
+    },
+    {
+      "epoch": 0.8001924927815207,
+      "grad_norm": 1.5830806493759155,
+      "learning_rate": 0.00013094196007094633,
+      "loss": 1.015,
+      "step": 20785
+    },
+    {
+      "epoch": 0.8003849855630414,
+      "grad_norm": 1.4669450521469116,
+      "learning_rate": 0.0001309132030410905,
+      "loss": 1.1081,
+      "step": 20790
+    },
+    {
+      "epoch": 0.800577478344562,
+      "grad_norm": 1.534693956375122,
+      "learning_rate": 0.00013088444318442704,
+      "loss": 1.1994,
+      "step": 20795
+    },
+    {
+      "epoch": 0.8007699711260827,
+      "grad_norm": 1.6019155979156494,
+      "learning_rate": 0.0001308556805035858,
+      "loss": 1.1598,
+      "step": 20800
+    },
+    {
+      "epoch": 0.8009624639076035,
+      "grad_norm": 2.0105199813842773,
+      "learning_rate": 0.00013082691500119703,
+      "loss": 1.0093,
+      "step": 20805
+    },
+    {
+      "epoch": 0.8011549566891242,
+      "grad_norm": 0.9521466493606567,
+      "learning_rate": 0.00013079814667989108,
+      "loss": 0.9551,
+      "step": 20810
+    },
+    {
+      "epoch": 0.8013474494706448,
+      "grad_norm": 1.155561089515686,
+      "learning_rate": 0.00013076937554229866,
+      "loss": 1.012,
+      "step": 20815
+    },
+    {
+      "epoch": 0.8015399422521655,
+      "grad_norm": 2.6006276607513428,
+      "learning_rate": 0.00013074060159105061,
+      "loss": 1.2012,
+      "step": 20820
+    },
+    {
+      "epoch": 0.8017324350336862,
+      "grad_norm": 1.318688154220581,
+      "learning_rate": 0.0001307118248287782,
+      "loss": 1.0862,
+      "step": 20825
+    },
+    {
+      "epoch": 0.801924927815207,
+      "grad_norm": 1.0366714000701904,
+      "learning_rate": 0.0001306830452581129,
+      "loss": 0.8681,
+      "step": 20830
+    },
+    {
+      "epoch": 0.8021174205967276,
+      "grad_norm": 1.463513731956482,
+      "learning_rate": 0.00013065426288168635,
+      "loss": 1.0677,
+      "step": 20835
+    },
+    {
+      "epoch": 0.8023099133782483,
+      "grad_norm": 1.2675501108169556,
+      "learning_rate": 0.00013062547770213053,
+      "loss": 1.0329,
+      "step": 20840
+    },
+    {
+      "epoch": 0.802502406159769,
+      "grad_norm": 1.9978399276733398,
+      "learning_rate": 0.00013059668972207762,
+      "loss": 1.106,
+      "step": 20845
+    },
+    {
+      "epoch": 0.8026948989412896,
+      "grad_norm": 0.831674337387085,
+      "learning_rate": 0.00013056789894416013,
+      "loss": 1.1393,
+      "step": 20850
+    },
+    {
+      "epoch": 0.8028873917228104,
+      "grad_norm": 1.492956280708313,
+      "learning_rate": 0.00013053910537101076,
+      "loss": 1.1106,
+      "step": 20855
+    },
+    {
+      "epoch": 0.8030798845043311,
+      "grad_norm": 1.5051852464675903,
+      "learning_rate": 0.00013051030900526256,
+      "loss": 1.1677,
+      "step": 20860
+    },
+    {
+      "epoch": 0.8032723772858518,
+      "grad_norm": 1.2556471824645996,
+      "learning_rate": 0.00013048150984954867,
+      "loss": 1.006,
+      "step": 20865
+    },
+    {
+      "epoch": 0.8034648700673724,
+      "grad_norm": 1.4677566289901733,
+      "learning_rate": 0.00013045270790650263,
+      "loss": 1.2083,
+      "step": 20870
+    },
+    {
+      "epoch": 0.8036573628488932,
+      "grad_norm": 1.2184150218963623,
+      "learning_rate": 0.00013042390317875822,
+      "loss": 1.0107,
+      "step": 20875
+    },
+    {
+      "epoch": 0.8038498556304139,
+      "grad_norm": 1.0294650793075562,
+      "learning_rate": 0.00013039509566894935,
+      "loss": 1.1263,
+      "step": 20880
+    },
+    {
+      "epoch": 0.8040423484119346,
+      "grad_norm": 1.0486441850662231,
+      "learning_rate": 0.00013036628537971037,
+      "loss": 1.2297,
+      "step": 20885
+    },
+    {
+      "epoch": 0.8042348411934552,
+      "grad_norm": 1.209472894668579,
+      "learning_rate": 0.00013033747231367573,
+      "loss": 1.0295,
+      "step": 20890
+    },
+    {
+      "epoch": 0.8044273339749759,
+      "grad_norm": 1.3195816278457642,
+      "learning_rate": 0.00013030865647348022,
+      "loss": 1.1223,
+      "step": 20895
+    },
+    {
+      "epoch": 0.8046198267564967,
+      "grad_norm": 1.127369999885559,
+      "learning_rate": 0.00013027983786175883,
+      "loss": 1.1293,
+      "step": 20900
+    },
+    {
+      "epoch": 0.8048123195380174,
+      "grad_norm": 1.7267670631408691,
+      "learning_rate": 0.00013025101648114692,
+      "loss": 1.1115,
+      "step": 20905
+    },
+    {
+      "epoch": 0.805004812319538,
+      "grad_norm": 1.1385564804077148,
+      "learning_rate": 0.0001302221923342799,
+      "loss": 1.2283,
+      "step": 20910
+    },
+    {
+      "epoch": 0.8051973051010587,
+      "grad_norm": 2.3153762817382812,
+      "learning_rate": 0.0001301933654237936,
+      "loss": 1.0369,
+      "step": 20915
+    },
+    {
+      "epoch": 0.8053897978825794,
+      "grad_norm": 2.7795004844665527,
+      "learning_rate": 0.000130164535752324,
+      "loss": 1.1383,
+      "step": 20920
+    },
+    {
+      "epoch": 0.8055822906641001,
+      "grad_norm": 1.4328944683074951,
+      "learning_rate": 0.00013013570332250752,
+      "loss": 1.0496,
+      "step": 20925
+    },
+    {
+      "epoch": 0.8057747834456208,
+      "grad_norm": 1.35623037815094,
+      "learning_rate": 0.00013010686813698055,
+      "loss": 1.034,
+      "step": 20930
+    },
+    {
+      "epoch": 0.8059672762271415,
+      "grad_norm": 2.241270065307617,
+      "learning_rate": 0.0001300780301983799,
+      "loss": 1.1334,
+      "step": 20935
+    },
+    {
+      "epoch": 0.8061597690086622,
+      "grad_norm": 1.0069741010665894,
+      "learning_rate": 0.00013004918950934268,
+      "loss": 1.0983,
+      "step": 20940
+    },
+    {
+      "epoch": 0.8063522617901828,
+      "grad_norm": 2.6190080642700195,
+      "learning_rate": 0.0001300203460725061,
+      "loss": 1.0864,
+      "step": 20945
+    },
+    {
+      "epoch": 0.8065447545717036,
+      "grad_norm": 1.0321654081344604,
+      "learning_rate": 0.00012999149989050777,
+      "loss": 1.1463,
+      "step": 20950
+    },
+    {
+      "epoch": 0.8067372473532243,
+      "grad_norm": 1.1893829107284546,
+      "learning_rate": 0.00012996265096598545,
+      "loss": 1.1653,
+      "step": 20955
+    },
+    {
+      "epoch": 0.8069297401347449,
+      "grad_norm": 1.0437567234039307,
+      "learning_rate": 0.00012993379930157717,
+      "loss": 1.0993,
+      "step": 20960
+    },
+    {
+      "epoch": 0.8071222329162656,
+      "grad_norm": 0.9310737252235413,
+      "learning_rate": 0.00012990494489992124,
+      "loss": 1.0351,
+      "step": 20965
+    },
+    {
+      "epoch": 0.8073147256977863,
+      "grad_norm": 1.0145392417907715,
+      "learning_rate": 0.00012987608776365622,
+      "loss": 1.0316,
+      "step": 20970
+    },
+    {
+      "epoch": 0.8075072184793071,
+      "grad_norm": 0.8872208595275879,
+      "learning_rate": 0.00012984722789542084,
+      "loss": 0.9912,
+      "step": 20975
+    },
+    {
+      "epoch": 0.8076997112608277,
+      "grad_norm": 1.4148224592208862,
+      "learning_rate": 0.00012981836529785422,
+      "loss": 1.1786,
+      "step": 20980
+    },
+    {
+      "epoch": 0.8078922040423484,
+      "grad_norm": 0.9235820770263672,
+      "learning_rate": 0.00012978949997359565,
+      "loss": 1.1649,
+      "step": 20985
+    },
+    {
+      "epoch": 0.8080846968238691,
+      "grad_norm": 1.932556390762329,
+      "learning_rate": 0.0001297606319252846,
+      "loss": 1.1976,
+      "step": 20990
+    },
+    {
+      "epoch": 0.8082771896053897,
+      "grad_norm": 1.1439324617385864,
+      "learning_rate": 0.00012973176115556092,
+      "loss": 0.9586,
+      "step": 20995
+    },
+    {
+      "epoch": 0.8084696823869105,
+      "grad_norm": 1.4145176410675049,
+      "learning_rate": 0.00012970288766706465,
+      "loss": 1.1235,
+      "step": 21000
+    },
+    {
+      "epoch": 0.8086621751684312,
+      "grad_norm": 1.2468702793121338,
+      "learning_rate": 0.00012967401146243604,
+      "loss": 1.2867,
+      "step": 21005
+    },
+    {
+      "epoch": 0.8088546679499519,
+      "grad_norm": 1.230528473854065,
+      "learning_rate": 0.00012964513254431567,
+      "loss": 1.1464,
+      "step": 21010
+    },
+    {
+      "epoch": 0.8090471607314725,
+      "grad_norm": 1.3284807205200195,
+      "learning_rate": 0.00012961625091534432,
+      "loss": 1.2665,
+      "step": 21015
+    },
+    {
+      "epoch": 0.8092396535129933,
+      "grad_norm": 1.5851751565933228,
+      "learning_rate": 0.000129587366578163,
+      "loss": 1.1016,
+      "step": 21020
+    },
+    {
+      "epoch": 0.809432146294514,
+      "grad_norm": 1.7818470001220703,
+      "learning_rate": 0.00012955847953541303,
+      "loss": 1.1393,
+      "step": 21025
+    },
+    {
+      "epoch": 0.8096246390760347,
+      "grad_norm": 2.6962087154388428,
+      "learning_rate": 0.00012952958978973585,
+      "loss": 1.1494,
+      "step": 21030
+    },
+    {
+      "epoch": 0.8098171318575553,
+      "grad_norm": 1.0526777505874634,
+      "learning_rate": 0.00012950069734377336,
+      "loss": 1.1411,
+      "step": 21035
+    },
+    {
+      "epoch": 0.810009624639076,
+      "grad_norm": 0.8282346129417419,
+      "learning_rate": 0.0001294718022001675,
+      "loss": 1.2306,
+      "step": 21040
+    },
+    {
+      "epoch": 0.8102021174205968,
+      "grad_norm": 0.9357605576515198,
+      "learning_rate": 0.00012944290436156056,
+      "loss": 1.0414,
+      "step": 21045
+    },
+    {
+      "epoch": 0.8103946102021174,
+      "grad_norm": 1.5050805807113647,
+      "learning_rate": 0.00012941400383059508,
+      "loss": 1.0919,
+      "step": 21050
+    },
+    {
+      "epoch": 0.8105871029836381,
+      "grad_norm": 1.8481179475784302,
+      "learning_rate": 0.00012938510060991377,
+      "loss": 1.2084,
+      "step": 21055
+    },
+    {
+      "epoch": 0.8107795957651588,
+      "grad_norm": 1.1937155723571777,
+      "learning_rate": 0.00012935619470215968,
+      "loss": 1.2939,
+      "step": 21060
+    },
+    {
+      "epoch": 0.8109720885466795,
+      "grad_norm": 1.077461838722229,
+      "learning_rate": 0.00012932728610997606,
+      "loss": 1.1172,
+      "step": 21065
+    },
+    {
+      "epoch": 0.8111645813282002,
+      "grad_norm": 1.2674756050109863,
+      "learning_rate": 0.00012929837483600638,
+      "loss": 1.0764,
+      "step": 21070
+    },
+    {
+      "epoch": 0.8113570741097209,
+      "grad_norm": 1.3384912014007568,
+      "learning_rate": 0.00012926946088289443,
+      "loss": 1.0626,
+      "step": 21075
+    },
+    {
+      "epoch": 0.8115495668912416,
+      "grad_norm": 1.9473553895950317,
+      "learning_rate": 0.00012924054425328415,
+      "loss": 0.9904,
+      "step": 21080
+    },
+    {
+      "epoch": 0.8117420596727623,
+      "grad_norm": 0.9565866589546204,
+      "learning_rate": 0.00012921162494981982,
+      "loss": 1.015,
+      "step": 21085
+    },
+    {
+      "epoch": 0.8119345524542829,
+      "grad_norm": 2.1136703491210938,
+      "learning_rate": 0.00012918270297514588,
+      "loss": 1.1105,
+      "step": 21090
+    },
+    {
+      "epoch": 0.8121270452358037,
+      "grad_norm": 1.2925735712051392,
+      "learning_rate": 0.0001291537783319071,
+      "loss": 1.1225,
+      "step": 21095
+    },
+    {
+      "epoch": 0.8123195380173244,
+      "grad_norm": 1.0134514570236206,
+      "learning_rate": 0.00012912485102274836,
+      "loss": 1.1333,
+      "step": 21100
+    },
+    {
+      "epoch": 0.812512030798845,
+      "grad_norm": 1.2157081365585327,
+      "learning_rate": 0.00012909592105031495,
+      "loss": 1.2096,
+      "step": 21105
+    },
+    {
+      "epoch": 0.8127045235803657,
+      "grad_norm": 1.0096231698989868,
+      "learning_rate": 0.00012906698841725234,
+      "loss": 1.1434,
+      "step": 21110
+    },
+    {
+      "epoch": 0.8128970163618864,
+      "grad_norm": 1.3600200414657593,
+      "learning_rate": 0.00012903805312620617,
+      "loss": 1.0407,
+      "step": 21115
+    },
+    {
+      "epoch": 0.8130895091434072,
+      "grad_norm": 1.7842247486114502,
+      "learning_rate": 0.00012900911517982238,
+      "loss": 1.0577,
+      "step": 21120
+    },
+    {
+      "epoch": 0.8132820019249278,
+      "grad_norm": 1.8753024339675903,
+      "learning_rate": 0.0001289801745807472,
+      "loss": 1.1792,
+      "step": 21125
+    },
+    {
+      "epoch": 0.8134744947064485,
+      "grad_norm": 1.000182867050171,
+      "learning_rate": 0.00012895123133162704,
+      "loss": 1.0124,
+      "step": 21130
+    },
+    {
+      "epoch": 0.8136669874879692,
+      "grad_norm": 0.9506773948669434,
+      "learning_rate": 0.00012892228543510856,
+      "loss": 1.2867,
+      "step": 21135
+    },
+    {
+      "epoch": 0.8138594802694898,
+      "grad_norm": 1.084410548210144,
+      "learning_rate": 0.00012889333689383863,
+      "loss": 1.2061,
+      "step": 21140
+    },
+    {
+      "epoch": 0.8140519730510106,
+      "grad_norm": 2.0268046855926514,
+      "learning_rate": 0.00012886438571046448,
+      "loss": 0.9941,
+      "step": 21145
+    },
+    {
+      "epoch": 0.8142444658325313,
+      "grad_norm": 1.1148359775543213,
+      "learning_rate": 0.00012883543188763345,
+      "loss": 1.3397,
+      "step": 21150
+    },
+    {
+      "epoch": 0.814436958614052,
+      "grad_norm": 1.990718126296997,
+      "learning_rate": 0.00012880647542799317,
+      "loss": 1.1802,
+      "step": 21155
+    },
+    {
+      "epoch": 0.8146294513955726,
+      "grad_norm": 0.9613543152809143,
+      "learning_rate": 0.00012877751633419158,
+      "loss": 1.0403,
+      "step": 21160
+    },
+    {
+      "epoch": 0.8148219441770933,
+      "grad_norm": 1.0529913902282715,
+      "learning_rate": 0.00012874855460887674,
+      "loss": 1.0797,
+      "step": 21165
+    },
+    {
+      "epoch": 0.8150144369586141,
+      "grad_norm": 1.083970069885254,
+      "learning_rate": 0.000128719590254697,
+      "loss": 1.0772,
+      "step": 21170
+    },
+    {
+      "epoch": 0.8152069297401348,
+      "grad_norm": 1.39369535446167,
+      "learning_rate": 0.000128690623274301,
+      "loss": 1.1004,
+      "step": 21175
+    },
+    {
+      "epoch": 0.8153994225216554,
+      "grad_norm": 1.5318938493728638,
+      "learning_rate": 0.00012866165367033755,
+      "loss": 1.0723,
+      "step": 21180
+    },
+    {
+      "epoch": 0.8155919153031761,
+      "grad_norm": 2.069422483444214,
+      "learning_rate": 0.00012863268144545575,
+      "loss": 1.1655,
+      "step": 21185
+    },
+    {
+      "epoch": 0.8157844080846969,
+      "grad_norm": 2.672980785369873,
+      "learning_rate": 0.0001286037066023049,
+      "loss": 1.1651,
+      "step": 21190
+    },
+    {
+      "epoch": 0.8159769008662175,
+      "grad_norm": 1.552042007446289,
+      "learning_rate": 0.00012857472914353456,
+      "loss": 1.1443,
+      "step": 21195
+    },
+    {
+      "epoch": 0.8161693936477382,
+      "grad_norm": 1.6215529441833496,
+      "learning_rate": 0.00012854574907179454,
+      "loss": 1.0127,
+      "step": 21200
+    },
+    {
+      "epoch": 0.8163618864292589,
+      "grad_norm": 1.9641019105911255,
+      "learning_rate": 0.00012851676638973486,
+      "loss": 1.1913,
+      "step": 21205
+    },
+    {
+      "epoch": 0.8165543792107796,
+      "grad_norm": 1.0362906455993652,
+      "learning_rate": 0.00012848778110000582,
+      "loss": 1.0131,
+      "step": 21210
+    },
+    {
+      "epoch": 0.8167468719923003,
+      "grad_norm": 1.3487643003463745,
+      "learning_rate": 0.00012845879320525787,
+      "loss": 1.3209,
+      "step": 21215
+    },
+    {
+      "epoch": 0.816939364773821,
+      "grad_norm": 1.2834949493408203,
+      "learning_rate": 0.00012842980270814182,
+      "loss": 0.9396,
+      "step": 21220
+    },
+    {
+      "epoch": 0.8171318575553417,
+      "grad_norm": 1.530375599861145,
+      "learning_rate": 0.00012840080961130864,
+      "loss": 0.9925,
+      "step": 21225
+    },
+    {
+      "epoch": 0.8173243503368623,
+      "grad_norm": 1.6313562393188477,
+      "learning_rate": 0.00012837181391740958,
+      "loss": 1.1684,
+      "step": 21230
+    },
+    {
+      "epoch": 0.817516843118383,
+      "grad_norm": 2.681976556777954,
+      "learning_rate": 0.0001283428156290961,
+      "loss": 1.0749,
+      "step": 21235
+    },
+    {
+      "epoch": 0.8177093358999038,
+      "grad_norm": 1.3075588941574097,
+      "learning_rate": 0.0001283138147490198,
+      "loss": 1.2704,
+      "step": 21240
+    },
+    {
+      "epoch": 0.8179018286814245,
+      "grad_norm": 1.4993854761123657,
+      "learning_rate": 0.0001282848112798328,
+      "loss": 1.1457,
+      "step": 21245
+    },
+    {
+      "epoch": 0.8180943214629451,
+      "grad_norm": 0.991303563117981,
+      "learning_rate": 0.00012825580522418716,
+      "loss": 1.0834,
+      "step": 21250
+    },
+    {
+      "epoch": 0.8182868142444658,
+      "grad_norm": 1.5046969652175903,
+      "learning_rate": 0.0001282267965847353,
+      "loss": 1.1156,
+      "step": 21255
+    },
+    {
+      "epoch": 0.8184793070259865,
+      "grad_norm": 2.711761713027954,
+      "learning_rate": 0.00012819778536412983,
+      "loss": 1.0736,
+      "step": 21260
+    },
+    {
+      "epoch": 0.8186717998075073,
+      "grad_norm": 2.3247745037078857,
+      "learning_rate": 0.00012816877156502374,
+      "loss": 1.0703,
+      "step": 21265
+    },
+    {
+      "epoch": 0.8188642925890279,
+      "grad_norm": 1.8839764595031738,
+      "learning_rate": 0.00012813975519007005,
+      "loss": 1.2301,
+      "step": 21270
+    },
+    {
+      "epoch": 0.8190567853705486,
+      "grad_norm": 1.3224703073501587,
+      "learning_rate": 0.00012811073624192218,
+      "loss": 0.942,
+      "step": 21275
+    },
+    {
+      "epoch": 0.8192492781520693,
+      "grad_norm": 1.2975091934204102,
+      "learning_rate": 0.00012808171472323366,
+      "loss": 1.0815,
+      "step": 21280
+    },
+    {
+      "epoch": 0.8194417709335899,
+      "grad_norm": 1.1744860410690308,
+      "learning_rate": 0.0001280526906366584,
+      "loss": 1.2635,
+      "step": 21285
+    },
+    {
+      "epoch": 0.8196342637151107,
+      "grad_norm": 1.3168641328811646,
+      "learning_rate": 0.0001280236639848504,
+      "loss": 1.1034,
+      "step": 21290
+    },
+    {
+      "epoch": 0.8198267564966314,
+      "grad_norm": 1.7554851770401,
+      "learning_rate": 0.00012799463477046396,
+      "loss": 1.215,
+      "step": 21295
+    },
+    {
+      "epoch": 0.8200192492781521,
+      "grad_norm": 1.458289384841919,
+      "learning_rate": 0.00012796560299615365,
+      "loss": 1.0511,
+      "step": 21300
+    },
+    {
+      "epoch": 0.8202117420596727,
+      "grad_norm": 1.4137495756149292,
+      "learning_rate": 0.00012793656866457416,
+      "loss": 0.9379,
+      "step": 21305
+    },
+    {
+      "epoch": 0.8204042348411934,
+      "grad_norm": 1.4759835004806519,
+      "learning_rate": 0.00012790753177838052,
+      "loss": 1.1523,
+      "step": 21310
+    },
+    {
+      "epoch": 0.8205967276227142,
+      "grad_norm": 1.7477853298187256,
+      "learning_rate": 0.00012787849234022802,
+      "loss": 1.1982,
+      "step": 21315
+    },
+    {
+      "epoch": 0.8207892204042349,
+      "grad_norm": 1.6621637344360352,
+      "learning_rate": 0.00012784945035277206,
+      "loss": 1.1225,
+      "step": 21320
+    },
+    {
+      "epoch": 0.8209817131857555,
+      "grad_norm": 1.2455633878707886,
+      "learning_rate": 0.0001278204058186683,
+      "loss": 1.1079,
+      "step": 21325
+    },
+    {
+      "epoch": 0.8211742059672762,
+      "grad_norm": 2.1116111278533936,
+      "learning_rate": 0.0001277913587405728,
+      "loss": 1.16,
+      "step": 21330
+    },
+    {
+      "epoch": 0.821366698748797,
+      "grad_norm": 1.7187306880950928,
+      "learning_rate": 0.00012776230912114157,
+      "loss": 1.0653,
+      "step": 21335
+    },
+    {
+      "epoch": 0.8215591915303176,
+      "grad_norm": 1.4385313987731934,
+      "learning_rate": 0.00012773325696303114,
+      "loss": 1.1132,
+      "step": 21340
+    },
+    {
+      "epoch": 0.8217516843118383,
+      "grad_norm": 1.1039965152740479,
+      "learning_rate": 0.00012770420226889802,
+      "loss": 1.072,
+      "step": 21345
+    },
+    {
+      "epoch": 0.821944177093359,
+      "grad_norm": 1.4439443349838257,
+      "learning_rate": 0.0001276751450413992,
+      "loss": 1.2279,
+      "step": 21350
+    },
+    {
+      "epoch": 0.8221366698748797,
+      "grad_norm": 1.0179024934768677,
+      "learning_rate": 0.00012764608528319165,
+      "loss": 1.0026,
+      "step": 21355
+    },
+    {
+      "epoch": 0.8223291626564004,
+      "grad_norm": 1.2904207706451416,
+      "learning_rate": 0.00012761702299693274,
+      "loss": 1.0848,
+      "step": 21360
+    },
+    {
+      "epoch": 0.8225216554379211,
+      "grad_norm": 0.8580986261367798,
+      "learning_rate": 0.00012758795818528003,
+      "loss": 1.0943,
+      "step": 21365
+    },
+    {
+      "epoch": 0.8227141482194418,
+      "grad_norm": 1.2557419538497925,
+      "learning_rate": 0.00012755889085089127,
+      "loss": 1.1489,
+      "step": 21370
+    },
+    {
+      "epoch": 0.8229066410009624,
+      "grad_norm": 1.0299915075302124,
+      "learning_rate": 0.00012752982099642455,
+      "loss": 1.0721,
+      "step": 21375
+    },
+    {
+      "epoch": 0.8230991337824831,
+      "grad_norm": 0.9946519136428833,
+      "learning_rate": 0.00012750074862453802,
+      "loss": 1.096,
+      "step": 21380
+    },
+    {
+      "epoch": 0.8232916265640039,
+      "grad_norm": 1.6168420314788818,
+      "learning_rate": 0.00012747167373789023,
+      "loss": 1.275,
+      "step": 21385
+    },
+    {
+      "epoch": 0.8234841193455246,
+      "grad_norm": 1.7367397546768188,
+      "learning_rate": 0.00012744259633913987,
+      "loss": 1.1666,
+      "step": 21390
+    },
+    {
+      "epoch": 0.8236766121270452,
+      "grad_norm": 1.6572107076644897,
+      "learning_rate": 0.00012741351643094583,
+      "loss": 1.0053,
+      "step": 21395
+    },
+    {
+      "epoch": 0.8238691049085659,
+      "grad_norm": 0.7522270083427429,
+      "learning_rate": 0.00012738443401596734,
+      "loss": 1.0007,
+      "step": 21400
+    },
+    {
+      "epoch": 0.8240615976900866,
+      "grad_norm": 1.8706845045089722,
+      "learning_rate": 0.00012735534909686375,
+      "loss": 1.2394,
+      "step": 21405
+    },
+    {
+      "epoch": 0.8242540904716074,
+      "grad_norm": 2.571775197982788,
+      "learning_rate": 0.0001273262616762947,
+      "loss": 1.4034,
+      "step": 21410
+    },
+    {
+      "epoch": 0.824446583253128,
+      "grad_norm": 1.1532241106033325,
+      "learning_rate": 0.00012729717175692006,
+      "loss": 1.095,
+      "step": 21415
+    },
+    {
+      "epoch": 0.8246390760346487,
+      "grad_norm": 1.6400253772735596,
+      "learning_rate": 0.00012726807934139987,
+      "loss": 1.1114,
+      "step": 21420
+    },
+    {
+      "epoch": 0.8248315688161694,
+      "grad_norm": 1.572492003440857,
+      "learning_rate": 0.00012723898443239447,
+      "loss": 1.0996,
+      "step": 21425
+    },
+    {
+      "epoch": 0.82502406159769,
+      "grad_norm": 1.266436219215393,
+      "learning_rate": 0.00012720988703256438,
+      "loss": 1.1825,
+      "step": 21430
+    },
+    {
+      "epoch": 0.8252165543792108,
+      "grad_norm": 1.0934292078018188,
+      "learning_rate": 0.00012718078714457035,
+      "loss": 1.1494,
+      "step": 21435
+    },
+    {
+      "epoch": 0.8254090471607315,
+      "grad_norm": 1.5081112384796143,
+      "learning_rate": 0.0001271516847710734,
+      "loss": 1.2056,
+      "step": 21440
+    },
+    {
+      "epoch": 0.8256015399422522,
+      "grad_norm": 0.9314339756965637,
+      "learning_rate": 0.00012712257991473477,
+      "loss": 1.105,
+      "step": 21445
+    },
+    {
+      "epoch": 0.8257940327237728,
+      "grad_norm": 1.483852744102478,
+      "learning_rate": 0.00012709347257821587,
+      "loss": 1.2253,
+      "step": 21450
+    },
+    {
+      "epoch": 0.8259865255052935,
+      "grad_norm": 0.9682250022888184,
+      "learning_rate": 0.00012706436276417837,
+      "loss": 1.0754,
+      "step": 21455
+    },
+    {
+      "epoch": 0.8261790182868143,
+      "grad_norm": 1.4954203367233276,
+      "learning_rate": 0.00012703525047528418,
+      "loss": 1.1777,
+      "step": 21460
+    },
+    {
+      "epoch": 0.826371511068335,
+      "grad_norm": 1.303179383277893,
+      "learning_rate": 0.00012700613571419543,
+      "loss": 1.1226,
+      "step": 21465
+    },
+    {
+      "epoch": 0.8265640038498556,
+      "grad_norm": 1.0216439962387085,
+      "learning_rate": 0.0001269770184835745,
+      "loss": 1.0702,
+      "step": 21470
+    },
+    {
+      "epoch": 0.8267564966313763,
+      "grad_norm": 1.5021692514419556,
+      "learning_rate": 0.00012694789878608386,
+      "loss": 1.1696,
+      "step": 21475
+    },
+    {
+      "epoch": 0.826948989412897,
+      "grad_norm": 1.1756644248962402,
+      "learning_rate": 0.00012691877662438642,
+      "loss": 1.1441,
+      "step": 21480
+    },
+    {
+      "epoch": 0.8271414821944177,
+      "grad_norm": 1.7375133037567139,
+      "learning_rate": 0.0001268896520011452,
+      "loss": 1.0854,
+      "step": 21485
+    },
+    {
+      "epoch": 0.8273339749759384,
+      "grad_norm": 0.894169270992279,
+      "learning_rate": 0.00012686052491902344,
+      "loss": 1.0767,
+      "step": 21490
+    },
+    {
+      "epoch": 0.8275264677574591,
+      "grad_norm": 2.3879435062408447,
+      "learning_rate": 0.00012683139538068457,
+      "loss": 1.0683,
+      "step": 21495
+    },
+    {
+      "epoch": 0.8277189605389798,
+      "grad_norm": 0.7615637183189392,
+      "learning_rate": 0.00012680226338879238,
+      "loss": 1.0482,
+      "step": 21500
+    },
+    {
+      "epoch": 0.8279114533205005,
+      "grad_norm": 1.056795597076416,
+      "learning_rate": 0.0001267731289460107,
+      "loss": 1.0303,
+      "step": 21505
+    },
+    {
+      "epoch": 0.8281039461020212,
+      "grad_norm": 1.099278211593628,
+      "learning_rate": 0.00012674399205500375,
+      "loss": 1.1923,
+      "step": 21510
+    },
+    {
+      "epoch": 0.8282964388835419,
+      "grad_norm": 1.1515510082244873,
+      "learning_rate": 0.0001267148527184359,
+      "loss": 1.1697,
+      "step": 21515
+    },
+    {
+      "epoch": 0.8284889316650625,
+      "grad_norm": 1.8237887620925903,
+      "learning_rate": 0.00012668571093897175,
+      "loss": 1.0131,
+      "step": 21520
+    },
+    {
+      "epoch": 0.8286814244465832,
+      "grad_norm": 1.1378989219665527,
+      "learning_rate": 0.0001266565667192761,
+      "loss": 1.1541,
+      "step": 21525
+    },
+    {
+      "epoch": 0.828873917228104,
+      "grad_norm": 1.1044882535934448,
+      "learning_rate": 0.000126627420062014,
+      "loss": 0.9193,
+      "step": 21530
+    },
+    {
+      "epoch": 0.8290664100096247,
+      "grad_norm": 1.0803862810134888,
+      "learning_rate": 0.00012659827096985073,
+      "loss": 1.0778,
+      "step": 21535
+    },
+    {
+      "epoch": 0.8292589027911453,
+      "grad_norm": 1.004387617111206,
+      "learning_rate": 0.00012656911944545178,
+      "loss": 1.0134,
+      "step": 21540
+    },
+    {
+      "epoch": 0.829451395572666,
+      "grad_norm": 1.0931042432785034,
+      "learning_rate": 0.00012653996549148285,
+      "loss": 1.0738,
+      "step": 21545
+    },
+    {
+      "epoch": 0.8296438883541867,
+      "grad_norm": 1.053134560585022,
+      "learning_rate": 0.00012651080911060993,
+      "loss": 0.9885,
+      "step": 21550
+    },
+    {
+      "epoch": 0.8298363811357075,
+      "grad_norm": 1.2585800886154175,
+      "learning_rate": 0.00012648165030549908,
+      "loss": 1.0844,
+      "step": 21555
+    },
+    {
+      "epoch": 0.8300288739172281,
+      "grad_norm": 1.301685094833374,
+      "learning_rate": 0.00012645248907881677,
+      "loss": 1.1168,
+      "step": 21560
+    },
+    {
+      "epoch": 0.8302213666987488,
+      "grad_norm": 2.356616973876953,
+      "learning_rate": 0.00012642332543322954,
+      "loss": 1.3397,
+      "step": 21565
+    },
+    {
+      "epoch": 0.8304138594802695,
+      "grad_norm": 1.0727852582931519,
+      "learning_rate": 0.00012639415937140428,
+      "loss": 1.1311,
+      "step": 21570
+    },
+    {
+      "epoch": 0.8306063522617901,
+      "grad_norm": 1.3422895669937134,
+      "learning_rate": 0.00012636499089600794,
+      "loss": 1.2474,
+      "step": 21575
+    },
+    {
+      "epoch": 0.8307988450433109,
+      "grad_norm": 1.1913752555847168,
+      "learning_rate": 0.00012633582000970788,
+      "loss": 1.1439,
+      "step": 21580
+    },
+    {
+      "epoch": 0.8309913378248316,
+      "grad_norm": 1.068538784980774,
+      "learning_rate": 0.0001263066467151715,
+      "loss": 1.0762,
+      "step": 21585
+    },
+    {
+      "epoch": 0.8311838306063523,
+      "grad_norm": 1.4798871278762817,
+      "learning_rate": 0.00012627747101506655,
+      "loss": 1.1836,
+      "step": 21590
+    },
+    {
+      "epoch": 0.8313763233878729,
+      "grad_norm": 0.9586134552955627,
+      "learning_rate": 0.00012624829291206098,
+      "loss": 1.154,
+      "step": 21595
+    },
+    {
+      "epoch": 0.8315688161693936,
+      "grad_norm": 1.3402488231658936,
+      "learning_rate": 0.00012621911240882287,
+      "loss": 1.0223,
+      "step": 21600
+    },
+    {
+      "epoch": 0.8317613089509144,
+      "grad_norm": 0.896938145160675,
+      "learning_rate": 0.00012618992950802062,
+      "loss": 1.0489,
+      "step": 21605
+    },
+    {
+      "epoch": 0.831953801732435,
+      "grad_norm": 1.466774821281433,
+      "learning_rate": 0.00012616074421232281,
+      "loss": 1.0963,
+      "step": 21610
+    },
+    {
+      "epoch": 0.8321462945139557,
+      "grad_norm": 1.714902639389038,
+      "learning_rate": 0.00012613155652439826,
+      "loss": 1.1946,
+      "step": 21615
+    },
+    {
+      "epoch": 0.8323387872954764,
+      "grad_norm": 1.3679184913635254,
+      "learning_rate": 0.00012610236644691592,
+      "loss": 1.1356,
+      "step": 21620
+    },
+    {
+      "epoch": 0.8325312800769971,
+      "grad_norm": 2.8363022804260254,
+      "learning_rate": 0.00012607317398254515,
+      "loss": 1.2324,
+      "step": 21625
+    },
+    {
+      "epoch": 0.8327237728585178,
+      "grad_norm": 1.7993220090866089,
+      "learning_rate": 0.00012604397913395528,
+      "loss": 1.0687,
+      "step": 21630
+    },
+    {
+      "epoch": 0.8329162656400385,
+      "grad_norm": 1.42396879196167,
+      "learning_rate": 0.00012601478190381608,
+      "loss": 1.1426,
+      "step": 21635
+    },
+    {
+      "epoch": 0.8331087584215592,
+      "grad_norm": 0.9980899691581726,
+      "learning_rate": 0.0001259855822947974,
+      "loss": 1.0239,
+      "step": 21640
+    },
+    {
+      "epoch": 0.8333012512030799,
+      "grad_norm": 1.641063928604126,
+      "learning_rate": 0.00012595638030956936,
+      "loss": 1.1006,
+      "step": 21645
+    },
+    {
+      "epoch": 0.8334937439846006,
+      "grad_norm": 0.7932726740837097,
+      "learning_rate": 0.00012592717595080226,
+      "loss": 0.9438,
+      "step": 21650
+    },
+    {
+      "epoch": 0.8336862367661213,
+      "grad_norm": 0.988121747970581,
+      "learning_rate": 0.0001258979692211667,
+      "loss": 1.0918,
+      "step": 21655
+    },
+    {
+      "epoch": 0.833878729547642,
+      "grad_norm": 2.024733066558838,
+      "learning_rate": 0.0001258687601233334,
+      "loss": 1.1982,
+      "step": 21660
+    },
+    {
+      "epoch": 0.8340712223291626,
+      "grad_norm": 1.4302330017089844,
+      "learning_rate": 0.00012583954865997337,
+      "loss": 1.0746,
+      "step": 21665
+    },
+    {
+      "epoch": 0.8342637151106833,
+      "grad_norm": 1.084583044052124,
+      "learning_rate": 0.00012581033483375777,
+      "loss": 1.173,
+      "step": 21670
+    },
+    {
+      "epoch": 0.8344562078922041,
+      "grad_norm": 1.3885962963104248,
+      "learning_rate": 0.00012578111864735802,
+      "loss": 1.0817,
+      "step": 21675
+    },
+    {
+      "epoch": 0.8346487006737248,
+      "grad_norm": 1.2596712112426758,
+      "learning_rate": 0.00012575190010344578,
+      "loss": 1.0657,
+      "step": 21680
+    },
+    {
+      "epoch": 0.8348411934552454,
+      "grad_norm": 1.376128911972046,
+      "learning_rate": 0.00012572267920469283,
+      "loss": 1.0961,
+      "step": 21685
+    },
+    {
+      "epoch": 0.8350336862367661,
+      "grad_norm": 1.203218936920166,
+      "learning_rate": 0.00012569345595377128,
+      "loss": 1.2123,
+      "step": 21690
+    },
+    {
+      "epoch": 0.8352261790182868,
+      "grad_norm": 2.2045395374298096,
+      "learning_rate": 0.00012566423035335338,
+      "loss": 1.2024,
+      "step": 21695
+    },
+    {
+      "epoch": 0.8354186717998076,
+      "grad_norm": 2.305617094039917,
+      "learning_rate": 0.00012563500240611166,
+      "loss": 1.027,
+      "step": 21700
+    },
+    {
+      "epoch": 0.8356111645813282,
+      "grad_norm": 0.8593419194221497,
+      "learning_rate": 0.00012560577211471875,
+      "loss": 1.0436,
+      "step": 21705
+    },
+    {
+      "epoch": 0.8358036573628489,
+      "grad_norm": 1.3252314329147339,
+      "learning_rate": 0.00012557653948184761,
+      "loss": 0.9099,
+      "step": 21710
+    },
+    {
+      "epoch": 0.8359961501443696,
+      "grad_norm": 1.569014072418213,
+      "learning_rate": 0.0001255473045101714,
+      "loss": 1.1202,
+      "step": 21715
+    },
+    {
+      "epoch": 0.8361886429258902,
+      "grad_norm": 0.9022922515869141,
+      "learning_rate": 0.00012551806720236338,
+      "loss": 1.1666,
+      "step": 21720
+    },
+    {
+      "epoch": 0.836381135707411,
+      "grad_norm": 1.5258845090866089,
+      "learning_rate": 0.00012548882756109717,
+      "loss": 1.1072,
+      "step": 21725
+    },
+    {
+      "epoch": 0.8365736284889317,
+      "grad_norm": 1.4463955163955688,
+      "learning_rate": 0.0001254595855890465,
+      "loss": 1.0444,
+      "step": 21730
+    },
+    {
+      "epoch": 0.8367661212704524,
+      "grad_norm": 1.6336601972579956,
+      "learning_rate": 0.00012543034128888544,
+      "loss": 1.2973,
+      "step": 21735
+    },
+    {
+      "epoch": 0.836958614051973,
+      "grad_norm": 1.2576451301574707,
+      "learning_rate": 0.0001254010946632881,
+      "loss": 1.0968,
+      "step": 21740
+    },
+    {
+      "epoch": 0.8371511068334937,
+      "grad_norm": 1.2530792951583862,
+      "learning_rate": 0.0001253718457149289,
+      "loss": 0.9215,
+      "step": 21745
+    },
+    {
+      "epoch": 0.8373435996150145,
+      "grad_norm": 1.3370553255081177,
+      "learning_rate": 0.0001253425944464825,
+      "loss": 1.1147,
+      "step": 21750
+    },
+    {
+      "epoch": 0.8375360923965351,
+      "grad_norm": 1.5159647464752197,
+      "learning_rate": 0.00012531334086062374,
+      "loss": 1.1326,
+      "step": 21755
+    },
+    {
+      "epoch": 0.8377285851780558,
+      "grad_norm": 1.4643933773040771,
+      "learning_rate": 0.0001252840849600276,
+      "loss": 0.9886,
+      "step": 21760
+    },
+    {
+      "epoch": 0.8379210779595765,
+      "grad_norm": 1.5805963277816772,
+      "learning_rate": 0.00012525482674736942,
+      "loss": 1.1807,
+      "step": 21765
+    },
+    {
+      "epoch": 0.8381135707410972,
+      "grad_norm": 1.3167610168457031,
+      "learning_rate": 0.00012522556622532462,
+      "loss": 0.93,
+      "step": 21770
+    },
+    {
+      "epoch": 0.8383060635226179,
+      "grad_norm": 1.4402116537094116,
+      "learning_rate": 0.00012519630339656888,
+      "loss": 1.1223,
+      "step": 21775
+    },
+    {
+      "epoch": 0.8384985563041386,
+      "grad_norm": 1.5734583139419556,
+      "learning_rate": 0.00012516703826377814,
+      "loss": 1.1476,
+      "step": 21780
+    },
+    {
+      "epoch": 0.8386910490856593,
+      "grad_norm": 1.581139326095581,
+      "learning_rate": 0.00012513777082962842,
+      "loss": 1.3,
+      "step": 21785
+    },
+    {
+      "epoch": 0.83888354186718,
+      "grad_norm": 1.0609793663024902,
+      "learning_rate": 0.0001251085010967961,
+      "loss": 1.1791,
+      "step": 21790
+    },
+    {
+      "epoch": 0.8390760346487006,
+      "grad_norm": 1.130085825920105,
+      "learning_rate": 0.00012507922906795772,
+      "loss": 1.0816,
+      "step": 21795
+    },
+    {
+      "epoch": 0.8392685274302214,
+      "grad_norm": 2.4827651977539062,
+      "learning_rate": 0.00012504995474578993,
+      "loss": 1.1677,
+      "step": 21800
+    },
+    {
+      "epoch": 0.8394610202117421,
+      "grad_norm": 1.187456727027893,
+      "learning_rate": 0.00012502067813296972,
+      "loss": 0.9263,
+      "step": 21805
+    },
+    {
+      "epoch": 0.8396535129932627,
+      "grad_norm": 1.375260353088379,
+      "learning_rate": 0.00012499139923217425,
+      "loss": 1.1236,
+      "step": 21810
+    },
+    {
+      "epoch": 0.8398460057747834,
+      "grad_norm": 1.0707255601882935,
+      "learning_rate": 0.00012496211804608089,
+      "loss": 1.0672,
+      "step": 21815
+    },
+    {
+      "epoch": 0.8400384985563042,
+      "grad_norm": 1.7111786603927612,
+      "learning_rate": 0.00012493283457736716,
+      "loss": 1.0458,
+      "step": 21820
+    },
+    {
+      "epoch": 0.8402309913378249,
+      "grad_norm": 0.9973586201667786,
+      "learning_rate": 0.0001249035488287109,
+      "loss": 0.9639,
+      "step": 21825
+    },
+    {
+      "epoch": 0.8404234841193455,
+      "grad_norm": 1.5254673957824707,
+      "learning_rate": 0.00012487426080279006,
+      "loss": 1.2324,
+      "step": 21830
+    },
+    {
+      "epoch": 0.8406159769008662,
+      "grad_norm": 1.4964022636413574,
+      "learning_rate": 0.00012484497050228285,
+      "loss": 1.1236,
+      "step": 21835
+    },
+    {
+      "epoch": 0.8408084696823869,
+      "grad_norm": 1.157464623451233,
+      "learning_rate": 0.00012481567792986767,
+      "loss": 1.0155,
+      "step": 21840
+    },
+    {
+      "epoch": 0.8410009624639077,
+      "grad_norm": 1.8712735176086426,
+      "learning_rate": 0.0001247863830882231,
+      "loss": 1.0772,
+      "step": 21845
+    },
+    {
+      "epoch": 0.8411934552454283,
+      "grad_norm": 1.2394201755523682,
+      "learning_rate": 0.00012475708598002805,
+      "loss": 1.2253,
+      "step": 21850
+    },
+    {
+      "epoch": 0.841385948026949,
+      "grad_norm": 1.3483343124389648,
+      "learning_rate": 0.00012472778660796145,
+      "loss": 1.0598,
+      "step": 21855
+    },
+    {
+      "epoch": 0.8415784408084697,
+      "grad_norm": 1.1623843908309937,
+      "learning_rate": 0.0001246984849747026,
+      "loss": 1.0261,
+      "step": 21860
+    },
+    {
+      "epoch": 0.8417709335899903,
+      "grad_norm": 5.310647964477539,
+      "learning_rate": 0.0001246691810829309,
+      "loss": 1.094,
+      "step": 21865
+    },
+    {
+      "epoch": 0.8419634263715111,
+      "grad_norm": 1.5004624128341675,
+      "learning_rate": 0.000124639874935326,
+      "loss": 1.1192,
+      "step": 21870
+    },
+    {
+      "epoch": 0.8421559191530318,
+      "grad_norm": 1.1160995960235596,
+      "learning_rate": 0.00012461056653456775,
+      "loss": 1.1409,
+      "step": 21875
+    },
+    {
+      "epoch": 0.8423484119345525,
+      "grad_norm": 1.3328487873077393,
+      "learning_rate": 0.0001245812558833362,
+      "loss": 0.7442,
+      "step": 21880
+    },
+    {
+      "epoch": 0.8425409047160731,
+      "grad_norm": 1.2014681100845337,
+      "learning_rate": 0.0001245519429843117,
+      "loss": 1.0333,
+      "step": 21885
+    },
+    {
+      "epoch": 0.8427333974975938,
+      "grad_norm": 0.916928231716156,
+      "learning_rate": 0.00012452262784017464,
+      "loss": 1.0076,
+      "step": 21890
+    },
+    {
+      "epoch": 0.8429258902791146,
+      "grad_norm": 1.1074447631835938,
+      "learning_rate": 0.0001244933104536057,
+      "loss": 1.1731,
+      "step": 21895
+    },
+    {
+      "epoch": 0.8431183830606352,
+      "grad_norm": 1.059792160987854,
+      "learning_rate": 0.00012446399082728578,
+      "loss": 1.0673,
+      "step": 21900
+    },
+    {
+      "epoch": 0.8433108758421559,
+      "grad_norm": 0.8817097544670105,
+      "learning_rate": 0.00012443466896389595,
+      "loss": 0.9764,
+      "step": 21905
+    },
+    {
+      "epoch": 0.8435033686236766,
+      "grad_norm": 1.0615062713623047,
+      "learning_rate": 0.0001244053448661175,
+      "loss": 1.162,
+      "step": 21910
+    },
+    {
+      "epoch": 0.8436958614051973,
+      "grad_norm": 1.173344612121582,
+      "learning_rate": 0.0001243760185366319,
+      "loss": 0.997,
+      "step": 21915
+    },
+    {
+      "epoch": 0.843888354186718,
+      "grad_norm": 1.5543683767318726,
+      "learning_rate": 0.00012434668997812092,
+      "loss": 1.1386,
+      "step": 21920
+    },
+    {
+      "epoch": 0.8440808469682387,
+      "grad_norm": 1.753692388534546,
+      "learning_rate": 0.00012431735919326645,
+      "loss": 1.24,
+      "step": 21925
+    },
+    {
+      "epoch": 0.8442733397497594,
+      "grad_norm": 1.3150075674057007,
+      "learning_rate": 0.00012428802618475053,
+      "loss": 1.0783,
+      "step": 21930
+    },
+    {
+      "epoch": 0.84446583253128,
+      "grad_norm": 1.4009346961975098,
+      "learning_rate": 0.00012425869095525552,
+      "loss": 0.9953,
+      "step": 21935
+    },
+    {
+      "epoch": 0.8446583253128007,
+      "grad_norm": 1.7203071117401123,
+      "learning_rate": 0.00012422935350746395,
+      "loss": 0.9115,
+      "step": 21940
+    },
+    {
+      "epoch": 0.8448508180943215,
+      "grad_norm": 1.1231714487075806,
+      "learning_rate": 0.00012420001384405845,
+      "loss": 1.1767,
+      "step": 21945
+    },
+    {
+      "epoch": 0.8450433108758422,
+      "grad_norm": 1.0938594341278076,
+      "learning_rate": 0.00012417067196772202,
+      "loss": 1.0433,
+      "step": 21950
+    },
+    {
+      "epoch": 0.8452358036573628,
+      "grad_norm": 1.9310616254806519,
+      "learning_rate": 0.00012414132788113777,
+      "loss": 1.0809,
+      "step": 21955
+    },
+    {
+      "epoch": 0.8454282964388835,
+      "grad_norm": 1.4712507724761963,
+      "learning_rate": 0.00012411198158698898,
+      "loss": 1.0663,
+      "step": 21960
+    },
+    {
+      "epoch": 0.8456207892204043,
+      "grad_norm": 0.9529402256011963,
+      "learning_rate": 0.00012408263308795922,
+      "loss": 1.0486,
+      "step": 21965
+    },
+    {
+      "epoch": 0.845813282001925,
+      "grad_norm": 0.8903969526290894,
+      "learning_rate": 0.0001240532823867322,
+      "loss": 1.0795,
+      "step": 21970
+    },
+    {
+      "epoch": 0.8460057747834456,
+      "grad_norm": 1.0439949035644531,
+      "learning_rate": 0.00012402392948599185,
+      "loss": 1.0456,
+      "step": 21975
+    },
+    {
+      "epoch": 0.8461982675649663,
+      "grad_norm": 1.5263192653656006,
+      "learning_rate": 0.00012399457438842226,
+      "loss": 1.004,
+      "step": 21980
+    },
+    {
+      "epoch": 0.846390760346487,
+      "grad_norm": 1.3312830924987793,
+      "learning_rate": 0.00012396521709670782,
+      "loss": 1.1062,
+      "step": 21985
+    },
+    {
+      "epoch": 0.8465832531280078,
+      "grad_norm": 1.6657592058181763,
+      "learning_rate": 0.00012393585761353302,
+      "loss": 1.019,
+      "step": 21990
+    },
+    {
+      "epoch": 0.8467757459095284,
+      "grad_norm": 0.9013869166374207,
+      "learning_rate": 0.0001239064959415826,
+      "loss": 1.3499,
+      "step": 21995
+    },
+    {
+      "epoch": 0.8469682386910491,
+      "grad_norm": 0.9990372061729431,
+      "learning_rate": 0.0001238771320835415,
+      "loss": 1.064,
+      "step": 22000
+    },
+    {
+      "epoch": 0.8471607314725698,
+      "grad_norm": 1.070052146911621,
+      "learning_rate": 0.00012384776604209484,
+      "loss": 1.0608,
+      "step": 22005
+    },
+    {
+      "epoch": 0.8473532242540904,
+      "grad_norm": 2.112750291824341,
+      "learning_rate": 0.00012381839781992797,
+      "loss": 1.1134,
+      "step": 22010
+    },
+    {
+      "epoch": 0.8475457170356112,
+      "grad_norm": 1.3599921464920044,
+      "learning_rate": 0.00012378902741972636,
+      "loss": 0.964,
+      "step": 22015
+    },
+    {
+      "epoch": 0.8477382098171319,
+      "grad_norm": 1.8712466955184937,
+      "learning_rate": 0.00012375965484417582,
+      "loss": 1.129,
+      "step": 22020
+    },
+    {
+      "epoch": 0.8479307025986526,
+      "grad_norm": 2.496819019317627,
+      "learning_rate": 0.00012373028009596222,
+      "loss": 1.1981,
+      "step": 22025
+    },
+    {
+      "epoch": 0.8481231953801732,
+      "grad_norm": 1.8566465377807617,
+      "learning_rate": 0.00012370090317777172,
+      "loss": 1.1229,
+      "step": 22030
+    },
+    {
+      "epoch": 0.8483156881616939,
+      "grad_norm": 1.0162265300750732,
+      "learning_rate": 0.0001236715240922906,
+      "loss": 1.2759,
+      "step": 22035
+    },
+    {
+      "epoch": 0.8485081809432147,
+      "grad_norm": 1.4011484384536743,
+      "learning_rate": 0.0001236421428422054,
+      "loss": 1.0725,
+      "step": 22040
+    },
+    {
+      "epoch": 0.8487006737247353,
+      "grad_norm": 0.9119108319282532,
+      "learning_rate": 0.0001236127594302029,
+      "loss": 1.0087,
+      "step": 22045
+    },
+    {
+      "epoch": 0.848893166506256,
+      "grad_norm": 1.4366172552108765,
+      "learning_rate": 0.00012358337385896993,
+      "loss": 0.9379,
+      "step": 22050
+    },
+    {
+      "epoch": 0.8490856592877767,
+      "grad_norm": 1.162611961364746,
+      "learning_rate": 0.00012355398613119367,
+      "loss": 1.0068,
+      "step": 22055
+    },
+    {
+      "epoch": 0.8492781520692974,
+      "grad_norm": 1.5164228677749634,
+      "learning_rate": 0.0001235245962495614,
+      "loss": 1.2239,
+      "step": 22060
+    },
+    {
+      "epoch": 0.8494706448508181,
+      "grad_norm": 1.0570234060287476,
+      "learning_rate": 0.00012349520421676066,
+      "loss": 1.0332,
+      "step": 22065
+    },
+    {
+      "epoch": 0.8496631376323388,
+      "grad_norm": 1.4047578573226929,
+      "learning_rate": 0.0001234658100354791,
+      "loss": 1.051,
+      "step": 22070
+    },
+    {
+      "epoch": 0.8498556304138595,
+      "grad_norm": 1.317559003829956,
+      "learning_rate": 0.00012343641370840465,
+      "loss": 1.0702,
+      "step": 22075
+    },
+    {
+      "epoch": 0.8500481231953801,
+      "grad_norm": 1.0683611631393433,
+      "learning_rate": 0.0001234070152382255,
+      "loss": 0.9898,
+      "step": 22080
+    },
+    {
+      "epoch": 0.8502406159769008,
+      "grad_norm": 0.960689902305603,
+      "learning_rate": 0.00012337761462762978,
+      "loss": 1.377,
+      "step": 22085
+    },
+    {
+      "epoch": 0.8504331087584216,
+      "grad_norm": 1.2345051765441895,
+      "learning_rate": 0.0001233482118793061,
+      "loss": 1.0417,
+      "step": 22090
+    },
+    {
+      "epoch": 0.8506256015399423,
+      "grad_norm": 1.9396436214447021,
+      "learning_rate": 0.0001233188069959431,
+      "loss": 1.204,
+      "step": 22095
+    },
+    {
+      "epoch": 0.8508180943214629,
+      "grad_norm": 1.8735712766647339,
+      "learning_rate": 0.0001232893999802297,
+      "loss": 0.9903,
+      "step": 22100
+    },
+    {
+      "epoch": 0.8510105871029836,
+      "grad_norm": 2.0837035179138184,
+      "learning_rate": 0.00012325999083485494,
+      "loss": 1.01,
+      "step": 22105
+    },
+    {
+      "epoch": 0.8512030798845043,
+      "grad_norm": 1.9637258052825928,
+      "learning_rate": 0.00012323057956250806,
+      "loss": 1.1652,
+      "step": 22110
+    },
+    {
+      "epoch": 0.8513955726660251,
+      "grad_norm": 1.3333475589752197,
+      "learning_rate": 0.0001232011661658786,
+      "loss": 0.9568,
+      "step": 22115
+    },
+    {
+      "epoch": 0.8515880654475457,
+      "grad_norm": 1.8034814596176147,
+      "learning_rate": 0.0001231717506476562,
+      "loss": 1.1863,
+      "step": 22120
+    },
+    {
+      "epoch": 0.8517805582290664,
+      "grad_norm": 0.7985187768936157,
+      "learning_rate": 0.00012314233301053068,
+      "loss": 0.9462,
+      "step": 22125
+    },
+    {
+      "epoch": 0.8519730510105871,
+      "grad_norm": 1.7575490474700928,
+      "learning_rate": 0.00012311291325719213,
+      "loss": 1.1151,
+      "step": 22130
+    },
+    {
+      "epoch": 0.8521655437921078,
+      "grad_norm": 1.058147668838501,
+      "learning_rate": 0.00012308349139033076,
+      "loss": 1.071,
+      "step": 22135
+    },
+    {
+      "epoch": 0.8523580365736285,
+      "grad_norm": 1.0178147554397583,
+      "learning_rate": 0.00012305406741263701,
+      "loss": 1.2246,
+      "step": 22140
+    },
+    {
+      "epoch": 0.8525505293551492,
+      "grad_norm": 1.0183746814727783,
+      "learning_rate": 0.0001230246413268015,
+      "loss": 1.076,
+      "step": 22145
+    },
+    {
+      "epoch": 0.8527430221366699,
+      "grad_norm": 1.4708818197250366,
+      "learning_rate": 0.0001229952131355151,
+      "loss": 1.2221,
+      "step": 22150
+    },
+    {
+      "epoch": 0.8529355149181905,
+      "grad_norm": 1.1673319339752197,
+      "learning_rate": 0.00012296578284146879,
+      "loss": 1.0985,
+      "step": 22155
+    },
+    {
+      "epoch": 0.8531280076997113,
+      "grad_norm": 1.2929120063781738,
+      "learning_rate": 0.00012293635044735373,
+      "loss": 1.1093,
+      "step": 22160
+    },
+    {
+      "epoch": 0.853320500481232,
+      "grad_norm": 1.097374677658081,
+      "learning_rate": 0.0001229069159558614,
+      "loss": 1.1361,
+      "step": 22165
+    },
+    {
+      "epoch": 0.8535129932627527,
+      "grad_norm": 1.5961623191833496,
+      "learning_rate": 0.00012287747936968335,
+      "loss": 1.0493,
+      "step": 22170
+    },
+    {
+      "epoch": 0.8537054860442733,
+      "grad_norm": 1.378281593322754,
+      "learning_rate": 0.00012284804069151133,
+      "loss": 1.0436,
+      "step": 22175
+    },
+    {
+      "epoch": 0.853897978825794,
+      "grad_norm": 0.4067041277885437,
+      "learning_rate": 0.00012281859992403736,
+      "loss": 1.0238,
+      "step": 22180
+    },
+    {
+      "epoch": 0.8540904716073148,
+      "grad_norm": 1.24373197555542,
+      "learning_rate": 0.00012278915706995358,
+      "loss": 1.0049,
+      "step": 22185
+    },
+    {
+      "epoch": 0.8542829643888354,
+      "grad_norm": 2.2100603580474854,
+      "learning_rate": 0.00012275971213195237,
+      "loss": 1.1251,
+      "step": 22190
+    },
+    {
+      "epoch": 0.8544754571703561,
+      "grad_norm": 1.0543677806854248,
+      "learning_rate": 0.00012273026511272622,
+      "loss": 1.0744,
+      "step": 22195
+    },
+    {
+      "epoch": 0.8546679499518768,
+      "grad_norm": 1.5290037393569946,
+      "learning_rate": 0.00012270081601496792,
+      "loss": 1.1105,
+      "step": 22200
+    },
+    {
+      "epoch": 0.8548604427333975,
+      "grad_norm": 1.0767719745635986,
+      "learning_rate": 0.00012267136484137042,
+      "loss": 1.0246,
+      "step": 22205
+    },
+    {
+      "epoch": 0.8550529355149182,
+      "grad_norm": 1.285439133644104,
+      "learning_rate": 0.00012264191159462674,
+      "loss": 0.9625,
+      "step": 22210
+    },
+    {
+      "epoch": 0.8552454282964389,
+      "grad_norm": 1.0118284225463867,
+      "learning_rate": 0.00012261245627743022,
+      "loss": 0.8992,
+      "step": 22215
+    },
+    {
+      "epoch": 0.8554379210779596,
+      "grad_norm": 1.4398612976074219,
+      "learning_rate": 0.0001225829988924744,
+      "loss": 1.2539,
+      "step": 22220
+    },
+    {
+      "epoch": 0.8556304138594802,
+      "grad_norm": 0.7380401492118835,
+      "learning_rate": 0.00012255353944245294,
+      "loss": 0.9303,
+      "step": 22225
+    },
+    {
+      "epoch": 0.8558229066410009,
+      "grad_norm": 1.2834627628326416,
+      "learning_rate": 0.00012252407793005964,
+      "loss": 1.0884,
+      "step": 22230
+    },
+    {
+      "epoch": 0.8560153994225217,
+      "grad_norm": 1.7945185899734497,
+      "learning_rate": 0.00012249461435798867,
+      "loss": 1.1546,
+      "step": 22235
+    },
+    {
+      "epoch": 0.8562078922040424,
+      "grad_norm": 1.4909627437591553,
+      "learning_rate": 0.00012246514872893424,
+      "loss": 1.097,
+      "step": 22240
+    },
+    {
+      "epoch": 0.856400384985563,
+      "grad_norm": 1.3793891668319702,
+      "learning_rate": 0.00012243568104559075,
+      "loss": 1.0008,
+      "step": 22245
+    },
+    {
+      "epoch": 0.8565928777670837,
+      "grad_norm": 1.5637750625610352,
+      "learning_rate": 0.00012240621131065287,
+      "loss": 1.0054,
+      "step": 22250
+    },
+    {
+      "epoch": 0.8567853705486044,
+      "grad_norm": 1.0803687572479248,
+      "learning_rate": 0.00012237673952681538,
+      "loss": 1.0349,
+      "step": 22255
+    },
+    {
+      "epoch": 0.8569778633301252,
+      "grad_norm": 1.220694661140442,
+      "learning_rate": 0.00012234726569677328,
+      "loss": 1.0886,
+      "step": 22260
+    },
+    {
+      "epoch": 0.8571703561116458,
+      "grad_norm": 1.5394152402877808,
+      "learning_rate": 0.0001223177898232218,
+      "loss": 1.1316,
+      "step": 22265
+    },
+    {
+      "epoch": 0.8573628488931665,
+      "grad_norm": 1.927856206893921,
+      "learning_rate": 0.00012228831190885627,
+      "loss": 1.1063,
+      "step": 22270
+    },
+    {
+      "epoch": 0.8575553416746872,
+      "grad_norm": 2.0876247882843018,
+      "learning_rate": 0.00012225883195637224,
+      "loss": 1.0356,
+      "step": 22275
+    },
+    {
+      "epoch": 0.857747834456208,
+      "grad_norm": 1.3493059873580933,
+      "learning_rate": 0.00012222934996846551,
+      "loss": 1.191,
+      "step": 22280
+    },
+    {
+      "epoch": 0.8579403272377286,
+      "grad_norm": 2.021245002746582,
+      "learning_rate": 0.00012219986594783194,
+      "loss": 1.2042,
+      "step": 22285
+    },
+    {
+      "epoch": 0.8581328200192493,
+      "grad_norm": 0.9374911189079285,
+      "learning_rate": 0.00012217037989716774,
+      "loss": 1.1407,
+      "step": 22290
+    },
+    {
+      "epoch": 0.85832531280077,
+      "grad_norm": 1.2450497150421143,
+      "learning_rate": 0.00012214089181916915,
+      "loss": 1.0325,
+      "step": 22295
+    },
+    {
+      "epoch": 0.8585178055822906,
+      "grad_norm": 1.8689814805984497,
+      "learning_rate": 0.00012211140171653265,
+      "loss": 1.086,
+      "step": 22300
+    },
+    {
+      "epoch": 0.8587102983638114,
+      "grad_norm": 1.4810429811477661,
+      "learning_rate": 0.00012208190959195496,
+      "loss": 1.1968,
+      "step": 22305
+    },
+    {
+      "epoch": 0.8589027911453321,
+      "grad_norm": 0.8686331510543823,
+      "learning_rate": 0.00012205241544813289,
+      "loss": 1.1492,
+      "step": 22310
+    },
+    {
+      "epoch": 0.8590952839268527,
+      "grad_norm": 1.288923740386963,
+      "learning_rate": 0.00012202291928776355,
+      "loss": 1.1514,
+      "step": 22315
+    },
+    {
+      "epoch": 0.8592877767083734,
+      "grad_norm": 1.5871915817260742,
+      "learning_rate": 0.00012199342111354411,
+      "loss": 0.9645,
+      "step": 22320
+    },
+    {
+      "epoch": 0.8594802694898941,
+      "grad_norm": 1.8444538116455078,
+      "learning_rate": 0.00012196392092817202,
+      "loss": 1.173,
+      "step": 22325
+    },
+    {
+      "epoch": 0.8596727622714149,
+      "grad_norm": 1.4887306690216064,
+      "learning_rate": 0.00012193441873434481,
+      "loss": 1.0281,
+      "step": 22330
+    },
+    {
+      "epoch": 0.8598652550529355,
+      "grad_norm": 1.3933312892913818,
+      "learning_rate": 0.00012190491453476036,
+      "loss": 0.9568,
+      "step": 22335
+    },
+    {
+      "epoch": 0.8600577478344562,
+      "grad_norm": 1.9519412517547607,
+      "learning_rate": 0.00012187540833211657,
+      "loss": 1.0294,
+      "step": 22340
+    },
+    {
+      "epoch": 0.8602502406159769,
+      "grad_norm": 1.5969189405441284,
+      "learning_rate": 0.00012184590012911158,
+      "loss": 1.111,
+      "step": 22345
+    },
+    {
+      "epoch": 0.8604427333974976,
+      "grad_norm": 1.068150281906128,
+      "learning_rate": 0.00012181638992844377,
+      "loss": 1.119,
+      "step": 22350
+    },
+    {
+      "epoch": 0.8606352261790183,
+      "grad_norm": 1.8505868911743164,
+      "learning_rate": 0.00012178687773281159,
+      "loss": 1.1762,
+      "step": 22355
+    },
+    {
+      "epoch": 0.860827718960539,
+      "grad_norm": 0.8779274225234985,
+      "learning_rate": 0.00012175736354491381,
+      "loss": 1.0762,
+      "step": 22360
+    },
+    {
+      "epoch": 0.8610202117420597,
+      "grad_norm": 0.7734150886535645,
+      "learning_rate": 0.0001217278473674492,
+      "loss": 1.139,
+      "step": 22365
+    },
+    {
+      "epoch": 0.8612127045235803,
+      "grad_norm": 1.2648577690124512,
+      "learning_rate": 0.00012169832920311693,
+      "loss": 0.9826,
+      "step": 22370
+    },
+    {
+      "epoch": 0.861405197305101,
+      "grad_norm": 1.2812930345535278,
+      "learning_rate": 0.00012166880905461618,
+      "loss": 1.0287,
+      "step": 22375
+    },
+    {
+      "epoch": 0.8615976900866218,
+      "grad_norm": 1.3110228776931763,
+      "learning_rate": 0.00012163928692464637,
+      "loss": 1.1135,
+      "step": 22380
+    },
+    {
+      "epoch": 0.8617901828681425,
+      "grad_norm": 1.1708343029022217,
+      "learning_rate": 0.00012160976281590713,
+      "loss": 1.2012,
+      "step": 22385
+    },
+    {
+      "epoch": 0.8619826756496631,
+      "grad_norm": 1.004078984260559,
+      "learning_rate": 0.00012158023673109824,
+      "loss": 0.978,
+      "step": 22390
+    },
+    {
+      "epoch": 0.8621751684311838,
+      "grad_norm": 1.4461795091629028,
+      "learning_rate": 0.00012155070867291965,
+      "loss": 1.2483,
+      "step": 22395
+    },
+    {
+      "epoch": 0.8623676612127045,
+      "grad_norm": 1.3903266191482544,
+      "learning_rate": 0.00012152117864407149,
+      "loss": 1.2962,
+      "step": 22400
+    },
+    {
+      "epoch": 0.8625601539942253,
+      "grad_norm": 1.88125479221344,
+      "learning_rate": 0.00012149164664725414,
+      "loss": 1.1184,
+      "step": 22405
+    },
+    {
+      "epoch": 0.8627526467757459,
+      "grad_norm": 1.5932334661483765,
+      "learning_rate": 0.00012146211268516805,
+      "loss": 1.117,
+      "step": 22410
+    },
+    {
+      "epoch": 0.8629451395572666,
+      "grad_norm": 1.5401246547698975,
+      "learning_rate": 0.00012143257676051394,
+      "loss": 1.1546,
+      "step": 22415
+    },
+    {
+      "epoch": 0.8631376323387873,
+      "grad_norm": 1.215257167816162,
+      "learning_rate": 0.00012140303887599269,
+      "loss": 1.139,
+      "step": 22420
+    },
+    {
+      "epoch": 0.8633301251203079,
+      "grad_norm": 1.1622364521026611,
+      "learning_rate": 0.00012137349903430528,
+      "loss": 1.1076,
+      "step": 22425
+    },
+    {
+      "epoch": 0.8635226179018287,
+      "grad_norm": 1.5525301694869995,
+      "learning_rate": 0.00012134395723815297,
+      "loss": 0.8998,
+      "step": 22430
+    },
+    {
+      "epoch": 0.8637151106833494,
+      "grad_norm": 1.1511106491088867,
+      "learning_rate": 0.00012131441349023718,
+      "loss": 1.0576,
+      "step": 22435
+    },
+    {
+      "epoch": 0.86390760346487,
+      "grad_norm": 1.3346484899520874,
+      "learning_rate": 0.00012128486779325947,
+      "loss": 1.1717,
+      "step": 22440
+    },
+    {
+      "epoch": 0.8641000962463907,
+      "grad_norm": 1.0038443803787231,
+      "learning_rate": 0.0001212553201499216,
+      "loss": 1.0569,
+      "step": 22445
+    },
+    {
+      "epoch": 0.8642925890279115,
+      "grad_norm": 1.3039393424987793,
+      "learning_rate": 0.00012122577056292548,
+      "loss": 1.117,
+      "step": 22450
+    },
+    {
+      "epoch": 0.8644850818094322,
+      "grad_norm": 1.9241334199905396,
+      "learning_rate": 0.00012119621903497329,
+      "loss": 1.3177,
+      "step": 22455
+    },
+    {
+      "epoch": 0.8646775745909528,
+      "grad_norm": 1.7553646564483643,
+      "learning_rate": 0.00012116666556876725,
+      "loss": 1.1536,
+      "step": 22460
+    },
+    {
+      "epoch": 0.8648700673724735,
+      "grad_norm": 1.4490256309509277,
+      "learning_rate": 0.00012113711016700986,
+      "loss": 1.0066,
+      "step": 22465
+    },
+    {
+      "epoch": 0.8650625601539942,
+      "grad_norm": 0.8614715337753296,
+      "learning_rate": 0.0001211075528324038,
+      "loss": 1.0534,
+      "step": 22470
+    },
+    {
+      "epoch": 0.865255052935515,
+      "grad_norm": 1.0804800987243652,
+      "learning_rate": 0.00012107799356765181,
+      "loss": 1.0743,
+      "step": 22475
+    },
+    {
+      "epoch": 0.8654475457170356,
+      "grad_norm": 1.186589241027832,
+      "learning_rate": 0.00012104843237545695,
+      "loss": 1.1916,
+      "step": 22480
+    },
+    {
+      "epoch": 0.8656400384985563,
+      "grad_norm": 1.1526156663894653,
+      "learning_rate": 0.00012101886925852242,
+      "loss": 1.069,
+      "step": 22485
+    },
+    {
+      "epoch": 0.865832531280077,
+      "grad_norm": 1.382684350013733,
+      "learning_rate": 0.0001209893042195515,
+      "loss": 1.1266,
+      "step": 22490
+    },
+    {
+      "epoch": 0.8660250240615976,
+      "grad_norm": 1.9565086364746094,
+      "learning_rate": 0.00012095973726124774,
+      "loss": 1.3461,
+      "step": 22495
+    },
+    {
+      "epoch": 0.8662175168431184,
+      "grad_norm": 0.820969820022583,
+      "learning_rate": 0.0001209301683863149,
+      "loss": 0.9658,
+      "step": 22500
+    },
+    {
+      "epoch": 0.8664100096246391,
+      "grad_norm": 1.1080729961395264,
+      "learning_rate": 0.00012090059759745677,
+      "loss": 1.0718,
+      "step": 22505
+    },
+    {
+      "epoch": 0.8666025024061598,
+      "grad_norm": 1.6945521831512451,
+      "learning_rate": 0.00012087102489737746,
+      "loss": 1.1481,
+      "step": 22510
+    },
+    {
+      "epoch": 0.8667949951876804,
+      "grad_norm": 1.9991236925125122,
+      "learning_rate": 0.00012084145028878117,
+      "loss": 1.06,
+      "step": 22515
+    },
+    {
+      "epoch": 0.8669874879692011,
+      "grad_norm": 1.1101261377334595,
+      "learning_rate": 0.00012081187377437233,
+      "loss": 1.1425,
+      "step": 22520
+    },
+    {
+      "epoch": 0.8671799807507219,
+      "grad_norm": 0.9426703453063965,
+      "learning_rate": 0.00012078229535685548,
+      "loss": 1.1868,
+      "step": 22525
+    },
+    {
+      "epoch": 0.8673724735322426,
+      "grad_norm": 1.0097079277038574,
+      "learning_rate": 0.0001207527150389354,
+      "loss": 1.1158,
+      "step": 22530
+    },
+    {
+      "epoch": 0.8675649663137632,
+      "grad_norm": 1.3798326253890991,
+      "learning_rate": 0.000120723132823317,
+      "loss": 1.217,
+      "step": 22535
+    },
+    {
+      "epoch": 0.8677574590952839,
+      "grad_norm": 1.17965567111969,
+      "learning_rate": 0.00012069354871270538,
+      "loss": 1.1504,
+      "step": 22540
+    },
+    {
+      "epoch": 0.8679499518768046,
+      "grad_norm": 0.9790163040161133,
+      "learning_rate": 0.00012066396270980583,
+      "loss": 1.1189,
+      "step": 22545
+    },
+    {
+      "epoch": 0.8681424446583254,
+      "grad_norm": 1.134660005569458,
+      "learning_rate": 0.00012063437481732375,
+      "loss": 1.1093,
+      "step": 22550
+    },
+    {
+      "epoch": 0.868334937439846,
+      "grad_norm": 2.053043842315674,
+      "learning_rate": 0.0001206047850379648,
+      "loss": 1.2373,
+      "step": 22555
+    },
+    {
+      "epoch": 0.8685274302213667,
+      "grad_norm": 1.0980230569839478,
+      "learning_rate": 0.00012057519337443477,
+      "loss": 1.0597,
+      "step": 22560
+    },
+    {
+      "epoch": 0.8687199230028874,
+      "grad_norm": 0.8968497514724731,
+      "learning_rate": 0.00012054559982943958,
+      "loss": 1.0963,
+      "step": 22565
+    },
+    {
+      "epoch": 0.868912415784408,
+      "grad_norm": 1.183481216430664,
+      "learning_rate": 0.00012051600440568538,
+      "loss": 1.0797,
+      "step": 22570
+    },
+    {
+      "epoch": 0.8691049085659288,
+      "grad_norm": 1.0228111743927002,
+      "learning_rate": 0.00012048640710587848,
+      "loss": 1.1906,
+      "step": 22575
+    },
+    {
+      "epoch": 0.8692974013474495,
+      "grad_norm": 1.070470929145813,
+      "learning_rate": 0.00012045680793272537,
+      "loss": 1.0835,
+      "step": 22580
+    },
+    {
+      "epoch": 0.8694898941289702,
+      "grad_norm": 1.7659672498703003,
+      "learning_rate": 0.0001204272068889327,
+      "loss": 1.0718,
+      "step": 22585
+    },
+    {
+      "epoch": 0.8696823869104908,
+      "grad_norm": 1.503348469734192,
+      "learning_rate": 0.00012039760397720727,
+      "loss": 1.052,
+      "step": 22590
+    },
+    {
+      "epoch": 0.8698748796920116,
+      "grad_norm": 1.4245517253875732,
+      "learning_rate": 0.00012036799920025607,
+      "loss": 1.1893,
+      "step": 22595
+    },
+    {
+      "epoch": 0.8700673724735323,
+      "grad_norm": 1.6989365816116333,
+      "learning_rate": 0.0001203383925607863,
+      "loss": 1.1963,
+      "step": 22600
+    },
+    {
+      "epoch": 0.870259865255053,
+      "grad_norm": 0.9346055388450623,
+      "learning_rate": 0.0001203087840615052,
+      "loss": 1.2617,
+      "step": 22605
+    },
+    {
+      "epoch": 0.8704523580365736,
+      "grad_norm": 1.197245717048645,
+      "learning_rate": 0.00012027917370512039,
+      "loss": 1.0629,
+      "step": 22610
+    },
+    {
+      "epoch": 0.8706448508180943,
+      "grad_norm": 1.3511501550674438,
+      "learning_rate": 0.00012024956149433946,
+      "loss": 1.2573,
+      "step": 22615
+    },
+    {
+      "epoch": 0.8708373435996151,
+      "grad_norm": 1.259965181350708,
+      "learning_rate": 0.00012021994743187026,
+      "loss": 1.03,
+      "step": 22620
+    },
+    {
+      "epoch": 0.8710298363811357,
+      "grad_norm": 0.90989089012146,
+      "learning_rate": 0.00012019033152042083,
+      "loss": 1.1222,
+      "step": 22625
+    },
+    {
+      "epoch": 0.8712223291626564,
+      "grad_norm": 1.5417457818984985,
+      "learning_rate": 0.00012016071376269932,
+      "loss": 1.2125,
+      "step": 22630
+    },
+    {
+      "epoch": 0.8714148219441771,
+      "grad_norm": 1.0979586839675903,
+      "learning_rate": 0.00012013109416141408,
+      "loss": 1.1149,
+      "step": 22635
+    },
+    {
+      "epoch": 0.8716073147256977,
+      "grad_norm": 1.3153108358383179,
+      "learning_rate": 0.00012010147271927367,
+      "loss": 0.9733,
+      "step": 22640
+    },
+    {
+      "epoch": 0.8717998075072185,
+      "grad_norm": 1.1540855169296265,
+      "learning_rate": 0.0001200718494389867,
+      "loss": 1.1337,
+      "step": 22645
+    },
+    {
+      "epoch": 0.8719923002887392,
+      "grad_norm": 2.1098320484161377,
+      "learning_rate": 0.00012004222432326211,
+      "loss": 1.2008,
+      "step": 22650
+    },
+    {
+      "epoch": 0.8721847930702599,
+      "grad_norm": 2.331286668777466,
+      "learning_rate": 0.00012001259737480883,
+      "loss": 1.3357,
+      "step": 22655
+    },
+    {
+      "epoch": 0.8723772858517805,
+      "grad_norm": 0.9849653244018555,
+      "learning_rate": 0.00011998296859633613,
+      "loss": 0.9834,
+      "step": 22660
+    },
+    {
+      "epoch": 0.8725697786333012,
+      "grad_norm": 1.9478217363357544,
+      "learning_rate": 0.0001199533379905533,
+      "loss": 1.1192,
+      "step": 22665
+    },
+    {
+      "epoch": 0.872762271414822,
+      "grad_norm": 1.009040355682373,
+      "learning_rate": 0.0001199237055601699,
+      "loss": 1.2274,
+      "step": 22670
+    },
+    {
+      "epoch": 0.8729547641963427,
+      "grad_norm": 1.4086748361587524,
+      "learning_rate": 0.00011989407130789558,
+      "loss": 1.028,
+      "step": 22675
+    },
+    {
+      "epoch": 0.8731472569778633,
+      "grad_norm": 1.224057674407959,
+      "learning_rate": 0.00011986443523644025,
+      "loss": 1.046,
+      "step": 22680
+    },
+    {
+      "epoch": 0.873339749759384,
+      "grad_norm": 1.7258949279785156,
+      "learning_rate": 0.00011983479734851393,
+      "loss": 1.1056,
+      "step": 22685
+    },
+    {
+      "epoch": 0.8735322425409047,
+      "grad_norm": 1.5146769285202026,
+      "learning_rate": 0.00011980515764682674,
+      "loss": 1.3491,
+      "step": 22690
+    },
+    {
+      "epoch": 0.8737247353224254,
+      "grad_norm": 1.7950037717819214,
+      "learning_rate": 0.0001197755161340891,
+      "loss": 1.0302,
+      "step": 22695
+    },
+    {
+      "epoch": 0.8739172281039461,
+      "grad_norm": 1.0561470985412598,
+      "learning_rate": 0.00011974587281301151,
+      "loss": 1.026,
+      "step": 22700
+    },
+    {
+      "epoch": 0.8741097208854668,
+      "grad_norm": 2.0204522609710693,
+      "learning_rate": 0.00011971622768630466,
+      "loss": 0.999,
+      "step": 22705
+    },
+    {
+      "epoch": 0.8743022136669875,
+      "grad_norm": 1.23675537109375,
+      "learning_rate": 0.00011968658075667938,
+      "loss": 0.9875,
+      "step": 22710
+    },
+    {
+      "epoch": 0.8744947064485081,
+      "grad_norm": 1.2208534479141235,
+      "learning_rate": 0.00011965693202684671,
+      "loss": 1.1022,
+      "step": 22715
+    },
+    {
+      "epoch": 0.8746871992300289,
+      "grad_norm": 1.2948905229568481,
+      "learning_rate": 0.00011962728149951785,
+      "loss": 1.1026,
+      "step": 22720
+    },
+    {
+      "epoch": 0.8748796920115496,
+      "grad_norm": 1.0135208368301392,
+      "learning_rate": 0.00011959762917740407,
+      "loss": 1.1093,
+      "step": 22725
+    },
+    {
+      "epoch": 0.8750721847930703,
+      "grad_norm": 1.3146969079971313,
+      "learning_rate": 0.00011956797506321695,
+      "loss": 1.0676,
+      "step": 22730
+    },
+    {
+      "epoch": 0.8752646775745909,
+      "grad_norm": 1.5172390937805176,
+      "learning_rate": 0.00011953831915966813,
+      "loss": 1.0981,
+      "step": 22735
+    },
+    {
+      "epoch": 0.8754571703561116,
+      "grad_norm": 1.0460716485977173,
+      "learning_rate": 0.00011950866146946946,
+      "loss": 1.1096,
+      "step": 22740
+    },
+    {
+      "epoch": 0.8756496631376324,
+      "grad_norm": 1.5505667924880981,
+      "learning_rate": 0.00011947900199533291,
+      "loss": 1.1211,
+      "step": 22745
+    },
+    {
+      "epoch": 0.875842155919153,
+      "grad_norm": 1.3323757648468018,
+      "learning_rate": 0.00011944934073997069,
+      "loss": 1.0103,
+      "step": 22750
+    },
+    {
+      "epoch": 0.8760346487006737,
+      "grad_norm": 1.1330286264419556,
+      "learning_rate": 0.0001194196777060951,
+      "loss": 1.0935,
+      "step": 22755
+    },
+    {
+      "epoch": 0.8762271414821944,
+      "grad_norm": 1.3528751134872437,
+      "learning_rate": 0.00011939001289641863,
+      "loss": 1.0812,
+      "step": 22760
+    },
+    {
+      "epoch": 0.8764196342637152,
+      "grad_norm": 1.3828253746032715,
+      "learning_rate": 0.0001193603463136539,
+      "loss": 1.0722,
+      "step": 22765
+    },
+    {
+      "epoch": 0.8766121270452358,
+      "grad_norm": 1.1661975383758545,
+      "learning_rate": 0.00011933067796051378,
+      "loss": 1.0631,
+      "step": 22770
+    },
+    {
+      "epoch": 0.8768046198267565,
+      "grad_norm": 1.3716634511947632,
+      "learning_rate": 0.00011930100783971118,
+      "loss": 1.0528,
+      "step": 22775
+    },
+    {
+      "epoch": 0.8769971126082772,
+      "grad_norm": 1.4899778366088867,
+      "learning_rate": 0.00011927133595395932,
+      "loss": 1.0373,
+      "step": 22780
+    },
+    {
+      "epoch": 0.8771896053897978,
+      "grad_norm": 1.0731201171875,
+      "learning_rate": 0.0001192416623059714,
+      "loss": 1.1288,
+      "step": 22785
+    },
+    {
+      "epoch": 0.8773820981713186,
+      "grad_norm": 1.1406874656677246,
+      "learning_rate": 0.00011921198689846094,
+      "loss": 1.0467,
+      "step": 22790
+    },
+    {
+      "epoch": 0.8775745909528393,
+      "grad_norm": 1.203275442123413,
+      "learning_rate": 0.00011918230973414157,
+      "loss": 1.1394,
+      "step": 22795
+    },
+    {
+      "epoch": 0.87776708373436,
+      "grad_norm": 1.0929608345031738,
+      "learning_rate": 0.000119152630815727,
+      "loss": 1.0472,
+      "step": 22800
+    },
+    {
+      "epoch": 0.8779595765158806,
+      "grad_norm": 0.9629132747650146,
+      "learning_rate": 0.00011912295014593124,
+      "loss": 1.0325,
+      "step": 22805
+    },
+    {
+      "epoch": 0.8781520692974013,
+      "grad_norm": 1.3864428997039795,
+      "learning_rate": 0.00011909326772746839,
+      "loss": 0.9702,
+      "step": 22810
+    },
+    {
+      "epoch": 0.8783445620789221,
+      "grad_norm": 2.243774175643921,
+      "learning_rate": 0.00011906358356305265,
+      "loss": 1.1504,
+      "step": 22815
+    },
+    {
+      "epoch": 0.8785370548604428,
+      "grad_norm": 1.1354432106018066,
+      "learning_rate": 0.0001190338976553985,
+      "loss": 1.0098,
+      "step": 22820
+    },
+    {
+      "epoch": 0.8787295476419634,
+      "grad_norm": 1.9807575941085815,
+      "learning_rate": 0.00011900421000722049,
+      "loss": 1.1305,
+      "step": 22825
+    },
+    {
+      "epoch": 0.8789220404234841,
+      "grad_norm": 1.0899982452392578,
+      "learning_rate": 0.00011897452062123338,
+      "loss": 1.0502,
+      "step": 22830
+    },
+    {
+      "epoch": 0.8791145332050048,
+      "grad_norm": 1.194968819618225,
+      "learning_rate": 0.00011894482950015202,
+      "loss": 1.0947,
+      "step": 22835
+    },
+    {
+      "epoch": 0.8793070259865255,
+      "grad_norm": 1.032558560371399,
+      "learning_rate": 0.00011891513664669152,
+      "loss": 1.3069,
+      "step": 22840
+    },
+    {
+      "epoch": 0.8794995187680462,
+      "grad_norm": 1.0530674457550049,
+      "learning_rate": 0.00011888544206356709,
+      "loss": 1.0373,
+      "step": 22845
+    },
+    {
+      "epoch": 0.8796920115495669,
+      "grad_norm": 2.1762609481811523,
+      "learning_rate": 0.00011885574575349407,
+      "loss": 1.1478,
+      "step": 22850
+    },
+    {
+      "epoch": 0.8798845043310876,
+      "grad_norm": 1.5496394634246826,
+      "learning_rate": 0.00011882604771918802,
+      "loss": 1.078,
+      "step": 22855
+    },
+    {
+      "epoch": 0.8800769971126082,
+      "grad_norm": 1.5198618173599243,
+      "learning_rate": 0.0001187963479633646,
+      "loss": 0.9666,
+      "step": 22860
+    },
+    {
+      "epoch": 0.880269489894129,
+      "grad_norm": 2.006168842315674,
+      "learning_rate": 0.00011876664648873969,
+      "loss": 1.0162,
+      "step": 22865
+    },
+    {
+      "epoch": 0.8804619826756497,
+      "grad_norm": 1.6335991621017456,
+      "learning_rate": 0.0001187369432980293,
+      "loss": 1.2807,
+      "step": 22870
+    },
+    {
+      "epoch": 0.8806544754571703,
+      "grad_norm": 1.4370218515396118,
+      "learning_rate": 0.00011870723839394953,
+      "loss": 1.0569,
+      "step": 22875
+    },
+    {
+      "epoch": 0.880846968238691,
+      "grad_norm": 1.4292523860931396,
+      "learning_rate": 0.00011867753177921675,
+      "loss": 1.0509,
+      "step": 22880
+    },
+    {
+      "epoch": 0.8810394610202117,
+      "grad_norm": 0.8672935366630554,
+      "learning_rate": 0.00011864782345654739,
+      "loss": 1.0674,
+      "step": 22885
+    },
+    {
+      "epoch": 0.8812319538017325,
+      "grad_norm": 2.000182628631592,
+      "learning_rate": 0.00011861811342865814,
+      "loss": 1.142,
+      "step": 22890
+    },
+    {
+      "epoch": 0.8814244465832531,
+      "grad_norm": 1.2838269472122192,
+      "learning_rate": 0.00011858840169826573,
+      "loss": 1.2496,
+      "step": 22895
+    },
+    {
+      "epoch": 0.8816169393647738,
+      "grad_norm": 0.8586519956588745,
+      "learning_rate": 0.00011855868826808712,
+      "loss": 1.0047,
+      "step": 22900
+    },
+    {
+      "epoch": 0.8818094321462945,
+      "grad_norm": 0.9435043334960938,
+      "learning_rate": 0.00011852897314083942,
+      "loss": 0.9603,
+      "step": 22905
+    },
+    {
+      "epoch": 0.8820019249278153,
+      "grad_norm": 1.5908135175704956,
+      "learning_rate": 0.00011849925631923986,
+      "loss": 1.1064,
+      "step": 22910
+    },
+    {
+      "epoch": 0.8821944177093359,
+      "grad_norm": 1.6797462701797485,
+      "learning_rate": 0.00011846953780600585,
+      "loss": 1.1746,
+      "step": 22915
+    },
+    {
+      "epoch": 0.8823869104908566,
+      "grad_norm": 1.5768578052520752,
+      "learning_rate": 0.00011843981760385496,
+      "loss": 1.153,
+      "step": 22920
+    },
+    {
+      "epoch": 0.8825794032723773,
+      "grad_norm": 1.136809229850769,
+      "learning_rate": 0.0001184100957155049,
+      "loss": 1.1523,
+      "step": 22925
+    },
+    {
+      "epoch": 0.8827718960538979,
+      "grad_norm": 1.0501734018325806,
+      "learning_rate": 0.00011838037214367354,
+      "loss": 1.0341,
+      "step": 22930
+    },
+    {
+      "epoch": 0.8829643888354187,
+      "grad_norm": 1.3567155599594116,
+      "learning_rate": 0.00011835064689107893,
+      "loss": 1.1462,
+      "step": 22935
+    },
+    {
+      "epoch": 0.8831568816169394,
+      "grad_norm": 2.095165729522705,
+      "learning_rate": 0.00011832091996043921,
+      "loss": 1.1365,
+      "step": 22940
+    },
+    {
+      "epoch": 0.8833493743984601,
+      "grad_norm": 1.119279384613037,
+      "learning_rate": 0.00011829119135447274,
+      "loss": 1.1334,
+      "step": 22945
+    },
+    {
+      "epoch": 0.8835418671799807,
+      "grad_norm": 0.8329470157623291,
+      "learning_rate": 0.00011826146107589795,
+      "loss": 0.9907,
+      "step": 22950
+    },
+    {
+      "epoch": 0.8837343599615014,
+      "grad_norm": 1.6559773683547974,
+      "learning_rate": 0.00011823172912743355,
+      "loss": 1.0937,
+      "step": 22955
+    },
+    {
+      "epoch": 0.8839268527430222,
+      "grad_norm": 1.1509395837783813,
+      "learning_rate": 0.00011820199551179827,
+      "loss": 1.213,
+      "step": 22960
+    },
+    {
+      "epoch": 0.8841193455245429,
+      "grad_norm": 1.926554799079895,
+      "learning_rate": 0.00011817226023171109,
+      "loss": 1.1967,
+      "step": 22965
+    },
+    {
+      "epoch": 0.8843118383060635,
+      "grad_norm": 1.3691015243530273,
+      "learning_rate": 0.00011814252328989111,
+      "loss": 1.2489,
+      "step": 22970
+    },
+    {
+      "epoch": 0.8845043310875842,
+      "grad_norm": 1.807802438735962,
+      "learning_rate": 0.00011811278468905753,
+      "loss": 1.2371,
+      "step": 22975
+    },
+    {
+      "epoch": 0.8846968238691049,
+      "grad_norm": 1.0577672719955444,
+      "learning_rate": 0.00011808304443192982,
+      "loss": 0.9763,
+      "step": 22980
+    },
+    {
+      "epoch": 0.8848893166506256,
+      "grad_norm": 1.69723379611969,
+      "learning_rate": 0.00011805330252122743,
+      "loss": 1.3159,
+      "step": 22985
+    },
+    {
+      "epoch": 0.8850818094321463,
+      "grad_norm": 2.4447522163391113,
+      "learning_rate": 0.00011802355895967017,
+      "loss": 1.1288,
+      "step": 22990
+    },
+    {
+      "epoch": 0.885274302213667,
+      "grad_norm": 1.1180788278579712,
+      "learning_rate": 0.0001179938137499778,
+      "loss": 1.2414,
+      "step": 22995
+    },
+    {
+      "epoch": 0.8854667949951877,
+      "grad_norm": 1.8794695138931274,
+      "learning_rate": 0.00011796406689487038,
+      "loss": 1.1329,
+      "step": 23000
+    },
+    {
+      "epoch": 0.8856592877767083,
+      "grad_norm": 1.1928154230117798,
+      "learning_rate": 0.00011793431839706803,
+      "loss": 1.0499,
+      "step": 23005
+    },
+    {
+      "epoch": 0.8858517805582291,
+      "grad_norm": 0.8071653842926025,
+      "learning_rate": 0.00011790456825929106,
+      "loss": 0.9116,
+      "step": 23010
+    },
+    {
+      "epoch": 0.8860442733397498,
+      "grad_norm": 1.4556511640548706,
+      "learning_rate": 0.00011787481648425995,
+      "loss": 0.9842,
+      "step": 23015
+    },
+    {
+      "epoch": 0.8862367661212704,
+      "grad_norm": 1.4011625051498413,
+      "learning_rate": 0.00011784506307469527,
+      "loss": 1.1377,
+      "step": 23020
+    },
+    {
+      "epoch": 0.8864292589027911,
+      "grad_norm": 1.596177339553833,
+      "learning_rate": 0.00011781530803331778,
+      "loss": 1.2584,
+      "step": 23025
+    },
+    {
+      "epoch": 0.8866217516843118,
+      "grad_norm": 1.4121986627578735,
+      "learning_rate": 0.00011778555136284839,
+      "loss": 1.0889,
+      "step": 23030
+    },
+    {
+      "epoch": 0.8868142444658326,
+      "grad_norm": 1.7569849491119385,
+      "learning_rate": 0.00011775579306600814,
+      "loss": 1.0882,
+      "step": 23035
+    },
+    {
+      "epoch": 0.8870067372473532,
+      "grad_norm": 2.0657474994659424,
+      "learning_rate": 0.00011772603314551825,
+      "loss": 1.0512,
+      "step": 23040
+    },
+    {
+      "epoch": 0.8871992300288739,
+      "grad_norm": 1.419628620147705,
+      "learning_rate": 0.00011769627160410006,
+      "loss": 1.2821,
+      "step": 23045
+    },
+    {
+      "epoch": 0.8873917228103946,
+      "grad_norm": 2.094897747039795,
+      "learning_rate": 0.00011766650844447505,
+      "loss": 1.0668,
+      "step": 23050
+    },
+    {
+      "epoch": 0.8875842155919152,
+      "grad_norm": 1.6981192827224731,
+      "learning_rate": 0.00011763674366936485,
+      "loss": 1.1257,
+      "step": 23055
+    },
+    {
+      "epoch": 0.887776708373436,
+      "grad_norm": 2.0729424953460693,
+      "learning_rate": 0.0001176069772814913,
+      "loss": 1.1388,
+      "step": 23060
+    },
+    {
+      "epoch": 0.8879692011549567,
+      "grad_norm": 1.7522398233413696,
+      "learning_rate": 0.0001175772092835763,
+      "loss": 1.3847,
+      "step": 23065
+    },
+    {
+      "epoch": 0.8881616939364774,
+      "grad_norm": 1.2705082893371582,
+      "learning_rate": 0.00011754743967834197,
+      "loss": 1.042,
+      "step": 23070
+    },
+    {
+      "epoch": 0.888354186717998,
+      "grad_norm": 1.9361348152160645,
+      "learning_rate": 0.0001175176684685105,
+      "loss": 1.1387,
+      "step": 23075
+    },
+    {
+      "epoch": 0.8885466794995188,
+      "grad_norm": 2.4146616458892822,
+      "learning_rate": 0.00011748789565680429,
+      "loss": 1.1017,
+      "step": 23080
+    },
+    {
+      "epoch": 0.8887391722810395,
+      "grad_norm": 1.01850163936615,
+      "learning_rate": 0.00011745812124594589,
+      "loss": 0.9676,
+      "step": 23085
+    },
+    {
+      "epoch": 0.8889316650625602,
+      "grad_norm": 1.8511348962783813,
+      "learning_rate": 0.00011742834523865796,
+      "loss": 1.0274,
+      "step": 23090
+    },
+    {
+      "epoch": 0.8891241578440808,
+      "grad_norm": 1.0579948425292969,
+      "learning_rate": 0.0001173985676376633,
+      "loss": 0.9528,
+      "step": 23095
+    },
+    {
+      "epoch": 0.8893166506256015,
+      "grad_norm": 1.9102215766906738,
+      "learning_rate": 0.00011736878844568486,
+      "loss": 1.0718,
+      "step": 23100
+    },
+    {
+      "epoch": 0.8895091434071223,
+      "grad_norm": 1.7665621042251587,
+      "learning_rate": 0.00011733900766544579,
+      "loss": 0.9721,
+      "step": 23105
+    },
+    {
+      "epoch": 0.889701636188643,
+      "grad_norm": 1.6327630281448364,
+      "learning_rate": 0.00011730922529966934,
+      "loss": 1.1314,
+      "step": 23110
+    },
+    {
+      "epoch": 0.8898941289701636,
+      "grad_norm": 1.567986011505127,
+      "learning_rate": 0.00011727944135107889,
+      "loss": 1.1881,
+      "step": 23115
+    },
+    {
+      "epoch": 0.8900866217516843,
+      "grad_norm": 1.3827725648880005,
+      "learning_rate": 0.00011724965582239798,
+      "loss": 1.0747,
+      "step": 23120
+    },
+    {
+      "epoch": 0.890279114533205,
+      "grad_norm": 1.1218315362930298,
+      "learning_rate": 0.00011721986871635034,
+      "loss": 1.0224,
+      "step": 23125
+    },
+    {
+      "epoch": 0.8904716073147257,
+      "grad_norm": 0.9669145941734314,
+      "learning_rate": 0.00011719008003565975,
+      "loss": 1.1214,
+      "step": 23130
+    },
+    {
+      "epoch": 0.8906641000962464,
+      "grad_norm": 2.2118616104125977,
+      "learning_rate": 0.00011716028978305023,
+      "loss": 1.0058,
+      "step": 23135
+    },
+    {
+      "epoch": 0.8908565928777671,
+      "grad_norm": 1.0324665307998657,
+      "learning_rate": 0.00011713049796124589,
+      "loss": 1.0645,
+      "step": 23140
+    },
+    {
+      "epoch": 0.8910490856592878,
+      "grad_norm": 0.9733279347419739,
+      "learning_rate": 0.00011710070457297097,
+      "loss": 1.1258,
+      "step": 23145
+    },
+    {
+      "epoch": 0.8912415784408084,
+      "grad_norm": 0.8568596243858337,
+      "learning_rate": 0.00011707090962094991,
+      "loss": 1.046,
+      "step": 23150
+    },
+    {
+      "epoch": 0.8914340712223292,
+      "grad_norm": 1.0557111501693726,
+      "learning_rate": 0.00011704111310790728,
+      "loss": 1.0856,
+      "step": 23155
+    },
+    {
+      "epoch": 0.8916265640038499,
+      "grad_norm": 1.436998724937439,
+      "learning_rate": 0.00011701131503656771,
+      "loss": 1.0475,
+      "step": 23160
+    },
+    {
+      "epoch": 0.8918190567853705,
+      "grad_norm": 1.3652065992355347,
+      "learning_rate": 0.00011698151540965607,
+      "loss": 1.2369,
+      "step": 23165
+    },
+    {
+      "epoch": 0.8920115495668912,
+      "grad_norm": 1.6808583736419678,
+      "learning_rate": 0.00011695171422989735,
+      "loss": 1.2898,
+      "step": 23170
+    },
+    {
+      "epoch": 0.8922040423484119,
+      "grad_norm": 1.468636155128479,
+      "learning_rate": 0.00011692191150001666,
+      "loss": 1.1976,
+      "step": 23175
+    },
+    {
+      "epoch": 0.8923965351299327,
+      "grad_norm": 1.6479716300964355,
+      "learning_rate": 0.00011689210722273925,
+      "loss": 0.9855,
+      "step": 23180
+    },
+    {
+      "epoch": 0.8925890279114533,
+      "grad_norm": 1.1517993211746216,
+      "learning_rate": 0.00011686230140079054,
+      "loss": 1.0893,
+      "step": 23185
+    },
+    {
+      "epoch": 0.892781520692974,
+      "grad_norm": 1.0213438272476196,
+      "learning_rate": 0.00011683249403689606,
+      "loss": 0.9994,
+      "step": 23190
+    },
+    {
+      "epoch": 0.8929740134744947,
+      "grad_norm": 0.9689280390739441,
+      "learning_rate": 0.00011680268513378152,
+      "loss": 0.9765,
+      "step": 23195
+    },
+    {
+      "epoch": 0.8931665062560153,
+      "grad_norm": 1.0091511011123657,
+      "learning_rate": 0.00011677287469417272,
+      "loss": 1.143,
+      "step": 23200
+    },
+    {
+      "epoch": 0.8933589990375361,
+      "grad_norm": 1.0775808095932007,
+      "learning_rate": 0.00011674306272079567,
+      "loss": 0.9407,
+      "step": 23205
+    },
+    {
+      "epoch": 0.8935514918190568,
+      "grad_norm": 2.0991709232330322,
+      "learning_rate": 0.00011671324921637641,
+      "loss": 1.1681,
+      "step": 23210
+    },
+    {
+      "epoch": 0.8937439846005775,
+      "grad_norm": 1.5666394233703613,
+      "learning_rate": 0.00011668343418364122,
+      "loss": 1.2306,
+      "step": 23215
+    },
+    {
+      "epoch": 0.8939364773820981,
+      "grad_norm": 1.256421685218811,
+      "learning_rate": 0.00011665361762531653,
+      "loss": 1.092,
+      "step": 23220
+    },
+    {
+      "epoch": 0.8941289701636189,
+      "grad_norm": 1.5252572298049927,
+      "learning_rate": 0.00011662379954412879,
+      "loss": 1.0102,
+      "step": 23225
+    },
+    {
+      "epoch": 0.8943214629451396,
+      "grad_norm": 1.1907771825790405,
+      "learning_rate": 0.00011659397994280472,
+      "loss": 1.0496,
+      "step": 23230
+    },
+    {
+      "epoch": 0.8945139557266603,
+      "grad_norm": 1.2852728366851807,
+      "learning_rate": 0.00011656415882407113,
+      "loss": 1.093,
+      "step": 23235
+    },
+    {
+      "epoch": 0.8947064485081809,
+      "grad_norm": 1.3600996732711792,
+      "learning_rate": 0.0001165343361906549,
+      "loss": 1.0328,
+      "step": 23240
+    },
+    {
+      "epoch": 0.8948989412897016,
+      "grad_norm": 1.609398603439331,
+      "learning_rate": 0.00011650451204528316,
+      "loss": 1.2157,
+      "step": 23245
+    },
+    {
+      "epoch": 0.8950914340712224,
+      "grad_norm": 0.9227400422096252,
+      "learning_rate": 0.00011647468639068318,
+      "loss": 1.0264,
+      "step": 23250
+    },
+    {
+      "epoch": 0.895283926852743,
+      "grad_norm": 1.7690601348876953,
+      "learning_rate": 0.00011644485922958222,
+      "loss": 1.3007,
+      "step": 23255
+    },
+    {
+      "epoch": 0.8954764196342637,
+      "grad_norm": 0.9615120887756348,
+      "learning_rate": 0.00011641503056470782,
+      "loss": 0.9696,
+      "step": 23260
+    },
+    {
+      "epoch": 0.8956689124157844,
+      "grad_norm": 1.3075053691864014,
+      "learning_rate": 0.00011638520039878762,
+      "loss": 1.0539,
+      "step": 23265
+    },
+    {
+      "epoch": 0.8958614051973051,
+      "grad_norm": 1.8990029096603394,
+      "learning_rate": 0.00011635536873454937,
+      "loss": 1.2406,
+      "step": 23270
+    },
+    {
+      "epoch": 0.8960538979788258,
+      "grad_norm": 1.562389850616455,
+      "learning_rate": 0.00011632553557472101,
+      "loss": 1.1273,
+      "step": 23275
+    },
+    {
+      "epoch": 0.8962463907603465,
+      "grad_norm": 0.8242204785346985,
+      "learning_rate": 0.00011629570092203059,
+      "loss": 0.991,
+      "step": 23280
+    },
+    {
+      "epoch": 0.8964388835418672,
+      "grad_norm": 1.8995356559753418,
+      "learning_rate": 0.00011626586477920625,
+      "loss": 1.3523,
+      "step": 23285
+    },
+    {
+      "epoch": 0.8966313763233879,
+      "grad_norm": 1.2772774696350098,
+      "learning_rate": 0.00011623602714897632,
+      "loss": 1.1427,
+      "step": 23290
+    },
+    {
+      "epoch": 0.8968238691049085,
+      "grad_norm": 1.5422502756118774,
+      "learning_rate": 0.00011620618803406929,
+      "loss": 1.0688,
+      "step": 23295
+    },
+    {
+      "epoch": 0.8970163618864293,
+      "grad_norm": 1.4164756536483765,
+      "learning_rate": 0.00011617634743721369,
+      "loss": 1.0575,
+      "step": 23300
+    },
+    {
+      "epoch": 0.89720885466795,
+      "grad_norm": 0.9892400503158569,
+      "learning_rate": 0.0001161465053611383,
+      "loss": 1.0522,
+      "step": 23305
+    },
+    {
+      "epoch": 0.8974013474494706,
+      "grad_norm": 1.217403531074524,
+      "learning_rate": 0.00011611666180857193,
+      "loss": 1.2071,
+      "step": 23310
+    },
+    {
+      "epoch": 0.8975938402309913,
+      "grad_norm": 1.9891669750213623,
+      "learning_rate": 0.00011608681678224362,
+      "loss": 1.0608,
+      "step": 23315
+    },
+    {
+      "epoch": 0.897786333012512,
+      "grad_norm": 1.7623512744903564,
+      "learning_rate": 0.00011605697028488244,
+      "loss": 1.0479,
+      "step": 23320
+    },
+    {
+      "epoch": 0.8979788257940328,
+      "grad_norm": 1.8815721273422241,
+      "learning_rate": 0.00011602712231921775,
+      "loss": 1.2632,
+      "step": 23325
+    },
+    {
+      "epoch": 0.8981713185755534,
+      "grad_norm": 1.2247025966644287,
+      "learning_rate": 0.00011599727288797888,
+      "loss": 1.0253,
+      "step": 23330
+    },
+    {
+      "epoch": 0.8983638113570741,
+      "grad_norm": 1.4108113050460815,
+      "learning_rate": 0.00011596742199389534,
+      "loss": 1.1154,
+      "step": 23335
+    },
+    {
+      "epoch": 0.8985563041385948,
+      "grad_norm": 1.6885722875595093,
+      "learning_rate": 0.00011593756963969686,
+      "loss": 1.0641,
+      "step": 23340
+    },
+    {
+      "epoch": 0.8987487969201154,
+      "grad_norm": 1.6542564630508423,
+      "learning_rate": 0.00011590771582811324,
+      "loss": 0.914,
+      "step": 23345
+    },
+    {
+      "epoch": 0.8989412897016362,
+      "grad_norm": 1.0524113178253174,
+      "learning_rate": 0.00011587786056187435,
+      "loss": 1.0983,
+      "step": 23350
+    },
+    {
+      "epoch": 0.8991337824831569,
+      "grad_norm": 1.223191499710083,
+      "learning_rate": 0.00011584800384371028,
+      "loss": 0.996,
+      "step": 23355
+    },
+    {
+      "epoch": 0.8993262752646776,
+      "grad_norm": 1.6196818351745605,
+      "learning_rate": 0.00011581814567635128,
+      "loss": 0.973,
+      "step": 23360
+    },
+    {
+      "epoch": 0.8995187680461982,
+      "grad_norm": 1.0416576862335205,
+      "learning_rate": 0.00011578828606252763,
+      "loss": 1.0055,
+      "step": 23365
+    },
+    {
+      "epoch": 0.8997112608277189,
+      "grad_norm": 0.8480489253997803,
+      "learning_rate": 0.00011575842500496979,
+      "loss": 1.039,
+      "step": 23370
+    },
+    {
+      "epoch": 0.8999037536092397,
+      "grad_norm": 1.4879505634307861,
+      "learning_rate": 0.0001157285625064084,
+      "loss": 1.1408,
+      "step": 23375
+    },
+    {
+      "epoch": 0.9000962463907604,
+      "grad_norm": 1.2672336101531982,
+      "learning_rate": 0.00011569869856957416,
+      "loss": 0.9923,
+      "step": 23380
+    },
+    {
+      "epoch": 0.900288739172281,
+      "grad_norm": 1.147540807723999,
+      "learning_rate": 0.0001156688331971979,
+      "loss": 1.04,
+      "step": 23385
+    },
+    {
+      "epoch": 0.9004812319538017,
+      "grad_norm": 1.0331413745880127,
+      "learning_rate": 0.00011563896639201068,
+      "loss": 1.0502,
+      "step": 23390
+    },
+    {
+      "epoch": 0.9006737247353225,
+      "grad_norm": 1.2713568210601807,
+      "learning_rate": 0.00011560909815674354,
+      "loss": 1.109,
+      "step": 23395
+    },
+    {
+      "epoch": 0.9008662175168431,
+      "grad_norm": 1.5601911544799805,
+      "learning_rate": 0.0001155792284941278,
+      "loss": 0.9567,
+      "step": 23400
+    },
+    {
+      "epoch": 0.9010587102983638,
+      "grad_norm": 2.2069153785705566,
+      "learning_rate": 0.00011554935740689483,
+      "loss": 1.1025,
+      "step": 23405
+    },
+    {
+      "epoch": 0.9012512030798845,
+      "grad_norm": 1.4782518148422241,
+      "learning_rate": 0.00011551948489777611,
+      "loss": 1.1219,
+      "step": 23410
+    },
+    {
+      "epoch": 0.9014436958614052,
+      "grad_norm": 1.1145074367523193,
+      "learning_rate": 0.00011548961096950332,
+      "loss": 0.9797,
+      "step": 23415
+    },
+    {
+      "epoch": 0.9016361886429259,
+      "grad_norm": 1.0069323778152466,
+      "learning_rate": 0.00011545973562480821,
+      "loss": 1.1028,
+      "step": 23420
+    },
+    {
+      "epoch": 0.9018286814244466,
+      "grad_norm": 1.3039555549621582,
+      "learning_rate": 0.00011542985886642268,
+      "loss": 0.9536,
+      "step": 23425
+    },
+    {
+      "epoch": 0.9020211742059673,
+      "grad_norm": 1.2636560201644897,
+      "learning_rate": 0.00011539998069707876,
+      "loss": 1.1687,
+      "step": 23430
+    },
+    {
+      "epoch": 0.902213666987488,
+      "grad_norm": 1.4051100015640259,
+      "learning_rate": 0.00011537010111950866,
+      "loss": 1.1032,
+      "step": 23435
+    },
+    {
+      "epoch": 0.9024061597690086,
+      "grad_norm": 1.3477351665496826,
+      "learning_rate": 0.00011534022013644459,
+      "loss": 1.0785,
+      "step": 23440
+    },
+    {
+      "epoch": 0.9025986525505294,
+      "grad_norm": 1.7776321172714233,
+      "learning_rate": 0.00011531033775061901,
+      "loss": 0.9295,
+      "step": 23445
+    },
+    {
+      "epoch": 0.9027911453320501,
+      "grad_norm": 1.9729288816452026,
+      "learning_rate": 0.0001152804539647645,
+      "loss": 1.1657,
+      "step": 23450
+    },
+    {
+      "epoch": 0.9029836381135707,
+      "grad_norm": 0.9185901284217834,
+      "learning_rate": 0.00011525056878161367,
+      "loss": 1.0089,
+      "step": 23455
+    },
+    {
+      "epoch": 0.9031761308950914,
+      "grad_norm": 1.6784918308258057,
+      "learning_rate": 0.00011522068220389934,
+      "loss": 1.0847,
+      "step": 23460
+    },
+    {
+      "epoch": 0.9033686236766121,
+      "grad_norm": 1.261899709701538,
+      "learning_rate": 0.00011519079423435449,
+      "loss": 1.0281,
+      "step": 23465
+    },
+    {
+      "epoch": 0.9035611164581329,
+      "grad_norm": 1.1247740983963013,
+      "learning_rate": 0.0001151609048757121,
+      "loss": 1.143,
+      "step": 23470
+    },
+    {
+      "epoch": 0.9037536092396535,
+      "grad_norm": 1.3260306119918823,
+      "learning_rate": 0.0001151310141307054,
+      "loss": 1.022,
+      "step": 23475
+    },
+    {
+      "epoch": 0.9039461020211742,
+      "grad_norm": 1.0106364488601685,
+      "learning_rate": 0.00011510112200206769,
+      "loss": 1.0949,
+      "step": 23480
+    },
+    {
+      "epoch": 0.9041385948026949,
+      "grad_norm": 1.2663025856018066,
+      "learning_rate": 0.00011507122849253243,
+      "loss": 1.1526,
+      "step": 23485
+    },
+    {
+      "epoch": 0.9043310875842155,
+      "grad_norm": 1.2372430562973022,
+      "learning_rate": 0.00011504731269249489,
+      "loss": 1.0767,
+      "step": 23490
+    },
+    {
+      "epoch": 0.9045235803657363,
+      "grad_norm": 1.1163527965545654,
+      "learning_rate": 0.00011501741670423267,
+      "loss": 1.0479,
+      "step": 23495
+    },
+    {
+      "epoch": 0.904716073147257,
+      "grad_norm": 1.766180157661438,
+      "learning_rate": 0.00011498751934272714,
+      "loss": 1.0545,
+      "step": 23500
+    },
+    {
+      "epoch": 0.9049085659287777,
+      "grad_norm": 1.2410964965820312,
+      "learning_rate": 0.00011495762061071227,
+      "loss": 1.0771,
+      "step": 23505
+    },
+    {
+      "epoch": 0.9051010587102983,
+      "grad_norm": 1.8052672147750854,
+      "learning_rate": 0.00011492772051092206,
+      "loss": 1.2158,
+      "step": 23510
+    },
+    {
+      "epoch": 0.905293551491819,
+      "grad_norm": 1.9236223697662354,
+      "learning_rate": 0.00011489781904609071,
+      "loss": 1.147,
+      "step": 23515
+    },
+    {
+      "epoch": 0.9054860442733398,
+      "grad_norm": 1.454115867614746,
+      "learning_rate": 0.0001148679162189525,
+      "loss": 1.2127,
+      "step": 23520
+    },
+    {
+      "epoch": 0.9056785370548605,
+      "grad_norm": 1.20502769947052,
+      "learning_rate": 0.00011483801203224185,
+      "loss": 1.2464,
+      "step": 23525
+    },
+    {
+      "epoch": 0.9058710298363811,
+      "grad_norm": 1.5403794050216675,
+      "learning_rate": 0.00011480810648869327,
+      "loss": 0.9593,
+      "step": 23530
+    },
+    {
+      "epoch": 0.9060635226179018,
+      "grad_norm": 0.9871656894683838,
+      "learning_rate": 0.00011477819959104145,
+      "loss": 1.0469,
+      "step": 23535
+    },
+    {
+      "epoch": 0.9062560153994226,
+      "grad_norm": 1.037180781364441,
+      "learning_rate": 0.00011474829134202121,
+      "loss": 0.8791,
+      "step": 23540
+    },
+    {
+      "epoch": 0.9064485081809432,
+      "grad_norm": 0.9241686463356018,
+      "learning_rate": 0.00011471838174436738,
+      "loss": 1.1382,
+      "step": 23545
+    },
+    {
+      "epoch": 0.9066410009624639,
+      "grad_norm": 1.891224980354309,
+      "learning_rate": 0.00011468847080081506,
+      "loss": 1.0493,
+      "step": 23550
+    },
+    {
+      "epoch": 0.9068334937439846,
+      "grad_norm": 1.9200764894485474,
+      "learning_rate": 0.00011465855851409939,
+      "loss": 1.0265,
+      "step": 23555
+    },
+    {
+      "epoch": 0.9070259865255053,
+      "grad_norm": 0.8929172158241272,
+      "learning_rate": 0.00011462864488695563,
+      "loss": 1.1777,
+      "step": 23560
+    },
+    {
+      "epoch": 0.907218479307026,
+      "grad_norm": 2.08207368850708,
+      "learning_rate": 0.00011459872992211922,
+      "loss": 0.9146,
+      "step": 23565
+    },
+    {
+      "epoch": 0.9074109720885467,
+      "grad_norm": 1.6728849411010742,
+      "learning_rate": 0.00011456881362232567,
+      "loss": 1.1298,
+      "step": 23570
+    },
+    {
+      "epoch": 0.9076034648700674,
+      "grad_norm": 2.403190851211548,
+      "learning_rate": 0.0001145388959903106,
+      "loss": 1.1824,
+      "step": 23575
+    },
+    {
+      "epoch": 0.907795957651588,
+      "grad_norm": 1.7039401531219482,
+      "learning_rate": 0.00011450897702880981,
+      "loss": 1.1583,
+      "step": 23580
+    },
+    {
+      "epoch": 0.9079884504331087,
+      "grad_norm": 1.2045179605484009,
+      "learning_rate": 0.0001144790567405592,
+      "loss": 1.2031,
+      "step": 23585
+    },
+    {
+      "epoch": 0.9081809432146295,
+      "grad_norm": 1.4418444633483887,
+      "learning_rate": 0.00011444913512829476,
+      "loss": 1.1499,
+      "step": 23590
+    },
+    {
+      "epoch": 0.9083734359961502,
+      "grad_norm": 2.100736379623413,
+      "learning_rate": 0.00011441921219475264,
+      "loss": 1.1585,
+      "step": 23595
+    },
+    {
+      "epoch": 0.9085659287776708,
+      "grad_norm": 1.10848867893219,
+      "learning_rate": 0.00011438928794266905,
+      "loss": 1.2494,
+      "step": 23600
+    },
+    {
+      "epoch": 0.9087584215591915,
+      "grad_norm": 1.8016185760498047,
+      "learning_rate": 0.00011435936237478045,
+      "loss": 1.061,
+      "step": 23605
+    },
+    {
+      "epoch": 0.9089509143407122,
+      "grad_norm": 1.1875159740447998,
+      "learning_rate": 0.00011432943549382325,
+      "loss": 1.1222,
+      "step": 23610
+    },
+    {
+      "epoch": 0.909143407122233,
+      "grad_norm": 1.8396415710449219,
+      "learning_rate": 0.00011429950730253407,
+      "loss": 1.1891,
+      "step": 23615
+    },
+    {
+      "epoch": 0.9093358999037536,
+      "grad_norm": 1.5621596574783325,
+      "learning_rate": 0.00011426957780364973,
+      "loss": 1.0585,
+      "step": 23620
+    },
+    {
+      "epoch": 0.9095283926852743,
+      "grad_norm": 1.253409504890442,
+      "learning_rate": 0.000114239646999907,
+      "loss": 1.1791,
+      "step": 23625
+    },
+    {
+      "epoch": 0.909720885466795,
+      "grad_norm": 2.3588509559631348,
+      "learning_rate": 0.00011420971489404287,
+      "loss": 1.1373,
+      "step": 23630
+    },
+    {
+      "epoch": 0.9099133782483156,
+      "grad_norm": 1.597448706626892,
+      "learning_rate": 0.00011417978148879445,
+      "loss": 1.2299,
+      "step": 23635
+    },
+    {
+      "epoch": 0.9101058710298364,
+      "grad_norm": 0.90355384349823,
+      "learning_rate": 0.00011414984678689895,
+      "loss": 0.9769,
+      "step": 23640
+    },
+    {
+      "epoch": 0.9102983638113571,
+      "grad_norm": 1.7792707681655884,
+      "learning_rate": 0.00011411991079109368,
+      "loss": 1.0331,
+      "step": 23645
+    },
+    {
+      "epoch": 0.9104908565928778,
+      "grad_norm": 1.5718694925308228,
+      "learning_rate": 0.00011408997350411614,
+      "loss": 1.1637,
+      "step": 23650
+    },
+    {
+      "epoch": 0.9106833493743984,
+      "grad_norm": 1.0493242740631104,
+      "learning_rate": 0.00011406003492870383,
+      "loss": 0.9587,
+      "step": 23655
+    },
+    {
+      "epoch": 0.9108758421559191,
+      "grad_norm": 1.3035274744033813,
+      "learning_rate": 0.00011403009506759445,
+      "loss": 1.0251,
+      "step": 23660
+    },
+    {
+      "epoch": 0.9110683349374399,
+      "grad_norm": 1.4814205169677734,
+      "learning_rate": 0.00011400015392352585,
+      "loss": 1.0466,
+      "step": 23665
+    },
+    {
+      "epoch": 0.9112608277189606,
+      "grad_norm": 1.178098440170288,
+      "learning_rate": 0.0001139702114992359,
+      "loss": 0.9776,
+      "step": 23670
+    },
+    {
+      "epoch": 0.9114533205004812,
+      "grad_norm": 1.1865342855453491,
+      "learning_rate": 0.00011394026779746267,
+      "loss": 1.0895,
+      "step": 23675
+    },
+    {
+      "epoch": 0.9116458132820019,
+      "grad_norm": 0.8956665992736816,
+      "learning_rate": 0.00011391032282094429,
+      "loss": 1.0819,
+      "step": 23680
+    },
+    {
+      "epoch": 0.9118383060635226,
+      "grad_norm": 1.0805063247680664,
+      "learning_rate": 0.00011388037657241904,
+      "loss": 0.9655,
+      "step": 23685
+    },
+    {
+      "epoch": 0.9120307988450433,
+      "grad_norm": 1.116804838180542,
+      "learning_rate": 0.00011385042905462526,
+      "loss": 0.945,
+      "step": 23690
+    },
+    {
+      "epoch": 0.912223291626564,
+      "grad_norm": 1.1748266220092773,
+      "learning_rate": 0.00011382048027030155,
+      "loss": 1.0803,
+      "step": 23695
+    },
+    {
+      "epoch": 0.9124157844080847,
+      "grad_norm": 1.5685302019119263,
+      "learning_rate": 0.00011379053022218645,
+      "loss": 1.1678,
+      "step": 23700
+    },
+    {
+      "epoch": 0.9126082771896054,
+      "grad_norm": 1.1496498584747314,
+      "learning_rate": 0.00011376057891301873,
+      "loss": 1.007,
+      "step": 23705
+    },
+    {
+      "epoch": 0.9128007699711261,
+      "grad_norm": 1.3414727449417114,
+      "learning_rate": 0.0001137306263455372,
+      "loss": 1.0309,
+      "step": 23710
+    },
+    {
+      "epoch": 0.9129932627526468,
+      "grad_norm": 0.9307123422622681,
+      "learning_rate": 0.00011370067252248085,
+      "loss": 1.2324,
+      "step": 23715
+    },
+    {
+      "epoch": 0.9131857555341675,
+      "grad_norm": 2.051027536392212,
+      "learning_rate": 0.00011367071744658875,
+      "loss": 1.2293,
+      "step": 23720
+    },
+    {
+      "epoch": 0.9133782483156881,
+      "grad_norm": 1.3794739246368408,
+      "learning_rate": 0.00011364076112060014,
+      "loss": 1.1828,
+      "step": 23725
+    },
+    {
+      "epoch": 0.9135707410972088,
+      "grad_norm": 1.082683801651001,
+      "learning_rate": 0.00011361080354725424,
+      "loss": 0.9434,
+      "step": 23730
+    },
+    {
+      "epoch": 0.9137632338787296,
+      "grad_norm": 2.3093974590301514,
+      "learning_rate": 0.00011358084472929054,
+      "loss": 1.1994,
+      "step": 23735
+    },
+    {
+      "epoch": 0.9139557266602503,
+      "grad_norm": 1.007186770439148,
+      "learning_rate": 0.00011355088466944855,
+      "loss": 1.0195,
+      "step": 23740
+    },
+    {
+      "epoch": 0.9141482194417709,
+      "grad_norm": 1.239039659500122,
+      "learning_rate": 0.00011352092337046791,
+      "loss": 1.1732,
+      "step": 23745
+    },
+    {
+      "epoch": 0.9143407122232916,
+      "grad_norm": 1.795091986656189,
+      "learning_rate": 0.00011349096083508841,
+      "loss": 1.1553,
+      "step": 23750
+    },
+    {
+      "epoch": 0.9145332050048123,
+      "grad_norm": 1.2720720767974854,
+      "learning_rate": 0.00011346099706604991,
+      "loss": 1.0455,
+      "step": 23755
+    },
+    {
+      "epoch": 0.9147256977863331,
+      "grad_norm": 1.3282806873321533,
+      "learning_rate": 0.00011343103206609241,
+      "loss": 0.9086,
+      "step": 23760
+    },
+    {
+      "epoch": 0.9149181905678537,
+      "grad_norm": 1.055714726448059,
+      "learning_rate": 0.00011340106583795597,
+      "loss": 1.2688,
+      "step": 23765
+    },
+    {
+      "epoch": 0.9151106833493744,
+      "grad_norm": 1.3557606935501099,
+      "learning_rate": 0.00011337109838438085,
+      "loss": 1.1387,
+      "step": 23770
+    },
+    {
+      "epoch": 0.9153031761308951,
+      "grad_norm": 1.2776196002960205,
+      "learning_rate": 0.00011334112970810737,
+      "loss": 0.9351,
+      "step": 23775
+    },
+    {
+      "epoch": 0.9154956689124157,
+      "grad_norm": 1.4037102460861206,
+      "learning_rate": 0.00011331115981187595,
+      "loss": 1.0117,
+      "step": 23780
+    },
+    {
+      "epoch": 0.9156881616939365,
+      "grad_norm": 0.8925278782844543,
+      "learning_rate": 0.00011328118869842712,
+      "loss": 1.108,
+      "step": 23785
+    },
+    {
+      "epoch": 0.9158806544754572,
+      "grad_norm": 2.050219774246216,
+      "learning_rate": 0.0001132512163705016,
+      "loss": 1.1364,
+      "step": 23790
+    },
+    {
+      "epoch": 0.9160731472569779,
+      "grad_norm": 1.0239925384521484,
+      "learning_rate": 0.0001132212428308401,
+      "loss": 1.0919,
+      "step": 23795
+    },
+    {
+      "epoch": 0.9162656400384985,
+      "grad_norm": 1.554990530014038,
+      "learning_rate": 0.00011319126808218355,
+      "loss": 1.1526,
+      "step": 23800
+    },
+    {
+      "epoch": 0.9164581328200192,
+      "grad_norm": 1.433929443359375,
+      "learning_rate": 0.00011316129212727292,
+      "loss": 1.1302,
+      "step": 23805
+    },
+    {
+      "epoch": 0.91665062560154,
+      "grad_norm": 1.6558189392089844,
+      "learning_rate": 0.00011313131496884927,
+      "loss": 1.1572,
+      "step": 23810
+    },
+    {
+      "epoch": 0.9168431183830607,
+      "grad_norm": 1.2385213375091553,
+      "learning_rate": 0.00011310133660965387,
+      "loss": 1.0693,
+      "step": 23815
+    },
+    {
+      "epoch": 0.9170356111645813,
+      "grad_norm": 1.636640191078186,
+      "learning_rate": 0.00011307135705242806,
+      "loss": 1.0358,
+      "step": 23820
+    },
+    {
+      "epoch": 0.917228103946102,
+      "grad_norm": 1.5593379735946655,
+      "learning_rate": 0.0001130413762999132,
+      "loss": 1.1899,
+      "step": 23825
+    },
+    {
+      "epoch": 0.9174205967276227,
+      "grad_norm": 1.4657301902770996,
+      "learning_rate": 0.00011301139435485086,
+      "loss": 1.0543,
+      "step": 23830
+    },
+    {
+      "epoch": 0.9176130895091434,
+      "grad_norm": 1.0801408290863037,
+      "learning_rate": 0.00011298141121998273,
+      "loss": 1.181,
+      "step": 23835
+    },
+    {
+      "epoch": 0.9178055822906641,
+      "grad_norm": 1.6339335441589355,
+      "learning_rate": 0.00011295142689805052,
+      "loss": 1.197,
+      "step": 23840
+    },
+    {
+      "epoch": 0.9179980750721848,
+      "grad_norm": 1.4383010864257812,
+      "learning_rate": 0.00011292144139179612,
+      "loss": 0.98,
+      "step": 23845
+    },
+    {
+      "epoch": 0.9181905678537055,
+      "grad_norm": 1.440516471862793,
+      "learning_rate": 0.00011289145470396152,
+      "loss": 1.1183,
+      "step": 23850
+    },
+    {
+      "epoch": 0.9183830606352262,
+      "grad_norm": 0.940071165561676,
+      "learning_rate": 0.00011286146683728876,
+      "loss": 1.0116,
+      "step": 23855
+    },
+    {
+      "epoch": 0.9185755534167469,
+      "grad_norm": 1.1691455841064453,
+      "learning_rate": 0.00011283147779452003,
+      "loss": 1.0608,
+      "step": 23860
+    },
+    {
+      "epoch": 0.9187680461982676,
+      "grad_norm": 1.218925952911377,
+      "learning_rate": 0.00011280148757839771,
+      "loss": 1.0606,
+      "step": 23865
+    },
+    {
+      "epoch": 0.9189605389797882,
+      "grad_norm": 0.890487790107727,
+      "learning_rate": 0.00011277149619166414,
+      "loss": 0.8924,
+      "step": 23870
+    },
+    {
+      "epoch": 0.9191530317613089,
+      "grad_norm": 1.4061552286148071,
+      "learning_rate": 0.00011274150363706182,
+      "loss": 1.1201,
+      "step": 23875
+    },
+    {
+      "epoch": 0.9193455245428297,
+      "grad_norm": 2.254596471786499,
+      "learning_rate": 0.00011271150991733341,
+      "loss": 1.1457,
+      "step": 23880
+    },
+    {
+      "epoch": 0.9195380173243504,
+      "grad_norm": 1.2079321146011353,
+      "learning_rate": 0.00011268151503522164,
+      "loss": 1.0976,
+      "step": 23885
+    },
+    {
+      "epoch": 0.919730510105871,
+      "grad_norm": 1.8789191246032715,
+      "learning_rate": 0.00011265151899346931,
+      "loss": 1.0205,
+      "step": 23890
+    },
+    {
+      "epoch": 0.9199230028873917,
+      "grad_norm": 1.5862555503845215,
+      "learning_rate": 0.00011262152179481938,
+      "loss": 1.1655,
+      "step": 23895
+    },
+    {
+      "epoch": 0.9201154956689124,
+      "grad_norm": 1.342571496963501,
+      "learning_rate": 0.00011259152344201492,
+      "loss": 1.1423,
+      "step": 23900
+    },
+    {
+      "epoch": 0.9203079884504332,
+      "grad_norm": 1.4979714155197144,
+      "learning_rate": 0.00011256152393779901,
+      "loss": 1.0377,
+      "step": 23905
+    },
+    {
+      "epoch": 0.9205004812319538,
+      "grad_norm": 2.0960614681243896,
+      "learning_rate": 0.00011253152328491496,
+      "loss": 1.0474,
+      "step": 23910
+    },
+    {
+      "epoch": 0.9206929740134745,
+      "grad_norm": 1.4610568284988403,
+      "learning_rate": 0.00011250152148610613,
+      "loss": 1.1822,
+      "step": 23915
+    },
+    {
+      "epoch": 0.9208854667949952,
+      "grad_norm": 1.4984396696090698,
+      "learning_rate": 0.000112471518544116,
+      "loss": 0.9195,
+      "step": 23920
+    },
+    {
+      "epoch": 0.9210779595765158,
+      "grad_norm": 1.0438648462295532,
+      "learning_rate": 0.00011244151446168807,
+      "loss": 0.9973,
+      "step": 23925
+    },
+    {
+      "epoch": 0.9212704523580366,
+      "grad_norm": 1.51102876663208,
+      "learning_rate": 0.00011241150924156609,
+      "loss": 1.1294,
+      "step": 23930
+    },
+    {
+      "epoch": 0.9214629451395573,
+      "grad_norm": 1.743852138519287,
+      "learning_rate": 0.00011238150288649381,
+      "loss": 1.0525,
+      "step": 23935
+    },
+    {
+      "epoch": 0.921655437921078,
+      "grad_norm": 1.6502209901809692,
+      "learning_rate": 0.00011235149539921508,
+      "loss": 1.0904,
+      "step": 23940
+    },
+    {
+      "epoch": 0.9218479307025986,
+      "grad_norm": 1.1279150247573853,
+      "learning_rate": 0.00011232148678247393,
+      "loss": 1.0176,
+      "step": 23945
+    },
+    {
+      "epoch": 0.9220404234841193,
+      "grad_norm": 1.2196617126464844,
+      "learning_rate": 0.00011229147703901444,
+      "loss": 0.8492,
+      "step": 23950
+    },
+    {
+      "epoch": 0.9222329162656401,
+      "grad_norm": 1.4295985698699951,
+      "learning_rate": 0.0001122614661715808,
+      "loss": 1.1331,
+      "step": 23955
+    },
+    {
+      "epoch": 0.9224254090471607,
+      "grad_norm": 1.8270727396011353,
+      "learning_rate": 0.00011223145418291731,
+      "loss": 1.0813,
+      "step": 23960
+    },
+    {
+      "epoch": 0.9226179018286814,
+      "grad_norm": 1.327848196029663,
+      "learning_rate": 0.00011220144107576834,
+      "loss": 0.9626,
+      "step": 23965
+    },
+    {
+      "epoch": 0.9228103946102021,
+      "grad_norm": 1.5928614139556885,
+      "learning_rate": 0.00011217142685287842,
+      "loss": 1.1328,
+      "step": 23970
+    },
+    {
+      "epoch": 0.9230028873917228,
+      "grad_norm": 1.3863136768341064,
+      "learning_rate": 0.00011214141151699215,
+      "loss": 1.1175,
+      "step": 23975
+    },
+    {
+      "epoch": 0.9231953801732435,
+      "grad_norm": 1.4480311870574951,
+      "learning_rate": 0.00011211139507085422,
+      "loss": 0.9841,
+      "step": 23980
+    },
+    {
+      "epoch": 0.9233878729547642,
+      "grad_norm": 1.1457321643829346,
+      "learning_rate": 0.00011208137751720944,
+      "loss": 1.0934,
+      "step": 23985
+    },
+    {
+      "epoch": 0.9235803657362849,
+      "grad_norm": 1.614317774772644,
+      "learning_rate": 0.00011205135885880272,
+      "loss": 1.0216,
+      "step": 23990
+    },
+    {
+      "epoch": 0.9237728585178056,
+      "grad_norm": 1.0232197046279907,
+      "learning_rate": 0.00011202133909837906,
+      "loss": 1.1024,
+      "step": 23995
+    },
+    {
+      "epoch": 0.9239653512993262,
+      "grad_norm": 1.317587971687317,
+      "learning_rate": 0.00011199131823868358,
+      "loss": 1.2039,
+      "step": 24000
+    },
+    {
+      "epoch": 0.924157844080847,
+      "grad_norm": 1.0058887004852295,
+      "learning_rate": 0.00011196129628246148,
+      "loss": 1.068,
+      "step": 24005
+    },
+    {
+      "epoch": 0.9243503368623677,
+      "grad_norm": 1.1201086044311523,
+      "learning_rate": 0.00011193127323245809,
+      "loss": 1.0745,
+      "step": 24010
+    },
+    {
+      "epoch": 0.9245428296438883,
+      "grad_norm": 1.1013514995574951,
+      "learning_rate": 0.00011190124909141877,
+      "loss": 1.1211,
+      "step": 24015
+    },
+    {
+      "epoch": 0.924735322425409,
+      "grad_norm": 1.3184977769851685,
+      "learning_rate": 0.00011187122386208908,
+      "loss": 1.2223,
+      "step": 24020
+    },
+    {
+      "epoch": 0.9249278152069298,
+      "grad_norm": 1.053260087966919,
+      "learning_rate": 0.0001118411975472146,
+      "loss": 0.9556,
+      "step": 24025
+    },
+    {
+      "epoch": 0.9251203079884505,
+      "grad_norm": 1.6878530979156494,
+      "learning_rate": 0.00011181117014954105,
+      "loss": 0.9915,
+      "step": 24030
+    },
+    {
+      "epoch": 0.9253128007699711,
+      "grad_norm": 0.6977512836456299,
+      "learning_rate": 0.00011178114167181423,
+      "loss": 1.0629,
+      "step": 24035
+    },
+    {
+      "epoch": 0.9255052935514918,
+      "grad_norm": 1.4478118419647217,
+      "learning_rate": 0.00011175111211678006,
+      "loss": 1.1585,
+      "step": 24040
+    },
+    {
+      "epoch": 0.9256977863330125,
+      "grad_norm": 1.1715469360351562,
+      "learning_rate": 0.0001117210814871845,
+      "loss": 0.9811,
+      "step": 24045
+    },
+    {
+      "epoch": 0.9258902791145333,
+      "grad_norm": 1.2581676244735718,
+      "learning_rate": 0.00011169104978577369,
+      "loss": 1.0722,
+      "step": 24050
+    },
+    {
+      "epoch": 0.9260827718960539,
+      "grad_norm": 1.064126968383789,
+      "learning_rate": 0.00011166101701529385,
+      "loss": 0.983,
+      "step": 24055
+    },
+    {
+      "epoch": 0.9262752646775746,
+      "grad_norm": 1.0658310651779175,
+      "learning_rate": 0.00011163098317849123,
+      "loss": 1.0436,
+      "step": 24060
+    },
+    {
+      "epoch": 0.9264677574590953,
+      "grad_norm": 1.1407309770584106,
+      "learning_rate": 0.00011160094827811223,
+      "loss": 0.957,
+      "step": 24065
+    },
+    {
+      "epoch": 0.9266602502406159,
+      "grad_norm": 1.3511934280395508,
+      "learning_rate": 0.0001115709123169034,
+      "loss": 1.0751,
+      "step": 24070
+    },
+    {
+      "epoch": 0.9268527430221367,
+      "grad_norm": 1.416675329208374,
+      "learning_rate": 0.00011154087529761125,
+      "loss": 1.2558,
+      "step": 24075
+    },
+    {
+      "epoch": 0.9270452358036574,
+      "grad_norm": 1.5210598707199097,
+      "learning_rate": 0.00011151083722298252,
+      "loss": 0.945,
+      "step": 24080
+    },
+    {
+      "epoch": 0.9272377285851781,
+      "grad_norm": 1.1831562519073486,
+      "learning_rate": 0.00011148079809576399,
+      "loss": 1.0632,
+      "step": 24085
+    },
+    {
+      "epoch": 0.9274302213666987,
+      "grad_norm": 1.1255745887756348,
+      "learning_rate": 0.00011145075791870252,
+      "loss": 1.0556,
+      "step": 24090
+    },
+    {
+      "epoch": 0.9276227141482194,
+      "grad_norm": 1.213739275932312,
+      "learning_rate": 0.00011142071669454507,
+      "loss": 1.1672,
+      "step": 24095
+    },
+    {
+      "epoch": 0.9278152069297402,
+      "grad_norm": 1.0633070468902588,
+      "learning_rate": 0.00011139067442603877,
+      "loss": 1.1084,
+      "step": 24100
+    },
+    {
+      "epoch": 0.9280076997112608,
+      "grad_norm": 0.8085178136825562,
+      "learning_rate": 0.00011136063111593073,
+      "loss": 0.9406,
+      "step": 24105
+    },
+    {
+      "epoch": 0.9282001924927815,
+      "grad_norm": 1.0284137725830078,
+      "learning_rate": 0.00011133058676696823,
+      "loss": 1.1271,
+      "step": 24110
+    },
+    {
+      "epoch": 0.9283926852743022,
+      "grad_norm": 1.3331536054611206,
+      "learning_rate": 0.00011130054138189863,
+      "loss": 1.1798,
+      "step": 24115
+    },
+    {
+      "epoch": 0.9285851780558229,
+      "grad_norm": 1.1535316705703735,
+      "learning_rate": 0.00011127049496346939,
+      "loss": 0.9878,
+      "step": 24120
+    },
+    {
+      "epoch": 0.9287776708373436,
+      "grad_norm": 1.4867428541183472,
+      "learning_rate": 0.00011124044751442803,
+      "loss": 1.0329,
+      "step": 24125
+    },
+    {
+      "epoch": 0.9289701636188643,
+      "grad_norm": 0.9730262756347656,
+      "learning_rate": 0.00011121039903752224,
+      "loss": 1.0598,
+      "step": 24130
+    },
+    {
+      "epoch": 0.929162656400385,
+      "grad_norm": 2.8743958473205566,
+      "learning_rate": 0.0001111803495354997,
+      "loss": 0.9686,
+      "step": 24135
+    },
+    {
+      "epoch": 0.9293551491819056,
+      "grad_norm": 2.40010404586792,
+      "learning_rate": 0.00011115029901110825,
+      "loss": 1.1852,
+      "step": 24140
+    },
+    {
+      "epoch": 0.9295476419634263,
+      "grad_norm": 1.7937008142471313,
+      "learning_rate": 0.00011112024746709581,
+      "loss": 1.0724,
+      "step": 24145
+    },
+    {
+      "epoch": 0.9297401347449471,
+      "grad_norm": 1.1320335865020752,
+      "learning_rate": 0.00011109019490621044,
+      "loss": 1.0574,
+      "step": 24150
+    },
+    {
+      "epoch": 0.9299326275264678,
+      "grad_norm": 1.4164587259292603,
+      "learning_rate": 0.0001110601413312002,
+      "loss": 1.1617,
+      "step": 24155
+    },
+    {
+      "epoch": 0.9301251203079884,
+      "grad_norm": 1.665640950202942,
+      "learning_rate": 0.00011103008674481328,
+      "loss": 1.1355,
+      "step": 24160
+    },
+    {
+      "epoch": 0.9303176130895091,
+      "grad_norm": 1.509874701499939,
+      "learning_rate": 0.00011100003114979802,
+      "loss": 1.0042,
+      "step": 24165
+    },
+    {
+      "epoch": 0.9305101058710299,
+      "grad_norm": 1.702763319015503,
+      "learning_rate": 0.00011096997454890275,
+      "loss": 1.1039,
+      "step": 24170
+    },
+    {
+      "epoch": 0.9307025986525506,
+      "grad_norm": 0.9513006210327148,
+      "learning_rate": 0.00011093991694487597,
+      "loss": 1.1699,
+      "step": 24175
+    },
+    {
+      "epoch": 0.9308950914340712,
+      "grad_norm": 0.8490985035896301,
+      "learning_rate": 0.0001109098583404663,
+      "loss": 1.0261,
+      "step": 24180
+    },
+    {
+      "epoch": 0.9310875842155919,
+      "grad_norm": 1.2935947179794312,
+      "learning_rate": 0.00011087979873842233,
+      "loss": 1.1481,
+      "step": 24185
+    },
+    {
+      "epoch": 0.9312800769971126,
+      "grad_norm": 0.9786511063575745,
+      "learning_rate": 0.00011084973814149284,
+      "loss": 0.9385,
+      "step": 24190
+    },
+    {
+      "epoch": 0.9314725697786334,
+      "grad_norm": 1.1575534343719482,
+      "learning_rate": 0.00011081967655242668,
+      "loss": 0.966,
+      "step": 24195
+    },
+    {
+      "epoch": 0.931665062560154,
+      "grad_norm": 1.2462104558944702,
+      "learning_rate": 0.00011078961397397276,
+      "loss": 1.0768,
+      "step": 24200
+    },
+    {
+      "epoch": 0.9318575553416747,
+      "grad_norm": 1.0118372440338135,
+      "learning_rate": 0.00011075955040888011,
+      "loss": 1.033,
+      "step": 24205
+    },
+    {
+      "epoch": 0.9320500481231954,
+      "grad_norm": 1.2791298627853394,
+      "learning_rate": 0.00011072948585989789,
+      "loss": 0.9917,
+      "step": 24210
+    },
+    {
+      "epoch": 0.932242540904716,
+      "grad_norm": 1.0876747369766235,
+      "learning_rate": 0.00011069942032977522,
+      "loss": 1.1617,
+      "step": 24215
+    },
+    {
+      "epoch": 0.9324350336862368,
+      "grad_norm": 2.205045223236084,
+      "learning_rate": 0.00011066935382126144,
+      "loss": 0.996,
+      "step": 24220
+    },
+    {
+      "epoch": 0.9326275264677575,
+      "grad_norm": 1.4552040100097656,
+      "learning_rate": 0.00011063928633710596,
+      "loss": 1.0561,
+      "step": 24225
+    },
+    {
+      "epoch": 0.9328200192492782,
+      "grad_norm": 1.6502306461334229,
+      "learning_rate": 0.0001106092178800582,
+      "loss": 1.0455,
+      "step": 24230
+    },
+    {
+      "epoch": 0.9330125120307988,
+      "grad_norm": 1.861480474472046,
+      "learning_rate": 0.00011057914845286777,
+      "loss": 0.9918,
+      "step": 24235
+    },
+    {
+      "epoch": 0.9332050048123195,
+      "grad_norm": 1.1051427125930786,
+      "learning_rate": 0.00011054907805828427,
+      "loss": 1.0833,
+      "step": 24240
+    },
+    {
+      "epoch": 0.9333974975938403,
+      "grad_norm": 1.3723036050796509,
+      "learning_rate": 0.00011051900669905748,
+      "loss": 1.1338,
+      "step": 24245
+    },
+    {
+      "epoch": 0.933589990375361,
+      "grad_norm": 2.1225595474243164,
+      "learning_rate": 0.00011048893437793721,
+      "loss": 1.0879,
+      "step": 24250
+    },
+    {
+      "epoch": 0.9337824831568816,
+      "grad_norm": 1.4194490909576416,
+      "learning_rate": 0.00011045886109767336,
+      "loss": 1.1349,
+      "step": 24255
+    },
+    {
+      "epoch": 0.9339749759384023,
+      "grad_norm": 1.5220705270767212,
+      "learning_rate": 0.00011042878686101597,
+      "loss": 1.0283,
+      "step": 24260
+    },
+    {
+      "epoch": 0.934167468719923,
+      "grad_norm": 1.7352094650268555,
+      "learning_rate": 0.00011039871167071507,
+      "loss": 1.2378,
+      "step": 24265
+    },
+    {
+      "epoch": 0.9343599615014437,
+      "grad_norm": 2.2150251865386963,
+      "learning_rate": 0.00011036863552952088,
+      "loss": 1.1452,
+      "step": 24270
+    },
+    {
+      "epoch": 0.9345524542829644,
+      "grad_norm": 0.9917442202568054,
+      "learning_rate": 0.00011033855844018368,
+      "loss": 0.8594,
+      "step": 24275
+    },
+    {
+      "epoch": 0.9347449470644851,
+      "grad_norm": 1.0165492296218872,
+      "learning_rate": 0.00011030848040545378,
+      "loss": 1.2048,
+      "step": 24280
+    },
+    {
+      "epoch": 0.9349374398460057,
+      "grad_norm": 1.1605945825576782,
+      "learning_rate": 0.00011027840142808163,
+      "loss": 1.0497,
+      "step": 24285
+    },
+    {
+      "epoch": 0.9351299326275264,
+      "grad_norm": 1.9988510608673096,
+      "learning_rate": 0.00011024832151081778,
+      "loss": 1.1581,
+      "step": 24290
+    },
+    {
+      "epoch": 0.9353224254090472,
+      "grad_norm": 1.6264874935150146,
+      "learning_rate": 0.0001102182406564128,
+      "loss": 1.089,
+      "step": 24295
+    },
+    {
+      "epoch": 0.9355149181905679,
+      "grad_norm": 1.006515622138977,
+      "learning_rate": 0.0001101881588676174,
+      "loss": 0.9477,
+      "step": 24300
+    },
+    {
+      "epoch": 0.9357074109720885,
+      "grad_norm": 1.7896440029144287,
+      "learning_rate": 0.00011015807614718236,
+      "loss": 0.9834,
+      "step": 24305
+    },
+    {
+      "epoch": 0.9358999037536092,
+      "grad_norm": 1.7942471504211426,
+      "learning_rate": 0.00011012799249785854,
+      "loss": 1.1321,
+      "step": 24310
+    },
+    {
+      "epoch": 0.9360923965351299,
+      "grad_norm": 1.4186915159225464,
+      "learning_rate": 0.00011009790792239692,
+      "loss": 1.1373,
+      "step": 24315
+    },
+    {
+      "epoch": 0.9362848893166507,
+      "grad_norm": 0.9895558953285217,
+      "learning_rate": 0.00011006782242354852,
+      "loss": 1.0627,
+      "step": 24320
+    },
+    {
+      "epoch": 0.9364773820981713,
+      "grad_norm": 1.438604474067688,
+      "learning_rate": 0.00011003773600406442,
+      "loss": 1.0535,
+      "step": 24325
+    },
+    {
+      "epoch": 0.936669874879692,
+      "grad_norm": 1.1675355434417725,
+      "learning_rate": 0.00011000764866669586,
+      "loss": 1.1195,
+      "step": 24330
+    },
+    {
+      "epoch": 0.9368623676612127,
+      "grad_norm": 1.4219048023223877,
+      "learning_rate": 0.00010997756041419416,
+      "loss": 1.0578,
+      "step": 24335
+    },
+    {
+      "epoch": 0.9370548604427335,
+      "grad_norm": 2.0969905853271484,
+      "learning_rate": 0.00010994747124931062,
+      "loss": 1.1122,
+      "step": 24340
+    },
+    {
+      "epoch": 0.9372473532242541,
+      "grad_norm": 1.433214545249939,
+      "learning_rate": 0.00010991738117479673,
+      "loss": 1.0633,
+      "step": 24345
+    },
+    {
+      "epoch": 0.9374398460057748,
+      "grad_norm": 1.5070579051971436,
+      "learning_rate": 0.00010988729019340407,
+      "loss": 0.9797,
+      "step": 24350
+    },
+    {
+      "epoch": 0.9376323387872955,
+      "grad_norm": 0.9930442571640015,
+      "learning_rate": 0.00010985719830788417,
+      "loss": 0.9648,
+      "step": 24355
+    },
+    {
+      "epoch": 0.9378248315688161,
+      "grad_norm": 1.7221930027008057,
+      "learning_rate": 0.00010982710552098883,
+      "loss": 0.9712,
+      "step": 24360
+    },
+    {
+      "epoch": 0.9380173243503369,
+      "grad_norm": 1.0316636562347412,
+      "learning_rate": 0.00010979701183546976,
+      "loss": 1.0854,
+      "step": 24365
+    },
+    {
+      "epoch": 0.9382098171318576,
+      "grad_norm": 1.5038282871246338,
+      "learning_rate": 0.00010976691725407886,
+      "loss": 1.1366,
+      "step": 24370
+    },
+    {
+      "epoch": 0.9384023099133783,
+      "grad_norm": 1.6560473442077637,
+      "learning_rate": 0.00010973682177956808,
+      "loss": 1.1497,
+      "step": 24375
+    },
+    {
+      "epoch": 0.9385948026948989,
+      "grad_norm": 1.3170095682144165,
+      "learning_rate": 0.00010970672541468943,
+      "loss": 1.1246,
+      "step": 24380
+    },
+    {
+      "epoch": 0.9387872954764196,
+      "grad_norm": 1.8806647062301636,
+      "learning_rate": 0.00010967662816219506,
+      "loss": 1.1005,
+      "step": 24385
+    },
+    {
+      "epoch": 0.9389797882579404,
+      "grad_norm": 1.4266135692596436,
+      "learning_rate": 0.00010964653002483713,
+      "loss": 1.0151,
+      "step": 24390
+    },
+    {
+      "epoch": 0.939172281039461,
+      "grad_norm": 2.0234365463256836,
+      "learning_rate": 0.00010961643100536794,
+      "loss": 1.1368,
+      "step": 24395
+    },
+    {
+      "epoch": 0.9393647738209817,
+      "grad_norm": 1.403439998626709,
+      "learning_rate": 0.00010958633110653987,
+      "loss": 1.2229,
+      "step": 24400
+    },
+    {
+      "epoch": 0.9395572666025024,
+      "grad_norm": 1.2399364709854126,
+      "learning_rate": 0.0001095562303311053,
+      "loss": 1.0342,
+      "step": 24405
+    },
+    {
+      "epoch": 0.939749759384023,
+      "grad_norm": 1.1271092891693115,
+      "learning_rate": 0.00010952612868181673,
+      "loss": 1.1703,
+      "step": 24410
+    },
+    {
+      "epoch": 0.9399422521655438,
+      "grad_norm": 1.1748933792114258,
+      "learning_rate": 0.00010949602616142685,
+      "loss": 1.1191,
+      "step": 24415
+    },
+    {
+      "epoch": 0.9401347449470645,
+      "grad_norm": 1.4523221254348755,
+      "learning_rate": 0.00010946592277268825,
+      "loss": 1.1226,
+      "step": 24420
+    },
+    {
+      "epoch": 0.9403272377285852,
+      "grad_norm": 2.0243780612945557,
+      "learning_rate": 0.00010943581851835373,
+      "loss": 1.1353,
+      "step": 24425
+    },
+    {
+      "epoch": 0.9405197305101058,
+      "grad_norm": 1.318712830543518,
+      "learning_rate": 0.00010940571340117613,
+      "loss": 1.1688,
+      "step": 24430
+    },
+    {
+      "epoch": 0.9407122232916265,
+      "grad_norm": 1.6720144748687744,
+      "learning_rate": 0.00010937560742390833,
+      "loss": 1.1818,
+      "step": 24435
+    },
+    {
+      "epoch": 0.9409047160731473,
+      "grad_norm": 0.8121243119239807,
+      "learning_rate": 0.00010934550058930336,
+      "loss": 0.9643,
+      "step": 24440
+    },
+    {
+      "epoch": 0.941097208854668,
+      "grad_norm": 1.427445650100708,
+      "learning_rate": 0.00010931539290011425,
+      "loss": 1.0619,
+      "step": 24445
+    },
+    {
+      "epoch": 0.9412897016361886,
+      "grad_norm": 1.4542162418365479,
+      "learning_rate": 0.00010928528435909415,
+      "loss": 1.278,
+      "step": 24450
+    },
+    {
+      "epoch": 0.9414821944177093,
+      "grad_norm": 0.9350261688232422,
+      "learning_rate": 0.00010925517496899633,
+      "loss": 1.1934,
+      "step": 24455
+    },
+    {
+      "epoch": 0.94167468719923,
+      "grad_norm": 0.9128903150558472,
+      "learning_rate": 0.00010922506473257408,
+      "loss": 0.991,
+      "step": 24460
+    },
+    {
+      "epoch": 0.9418671799807508,
+      "grad_norm": 2.040079355239868,
+      "learning_rate": 0.00010919495365258077,
+      "loss": 1.2565,
+      "step": 24465
+    },
+    {
+      "epoch": 0.9420596727622714,
+      "grad_norm": 1.4088531732559204,
+      "learning_rate": 0.00010916484173176984,
+      "loss": 1.0567,
+      "step": 24470
+    },
+    {
+      "epoch": 0.9422521655437921,
+      "grad_norm": 1.0955448150634766,
+      "learning_rate": 0.00010913472897289485,
+      "loss": 0.9855,
+      "step": 24475
+    },
+    {
+      "epoch": 0.9424446583253128,
+      "grad_norm": 1.6121997833251953,
+      "learning_rate": 0.00010910461537870942,
+      "loss": 1.0122,
+      "step": 24480
+    },
+    {
+      "epoch": 0.9426371511068335,
+      "grad_norm": 1.548582911491394,
+      "learning_rate": 0.00010907450095196718,
+      "loss": 0.9924,
+      "step": 24485
+    },
+    {
+      "epoch": 0.9428296438883542,
+      "grad_norm": 1.2104709148406982,
+      "learning_rate": 0.00010904438569542202,
+      "loss": 1.0338,
+      "step": 24490
+    },
+    {
+      "epoch": 0.9430221366698749,
+      "grad_norm": 1.3403939008712769,
+      "learning_rate": 0.00010901426961182764,
+      "loss": 1.0474,
+      "step": 24495
+    },
+    {
+      "epoch": 0.9432146294513956,
+      "grad_norm": 1.8973404169082642,
+      "learning_rate": 0.00010898415270393802,
+      "loss": 0.9729,
+      "step": 24500
+    },
+    {
+      "epoch": 0.9434071222329162,
+      "grad_norm": 0.9071372747421265,
+      "learning_rate": 0.00010895403497450716,
+      "loss": 1.2973,
+      "step": 24505
+    },
+    {
+      "epoch": 0.943599615014437,
+      "grad_norm": 2.5191125869750977,
+      "learning_rate": 0.00010892391642628912,
+      "loss": 1.1838,
+      "step": 24510
+    },
+    {
+      "epoch": 0.9437921077959577,
+      "grad_norm": 1.3101541996002197,
+      "learning_rate": 0.00010889379706203804,
+      "loss": 1.0826,
+      "step": 24515
+    },
+    {
+      "epoch": 0.9439846005774783,
+      "grad_norm": 1.1167513132095337,
+      "learning_rate": 0.00010886367688450811,
+      "loss": 0.9787,
+      "step": 24520
+    },
+    {
+      "epoch": 0.944177093358999,
+      "grad_norm": 1.8214826583862305,
+      "learning_rate": 0.0001088335558964537,
+      "loss": 1.1172,
+      "step": 24525
+    },
+    {
+      "epoch": 0.9443695861405197,
+      "grad_norm": 0.8352447748184204,
+      "learning_rate": 0.00010880343410062908,
+      "loss": 0.9896,
+      "step": 24530
+    },
+    {
+      "epoch": 0.9445620789220405,
+      "grad_norm": 1.6729201078414917,
+      "learning_rate": 0.00010877331149978873,
+      "loss": 1.1613,
+      "step": 24535
+    },
+    {
+      "epoch": 0.9447545717035611,
+      "grad_norm": 1.173158049583435,
+      "learning_rate": 0.00010874318809668718,
+      "loss": 1.206,
+      "step": 24540
+    },
+    {
+      "epoch": 0.9449470644850818,
+      "grad_norm": 2.4930291175842285,
+      "learning_rate": 0.00010871306389407898,
+      "loss": 1.2144,
+      "step": 24545
+    },
+    {
+      "epoch": 0.9451395572666025,
+      "grad_norm": 0.8936457633972168,
+      "learning_rate": 0.00010868293889471881,
+      "loss": 1.0495,
+      "step": 24550
+    },
+    {
+      "epoch": 0.9453320500481232,
+      "grad_norm": 1.9174963235855103,
+      "learning_rate": 0.00010865281310136142,
+      "loss": 1.2971,
+      "step": 24555
+    },
+    {
+      "epoch": 0.9455245428296439,
+      "grad_norm": 0.8766136169433594,
+      "learning_rate": 0.00010862268651676155,
+      "loss": 1.3837,
+      "step": 24560
+    },
+    {
+      "epoch": 0.9457170356111646,
+      "grad_norm": 1.577832818031311,
+      "learning_rate": 0.00010859255914367414,
+      "loss": 0.9897,
+      "step": 24565
+    },
+    {
+      "epoch": 0.9459095283926853,
+      "grad_norm": 1.1107361316680908,
+      "learning_rate": 0.00010856243098485412,
+      "loss": 1.0474,
+      "step": 24570
+    },
+    {
+      "epoch": 0.9461020211742059,
+      "grad_norm": 1.0004538297653198,
+      "learning_rate": 0.00010853230204305651,
+      "loss": 1.0261,
+      "step": 24575
+    },
+    {
+      "epoch": 0.9462945139557266,
+      "grad_norm": 0.9005370736122131,
+      "learning_rate": 0.00010850217232103639,
+      "loss": 0.9002,
+      "step": 24580
+    },
+    {
+      "epoch": 0.9464870067372474,
+      "grad_norm": 0.3695490062236786,
+      "learning_rate": 0.00010847204182154895,
+      "loss": 0.9869,
+      "step": 24585
+    },
+    {
+      "epoch": 0.9466794995187681,
+      "grad_norm": 0.9990583658218384,
+      "learning_rate": 0.00010844191054734938,
+      "loss": 1.0765,
+      "step": 24590
+    },
+    {
+      "epoch": 0.9468719923002887,
+      "grad_norm": 1.0551097393035889,
+      "learning_rate": 0.00010841177850119301,
+      "loss": 1.0853,
+      "step": 24595
+    },
+    {
+      "epoch": 0.9470644850818094,
+      "grad_norm": 1.0011540651321411,
+      "learning_rate": 0.00010838164568583526,
+      "loss": 1.2651,
+      "step": 24600
+    },
+    {
+      "epoch": 0.9472569778633301,
+      "grad_norm": 1.1266576051712036,
+      "learning_rate": 0.0001083515121040315,
+      "loss": 0.9916,
+      "step": 24605
+    },
+    {
+      "epoch": 0.9474494706448509,
+      "grad_norm": 1.543762445449829,
+      "learning_rate": 0.00010832137775853728,
+      "loss": 1.2188,
+      "step": 24610
+    },
+    {
+      "epoch": 0.9476419634263715,
+      "grad_norm": 1.1634615659713745,
+      "learning_rate": 0.00010829124265210822,
+      "loss": 0.953,
+      "step": 24615
+    },
+    {
+      "epoch": 0.9478344562078922,
+      "grad_norm": 2.0008814334869385,
+      "learning_rate": 0.00010826110678749992,
+      "loss": 1.1433,
+      "step": 24620
+    },
+    {
+      "epoch": 0.9480269489894129,
+      "grad_norm": 1.6230239868164062,
+      "learning_rate": 0.00010823097016746813,
+      "loss": 1.0228,
+      "step": 24625
+    },
+    {
+      "epoch": 0.9482194417709335,
+      "grad_norm": 0.9347220659255981,
+      "learning_rate": 0.00010820083279476865,
+      "loss": 1.0337,
+      "step": 24630
+    },
+    {
+      "epoch": 0.9484119345524543,
+      "grad_norm": 1.8142660856246948,
+      "learning_rate": 0.00010817069467215732,
+      "loss": 1.0862,
+      "step": 24635
+    },
+    {
+      "epoch": 0.948604427333975,
+      "grad_norm": 1.6012142896652222,
+      "learning_rate": 0.0001081405558023901,
+      "loss": 1.0592,
+      "step": 24640
+    },
+    {
+      "epoch": 0.9487969201154957,
+      "grad_norm": 2.480301856994629,
+      "learning_rate": 0.00010811041618822297,
+      "loss": 1.1688,
+      "step": 24645
+    },
+    {
+      "epoch": 0.9489894128970163,
+      "grad_norm": 1.339879035949707,
+      "learning_rate": 0.00010808027583241203,
+      "loss": 1.1535,
+      "step": 24650
+    },
+    {
+      "epoch": 0.9491819056785371,
+      "grad_norm": 2.2561163902282715,
+      "learning_rate": 0.00010805013473771337,
+      "loss": 1.1379,
+      "step": 24655
+    },
+    {
+      "epoch": 0.9493743984600578,
+      "grad_norm": 0.9310626983642578,
+      "learning_rate": 0.00010801999290688323,
+      "loss": 1.117,
+      "step": 24660
+    },
+    {
+      "epoch": 0.9495668912415784,
+      "grad_norm": 1.099223256111145,
+      "learning_rate": 0.00010798985034267786,
+      "loss": 0.9384,
+      "step": 24665
+    },
+    {
+      "epoch": 0.9497593840230991,
+      "grad_norm": 1.5492980480194092,
+      "learning_rate": 0.0001079597070478536,
+      "loss": 1.0908,
+      "step": 24670
+    },
+    {
+      "epoch": 0.9499518768046198,
+      "grad_norm": 0.9033131003379822,
+      "learning_rate": 0.00010792956302516688,
+      "loss": 1.0303,
+      "step": 24675
+    },
+    {
+      "epoch": 0.9501443695861406,
+      "grad_norm": 1.0277695655822754,
+      "learning_rate": 0.00010789941827737411,
+      "loss": 1.1463,
+      "step": 24680
+    },
+    {
+      "epoch": 0.9503368623676612,
+      "grad_norm": 1.6218613386154175,
+      "learning_rate": 0.00010786927280723192,
+      "loss": 1.0647,
+      "step": 24685
+    },
+    {
+      "epoch": 0.9505293551491819,
+      "grad_norm": 0.9777045249938965,
+      "learning_rate": 0.00010783912661749682,
+      "loss": 0.9672,
+      "step": 24690
+    },
+    {
+      "epoch": 0.9507218479307026,
+      "grad_norm": 1.7008765935897827,
+      "learning_rate": 0.00010780897971092554,
+      "loss": 1.0837,
+      "step": 24695
+    },
+    {
+      "epoch": 0.9509143407122232,
+      "grad_norm": 1.286529779434204,
+      "learning_rate": 0.00010777883209027477,
+      "loss": 1.0237,
+      "step": 24700
+    },
+    {
+      "epoch": 0.951106833493744,
+      "grad_norm": 1.0446587800979614,
+      "learning_rate": 0.00010774868375830133,
+      "loss": 1.1015,
+      "step": 24705
+    },
+    {
+      "epoch": 0.9512993262752647,
+      "grad_norm": 1.1524755954742432,
+      "learning_rate": 0.00010771853471776215,
+      "loss": 1.0399,
+      "step": 24710
+    },
+    {
+      "epoch": 0.9514918190567854,
+      "grad_norm": 1.3211500644683838,
+      "learning_rate": 0.00010768838497141404,
+      "loss": 1.1109,
+      "step": 24715
+    },
+    {
+      "epoch": 0.951684311838306,
+      "grad_norm": 1.840817928314209,
+      "learning_rate": 0.00010765823452201406,
+      "loss": 1.1012,
+      "step": 24720
+    },
+    {
+      "epoch": 0.9518768046198267,
+      "grad_norm": 1.2010953426361084,
+      "learning_rate": 0.00010762808337231931,
+      "loss": 0.9257,
+      "step": 24725
+    },
+    {
+      "epoch": 0.9520692974013475,
+      "grad_norm": 1.1567620038986206,
+      "learning_rate": 0.00010759793152508684,
+      "loss": 1.1649,
+      "step": 24730
+    },
+    {
+      "epoch": 0.9522617901828682,
+      "grad_norm": 1.1958009004592896,
+      "learning_rate": 0.00010756777898307384,
+      "loss": 1.0035,
+      "step": 24735
+    },
+    {
+      "epoch": 0.9524542829643888,
+      "grad_norm": 1.7795593738555908,
+      "learning_rate": 0.00010753762574903763,
+      "loss": 1.1245,
+      "step": 24740
+    },
+    {
+      "epoch": 0.9526467757459095,
+      "grad_norm": 1.2421553134918213,
+      "learning_rate": 0.00010750747182573544,
+      "loss": 1.1169,
+      "step": 24745
+    },
+    {
+      "epoch": 0.9528392685274302,
+      "grad_norm": 2.212799549102783,
+      "learning_rate": 0.00010747731721592469,
+      "loss": 1.1195,
+      "step": 24750
+    },
+    {
+      "epoch": 0.953031761308951,
+      "grad_norm": 1.3483097553253174,
+      "learning_rate": 0.00010744716192236284,
+      "loss": 1.0309,
+      "step": 24755
+    },
+    {
+      "epoch": 0.9532242540904716,
+      "grad_norm": 0.8752015233039856,
+      "learning_rate": 0.00010741700594780734,
+      "loss": 0.9972,
+      "step": 24760
+    },
+    {
+      "epoch": 0.9534167468719923,
+      "grad_norm": 0.778307318687439,
+      "learning_rate": 0.00010738684929501576,
+      "loss": 0.9989,
+      "step": 24765
+    },
+    {
+      "epoch": 0.953609239653513,
+      "grad_norm": 1.3148306608200073,
+      "learning_rate": 0.00010735669196674578,
+      "loss": 1.0519,
+      "step": 24770
+    },
+    {
+      "epoch": 0.9538017324350336,
+      "grad_norm": 1.7633293867111206,
+      "learning_rate": 0.00010732653396575504,
+      "loss": 1.0377,
+      "step": 24775
+    },
+    {
+      "epoch": 0.9539942252165544,
+      "grad_norm": 1.5333178043365479,
+      "learning_rate": 0.00010729637529480132,
+      "loss": 0.9923,
+      "step": 24780
+    },
+    {
+      "epoch": 0.9541867179980751,
+      "grad_norm": 1.2737149000167847,
+      "learning_rate": 0.0001072662159566424,
+      "loss": 1.2048,
+      "step": 24785
+    },
+    {
+      "epoch": 0.9543792107795958,
+      "grad_norm": 1.3000686168670654,
+      "learning_rate": 0.00010723605595403616,
+      "loss": 0.8916,
+      "step": 24790
+    },
+    {
+      "epoch": 0.9545717035611164,
+      "grad_norm": 1.2083548307418823,
+      "learning_rate": 0.00010720589528974056,
+      "loss": 1.0176,
+      "step": 24795
+    },
+    {
+      "epoch": 0.9547641963426372,
+      "grad_norm": 1.8047492504119873,
+      "learning_rate": 0.00010717573396651355,
+      "loss": 1.094,
+      "step": 24800
+    },
+    {
+      "epoch": 0.9549566891241579,
+      "grad_norm": 1.2499585151672363,
+      "learning_rate": 0.00010714557198711321,
+      "loss": 1.2239,
+      "step": 24805
+    },
+    {
+      "epoch": 0.9551491819056785,
+      "grad_norm": 1.58943510055542,
+      "learning_rate": 0.00010711540935429764,
+      "loss": 1.1283,
+      "step": 24810
+    },
+    {
+      "epoch": 0.9553416746871992,
+      "grad_norm": 1.0315110683441162,
+      "learning_rate": 0.00010708524607082502,
+      "loss": 0.9449,
+      "step": 24815
+    },
+    {
+      "epoch": 0.9555341674687199,
+      "grad_norm": 1.182955265045166,
+      "learning_rate": 0.00010705508213945362,
+      "loss": 0.9075,
+      "step": 24820
+    },
+    {
+      "epoch": 0.9557266602502407,
+      "grad_norm": 1.0133094787597656,
+      "learning_rate": 0.00010702491756294164,
+      "loss": 1.0881,
+      "step": 24825
+    },
+    {
+      "epoch": 0.9559191530317613,
+      "grad_norm": 1.5667768716812134,
+      "learning_rate": 0.00010699475234404749,
+      "loss": 1.0341,
+      "step": 24830
+    },
+    {
+      "epoch": 0.956111645813282,
+      "grad_norm": 1.2236697673797607,
+      "learning_rate": 0.0001069645864855296,
+      "loss": 1.1377,
+      "step": 24835
+    },
+    {
+      "epoch": 0.9563041385948027,
+      "grad_norm": 1.6470654010772705,
+      "learning_rate": 0.0001069344199901464,
+      "loss": 1.1478,
+      "step": 24840
+    },
+    {
+      "epoch": 0.9564966313763233,
+      "grad_norm": 0.8241544365882874,
+      "learning_rate": 0.00010690425286065642,
+      "loss": 1.018,
+      "step": 24845
+    },
+    {
+      "epoch": 0.9566891241578441,
+      "grad_norm": 1.580249547958374,
+      "learning_rate": 0.00010687408509981827,
+      "loss": 1.2341,
+      "step": 24850
+    },
+    {
+      "epoch": 0.9568816169393648,
+      "grad_norm": 2.250305414199829,
+      "learning_rate": 0.00010684391671039056,
+      "loss": 1.0795,
+      "step": 24855
+    },
+    {
+      "epoch": 0.9570741097208855,
+      "grad_norm": 1.0578948259353638,
+      "learning_rate": 0.000106813747695132,
+      "loss": 1.0914,
+      "step": 24860
+    },
+    {
+      "epoch": 0.9572666025024061,
+      "grad_norm": 1.2390888929367065,
+      "learning_rate": 0.00010678357805680137,
+      "loss": 1.2108,
+      "step": 24865
+    },
+    {
+      "epoch": 0.9574590952839268,
+      "grad_norm": 2.08738112449646,
+      "learning_rate": 0.00010675340779815745,
+      "loss": 1.064,
+      "step": 24870
+    },
+    {
+      "epoch": 0.9576515880654476,
+      "grad_norm": 1.882812738418579,
+      "learning_rate": 0.00010672323692195912,
+      "loss": 0.9825,
+      "step": 24875
+    },
+    {
+      "epoch": 0.9578440808469683,
+      "grad_norm": 1.5878349542617798,
+      "learning_rate": 0.00010669306543096534,
+      "loss": 1.1894,
+      "step": 24880
+    },
+    {
+      "epoch": 0.9580365736284889,
+      "grad_norm": 1.1730328798294067,
+      "learning_rate": 0.00010666289332793503,
+      "loss": 1.0882,
+      "step": 24885
+    },
+    {
+      "epoch": 0.9582290664100096,
+      "grad_norm": 1.14130699634552,
+      "learning_rate": 0.00010663272061562726,
+      "loss": 0.9501,
+      "step": 24890
+    },
+    {
+      "epoch": 0.9584215591915303,
+      "grad_norm": 1.1763651371002197,
+      "learning_rate": 0.00010660254729680117,
+      "loss": 1.1355,
+      "step": 24895
+    },
+    {
+      "epoch": 0.958614051973051,
+      "grad_norm": 1.5443150997161865,
+      "learning_rate": 0.00010657237337421582,
+      "loss": 0.8956,
+      "step": 24900
+    },
+    {
+      "epoch": 0.9588065447545717,
+      "grad_norm": 1.0410224199295044,
+      "learning_rate": 0.0001065421988506305,
+      "loss": 1.0776,
+      "step": 24905
+    },
+    {
+      "epoch": 0.9589990375360924,
+      "grad_norm": 0.9984595775604248,
+      "learning_rate": 0.0001065120237288044,
+      "loss": 1.1329,
+      "step": 24910
+    },
+    {
+      "epoch": 0.9591915303176131,
+      "grad_norm": 1.8508329391479492,
+      "learning_rate": 0.00010648184801149689,
+      "loss": 1.0724,
+      "step": 24915
+    },
+    {
+      "epoch": 0.9593840230991337,
+      "grad_norm": 2.0840296745300293,
+      "learning_rate": 0.00010645167170146733,
+      "loss": 1.0316,
+      "step": 24920
+    },
+    {
+      "epoch": 0.9595765158806545,
+      "grad_norm": 1.6280990839004517,
+      "learning_rate": 0.00010642149480147509,
+      "loss": 0.9811,
+      "step": 24925
+    },
+    {
+      "epoch": 0.9597690086621752,
+      "grad_norm": 1.7951642274856567,
+      "learning_rate": 0.00010639131731427974,
+      "loss": 1.0428,
+      "step": 24930
+    },
+    {
+      "epoch": 0.9599615014436959,
+      "grad_norm": 1.0857096910476685,
+      "learning_rate": 0.00010636113924264073,
+      "loss": 0.9805,
+      "step": 24935
+    },
+    {
+      "epoch": 0.9601539942252165,
+      "grad_norm": 1.919179916381836,
+      "learning_rate": 0.00010633096058931766,
+      "loss": 1.0721,
+      "step": 24940
+    },
+    {
+      "epoch": 0.9603464870067372,
+      "grad_norm": 0.9968999028205872,
+      "learning_rate": 0.0001063007813570702,
+      "loss": 1.0928,
+      "step": 24945
+    },
+    {
+      "epoch": 0.960538979788258,
+      "grad_norm": 1.4042326211929321,
+      "learning_rate": 0.00010627060154865802,
+      "loss": 1.1576,
+      "step": 24950
+    },
+    {
+      "epoch": 0.9607314725697786,
+      "grad_norm": 1.0694444179534912,
+      "learning_rate": 0.00010624042116684088,
+      "loss": 1.0468,
+      "step": 24955
+    },
+    {
+      "epoch": 0.9609239653512993,
+      "grad_norm": 1.127734899520874,
+      "learning_rate": 0.00010621024021437855,
+      "loss": 1.1967,
+      "step": 24960
+    },
+    {
+      "epoch": 0.96111645813282,
+      "grad_norm": 0.9498047232627869,
+      "learning_rate": 0.0001061800586940309,
+      "loss": 1.0848,
+      "step": 24965
+    },
+    {
+      "epoch": 0.9613089509143408,
+      "grad_norm": 1.6244068145751953,
+      "learning_rate": 0.0001061498766085578,
+      "loss": 1.1116,
+      "step": 24970
+    },
+    {
+      "epoch": 0.9615014436958614,
+      "grad_norm": 1.0797539949417114,
+      "learning_rate": 0.00010611969396071926,
+      "loss": 1.025,
+      "step": 24975
+    },
+    {
+      "epoch": 0.9616939364773821,
+      "grad_norm": 1.1946651935577393,
+      "learning_rate": 0.00010608951075327522,
+      "loss": 1.3113,
+      "step": 24980
+    },
+    {
+      "epoch": 0.9618864292589028,
+      "grad_norm": 0.915337324142456,
+      "learning_rate": 0.00010605932698898576,
+      "loss": 1.1864,
+      "step": 24985
+    },
+    {
+      "epoch": 0.9620789220404234,
+      "grad_norm": 0.821142315864563,
+      "learning_rate": 0.00010602914267061101,
+      "loss": 0.9772,
+      "step": 24990
+    },
+    {
+      "epoch": 0.9622714148219442,
+      "grad_norm": 1.0821641683578491,
+      "learning_rate": 0.00010599895780091106,
+      "loss": 0.9488,
+      "step": 24995
+    },
+    {
+      "epoch": 0.9624639076034649,
+      "grad_norm": 1.600436806678772,
+      "learning_rate": 0.0001059687723826462,
+      "loss": 1.1065,
+      "step": 25000
+    },
+    {
+      "epoch": 0.9626564003849856,
+      "grad_norm": 1.296973466873169,
+      "learning_rate": 0.00010593858641857664,
+      "loss": 1.0329,
+      "step": 25005
+    },
+    {
+      "epoch": 0.9628488931665062,
+      "grad_norm": 1.2150559425354004,
+      "learning_rate": 0.00010590839991146269,
+      "loss": 1.176,
+      "step": 25010
+    },
+    {
+      "epoch": 0.9630413859480269,
+      "grad_norm": 1.5039284229278564,
+      "learning_rate": 0.00010587821286406469,
+      "loss": 1.055,
+      "step": 25015
+    },
+    {
+      "epoch": 0.9632338787295477,
+      "grad_norm": 1.5378166437149048,
+      "learning_rate": 0.00010584802527914308,
+      "loss": 1.1067,
+      "step": 25020
+    },
+    {
+      "epoch": 0.9634263715110684,
+      "grad_norm": 1.1952167749404907,
+      "learning_rate": 0.0001058178371594583,
+      "loss": 1.1889,
+      "step": 25025
+    },
+    {
+      "epoch": 0.963618864292589,
+      "grad_norm": 1.8712360858917236,
+      "learning_rate": 0.00010578764850777084,
+      "loss": 1.2401,
+      "step": 25030
+    },
+    {
+      "epoch": 0.9638113570741097,
+      "grad_norm": 1.0512574911117554,
+      "learning_rate": 0.00010575745932684125,
+      "loss": 1.1021,
+      "step": 25035
+    },
+    {
+      "epoch": 0.9640038498556304,
+      "grad_norm": 1.1200644969940186,
+      "learning_rate": 0.00010572726961943017,
+      "loss": 1.1023,
+      "step": 25040
+    },
+    {
+      "epoch": 0.9641963426371511,
+      "grad_norm": 1.448231816291809,
+      "learning_rate": 0.00010569707938829821,
+      "loss": 0.979,
+      "step": 25045
+    },
+    {
+      "epoch": 0.9643888354186718,
+      "grad_norm": 2.4415132999420166,
+      "learning_rate": 0.00010566688863620608,
+      "loss": 1.037,
+      "step": 25050
+    },
+    {
+      "epoch": 0.9645813282001925,
+      "grad_norm": 1.2661056518554688,
+      "learning_rate": 0.00010563669736591453,
+      "loss": 1.2452,
+      "step": 25055
+    },
+    {
+      "epoch": 0.9647738209817132,
+      "grad_norm": 1.1040503978729248,
+      "learning_rate": 0.00010560650558018434,
+      "loss": 1.0118,
+      "step": 25060
+    },
+    {
+      "epoch": 0.9649663137632338,
+      "grad_norm": 1.2488709688186646,
+      "learning_rate": 0.00010557631328177636,
+      "loss": 1.0467,
+      "step": 25065
+    },
+    {
+      "epoch": 0.9651588065447546,
+      "grad_norm": 1.267828106880188,
+      "learning_rate": 0.00010554612047345147,
+      "loss": 0.9722,
+      "step": 25070
+    },
+    {
+      "epoch": 0.9653512993262753,
+      "grad_norm": 1.5606194734573364,
+      "learning_rate": 0.00010551592715797058,
+      "loss": 1.0212,
+      "step": 25075
+    },
+    {
+      "epoch": 0.965543792107796,
+      "grad_norm": 1.2547963857650757,
+      "learning_rate": 0.0001054857333380947,
+      "loss": 0.9649,
+      "step": 25080
+    },
+    {
+      "epoch": 0.9657362848893166,
+      "grad_norm": 1.1586534976959229,
+      "learning_rate": 0.00010545553901658486,
+      "loss": 1.1205,
+      "step": 25085
+    },
+    {
+      "epoch": 0.9659287776708373,
+      "grad_norm": 1.03237783908844,
+      "learning_rate": 0.00010542534419620214,
+      "loss": 0.9806,
+      "step": 25090
+    },
+    {
+      "epoch": 0.9661212704523581,
+      "grad_norm": 1.0828287601470947,
+      "learning_rate": 0.00010539514887970758,
+      "loss": 1.1297,
+      "step": 25095
+    },
+    {
+      "epoch": 0.9663137632338787,
+      "grad_norm": 1.0693068504333496,
+      "learning_rate": 0.00010536495306986243,
+      "loss": 0.9167,
+      "step": 25100
+    },
+    {
+      "epoch": 0.9665062560153994,
+      "grad_norm": 0.9770272374153137,
+      "learning_rate": 0.00010533475676942785,
+      "loss": 1.1324,
+      "step": 25105
+    },
+    {
+      "epoch": 0.9666987487969201,
+      "grad_norm": 1.3809349536895752,
+      "learning_rate": 0.00010530455998116511,
+      "loss": 1.2411,
+      "step": 25110
+    },
+    {
+      "epoch": 0.9668912415784409,
+      "grad_norm": 1.8168786764144897,
+      "learning_rate": 0.00010527436270783551,
+      "loss": 1.0129,
+      "step": 25115
+    },
+    {
+      "epoch": 0.9670837343599615,
+      "grad_norm": 1.3123118877410889,
+      "learning_rate": 0.00010524416495220035,
+      "loss": 1.0677,
+      "step": 25120
+    },
+    {
+      "epoch": 0.9672762271414822,
+      "grad_norm": 1.329199194908142,
+      "learning_rate": 0.00010521396671702106,
+      "loss": 1.0804,
+      "step": 25125
+    },
+    {
+      "epoch": 0.9674687199230029,
+      "grad_norm": 1.5670257806777954,
+      "learning_rate": 0.00010518376800505907,
+      "loss": 1.1608,
+      "step": 25130
+    },
+    {
+      "epoch": 0.9676612127045235,
+      "grad_norm": 1.2058695554733276,
+      "learning_rate": 0.00010515356881907581,
+      "loss": 1.2903,
+      "step": 25135
+    },
+    {
+      "epoch": 0.9678537054860443,
+      "grad_norm": 1.4438813924789429,
+      "learning_rate": 0.0001051233691618328,
+      "loss": 1.2552,
+      "step": 25140
+    },
+    {
+      "epoch": 0.968046198267565,
+      "grad_norm": 1.1188582181930542,
+      "learning_rate": 0.00010509316903609167,
+      "loss": 1.1011,
+      "step": 25145
+    },
+    {
+      "epoch": 0.9682386910490857,
+      "grad_norm": 1.0020146369934082,
+      "learning_rate": 0.00010506296844461394,
+      "loss": 1.3431,
+      "step": 25150
+    },
+    {
+      "epoch": 0.9684311838306063,
+      "grad_norm": 1.0120351314544678,
+      "learning_rate": 0.00010503276739016128,
+      "loss": 1.0744,
+      "step": 25155
+    },
+    {
+      "epoch": 0.968623676612127,
+      "grad_norm": 1.0021151304244995,
+      "learning_rate": 0.0001050025658754954,
+      "loss": 1.0356,
+      "step": 25160
+    },
+    {
+      "epoch": 0.9688161693936478,
+      "grad_norm": 1.2012884616851807,
+      "learning_rate": 0.00010497236390337801,
+      "loss": 1.0556,
+      "step": 25165
+    },
+    {
+      "epoch": 0.9690086621751685,
+      "grad_norm": 2.0097708702087402,
+      "learning_rate": 0.00010494216147657086,
+      "loss": 1.1924,
+      "step": 25170
+    },
+    {
+      "epoch": 0.9692011549566891,
+      "grad_norm": 1.8217684030532837,
+      "learning_rate": 0.0001049119585978358,
+      "loss": 1.1365,
+      "step": 25175
+    },
+    {
+      "epoch": 0.9693936477382098,
+      "grad_norm": 2.422804832458496,
+      "learning_rate": 0.00010488175526993466,
+      "loss": 1.1675,
+      "step": 25180
+    },
+    {
+      "epoch": 0.9695861405197305,
+      "grad_norm": 1.59537672996521,
+      "learning_rate": 0.00010485155149562933,
+      "loss": 1.018,
+      "step": 25185
+    },
+    {
+      "epoch": 0.9697786333012512,
+      "grad_norm": 1.0549684762954712,
+      "learning_rate": 0.00010482134727768175,
+      "loss": 1.076,
+      "step": 25190
+    },
+    {
+      "epoch": 0.9699711260827719,
+      "grad_norm": 1.1994887590408325,
+      "learning_rate": 0.00010479114261885395,
+      "loss": 1.0012,
+      "step": 25195
+    },
+    {
+      "epoch": 0.9701636188642926,
+      "grad_norm": 1.6719814538955688,
+      "learning_rate": 0.00010476093752190784,
+      "loss": 1.1797,
+      "step": 25200
+    },
+    {
+      "epoch": 0.9703561116458133,
+      "grad_norm": 1.5435569286346436,
+      "learning_rate": 0.00010473073198960555,
+      "loss": 1.0019,
+      "step": 25205
+    },
+    {
+      "epoch": 0.9705486044273339,
+      "grad_norm": 1.0406394004821777,
+      "learning_rate": 0.00010470052602470917,
+      "loss": 1.0301,
+      "step": 25210
+    },
+    {
+      "epoch": 0.9707410972088547,
+      "grad_norm": 0.8994914889335632,
+      "learning_rate": 0.0001046703196299808,
+      "loss": 0.9683,
+      "step": 25215
+    },
+    {
+      "epoch": 0.9709335899903754,
+      "grad_norm": 1.6628937721252441,
+      "learning_rate": 0.00010464011280818266,
+      "loss": 1.2508,
+      "step": 25220
+    },
+    {
+      "epoch": 0.971126082771896,
+      "grad_norm": 1.7515249252319336,
+      "learning_rate": 0.00010460990556207693,
+      "loss": 1.3141,
+      "step": 25225
+    },
+    {
+      "epoch": 0.9713185755534167,
+      "grad_norm": 1.4734959602355957,
+      "learning_rate": 0.00010457969789442587,
+      "loss": 1.0589,
+      "step": 25230
+    },
+    {
+      "epoch": 0.9715110683349374,
+      "grad_norm": 1.102239966392517,
+      "learning_rate": 0.00010454948980799179,
+      "loss": 1.0115,
+      "step": 25235
+    },
+    {
+      "epoch": 0.9717035611164582,
+      "grad_norm": 1.9879227876663208,
+      "learning_rate": 0.000104519281305537,
+      "loss": 0.9781,
+      "step": 25240
+    },
+    {
+      "epoch": 0.9718960538979788,
+      "grad_norm": 1.2674641609191895,
+      "learning_rate": 0.00010448907238982387,
+      "loss": 1.1459,
+      "step": 25245
+    },
+    {
+      "epoch": 0.9720885466794995,
+      "grad_norm": 1.922810673713684,
+      "learning_rate": 0.00010445886306361479,
+      "loss": 1.0625,
+      "step": 25250
+    },
+    {
+      "epoch": 0.9722810394610202,
+      "grad_norm": 1.0224851369857788,
+      "learning_rate": 0.00010442865332967225,
+      "loss": 0.9894,
+      "step": 25255
+    },
+    {
+      "epoch": 0.9724735322425409,
+      "grad_norm": 1.3521889448165894,
+      "learning_rate": 0.00010439844319075868,
+      "loss": 1.0474,
+      "step": 25260
+    },
+    {
+      "epoch": 0.9726660250240616,
+      "grad_norm": 1.4089123010635376,
+      "learning_rate": 0.00010436823264963662,
+      "loss": 1.0874,
+      "step": 25265
+    },
+    {
+      "epoch": 0.9728585178055823,
+      "grad_norm": 1.0368316173553467,
+      "learning_rate": 0.00010433802170906863,
+      "loss": 1.0466,
+      "step": 25270
+    },
+    {
+      "epoch": 0.973051010587103,
+      "grad_norm": 1.541629672050476,
+      "learning_rate": 0.00010430781037181727,
+      "loss": 1.0565,
+      "step": 25275
+    },
+    {
+      "epoch": 0.9732435033686236,
+      "grad_norm": 0.7525898814201355,
+      "learning_rate": 0.00010427759864064521,
+      "loss": 0.9438,
+      "step": 25280
+    },
+    {
+      "epoch": 0.9734359961501444,
+      "grad_norm": 1.5569766759872437,
+      "learning_rate": 0.00010424738651831507,
+      "loss": 1.1061,
+      "step": 25285
+    },
+    {
+      "epoch": 0.9736284889316651,
+      "grad_norm": 1.4111804962158203,
+      "learning_rate": 0.0001042171740075896,
+      "loss": 1.2118,
+      "step": 25290
+    },
+    {
+      "epoch": 0.9738209817131858,
+      "grad_norm": 1.2276591062545776,
+      "learning_rate": 0.00010418696111123148,
+      "loss": 1.1494,
+      "step": 25295
+    },
+    {
+      "epoch": 0.9740134744947064,
+      "grad_norm": 1.320253849029541,
+      "learning_rate": 0.00010415674783200349,
+      "loss": 1.1241,
+      "step": 25300
+    },
+    {
+      "epoch": 0.9742059672762271,
+      "grad_norm": 1.6126363277435303,
+      "learning_rate": 0.00010412653417266849,
+      "loss": 1.068,
+      "step": 25305
+    },
+    {
+      "epoch": 0.9743984600577479,
+      "grad_norm": 1.2316113710403442,
+      "learning_rate": 0.00010409632013598924,
+      "loss": 1.124,
+      "step": 25310
+    },
+    {
+      "epoch": 0.9745909528392686,
+      "grad_norm": 1.3497854471206665,
+      "learning_rate": 0.00010406610572472866,
+      "loss": 1.0544,
+      "step": 25315
+    },
+    {
+      "epoch": 0.9747834456207892,
+      "grad_norm": 1.1057459115982056,
+      "learning_rate": 0.00010403589094164966,
+      "loss": 0.982,
+      "step": 25320
+    },
+    {
+      "epoch": 0.9749759384023099,
+      "grad_norm": 1.654336929321289,
+      "learning_rate": 0.00010400567578951515,
+      "loss": 1.1425,
+      "step": 25325
+    },
+    {
+      "epoch": 0.9751684311838306,
+      "grad_norm": 0.8302839398384094,
+      "learning_rate": 0.00010397546027108814,
+      "loss": 1.0543,
+      "step": 25330
+    },
+    {
+      "epoch": 0.9753609239653513,
+      "grad_norm": 3.0035483837127686,
+      "learning_rate": 0.00010394524438913161,
+      "loss": 1.0277,
+      "step": 25335
+    },
+    {
+      "epoch": 0.975553416746872,
+      "grad_norm": 1.4531986713409424,
+      "learning_rate": 0.00010391502814640864,
+      "loss": 1.1209,
+      "step": 25340
+    },
+    {
+      "epoch": 0.9757459095283927,
+      "grad_norm": 1.7825361490249634,
+      "learning_rate": 0.00010388481154568224,
+      "loss": 0.9739,
+      "step": 25345
+    },
+    {
+      "epoch": 0.9759384023099134,
+      "grad_norm": 0.9865066409111023,
+      "learning_rate": 0.00010385459458971558,
+      "loss": 1.0587,
+      "step": 25350
+    },
+    {
+      "epoch": 0.976130895091434,
+      "grad_norm": 1.5732641220092773,
+      "learning_rate": 0.00010382437728127176,
+      "loss": 1.1509,
+      "step": 25355
+    },
+    {
+      "epoch": 0.9763233878729548,
+      "grad_norm": 1.3935140371322632,
+      "learning_rate": 0.000103794159623114,
+      "loss": 1.0582,
+      "step": 25360
+    },
+    {
+      "epoch": 0.9765158806544755,
+      "grad_norm": 1.156906247138977,
+      "learning_rate": 0.0001037639416180055,
+      "loss": 0.9355,
+      "step": 25365
+    },
+    {
+      "epoch": 0.9767083734359961,
+      "grad_norm": 1.3732504844665527,
+      "learning_rate": 0.0001037337232687094,
+      "loss": 1.085,
+      "step": 25370
+    },
+    {
+      "epoch": 0.9769008662175168,
+      "grad_norm": 1.41510009765625,
+      "learning_rate": 0.00010370350457798907,
+      "loss": 1.0622,
+      "step": 25375
+    },
+    {
+      "epoch": 0.9770933589990375,
+      "grad_norm": 2.79821515083313,
+      "learning_rate": 0.00010367328554860783,
+      "loss": 1.1103,
+      "step": 25380
+    },
+    {
+      "epoch": 0.9772858517805583,
+      "grad_norm": 0.6538259387016296,
+      "learning_rate": 0.00010364306618332889,
+      "loss": 0.8355,
+      "step": 25385
+    },
+    {
+      "epoch": 0.9774783445620789,
+      "grad_norm": 1.126719355583191,
+      "learning_rate": 0.00010361284648491571,
+      "loss": 1.0243,
+      "step": 25390
+    },
+    {
+      "epoch": 0.9776708373435996,
+      "grad_norm": 1.7321900129318237,
+      "learning_rate": 0.00010358262645613166,
+      "loss": 1.1714,
+      "step": 25395
+    },
+    {
+      "epoch": 0.9778633301251203,
+      "grad_norm": 1.1369127035140991,
+      "learning_rate": 0.00010355240609974015,
+      "loss": 0.9301,
+      "step": 25400
+    },
+    {
+      "epoch": 0.978055822906641,
+      "grad_norm": 1.0569565296173096,
+      "learning_rate": 0.00010352218541850461,
+      "loss": 1.1441,
+      "step": 25405
+    },
+    {
+      "epoch": 0.9782483156881617,
+      "grad_norm": 1.2385330200195312,
+      "learning_rate": 0.00010349196441518855,
+      "loss": 1.0675,
+      "step": 25410
+    },
+    {
+      "epoch": 0.9784408084696824,
+      "grad_norm": 1.161966323852539,
+      "learning_rate": 0.00010346174309255552,
+      "loss": 1.0511,
+      "step": 25415
+    },
+    {
+      "epoch": 0.9786333012512031,
+      "grad_norm": 1.3913847208023071,
+      "learning_rate": 0.00010343152145336899,
+      "loss": 1.2011,
+      "step": 25420
+    },
+    {
+      "epoch": 0.9788257940327237,
+      "grad_norm": 2.150585412979126,
+      "learning_rate": 0.00010340129950039253,
+      "loss": 0.9328,
+      "step": 25425
+    },
+    {
+      "epoch": 0.9790182868142445,
+      "grad_norm": 0.9360420107841492,
+      "learning_rate": 0.00010337107723638979,
+      "loss": 1.0759,
+      "step": 25430
+    },
+    {
+      "epoch": 0.9792107795957652,
+      "grad_norm": 1.2746261358261108,
+      "learning_rate": 0.00010334085466412435,
+      "loss": 1.0884,
+      "step": 25435
+    },
+    {
+      "epoch": 0.9794032723772859,
+      "grad_norm": 1.4159737825393677,
+      "learning_rate": 0.00010331063178635991,
+      "loss": 0.8767,
+      "step": 25440
+    },
+    {
+      "epoch": 0.9795957651588065,
+      "grad_norm": 1.0822529792785645,
+      "learning_rate": 0.00010328040860586013,
+      "loss": 1.0087,
+      "step": 25445
+    },
+    {
+      "epoch": 0.9797882579403272,
+      "grad_norm": 1.3765298128128052,
+      "learning_rate": 0.00010325018512538868,
+      "loss": 1.1506,
+      "step": 25450
+    },
+    {
+      "epoch": 0.979980750721848,
+      "grad_norm": 1.1696947813034058,
+      "learning_rate": 0.00010321996134770935,
+      "loss": 1.1162,
+      "step": 25455
+    },
+    {
+      "epoch": 0.9801732435033687,
+      "grad_norm": 1.0629351139068604,
+      "learning_rate": 0.0001031897372755859,
+      "loss": 1.0682,
+      "step": 25460
+    },
+    {
+      "epoch": 0.9803657362848893,
+      "grad_norm": 1.7273883819580078,
+      "learning_rate": 0.00010315951291178208,
+      "loss": 1.0015,
+      "step": 25465
+    },
+    {
+      "epoch": 0.98055822906641,
+      "grad_norm": 1.154524803161621,
+      "learning_rate": 0.00010312928825906172,
+      "loss": 1.0113,
+      "step": 25470
+    },
+    {
+      "epoch": 0.9807507218479307,
+      "grad_norm": 1.5868383646011353,
+      "learning_rate": 0.0001030990633201887,
+      "loss": 1.1021,
+      "step": 25475
+    },
+    {
+      "epoch": 0.9809432146294514,
+      "grad_norm": 1.425208568572998,
+      "learning_rate": 0.00010306883809792687,
+      "loss": 1.0993,
+      "step": 25480
+    },
+    {
+      "epoch": 0.9811357074109721,
+      "grad_norm": 1.5670615434646606,
+      "learning_rate": 0.00010303861259504011,
+      "loss": 1.1032,
+      "step": 25485
+    },
+    {
+      "epoch": 0.9813282001924928,
+      "grad_norm": 0.9901431202888489,
+      "learning_rate": 0.00010300838681429239,
+      "loss": 1.0594,
+      "step": 25490
+    },
+    {
+      "epoch": 0.9815206929740135,
+      "grad_norm": 1.2887781858444214,
+      "learning_rate": 0.0001029781607584476,
+      "loss": 1.0748,
+      "step": 25495
+    },
+    {
+      "epoch": 0.9817131857555341,
+      "grad_norm": 1.1845282316207886,
+      "learning_rate": 0.00010294793443026974,
+      "loss": 1.0229,
+      "step": 25500
+    },
+    {
+      "epoch": 0.9819056785370549,
+      "grad_norm": 1.3194280862808228,
+      "learning_rate": 0.0001029177078325228,
+      "loss": 1.0796,
+      "step": 25505
+    },
+    {
+      "epoch": 0.9820981713185756,
+      "grad_norm": 1.5246963500976562,
+      "learning_rate": 0.0001028874809679708,
+      "loss": 1.2291,
+      "step": 25510
+    },
+    {
+      "epoch": 0.9822906641000962,
+      "grad_norm": 1.210331678390503,
+      "learning_rate": 0.00010285725383937782,
+      "loss": 0.9774,
+      "step": 25515
+    },
+    {
+      "epoch": 0.9824831568816169,
+      "grad_norm": 1.1501210927963257,
+      "learning_rate": 0.00010282702644950788,
+      "loss": 0.8948,
+      "step": 25520
+    },
+    {
+      "epoch": 0.9826756496631376,
+      "grad_norm": 1.2343779802322388,
+      "learning_rate": 0.0001027967988011251,
+      "loss": 1.1353,
+      "step": 25525
+    },
+    {
+      "epoch": 0.9828681424446584,
+      "grad_norm": 0.9154959321022034,
+      "learning_rate": 0.00010276657089699359,
+      "loss": 1.1039,
+      "step": 25530
+    },
+    {
+      "epoch": 0.983060635226179,
+      "grad_norm": 2.102421283721924,
+      "learning_rate": 0.00010273634273987754,
+      "loss": 1.1101,
+      "step": 25535
+    },
+    {
+      "epoch": 0.9832531280076997,
+      "grad_norm": 1.5680655241012573,
+      "learning_rate": 0.00010270611433254102,
+      "loss": 0.9902,
+      "step": 25540
+    },
+    {
+      "epoch": 0.9834456207892204,
+      "grad_norm": 0.9244110584259033,
+      "learning_rate": 0.0001026758856777483,
+      "loss": 0.9179,
+      "step": 25545
+    },
+    {
+      "epoch": 0.983638113570741,
+      "grad_norm": 0.8917207717895508,
+      "learning_rate": 0.00010264565677826356,
+      "loss": 0.987,
+      "step": 25550
+    },
+    {
+      "epoch": 0.9838306063522618,
+      "grad_norm": 1.52036452293396,
+      "learning_rate": 0.00010261542763685104,
+      "loss": 0.9879,
+      "step": 25555
+    },
+    {
+      "epoch": 0.9840230991337825,
+      "grad_norm": 1.535866618156433,
+      "learning_rate": 0.000102585198256275,
+      "loss": 1.121,
+      "step": 25560
+    },
+    {
+      "epoch": 0.9842155919153032,
+      "grad_norm": 1.3526102304458618,
+      "learning_rate": 0.00010255496863929965,
+      "loss": 1.0613,
+      "step": 25565
+    },
+    {
+      "epoch": 0.9844080846968238,
+      "grad_norm": 1.4781625270843506,
+      "learning_rate": 0.0001025247387886894,
+      "loss": 1.0461,
+      "step": 25570
+    },
+    {
+      "epoch": 0.9846005774783445,
+      "grad_norm": 1.0873816013336182,
+      "learning_rate": 0.00010249450870720849,
+      "loss": 0.9743,
+      "step": 25575
+    },
+    {
+      "epoch": 0.9847930702598653,
+      "grad_norm": 1.3945918083190918,
+      "learning_rate": 0.00010246427839762127,
+      "loss": 1.1183,
+      "step": 25580
+    },
+    {
+      "epoch": 0.984985563041386,
+      "grad_norm": 2.6019561290740967,
+      "learning_rate": 0.00010243404786269215,
+      "loss": 1.0604,
+      "step": 25585
+    },
+    {
+      "epoch": 0.9851780558229066,
+      "grad_norm": 1.1407113075256348,
+      "learning_rate": 0.00010240381710518542,
+      "loss": 0.9804,
+      "step": 25590
+    },
+    {
+      "epoch": 0.9853705486044273,
+      "grad_norm": 1.5416691303253174,
+      "learning_rate": 0.00010237358612786558,
+      "loss": 1.097,
+      "step": 25595
+    },
+    {
+      "epoch": 0.9855630413859481,
+      "grad_norm": 1.0081915855407715,
+      "learning_rate": 0.00010234335493349703,
+      "loss": 1.0342,
+      "step": 25600
+    },
+    {
+      "epoch": 0.9857555341674687,
+      "grad_norm": 1.0994783639907837,
+      "learning_rate": 0.00010231312352484417,
+      "loss": 1.2357,
+      "step": 25605
+    },
+    {
+      "epoch": 0.9859480269489894,
+      "grad_norm": 1.106605887413025,
+      "learning_rate": 0.00010228289190467146,
+      "loss": 1.0279,
+      "step": 25610
+    },
+    {
+      "epoch": 0.9861405197305101,
+      "grad_norm": 0.9887366890907288,
+      "learning_rate": 0.00010225266007574345,
+      "loss": 1.1127,
+      "step": 25615
+    },
+    {
+      "epoch": 0.9863330125120308,
+      "grad_norm": 1.079148530960083,
+      "learning_rate": 0.00010222242804082458,
+      "loss": 1.1353,
+      "step": 25620
+    },
+    {
+      "epoch": 0.9865255052935515,
+      "grad_norm": 1.2472405433654785,
+      "learning_rate": 0.00010219219580267938,
+      "loss": 1.1579,
+      "step": 25625
+    },
+    {
+      "epoch": 0.9867179980750722,
+      "grad_norm": 1.0981230735778809,
+      "learning_rate": 0.00010216196336407242,
+      "loss": 1.1264,
+      "step": 25630
+    },
+    {
+      "epoch": 0.9869104908565929,
+      "grad_norm": 1.4295231103897095,
+      "learning_rate": 0.00010213173072776823,
+      "loss": 1.0099,
+      "step": 25635
+    },
+    {
+      "epoch": 0.9871029836381136,
+      "grad_norm": 1.3902168273925781,
+      "learning_rate": 0.00010210149789653137,
+      "loss": 1.1308,
+      "step": 25640
+    },
+    {
+      "epoch": 0.9872954764196342,
+      "grad_norm": 1.2584228515625,
+      "learning_rate": 0.00010207126487312646,
+      "loss": 1.0638,
+      "step": 25645
+    },
+    {
+      "epoch": 0.987487969201155,
+      "grad_norm": 1.27899968624115,
+      "learning_rate": 0.00010204103166031809,
+      "loss": 1.0931,
+      "step": 25650
+    },
+    {
+      "epoch": 0.9876804619826757,
+      "grad_norm": 2.330070972442627,
+      "learning_rate": 0.00010201079826087088,
+      "loss": 1.1429,
+      "step": 25655
+    },
+    {
+      "epoch": 0.9878729547641963,
+      "grad_norm": 1.0945899486541748,
+      "learning_rate": 0.00010198056467754953,
+      "loss": 1.0337,
+      "step": 25660
+    },
+    {
+      "epoch": 0.988065447545717,
+      "grad_norm": 1.3564465045928955,
+      "learning_rate": 0.00010195033091311866,
+      "loss": 1.1632,
+      "step": 25665
+    },
+    {
+      "epoch": 0.9882579403272377,
+      "grad_norm": 1.0468459129333496,
+      "learning_rate": 0.0001019200969703429,
+      "loss": 1.1247,
+      "step": 25670
+    },
+    {
+      "epoch": 0.9884504331087585,
+      "grad_norm": 1.7445849180221558,
+      "learning_rate": 0.00010188986285198703,
+      "loss": 0.9937,
+      "step": 25675
+    },
+    {
+      "epoch": 0.9886429258902791,
+      "grad_norm": 0.9142084121704102,
+      "learning_rate": 0.0001018596285608157,
+      "loss": 1.1336,
+      "step": 25680
+    },
+    {
+      "epoch": 0.9888354186717998,
+      "grad_norm": 1.7198858261108398,
+      "learning_rate": 0.00010182939409959366,
+      "loss": 1.1145,
+      "step": 25685
+    },
+    {
+      "epoch": 0.9890279114533205,
+      "grad_norm": 1.204819679260254,
+      "learning_rate": 0.00010179915947108565,
+      "loss": 1.0323,
+      "step": 25690
+    },
+    {
+      "epoch": 0.9892204042348411,
+      "grad_norm": 0.8539313673973083,
+      "learning_rate": 0.00010176892467805646,
+      "loss": 0.8969,
+      "step": 25695
+    },
+    {
+      "epoch": 0.9894128970163619,
+      "grad_norm": 1.1470470428466797,
+      "learning_rate": 0.00010173868972327079,
+      "loss": 1.1075,
+      "step": 25700
+    },
+    {
+      "epoch": 0.9896053897978826,
+      "grad_norm": 1.3425363302230835,
+      "learning_rate": 0.00010170845460949345,
+      "loss": 1.178,
+      "step": 25705
+    },
+    {
+      "epoch": 0.9897978825794033,
+      "grad_norm": 1.2167271375656128,
+      "learning_rate": 0.00010167821933948929,
+      "loss": 0.9844,
+      "step": 25710
+    },
+    {
+      "epoch": 0.9899903753609239,
+      "grad_norm": 1.0317187309265137,
+      "learning_rate": 0.00010164798391602306,
+      "loss": 1.1865,
+      "step": 25715
+    },
+    {
+      "epoch": 0.9901828681424446,
+      "grad_norm": 1.983253836631775,
+      "learning_rate": 0.00010161774834185962,
+      "loss": 1.1188,
+      "step": 25720
+    },
+    {
+      "epoch": 0.9903753609239654,
+      "grad_norm": 0.8479856848716736,
+      "learning_rate": 0.00010158751261976382,
+      "loss": 1.1431,
+      "step": 25725
+    },
+    {
+      "epoch": 0.9905678537054861,
+      "grad_norm": 1.3559181690216064,
+      "learning_rate": 0.0001015572767525005,
+      "loss": 1.1233,
+      "step": 25730
+    },
+    {
+      "epoch": 0.9907603464870067,
+      "grad_norm": 0.8876566290855408,
+      "learning_rate": 0.00010152704074283454,
+      "loss": 1.0564,
+      "step": 25735
+    },
+    {
+      "epoch": 0.9909528392685274,
+      "grad_norm": 1.189932942390442,
+      "learning_rate": 0.00010149680459353083,
+      "loss": 1.0271,
+      "step": 25740
+    },
+    {
+      "epoch": 0.9911453320500482,
+      "grad_norm": 2.000798225402832,
+      "learning_rate": 0.00010146656830735424,
+      "loss": 1.3183,
+      "step": 25745
+    },
+    {
+      "epoch": 0.9913378248315688,
+      "grad_norm": 1.558982253074646,
+      "learning_rate": 0.00010143633188706969,
+      "loss": 0.9362,
+      "step": 25750
+    },
+    {
+      "epoch": 0.9915303176130895,
+      "grad_norm": 1.8294328451156616,
+      "learning_rate": 0.00010140609533544215,
+      "loss": 0.94,
+      "step": 25755
+    },
+    {
+      "epoch": 0.9917228103946102,
+      "grad_norm": 1.833295226097107,
+      "learning_rate": 0.00010137585865523644,
+      "loss": 0.8496,
+      "step": 25760
+    },
+    {
+      "epoch": 0.9919153031761309,
+      "grad_norm": 0.8973735570907593,
+      "learning_rate": 0.00010134562184921761,
+      "loss": 1.2634,
+      "step": 25765
+    },
+    {
+      "epoch": 0.9921077959576516,
+      "grad_norm": 1.6813448667526245,
+      "learning_rate": 0.00010131538492015056,
+      "loss": 1.1065,
+      "step": 25770
+    },
+    {
+      "epoch": 0.9923002887391723,
+      "grad_norm": 1.0313218832015991,
+      "learning_rate": 0.0001012851478708003,
+      "loss": 1.026,
+      "step": 25775
+    },
+    {
+      "epoch": 0.992492781520693,
+      "grad_norm": 1.324073314666748,
+      "learning_rate": 0.00010125491070393176,
+      "loss": 1.168,
+      "step": 25780
+    },
+    {
+      "epoch": 0.9926852743022136,
+      "grad_norm": 1.5872528553009033,
+      "learning_rate": 0.00010122467342230997,
+      "loss": 0.9647,
+      "step": 25785
+    },
+    {
+      "epoch": 0.9928777670837343,
+      "grad_norm": 1.108878254890442,
+      "learning_rate": 0.00010119443602869987,
+      "loss": 0.9983,
+      "step": 25790
+    },
+    {
+      "epoch": 0.9930702598652551,
+      "grad_norm": 0.7914887070655823,
+      "learning_rate": 0.00010116419852586652,
+      "loss": 0.974,
+      "step": 25795
+    },
+    {
+      "epoch": 0.9932627526467758,
+      "grad_norm": 1.9735599756240845,
+      "learning_rate": 0.00010113396091657492,
+      "loss": 1.1217,
+      "step": 25800
+    },
+    {
+      "epoch": 0.9934552454282964,
+      "grad_norm": 1.5698872804641724,
+      "learning_rate": 0.00010110372320359014,
+      "loss": 1.1984,
+      "step": 25805
+    },
+    {
+      "epoch": 0.9936477382098171,
+      "grad_norm": 1.5581351518630981,
+      "learning_rate": 0.00010107348538967714,
+      "loss": 1.0301,
+      "step": 25810
+    },
+    {
+      "epoch": 0.9938402309913378,
+      "grad_norm": 0.868083655834198,
+      "learning_rate": 0.00010104324747760104,
+      "loss": 1.1439,
+      "step": 25815
+    },
+    {
+      "epoch": 0.9940327237728586,
+      "grad_norm": 1.065421462059021,
+      "learning_rate": 0.00010101300947012686,
+      "loss": 1.0275,
+      "step": 25820
+    },
+    {
+      "epoch": 0.9942252165543792,
+      "grad_norm": 1.3174570798873901,
+      "learning_rate": 0.00010098277137001967,
+      "loss": 1.0925,
+      "step": 25825
+    },
+    {
+      "epoch": 0.9944177093358999,
+      "grad_norm": 1.1705584526062012,
+      "learning_rate": 0.00010095253318004457,
+      "loss": 1.1513,
+      "step": 25830
+    },
+    {
+      "epoch": 0.9946102021174206,
+      "grad_norm": 1.9153684377670288,
+      "learning_rate": 0.00010092229490296661,
+      "loss": 0.9746,
+      "step": 25835
+    },
+    {
+      "epoch": 0.9948026948989412,
+      "grad_norm": 1.1694215536117554,
+      "learning_rate": 0.00010089205654155087,
+      "loss": 1.1149,
+      "step": 25840
+    },
+    {
+      "epoch": 0.994995187680462,
+      "grad_norm": 1.2398862838745117,
+      "learning_rate": 0.00010086181809856248,
+      "loss": 1.0225,
+      "step": 25845
+    },
+    {
+      "epoch": 0.9951876804619827,
+      "grad_norm": 0.803766131401062,
+      "learning_rate": 0.00010083157957676657,
+      "loss": 0.9342,
+      "step": 25850
+    },
+    {
+      "epoch": 0.9953801732435034,
+      "grad_norm": 1.1521053314208984,
+      "learning_rate": 0.00010080134097892817,
+      "loss": 1.1817,
+      "step": 25855
+    },
+    {
+      "epoch": 0.995572666025024,
+      "grad_norm": 1.6817054748535156,
+      "learning_rate": 0.00010077110230781246,
+      "loss": 0.9607,
+      "step": 25860
+    },
+    {
+      "epoch": 0.9957651588065447,
+      "grad_norm": 1.8072839975357056,
+      "learning_rate": 0.00010074086356618457,
+      "loss": 1.2474,
+      "step": 25865
+    },
+    {
+      "epoch": 0.9959576515880655,
+      "grad_norm": 1.1039899587631226,
+      "learning_rate": 0.00010071062475680959,
+      "loss": 1.0931,
+      "step": 25870
+    },
+    {
+      "epoch": 0.9961501443695862,
+      "grad_norm": 1.3213703632354736,
+      "learning_rate": 0.0001006803858824527,
+      "loss": 0.9709,
+      "step": 25875
+    },
+    {
+      "epoch": 0.9963426371511068,
+      "grad_norm": 2.1169626712799072,
+      "learning_rate": 0.00010065014694587902,
+      "loss": 1.0185,
+      "step": 25880
+    },
+    {
+      "epoch": 0.9965351299326275,
+      "grad_norm": 1.997347116470337,
+      "learning_rate": 0.00010061990794985372,
+      "loss": 0.9641,
+      "step": 25885
+    },
+    {
+      "epoch": 0.9967276227141482,
+      "grad_norm": 2.4005537033081055,
+      "learning_rate": 0.00010058966889714192,
+      "loss": 0.956,
+      "step": 25890
+    },
+    {
+      "epoch": 0.996920115495669,
+      "grad_norm": 1.1593470573425293,
+      "learning_rate": 0.00010055942979050886,
+      "loss": 0.9971,
+      "step": 25895
+    },
+    {
+      "epoch": 0.9971126082771896,
+      "grad_norm": 1.5145149230957031,
+      "learning_rate": 0.0001005291906327196,
+      "loss": 1.029,
+      "step": 25900
+    },
+    {
+      "epoch": 0.9973051010587103,
+      "grad_norm": 1.6127921342849731,
+      "learning_rate": 0.00010049895142653936,
+      "loss": 1.0643,
+      "step": 25905
+    },
+    {
+      "epoch": 0.997497593840231,
+      "grad_norm": 1.326854944229126,
+      "learning_rate": 0.00010046871217473334,
+      "loss": 1.0119,
+      "step": 25910
+    },
+    {
+      "epoch": 0.9976900866217517,
+      "grad_norm": 1.0918645858764648,
+      "learning_rate": 0.00010043847288006666,
+      "loss": 1.2797,
+      "step": 25915
+    },
+    {
+      "epoch": 0.9978825794032724,
+      "grad_norm": 1.0558148622512817,
+      "learning_rate": 0.00010040823354530457,
+      "loss": 1.0526,
+      "step": 25920
+    },
+    {
+      "epoch": 0.9980750721847931,
+      "grad_norm": 1.1978880167007446,
+      "learning_rate": 0.00010037799417321222,
+      "loss": 1.0564,
+      "step": 25925
+    },
+    {
+      "epoch": 0.9982675649663137,
+      "grad_norm": 1.9178636074066162,
+      "learning_rate": 0.00010034775476655482,
+      "loss": 1.068,
+      "step": 25930
+    },
+    {
+      "epoch": 0.9984600577478344,
+      "grad_norm": 1.7356677055358887,
+      "learning_rate": 0.00010031751532809755,
+      "loss": 1.2817,
+      "step": 25935
+    },
+    {
+      "epoch": 0.9986525505293552,
+      "grad_norm": 1.0780870914459229,
+      "learning_rate": 0.00010028727586060558,
+      "loss": 1.0834,
+      "step": 25940
+    },
+    {
+      "epoch": 0.9988450433108759,
+      "grad_norm": 1.4201254844665527,
+      "learning_rate": 0.00010025703636684416,
+      "loss": 0.9698,
+      "step": 25945
+    },
+    {
+      "epoch": 0.9990375360923965,
+      "grad_norm": 1.0446133613586426,
+      "learning_rate": 0.00010022679684957845,
+      "loss": 1.1268,
+      "step": 25950
+    },
+    {
+      "epoch": 0.9992300288739172,
+      "grad_norm": 1.2714285850524902,
+      "learning_rate": 0.0001001965573115737,
+      "loss": 1.2019,
+      "step": 25955
+    },
+    {
+      "epoch": 0.9994225216554379,
+      "grad_norm": 0.9263586401939392,
+      "learning_rate": 0.00010016631775559506,
+      "loss": 1.1232,
+      "step": 25960
+    },
+    {
+      "epoch": 0.9996150144369587,
+      "grad_norm": 2.6559371948242188,
+      "learning_rate": 0.00010013607818440775,
+      "loss": 1.1367,
+      "step": 25965
+    },
+    {
+      "epoch": 0.9998075072184793,
+      "grad_norm": 1.6373353004455566,
+      "learning_rate": 0.00010010583860077703,
+      "loss": 1.002,
+      "step": 25970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.38175630569458,
+      "learning_rate": 0.00010007559900746805,
+      "loss": 1.198,
+      "step": 25975
+    },
+    {
+      "epoch": 1.0001924927815207,
+      "grad_norm": 0.9677699208259583,
+      "learning_rate": 0.00010004535940724604,
+      "loss": 0.9076,
+      "step": 25980
+    },
+    {
+      "epoch": 1.0003849855630413,
+      "grad_norm": 1.5017547607421875,
+      "learning_rate": 0.00010001511980287623,
+      "loss": 0.9611,
+      "step": 25985
+    },
+    {
+      "epoch": 1.000577478344562,
+      "grad_norm": 0.8973893523216248,
+      "learning_rate": 9.998488019712379e-05,
+      "loss": 0.8939,
+      "step": 25990
+    },
+    {
+      "epoch": 1.0007699711260827,
+      "grad_norm": 1.614953875541687,
+      "learning_rate": 9.995464059275396e-05,
+      "loss": 0.6977,
+      "step": 25995
+    },
+    {
+      "epoch": 1.0009624639076036,
+      "grad_norm": 2.5739526748657227,
+      "learning_rate": 9.992440099253196e-05,
+      "loss": 0.967,
+      "step": 26000
+    },
+    {
+      "epoch": 1.0011549566891242,
+      "grad_norm": 1.2019327878952026,
+      "learning_rate": 9.989416139922301e-05,
+      "loss": 0.8387,
+      "step": 26005
+    },
+    {
+      "epoch": 1.001347449470645,
+      "grad_norm": 1.8093628883361816,
+      "learning_rate": 9.986392181559223e-05,
+      "loss": 0.9027,
+      "step": 26010
+    },
+    {
+      "epoch": 1.0015399422521656,
+      "grad_norm": 1.4864177703857422,
+      "learning_rate": 9.983368224440496e-05,
+      "loss": 0.7639,
+      "step": 26015
+    },
+    {
+      "epoch": 1.0017324350336863,
+      "grad_norm": 1.123429775238037,
+      "learning_rate": 9.980344268842634e-05,
+      "loss": 0.7783,
+      "step": 26020
+    },
+    {
+      "epoch": 1.001924927815207,
+      "grad_norm": 1.5506818294525146,
+      "learning_rate": 9.977320315042154e-05,
+      "loss": 0.7143,
+      "step": 26025
+    },
+    {
+      "epoch": 1.0021174205967276,
+      "grad_norm": 2.1665854454040527,
+      "learning_rate": 9.974296363315585e-05,
+      "loss": 0.9101,
+      "step": 26030
+    },
+    {
+      "epoch": 1.0023099133782483,
+      "grad_norm": 1.8006185293197632,
+      "learning_rate": 9.971272413939444e-05,
+      "loss": 0.9212,
+      "step": 26035
+    },
+    {
+      "epoch": 1.002502406159769,
+      "grad_norm": 1.2385761737823486,
+      "learning_rate": 9.968248467190245e-05,
+      "loss": 0.8715,
+      "step": 26040
+    },
+    {
+      "epoch": 1.0026948989412896,
+      "grad_norm": 1.4337763786315918,
+      "learning_rate": 9.965224523344519e-05,
+      "loss": 0.6949,
+      "step": 26045
+    },
+    {
+      "epoch": 1.0028873917228105,
+      "grad_norm": 1.586151123046875,
+      "learning_rate": 9.96220058267878e-05,
+      "loss": 0.913,
+      "step": 26050
+    },
+    {
+      "epoch": 1.0030798845043312,
+      "grad_norm": 1.3005743026733398,
+      "learning_rate": 9.959176645469542e-05,
+      "loss": 0.8218,
+      "step": 26055
+    },
+    {
+      "epoch": 1.0032723772858518,
+      "grad_norm": 0.735365629196167,
+      "learning_rate": 9.956152711993335e-05,
+      "loss": 0.7206,
+      "step": 26060
+    },
+    {
+      "epoch": 1.0034648700673725,
+      "grad_norm": 1.7162381410598755,
+      "learning_rate": 9.95312878252667e-05,
+      "loss": 0.7916,
+      "step": 26065
+    },
+    {
+      "epoch": 1.0036573628488932,
+      "grad_norm": 0.894173800945282,
+      "learning_rate": 9.950104857346064e-05,
+      "loss": 0.9683,
+      "step": 26070
+    },
+    {
+      "epoch": 1.0038498556304138,
+      "grad_norm": 1.6856434345245361,
+      "learning_rate": 9.947080936728044e-05,
+      "loss": 0.7997,
+      "step": 26075
+    },
+    {
+      "epoch": 1.0040423484119345,
+      "grad_norm": 1.2891976833343506,
+      "learning_rate": 9.94405702094912e-05,
+      "loss": 0.802,
+      "step": 26080
+    },
+    {
+      "epoch": 1.0042348411934552,
+      "grad_norm": 1.5862544775009155,
+      "learning_rate": 9.941033110285809e-05,
+      "loss": 0.857,
+      "step": 26085
+    },
+    {
+      "epoch": 1.0044273339749759,
+      "grad_norm": 2.165689706802368,
+      "learning_rate": 9.93800920501463e-05,
+      "loss": 1.01,
+      "step": 26090
+    },
+    {
+      "epoch": 1.0046198267564965,
+      "grad_norm": 1.3247888088226318,
+      "learning_rate": 9.9349853054121e-05,
+      "loss": 0.872,
+      "step": 26095
+    },
+    {
+      "epoch": 1.0048123195380174,
+      "grad_norm": 1.7285022735595703,
+      "learning_rate": 9.931961411754732e-05,
+      "loss": 0.913,
+      "step": 26100
+    },
+    {
+      "epoch": 1.005004812319538,
+      "grad_norm": 1.504143238067627,
+      "learning_rate": 9.928937524319043e-05,
+      "loss": 0.7396,
+      "step": 26105
+    },
+    {
+      "epoch": 1.0051973051010588,
+      "grad_norm": 1.633792519569397,
+      "learning_rate": 9.925913643381546e-05,
+      "loss": 0.8946,
+      "step": 26110
+    },
+    {
+      "epoch": 1.0053897978825794,
+      "grad_norm": 1.8839080333709717,
+      "learning_rate": 9.922889769218755e-05,
+      "loss": 0.8844,
+      "step": 26115
+    },
+    {
+      "epoch": 1.0055822906641,
+      "grad_norm": 0.9367867708206177,
+      "learning_rate": 9.919865902107185e-05,
+      "loss": 0.8064,
+      "step": 26120
+    },
+    {
+      "epoch": 1.0057747834456208,
+      "grad_norm": 1.0992298126220703,
+      "learning_rate": 9.916842042323348e-05,
+      "loss": 0.8674,
+      "step": 26125
+    },
+    {
+      "epoch": 1.0059672762271414,
+      "grad_norm": 1.031310796737671,
+      "learning_rate": 9.913818190143751e-05,
+      "loss": 0.8082,
+      "step": 26130
+    },
+    {
+      "epoch": 1.006159769008662,
+      "grad_norm": 1.5799235105514526,
+      "learning_rate": 9.910794345844914e-05,
+      "loss": 0.7128,
+      "step": 26135
+    },
+    {
+      "epoch": 1.0063522617901828,
+      "grad_norm": 1.5742192268371582,
+      "learning_rate": 9.907770509703344e-05,
+      "loss": 0.8554,
+      "step": 26140
+    },
+    {
+      "epoch": 1.0065447545717037,
+      "grad_norm": 1.3698316812515259,
+      "learning_rate": 9.904746681995544e-05,
+      "loss": 0.6932,
+      "step": 26145
+    },
+    {
+      "epoch": 1.0067372473532243,
+      "grad_norm": 1.2731399536132812,
+      "learning_rate": 9.901722862998034e-05,
+      "loss": 0.9624,
+      "step": 26150
+    },
+    {
+      "epoch": 1.006929740134745,
+      "grad_norm": 1.0184792280197144,
+      "learning_rate": 9.898699052987318e-05,
+      "loss": 0.8248,
+      "step": 26155
+    },
+    {
+      "epoch": 1.0071222329162657,
+      "grad_norm": 1.1166132688522339,
+      "learning_rate": 9.895675252239896e-05,
+      "loss": 0.7907,
+      "step": 26160
+    },
+    {
+      "epoch": 1.0073147256977864,
+      "grad_norm": 2.4963834285736084,
+      "learning_rate": 9.892651461032287e-05,
+      "loss": 0.8758,
+      "step": 26165
+    },
+    {
+      "epoch": 1.007507218479307,
+      "grad_norm": 2.192413091659546,
+      "learning_rate": 9.889627679640991e-05,
+      "loss": 0.9262,
+      "step": 26170
+    },
+    {
+      "epoch": 1.0076997112608277,
+      "grad_norm": 1.760149598121643,
+      "learning_rate": 9.886603908342508e-05,
+      "loss": 0.8604,
+      "step": 26175
+    },
+    {
+      "epoch": 1.0078922040423484,
+      "grad_norm": 1.2598868608474731,
+      "learning_rate": 9.88358014741335e-05,
+      "loss": 0.9674,
+      "step": 26180
+    },
+    {
+      "epoch": 1.008084696823869,
+      "grad_norm": 1.29741632938385,
+      "learning_rate": 9.880556397130018e-05,
+      "loss": 0.8107,
+      "step": 26185
+    },
+    {
+      "epoch": 1.0082771896053897,
+      "grad_norm": 1.4447206258773804,
+      "learning_rate": 9.877532657769006e-05,
+      "loss": 0.9182,
+      "step": 26190
+    },
+    {
+      "epoch": 1.0084696823869106,
+      "grad_norm": 1.8422036170959473,
+      "learning_rate": 9.874508929606827e-05,
+      "loss": 1.0373,
+      "step": 26195
+    },
+    {
+      "epoch": 1.0086621751684313,
+      "grad_norm": 1.393976092338562,
+      "learning_rate": 9.871485212919974e-05,
+      "loss": 0.8484,
+      "step": 26200
+    },
+    {
+      "epoch": 1.008854667949952,
+      "grad_norm": 1.280700922012329,
+      "learning_rate": 9.868461507984945e-05,
+      "loss": 0.7288,
+      "step": 26205
+    },
+    {
+      "epoch": 1.0090471607314726,
+      "grad_norm": 1.8476749658584595,
+      "learning_rate": 9.86543781507824e-05,
+      "loss": 0.9356,
+      "step": 26210
+    },
+    {
+      "epoch": 1.0092396535129933,
+      "grad_norm": 1.1410245895385742,
+      "learning_rate": 9.862414134476358e-05,
+      "loss": 0.7967,
+      "step": 26215
+    },
+    {
+      "epoch": 1.009432146294514,
+      "grad_norm": 1.401632308959961,
+      "learning_rate": 9.859390466455789e-05,
+      "loss": 0.825,
+      "step": 26220
+    },
+    {
+      "epoch": 1.0096246390760346,
+      "grad_norm": 1.705074667930603,
+      "learning_rate": 9.856366811293033e-05,
+      "loss": 1.0509,
+      "step": 26225
+    },
+    {
+      "epoch": 1.0098171318575553,
+      "grad_norm": 1.788540244102478,
+      "learning_rate": 9.853343169264581e-05,
+      "loss": 0.9138,
+      "step": 26230
+    },
+    {
+      "epoch": 1.010009624639076,
+      "grad_norm": 1.9128284454345703,
+      "learning_rate": 9.850319540646919e-05,
+      "loss": 0.931,
+      "step": 26235
+    },
+    {
+      "epoch": 1.0102021174205966,
+      "grad_norm": 1.1917674541473389,
+      "learning_rate": 9.847295925716548e-05,
+      "loss": 0.7869,
+      "step": 26240
+    },
+    {
+      "epoch": 1.0103946102021175,
+      "grad_norm": 1.351189136505127,
+      "learning_rate": 9.844272324749955e-05,
+      "loss": 0.868,
+      "step": 26245
+    },
+    {
+      "epoch": 1.0105871029836382,
+      "grad_norm": 1.3934807777404785,
+      "learning_rate": 9.841248738023619e-05,
+      "loss": 0.9278,
+      "step": 26250
+    },
+    {
+      "epoch": 1.0107795957651589,
+      "grad_norm": 1.337436556816101,
+      "learning_rate": 9.83822516581404e-05,
+      "loss": 0.9855,
+      "step": 26255
+    },
+    {
+      "epoch": 1.0109720885466795,
+      "grad_norm": 1.5759284496307373,
+      "learning_rate": 9.835201608397695e-05,
+      "loss": 0.6804,
+      "step": 26260
+    },
+    {
+      "epoch": 1.0111645813282002,
+      "grad_norm": 1.530190110206604,
+      "learning_rate": 9.832178066051075e-05,
+      "loss": 1.1206,
+      "step": 26265
+    },
+    {
+      "epoch": 1.0113570741097209,
+      "grad_norm": 1.9554476737976074,
+      "learning_rate": 9.829154539050657e-05,
+      "loss": 1.0103,
+      "step": 26270
+    },
+    {
+      "epoch": 1.0115495668912415,
+      "grad_norm": 1.9727224111557007,
+      "learning_rate": 9.826131027672922e-05,
+      "loss": 0.9634,
+      "step": 26275
+    },
+    {
+      "epoch": 1.0117420596727622,
+      "grad_norm": 1.2419236898422241,
+      "learning_rate": 9.823107532194358e-05,
+      "loss": 0.6752,
+      "step": 26280
+    },
+    {
+      "epoch": 1.0119345524542829,
+      "grad_norm": 1.707638144493103,
+      "learning_rate": 9.820084052891436e-05,
+      "loss": 0.8883,
+      "step": 26285
+    },
+    {
+      "epoch": 1.0121270452358035,
+      "grad_norm": 1.57963228225708,
+      "learning_rate": 9.817060590040633e-05,
+      "loss": 0.8674,
+      "step": 26290
+    },
+    {
+      "epoch": 1.0123195380173244,
+      "grad_norm": 0.9437096118927002,
+      "learning_rate": 9.81403714391843e-05,
+      "loss": 0.7648,
+      "step": 26295
+    },
+    {
+      "epoch": 1.012512030798845,
+      "grad_norm": 1.6244466304779053,
+      "learning_rate": 9.8110137148013e-05,
+      "loss": 0.8088,
+      "step": 26300
+    },
+    {
+      "epoch": 1.0127045235803658,
+      "grad_norm": 1.200982689857483,
+      "learning_rate": 9.807990302965712e-05,
+      "loss": 0.9869,
+      "step": 26305
+    },
+    {
+      "epoch": 1.0128970163618864,
+      "grad_norm": 1.7576247453689575,
+      "learning_rate": 9.804966908688137e-05,
+      "loss": 0.7445,
+      "step": 26310
+    },
+    {
+      "epoch": 1.0130895091434071,
+      "grad_norm": 1.2843434810638428,
+      "learning_rate": 9.801943532245049e-05,
+      "loss": 0.7024,
+      "step": 26315
+    },
+    {
+      "epoch": 1.0132820019249278,
+      "grad_norm": 1.0596975088119507,
+      "learning_rate": 9.79892017391291e-05,
+      "loss": 0.8457,
+      "step": 26320
+    },
+    {
+      "epoch": 1.0134744947064485,
+      "grad_norm": 1.8278613090515137,
+      "learning_rate": 9.795896833968193e-05,
+      "loss": 0.9703,
+      "step": 26325
+    },
+    {
+      "epoch": 1.0136669874879691,
+      "grad_norm": 1.732739806175232,
+      "learning_rate": 9.792873512687359e-05,
+      "loss": 0.8173,
+      "step": 26330
+    },
+    {
+      "epoch": 1.0138594802694898,
+      "grad_norm": 0.9648169279098511,
+      "learning_rate": 9.789850210346864e-05,
+      "loss": 0.8673,
+      "step": 26335
+    },
+    {
+      "epoch": 1.0140519730510107,
+      "grad_norm": 1.4606342315673828,
+      "learning_rate": 9.78682692722318e-05,
+      "loss": 0.9835,
+      "step": 26340
+    },
+    {
+      "epoch": 1.0142444658325314,
+      "grad_norm": 1.4847335815429688,
+      "learning_rate": 9.783803663592762e-05,
+      "loss": 0.9562,
+      "step": 26345
+    },
+    {
+      "epoch": 1.014436958614052,
+      "grad_norm": 1.0714442729949951,
+      "learning_rate": 9.78078041973206e-05,
+      "loss": 0.8459,
+      "step": 26350
+    },
+    {
+      "epoch": 1.0146294513955727,
+      "grad_norm": 1.2830302715301514,
+      "learning_rate": 9.777757195917544e-05,
+      "loss": 1.0389,
+      "step": 26355
+    },
+    {
+      "epoch": 1.0148219441770934,
+      "grad_norm": 1.007949948310852,
+      "learning_rate": 9.774733992425659e-05,
+      "loss": 0.8296,
+      "step": 26360
+    },
+    {
+      "epoch": 1.015014436958614,
+      "grad_norm": 1.4365791082382202,
+      "learning_rate": 9.771710809532853e-05,
+      "loss": 0.8113,
+      "step": 26365
+    },
+    {
+      "epoch": 1.0152069297401347,
+      "grad_norm": 1.213368535041809,
+      "learning_rate": 9.768687647515587e-05,
+      "loss": 0.829,
+      "step": 26370
+    },
+    {
+      "epoch": 1.0153994225216554,
+      "grad_norm": 1.6301283836364746,
+      "learning_rate": 9.765664506650302e-05,
+      "loss": 0.8657,
+      "step": 26375
+    },
+    {
+      "epoch": 1.015591915303176,
+      "grad_norm": 1.4944705963134766,
+      "learning_rate": 9.762641387213442e-05,
+      "loss": 0.8043,
+      "step": 26380
+    },
+    {
+      "epoch": 1.0157844080846967,
+      "grad_norm": 1.2845187187194824,
+      "learning_rate": 9.759618289481459e-05,
+      "loss": 0.6866,
+      "step": 26385
+    },
+    {
+      "epoch": 1.0159769008662176,
+      "grad_norm": 1.635873556137085,
+      "learning_rate": 9.75659521373079e-05,
+      "loss": 0.9432,
+      "step": 26390
+    },
+    {
+      "epoch": 1.0161693936477383,
+      "grad_norm": 0.9494144320487976,
+      "learning_rate": 9.753572160237873e-05,
+      "loss": 0.7428,
+      "step": 26395
+    },
+    {
+      "epoch": 1.016361886429259,
+      "grad_norm": 1.2726250886917114,
+      "learning_rate": 9.750549129279153e-05,
+      "loss": 0.6935,
+      "step": 26400
+    },
+    {
+      "epoch": 1.0165543792107796,
+      "grad_norm": 1.4167382717132568,
+      "learning_rate": 9.747526121131064e-05,
+      "loss": 1.0576,
+      "step": 26405
+    },
+    {
+      "epoch": 1.0167468719923003,
+      "grad_norm": 1.8464453220367432,
+      "learning_rate": 9.744503136070036e-05,
+      "loss": 0.8568,
+      "step": 26410
+    },
+    {
+      "epoch": 1.016939364773821,
+      "grad_norm": 1.164482593536377,
+      "learning_rate": 9.741480174372504e-05,
+      "loss": 0.8255,
+      "step": 26415
+    },
+    {
+      "epoch": 1.0171318575553416,
+      "grad_norm": 0.8602749109268188,
+      "learning_rate": 9.738457236314898e-05,
+      "loss": 0.7958,
+      "step": 26420
+    },
+    {
+      "epoch": 1.0173243503368623,
+      "grad_norm": 1.5463776588439941,
+      "learning_rate": 9.735434322173645e-05,
+      "loss": 0.79,
+      "step": 26425
+    },
+    {
+      "epoch": 1.017516843118383,
+      "grad_norm": 1.2181867361068726,
+      "learning_rate": 9.732411432225173e-05,
+      "loss": 0.8152,
+      "step": 26430
+    },
+    {
+      "epoch": 1.0177093358999039,
+      "grad_norm": 1.4031577110290527,
+      "learning_rate": 9.729388566745899e-05,
+      "loss": 0.8809,
+      "step": 26435
+    },
+    {
+      "epoch": 1.0179018286814245,
+      "grad_norm": 0.7895151376724243,
+      "learning_rate": 9.726365726012249e-05,
+      "loss": 0.8356,
+      "step": 26440
+    },
+    {
+      "epoch": 1.0180943214629452,
+      "grad_norm": 1.0852110385894775,
+      "learning_rate": 9.723342910300642e-05,
+      "loss": 0.71,
+      "step": 26445
+    },
+    {
+      "epoch": 1.0182868142444659,
+      "grad_norm": 1.8769862651824951,
+      "learning_rate": 9.720320119887494e-05,
+      "loss": 0.8581,
+      "step": 26450
+    },
+    {
+      "epoch": 1.0184793070259865,
+      "grad_norm": 1.0951234102249146,
+      "learning_rate": 9.717297355049215e-05,
+      "loss": 0.9184,
+      "step": 26455
+    },
+    {
+      "epoch": 1.0186717998075072,
+      "grad_norm": 1.4983422756195068,
+      "learning_rate": 9.714274616062222e-05,
+      "loss": 0.9141,
+      "step": 26460
+    },
+    {
+      "epoch": 1.0188642925890279,
+      "grad_norm": 1.0044662952423096,
+      "learning_rate": 9.711251903202923e-05,
+      "loss": 0.77,
+      "step": 26465
+    },
+    {
+      "epoch": 1.0190567853705486,
+      "grad_norm": 1.339004635810852,
+      "learning_rate": 9.708229216747721e-05,
+      "loss": 1.0225,
+      "step": 26470
+    },
+    {
+      "epoch": 1.0192492781520692,
+      "grad_norm": 1.5618289709091187,
+      "learning_rate": 9.70520655697303e-05,
+      "loss": 0.9058,
+      "step": 26475
+    },
+    {
+      "epoch": 1.01944177093359,
+      "grad_norm": 0.7431265115737915,
+      "learning_rate": 9.702183924155246e-05,
+      "loss": 0.8251,
+      "step": 26480
+    },
+    {
+      "epoch": 1.0196342637151108,
+      "grad_norm": 1.1725512742996216,
+      "learning_rate": 9.699161318570765e-05,
+      "loss": 0.8755,
+      "step": 26485
+    },
+    {
+      "epoch": 1.0198267564966315,
+      "grad_norm": 2.5678439140319824,
+      "learning_rate": 9.696138740495991e-05,
+      "loss": 0.9383,
+      "step": 26490
+    },
+    {
+      "epoch": 1.0200192492781521,
+      "grad_norm": 1.203439474105835,
+      "learning_rate": 9.693116190207318e-05,
+      "loss": 0.8122,
+      "step": 26495
+    },
+    {
+      "epoch": 1.0202117420596728,
+      "grad_norm": 1.7654101848602295,
+      "learning_rate": 9.690093667981132e-05,
+      "loss": 0.8668,
+      "step": 26500
+    },
+    {
+      "epoch": 1.0204042348411935,
+      "grad_norm": 1.5297460556030273,
+      "learning_rate": 9.687071174093831e-05,
+      "loss": 0.9342,
+      "step": 26505
+    },
+    {
+      "epoch": 1.0205967276227141,
+      "grad_norm": 1.2737789154052734,
+      "learning_rate": 9.684048708821796e-05,
+      "loss": 0.8387,
+      "step": 26510
+    },
+    {
+      "epoch": 1.0207892204042348,
+      "grad_norm": 1.0593212842941284,
+      "learning_rate": 9.681026272441414e-05,
+      "loss": 0.9148,
+      "step": 26515
+    },
+    {
+      "epoch": 1.0209817131857555,
+      "grad_norm": 1.3698254823684692,
+      "learning_rate": 9.678003865229067e-05,
+      "loss": 0.9545,
+      "step": 26520
+    },
+    {
+      "epoch": 1.0211742059672761,
+      "grad_norm": 1.5303738117218018,
+      "learning_rate": 9.674981487461133e-05,
+      "loss": 0.98,
+      "step": 26525
+    },
+    {
+      "epoch": 1.0213666987487968,
+      "grad_norm": 0.8624149560928345,
+      "learning_rate": 9.671959139413991e-05,
+      "loss": 0.8921,
+      "step": 26530
+    },
+    {
+      "epoch": 1.0215591915303177,
+      "grad_norm": 0.9601109027862549,
+      "learning_rate": 9.66893682136401e-05,
+      "loss": 0.8469,
+      "step": 26535
+    },
+    {
+      "epoch": 1.0217516843118384,
+      "grad_norm": 1.3074990510940552,
+      "learning_rate": 9.665914533587563e-05,
+      "loss": 0.8174,
+      "step": 26540
+    },
+    {
+      "epoch": 1.021944177093359,
+      "grad_norm": 1.4614131450653076,
+      "learning_rate": 9.662892276361022e-05,
+      "loss": 1.0461,
+      "step": 26545
+    },
+    {
+      "epoch": 1.0221366698748797,
+      "grad_norm": 2.2266509532928467,
+      "learning_rate": 9.65987004996075e-05,
+      "loss": 0.9664,
+      "step": 26550
+    },
+    {
+      "epoch": 1.0223291626564004,
+      "grad_norm": 1.8726714849472046,
+      "learning_rate": 9.656847854663102e-05,
+      "loss": 0.9988,
+      "step": 26555
+    },
+    {
+      "epoch": 1.022521655437921,
+      "grad_norm": 1.4316071271896362,
+      "learning_rate": 9.65382569074445e-05,
+      "loss": 0.9426,
+      "step": 26560
+    },
+    {
+      "epoch": 1.0227141482194417,
+      "grad_norm": 0.9609633088111877,
+      "learning_rate": 9.650803558481146e-05,
+      "loss": 0.8867,
+      "step": 26565
+    },
+    {
+      "epoch": 1.0229066410009624,
+      "grad_norm": 2.933589458465576,
+      "learning_rate": 9.647781458149539e-05,
+      "loss": 0.9714,
+      "step": 26570
+    },
+    {
+      "epoch": 1.023099133782483,
+      "grad_norm": 1.019455075263977,
+      "learning_rate": 9.644759390025988e-05,
+      "loss": 0.7742,
+      "step": 26575
+    },
+    {
+      "epoch": 1.0232916265640037,
+      "grad_norm": 1.2443771362304688,
+      "learning_rate": 9.641737354386839e-05,
+      "loss": 0.7977,
+      "step": 26580
+    },
+    {
+      "epoch": 1.0234841193455246,
+      "grad_norm": 1.4607973098754883,
+      "learning_rate": 9.63871535150843e-05,
+      "loss": 0.9563,
+      "step": 26585
+    },
+    {
+      "epoch": 1.0236766121270453,
+      "grad_norm": 1.0155866146087646,
+      "learning_rate": 9.635693381667112e-05,
+      "loss": 0.8839,
+      "step": 26590
+    },
+    {
+      "epoch": 1.023869104908566,
+      "grad_norm": 1.453011155128479,
+      "learning_rate": 9.632671445139223e-05,
+      "loss": 0.8905,
+      "step": 26595
+    },
+    {
+      "epoch": 1.0240615976900866,
+      "grad_norm": 1.598872184753418,
+      "learning_rate": 9.629649542201091e-05,
+      "loss": 0.8839,
+      "step": 26600
+    },
+    {
+      "epoch": 1.0242540904716073,
+      "grad_norm": 1.3655000925064087,
+      "learning_rate": 9.626627673129062e-05,
+      "loss": 0.8606,
+      "step": 26605
+    },
+    {
+      "epoch": 1.024446583253128,
+      "grad_norm": 0.9747033715248108,
+      "learning_rate": 9.623605838199457e-05,
+      "loss": 0.9387,
+      "step": 26610
+    },
+    {
+      "epoch": 1.0246390760346487,
+      "grad_norm": 1.2741761207580566,
+      "learning_rate": 9.6205840376886e-05,
+      "loss": 0.8915,
+      "step": 26615
+    },
+    {
+      "epoch": 1.0248315688161693,
+      "grad_norm": 1.0303744077682495,
+      "learning_rate": 9.617562271872825e-05,
+      "loss": 0.9422,
+      "step": 26620
+    },
+    {
+      "epoch": 1.02502406159769,
+      "grad_norm": 1.2387011051177979,
+      "learning_rate": 9.614540541028445e-05,
+      "loss": 0.9212,
+      "step": 26625
+    },
+    {
+      "epoch": 1.0252165543792109,
+      "grad_norm": 1.31088387966156,
+      "learning_rate": 9.611518845431778e-05,
+      "loss": 0.8258,
+      "step": 26630
+    },
+    {
+      "epoch": 1.0254090471607316,
+      "grad_norm": 1.3650875091552734,
+      "learning_rate": 9.60849718535914e-05,
+      "loss": 0.8836,
+      "step": 26635
+    },
+    {
+      "epoch": 1.0256015399422522,
+      "grad_norm": 1.539516568183899,
+      "learning_rate": 9.605475561086842e-05,
+      "loss": 0.8616,
+      "step": 26640
+    },
+    {
+      "epoch": 1.025794032723773,
+      "grad_norm": 1.0243744850158691,
+      "learning_rate": 9.602453972891189e-05,
+      "loss": 0.7763,
+      "step": 26645
+    },
+    {
+      "epoch": 1.0259865255052936,
+      "grad_norm": 1.4782140254974365,
+      "learning_rate": 9.599432421048488e-05,
+      "loss": 0.8951,
+      "step": 26650
+    },
+    {
+      "epoch": 1.0261790182868142,
+      "grad_norm": 1.0604002475738525,
+      "learning_rate": 9.596410905835037e-05,
+      "loss": 0.8381,
+      "step": 26655
+    },
+    {
+      "epoch": 1.026371511068335,
+      "grad_norm": 1.6597970724105835,
+      "learning_rate": 9.593389427527135e-05,
+      "loss": 0.9625,
+      "step": 26660
+    },
+    {
+      "epoch": 1.0265640038498556,
+      "grad_norm": 0.9541493654251099,
+      "learning_rate": 9.590367986401078e-05,
+      "loss": 0.8285,
+      "step": 26665
+    },
+    {
+      "epoch": 1.0267564966313762,
+      "grad_norm": 1.096489667892456,
+      "learning_rate": 9.587346582733155e-05,
+      "loss": 0.8462,
+      "step": 26670
+    },
+    {
+      "epoch": 1.026948989412897,
+      "grad_norm": 1.7347831726074219,
+      "learning_rate": 9.58432521679965e-05,
+      "loss": 0.9057,
+      "step": 26675
+    },
+    {
+      "epoch": 1.0271414821944178,
+      "grad_norm": 1.02500319480896,
+      "learning_rate": 9.581303888876855e-05,
+      "loss": 0.8081,
+      "step": 26680
+    },
+    {
+      "epoch": 1.0273339749759385,
+      "grad_norm": 1.090671181678772,
+      "learning_rate": 9.578282599241044e-05,
+      "loss": 0.829,
+      "step": 26685
+    },
+    {
+      "epoch": 1.0275264677574591,
+      "grad_norm": 1.345054030418396,
+      "learning_rate": 9.575261348168493e-05,
+      "loss": 0.9518,
+      "step": 26690
+    },
+    {
+      "epoch": 1.0277189605389798,
+      "grad_norm": 1.129591703414917,
+      "learning_rate": 9.572240135935483e-05,
+      "loss": 0.8178,
+      "step": 26695
+    },
+    {
+      "epoch": 1.0279114533205005,
+      "grad_norm": 1.1437878608703613,
+      "learning_rate": 9.569218962818276e-05,
+      "loss": 0.8266,
+      "step": 26700
+    },
+    {
+      "epoch": 1.0281039461020212,
+      "grad_norm": 1.2581342458724976,
+      "learning_rate": 9.566197829093138e-05,
+      "loss": 0.8247,
+      "step": 26705
+    },
+    {
+      "epoch": 1.0282964388835418,
+      "grad_norm": 1.0760700702667236,
+      "learning_rate": 9.56317673503634e-05,
+      "loss": 1.003,
+      "step": 26710
+    },
+    {
+      "epoch": 1.0284889316650625,
+      "grad_norm": 1.10942804813385,
+      "learning_rate": 9.560155680924137e-05,
+      "loss": 0.7992,
+      "step": 26715
+    },
+    {
+      "epoch": 1.0286814244465832,
+      "grad_norm": 1.9202826023101807,
+      "learning_rate": 9.557134667032776e-05,
+      "loss": 0.989,
+      "step": 26720
+    },
+    {
+      "epoch": 1.0288739172281038,
+      "grad_norm": 2.0237245559692383,
+      "learning_rate": 9.554113693638522e-05,
+      "loss": 0.9404,
+      "step": 26725
+    },
+    {
+      "epoch": 1.0290664100096247,
+      "grad_norm": 2.128678798675537,
+      "learning_rate": 9.551092761017618e-05,
+      "loss": 1.0402,
+      "step": 26730
+    },
+    {
+      "epoch": 1.0292589027911454,
+      "grad_norm": 0.9286008477210999,
+      "learning_rate": 9.5480718694463e-05,
+      "loss": 0.9927,
+      "step": 26735
+    },
+    {
+      "epoch": 1.029451395572666,
+      "grad_norm": 1.2808765172958374,
+      "learning_rate": 9.545051019200823e-05,
+      "loss": 0.9026,
+      "step": 26740
+    },
+    {
+      "epoch": 1.0296438883541867,
+      "grad_norm": 1.4116573333740234,
+      "learning_rate": 9.542030210557415e-05,
+      "loss": 0.8285,
+      "step": 26745
+    },
+    {
+      "epoch": 1.0298363811357074,
+      "grad_norm": 2.100184440612793,
+      "learning_rate": 9.539009443792309e-05,
+      "loss": 0.925,
+      "step": 26750
+    },
+    {
+      "epoch": 1.030028873917228,
+      "grad_norm": 1.3617011308670044,
+      "learning_rate": 9.535988719181736e-05,
+      "loss": 0.8356,
+      "step": 26755
+    },
+    {
+      "epoch": 1.0302213666987488,
+      "grad_norm": 0.9685754776000977,
+      "learning_rate": 9.532968037001923e-05,
+      "loss": 0.8056,
+      "step": 26760
+    },
+    {
+      "epoch": 1.0304138594802694,
+      "grad_norm": 1.2491365671157837,
+      "learning_rate": 9.529947397529086e-05,
+      "loss": 0.9138,
+      "step": 26765
+    },
+    {
+      "epoch": 1.03060635226179,
+      "grad_norm": 0.8603072166442871,
+      "learning_rate": 9.526926801039449e-05,
+      "loss": 0.9497,
+      "step": 26770
+    },
+    {
+      "epoch": 1.030798845043311,
+      "grad_norm": 0.8559325337409973,
+      "learning_rate": 9.523906247809216e-05,
+      "loss": 0.6938,
+      "step": 26775
+    },
+    {
+      "epoch": 1.0309913378248317,
+      "grad_norm": 1.330196738243103,
+      "learning_rate": 9.520885738114609e-05,
+      "loss": 0.8141,
+      "step": 26780
+    },
+    {
+      "epoch": 1.0311838306063523,
+      "grad_norm": 1.7451072931289673,
+      "learning_rate": 9.517865272231827e-05,
+      "loss": 0.8798,
+      "step": 26785
+    },
+    {
+      "epoch": 1.031376323387873,
+      "grad_norm": 2.3189198970794678,
+      "learning_rate": 9.514844850437066e-05,
+      "loss": 0.9944,
+      "step": 26790
+    },
+    {
+      "epoch": 1.0315688161693937,
+      "grad_norm": 1.4511535167694092,
+      "learning_rate": 9.511824473006535e-05,
+      "loss": 0.8759,
+      "step": 26795
+    },
+    {
+      "epoch": 1.0317613089509143,
+      "grad_norm": 1.5764187574386597,
+      "learning_rate": 9.508804140216423e-05,
+      "loss": 0.8792,
+      "step": 26800
+    },
+    {
+      "epoch": 1.031953801732435,
+      "grad_norm": 1.368013620376587,
+      "learning_rate": 9.505783852342913e-05,
+      "loss": 0.8946,
+      "step": 26805
+    },
+    {
+      "epoch": 1.0321462945139557,
+      "grad_norm": 1.747062087059021,
+      "learning_rate": 9.502763609662201e-05,
+      "loss": 0.9188,
+      "step": 26810
+    },
+    {
+      "epoch": 1.0323387872954763,
+      "grad_norm": 1.7886524200439453,
+      "learning_rate": 9.499743412450463e-05,
+      "loss": 0.8947,
+      "step": 26815
+    },
+    {
+      "epoch": 1.032531280076997,
+      "grad_norm": 1.3724089860916138,
+      "learning_rate": 9.496723260983871e-05,
+      "loss": 0.909,
+      "step": 26820
+    },
+    {
+      "epoch": 1.032723772858518,
+      "grad_norm": 1.7624560594558716,
+      "learning_rate": 9.493703155538607e-05,
+      "loss": 0.6966,
+      "step": 26825
+    },
+    {
+      "epoch": 1.0329162656400386,
+      "grad_norm": 1.3632899522781372,
+      "learning_rate": 9.490683096390836e-05,
+      "loss": 0.7867,
+      "step": 26830
+    },
+    {
+      "epoch": 1.0331087584215592,
+      "grad_norm": 0.9639708399772644,
+      "learning_rate": 9.487663083816718e-05,
+      "loss": 1.0168,
+      "step": 26835
+    },
+    {
+      "epoch": 1.03330125120308,
+      "grad_norm": 1.6078749895095825,
+      "learning_rate": 9.484643118092422e-05,
+      "loss": 0.8775,
+      "step": 26840
+    },
+    {
+      "epoch": 1.0334937439846006,
+      "grad_norm": 0.9812796711921692,
+      "learning_rate": 9.481623199494098e-05,
+      "loss": 0.8573,
+      "step": 26845
+    },
+    {
+      "epoch": 1.0336862367661213,
+      "grad_norm": 1.926906943321228,
+      "learning_rate": 9.478603328297896e-05,
+      "loss": 0.8621,
+      "step": 26850
+    },
+    {
+      "epoch": 1.033878729547642,
+      "grad_norm": 1.4582041501998901,
+      "learning_rate": 9.475583504779966e-05,
+      "loss": 0.8978,
+      "step": 26855
+    },
+    {
+      "epoch": 1.0340712223291626,
+      "grad_norm": 1.22900390625,
+      "learning_rate": 9.472563729216453e-05,
+      "loss": 0.8079,
+      "step": 26860
+    },
+    {
+      "epoch": 1.0342637151106833,
+      "grad_norm": 1.8301913738250732,
+      "learning_rate": 9.46954400188349e-05,
+      "loss": 0.8844,
+      "step": 26865
+    },
+    {
+      "epoch": 1.034456207892204,
+      "grad_norm": 0.9074123501777649,
+      "learning_rate": 9.466524323057217e-05,
+      "loss": 0.8437,
+      "step": 26870
+    },
+    {
+      "epoch": 1.0346487006737248,
+      "grad_norm": 1.1448841094970703,
+      "learning_rate": 9.463504693013761e-05,
+      "loss": 0.9197,
+      "step": 26875
+    },
+    {
+      "epoch": 1.0348411934552455,
+      "grad_norm": 1.2825102806091309,
+      "learning_rate": 9.460485112029241e-05,
+      "loss": 0.8754,
+      "step": 26880
+    },
+    {
+      "epoch": 1.0350336862367662,
+      "grad_norm": 1.4492766857147217,
+      "learning_rate": 9.45746558037979e-05,
+      "loss": 1.068,
+      "step": 26885
+    },
+    {
+      "epoch": 1.0352261790182868,
+      "grad_norm": 1.4964579343795776,
+      "learning_rate": 9.454446098341516e-05,
+      "loss": 0.7369,
+      "step": 26890
+    },
+    {
+      "epoch": 1.0354186717998075,
+      "grad_norm": 1.2019277811050415,
+      "learning_rate": 9.45142666619053e-05,
+      "loss": 0.8437,
+      "step": 26895
+    },
+    {
+      "epoch": 1.0356111645813282,
+      "grad_norm": 1.22611665725708,
+      "learning_rate": 9.448407284202943e-05,
+      "loss": 0.7403,
+      "step": 26900
+    },
+    {
+      "epoch": 1.0358036573628489,
+      "grad_norm": 1.3472671508789062,
+      "learning_rate": 9.445387952654857e-05,
+      "loss": 0.8897,
+      "step": 26905
+    },
+    {
+      "epoch": 1.0359961501443695,
+      "grad_norm": 0.9485424160957336,
+      "learning_rate": 9.442368671822364e-05,
+      "loss": 1.0716,
+      "step": 26910
+    },
+    {
+      "epoch": 1.0361886429258902,
+      "grad_norm": 1.5943937301635742,
+      "learning_rate": 9.439349441981569e-05,
+      "loss": 0.797,
+      "step": 26915
+    },
+    {
+      "epoch": 1.036381135707411,
+      "grad_norm": 1.312227487564087,
+      "learning_rate": 9.436330263408551e-05,
+      "loss": 0.7671,
+      "step": 26920
+    },
+    {
+      "epoch": 1.0365736284889318,
+      "grad_norm": 1.4809763431549072,
+      "learning_rate": 9.433311136379393e-05,
+      "loss": 0.958,
+      "step": 26925
+    },
+    {
+      "epoch": 1.0367661212704524,
+      "grad_norm": 1.0630615949630737,
+      "learning_rate": 9.430292061170181e-05,
+      "loss": 0.8594,
+      "step": 26930
+    },
+    {
+      "epoch": 1.036958614051973,
+      "grad_norm": 2.018148422241211,
+      "learning_rate": 9.427273038056988e-05,
+      "loss": 0.8566,
+      "step": 26935
+    },
+    {
+      "epoch": 1.0371511068334938,
+      "grad_norm": 1.4871729612350464,
+      "learning_rate": 9.424254067315875e-05,
+      "loss": 1.0174,
+      "step": 26940
+    },
+    {
+      "epoch": 1.0373435996150144,
+      "grad_norm": 1.2022002935409546,
+      "learning_rate": 9.421235149222919e-05,
+      "loss": 0.7991,
+      "step": 26945
+    },
+    {
+      "epoch": 1.037536092396535,
+      "grad_norm": 1.7707716226577759,
+      "learning_rate": 9.418216284054174e-05,
+      "loss": 0.881,
+      "step": 26950
+    },
+    {
+      "epoch": 1.0377285851780558,
+      "grad_norm": 1.1901010274887085,
+      "learning_rate": 9.415197472085693e-05,
+      "loss": 0.7872,
+      "step": 26955
+    },
+    {
+      "epoch": 1.0379210779595764,
+      "grad_norm": 1.460284948348999,
+      "learning_rate": 9.412178713593532e-05,
+      "loss": 0.9017,
+      "step": 26960
+    },
+    {
+      "epoch": 1.0381135707410971,
+      "grad_norm": 1.8410388231277466,
+      "learning_rate": 9.409160008853735e-05,
+      "loss": 0.8422,
+      "step": 26965
+    },
+    {
+      "epoch": 1.038306063522618,
+      "grad_norm": 1.3502047061920166,
+      "learning_rate": 9.406141358142339e-05,
+      "loss": 0.6949,
+      "step": 26970
+    },
+    {
+      "epoch": 1.0384985563041387,
+      "grad_norm": 1.433706521987915,
+      "learning_rate": 9.403122761735382e-05,
+      "loss": 0.8941,
+      "step": 26975
+    },
+    {
+      "epoch": 1.0386910490856593,
+      "grad_norm": 1.3440196514129639,
+      "learning_rate": 9.400104219908895e-05,
+      "loss": 1.1485,
+      "step": 26980
+    },
+    {
+      "epoch": 1.03888354186718,
+      "grad_norm": 1.2331663370132446,
+      "learning_rate": 9.397085732938902e-05,
+      "loss": 0.7441,
+      "step": 26985
+    },
+    {
+      "epoch": 1.0390760346487007,
+      "grad_norm": 0.953507125377655,
+      "learning_rate": 9.394067301101425e-05,
+      "loss": 0.9023,
+      "step": 26990
+    },
+    {
+      "epoch": 1.0392685274302214,
+      "grad_norm": 1.993162751197815,
+      "learning_rate": 9.391048924672483e-05,
+      "loss": 0.9112,
+      "step": 26995
+    },
+    {
+      "epoch": 1.039461020211742,
+      "grad_norm": 1.193826675415039,
+      "learning_rate": 9.388030603928077e-05,
+      "loss": 0.9214,
+      "step": 27000
+    },
+    {
+      "epoch": 1.0396535129932627,
+      "grad_norm": 1.5678527355194092,
+      "learning_rate": 9.385012339144221e-05,
+      "loss": 0.8798,
+      "step": 27005
+    },
+    {
+      "epoch": 1.0398460057747834,
+      "grad_norm": 1.09235680103302,
+      "learning_rate": 9.381994130596916e-05,
+      "loss": 0.9111,
+      "step": 27010
+    },
+    {
+      "epoch": 1.040038498556304,
+      "grad_norm": 1.3763002157211304,
+      "learning_rate": 9.378975978562147e-05,
+      "loss": 0.8024,
+      "step": 27015
+    },
+    {
+      "epoch": 1.040230991337825,
+      "grad_norm": 2.1225831508636475,
+      "learning_rate": 9.375957883315916e-05,
+      "loss": 0.8644,
+      "step": 27020
+    },
+    {
+      "epoch": 1.0404234841193456,
+      "grad_norm": 1.7183401584625244,
+      "learning_rate": 9.372939845134197e-05,
+      "loss": 0.6908,
+      "step": 27025
+    },
+    {
+      "epoch": 1.0406159769008663,
+      "grad_norm": 1.1370265483856201,
+      "learning_rate": 9.369921864292981e-05,
+      "loss": 0.7209,
+      "step": 27030
+    },
+    {
+      "epoch": 1.040808469682387,
+      "grad_norm": 1.4425493478775024,
+      "learning_rate": 9.366903941068237e-05,
+      "loss": 0.9938,
+      "step": 27035
+    },
+    {
+      "epoch": 1.0410009624639076,
+      "grad_norm": 1.009415864944458,
+      "learning_rate": 9.363886075735928e-05,
+      "loss": 0.822,
+      "step": 27040
+    },
+    {
+      "epoch": 1.0411934552454283,
+      "grad_norm": 1.7961349487304688,
+      "learning_rate": 9.36086826857203e-05,
+      "loss": 0.8856,
+      "step": 27045
+    },
+    {
+      "epoch": 1.041385948026949,
+      "grad_norm": 1.6287988424301147,
+      "learning_rate": 9.357850519852492e-05,
+      "loss": 0.6491,
+      "step": 27050
+    },
+    {
+      "epoch": 1.0415784408084696,
+      "grad_norm": 0.8782758116722107,
+      "learning_rate": 9.354832829853268e-05,
+      "loss": 0.7966,
+      "step": 27055
+    },
+    {
+      "epoch": 1.0417709335899903,
+      "grad_norm": 2.51961612701416,
+      "learning_rate": 9.351815198850312e-05,
+      "loss": 0.7717,
+      "step": 27060
+    },
+    {
+      "epoch": 1.041963426371511,
+      "grad_norm": 1.5812909603118896,
+      "learning_rate": 9.348797627119561e-05,
+      "loss": 0.8912,
+      "step": 27065
+    },
+    {
+      "epoch": 1.0421559191530319,
+      "grad_norm": 2.106905460357666,
+      "learning_rate": 9.345780114936951e-05,
+      "loss": 0.8447,
+      "step": 27070
+    },
+    {
+      "epoch": 1.0423484119345525,
+      "grad_norm": 1.7063692808151245,
+      "learning_rate": 9.342762662578419e-05,
+      "loss": 0.9389,
+      "step": 27075
+    },
+    {
+      "epoch": 1.0425409047160732,
+      "grad_norm": 1.5984817743301392,
+      "learning_rate": 9.339745270319886e-05,
+      "loss": 0.9812,
+      "step": 27080
+    },
+    {
+      "epoch": 1.0427333974975939,
+      "grad_norm": 1.0404373407363892,
+      "learning_rate": 9.336727938437274e-05,
+      "loss": 0.8982,
+      "step": 27085
+    },
+    {
+      "epoch": 1.0429258902791145,
+      "grad_norm": 1.7776751518249512,
+      "learning_rate": 9.333710667206498e-05,
+      "loss": 0.7817,
+      "step": 27090
+    },
+    {
+      "epoch": 1.0431183830606352,
+      "grad_norm": 1.092366099357605,
+      "learning_rate": 9.330693456903471e-05,
+      "loss": 0.8735,
+      "step": 27095
+    },
+    {
+      "epoch": 1.0433108758421559,
+      "grad_norm": 1.3183677196502686,
+      "learning_rate": 9.327676307804088e-05,
+      "loss": 0.855,
+      "step": 27100
+    },
+    {
+      "epoch": 1.0435033686236765,
+      "grad_norm": 1.6006138324737549,
+      "learning_rate": 9.324659220184256e-05,
+      "loss": 0.9884,
+      "step": 27105
+    },
+    {
+      "epoch": 1.0436958614051972,
+      "grad_norm": 2.0356695652008057,
+      "learning_rate": 9.321642194319867e-05,
+      "loss": 0.9388,
+      "step": 27110
+    },
+    {
+      "epoch": 1.043888354186718,
+      "grad_norm": 1.6827491521835327,
+      "learning_rate": 9.3186252304868e-05,
+      "loss": 0.8646,
+      "step": 27115
+    },
+    {
+      "epoch": 1.0440808469682388,
+      "grad_norm": 1.6442832946777344,
+      "learning_rate": 9.315608328960946e-05,
+      "loss": 0.9892,
+      "step": 27120
+    },
+    {
+      "epoch": 1.0442733397497594,
+      "grad_norm": 2.0704457759857178,
+      "learning_rate": 9.312591490018176e-05,
+      "loss": 0.7849,
+      "step": 27125
+    },
+    {
+      "epoch": 1.0444658325312801,
+      "grad_norm": 1.5209687948226929,
+      "learning_rate": 9.309574713934359e-05,
+      "loss": 0.7079,
+      "step": 27130
+    },
+    {
+      "epoch": 1.0446583253128008,
+      "grad_norm": 1.6896909475326538,
+      "learning_rate": 9.306558000985363e-05,
+      "loss": 0.8909,
+      "step": 27135
+    },
+    {
+      "epoch": 1.0448508180943215,
+      "grad_norm": 1.2594019174575806,
+      "learning_rate": 9.303541351447043e-05,
+      "loss": 0.9134,
+      "step": 27140
+    },
+    {
+      "epoch": 1.0450433108758421,
+      "grad_norm": 1.4927109479904175,
+      "learning_rate": 9.300524765595252e-05,
+      "loss": 0.9467,
+      "step": 27145
+    },
+    {
+      "epoch": 1.0452358036573628,
+      "grad_norm": 2.307347059249878,
+      "learning_rate": 9.297508243705838e-05,
+      "loss": 0.9415,
+      "step": 27150
+    },
+    {
+      "epoch": 1.0454282964388835,
+      "grad_norm": 1.2363544702529907,
+      "learning_rate": 9.294491786054645e-05,
+      "loss": 0.8342,
+      "step": 27155
+    },
+    {
+      "epoch": 1.0456207892204041,
+      "grad_norm": 1.7037842273712158,
+      "learning_rate": 9.291475392917498e-05,
+      "loss": 0.8995,
+      "step": 27160
+    },
+    {
+      "epoch": 1.045813282001925,
+      "grad_norm": 1.9591143131256104,
+      "learning_rate": 9.288459064570238e-05,
+      "loss": 0.9448,
+      "step": 27165
+    },
+    {
+      "epoch": 1.0460057747834457,
+      "grad_norm": 1.5714846849441528,
+      "learning_rate": 9.285442801288681e-05,
+      "loss": 0.7211,
+      "step": 27170
+    },
+    {
+      "epoch": 1.0461982675649664,
+      "grad_norm": 1.2620651721954346,
+      "learning_rate": 9.282426603348648e-05,
+      "loss": 0.8678,
+      "step": 27175
+    },
+    {
+      "epoch": 1.046390760346487,
+      "grad_norm": 1.3091076612472534,
+      "learning_rate": 9.279410471025946e-05,
+      "loss": 0.947,
+      "step": 27180
+    },
+    {
+      "epoch": 1.0465832531280077,
+      "grad_norm": 1.055066704750061,
+      "learning_rate": 9.276394404596384e-05,
+      "loss": 0.8502,
+      "step": 27185
+    },
+    {
+      "epoch": 1.0467757459095284,
+      "grad_norm": 1.4299228191375732,
+      "learning_rate": 9.27337840433576e-05,
+      "loss": 0.8472,
+      "step": 27190
+    },
+    {
+      "epoch": 1.046968238691049,
+      "grad_norm": 1.315097451210022,
+      "learning_rate": 9.27036247051987e-05,
+      "loss": 0.9291,
+      "step": 27195
+    },
+    {
+      "epoch": 1.0471607314725697,
+      "grad_norm": 1.491268515586853,
+      "learning_rate": 9.267346603424497e-05,
+      "loss": 0.8054,
+      "step": 27200
+    },
+    {
+      "epoch": 1.0473532242540904,
+      "grad_norm": 1.2065061330795288,
+      "learning_rate": 9.264330803325422e-05,
+      "loss": 0.7778,
+      "step": 27205
+    },
+    {
+      "epoch": 1.0475457170356113,
+      "grad_norm": 1.1134014129638672,
+      "learning_rate": 9.261315070498425e-05,
+      "loss": 0.9711,
+      "step": 27210
+    },
+    {
+      "epoch": 1.047738209817132,
+      "grad_norm": 1.3288688659667969,
+      "learning_rate": 9.258299405219271e-05,
+      "loss": 0.833,
+      "step": 27215
+    },
+    {
+      "epoch": 1.0479307025986526,
+      "grad_norm": 1.440891146659851,
+      "learning_rate": 9.255283807763719e-05,
+      "loss": 0.9198,
+      "step": 27220
+    },
+    {
+      "epoch": 1.0481231953801733,
+      "grad_norm": 1.3492053747177124,
+      "learning_rate": 9.252268278407532e-05,
+      "loss": 0.7503,
+      "step": 27225
+    },
+    {
+      "epoch": 1.048315688161694,
+      "grad_norm": 0.9512596726417542,
+      "learning_rate": 9.249252817426459e-05,
+      "loss": 0.8406,
+      "step": 27230
+    },
+    {
+      "epoch": 1.0485081809432146,
+      "grad_norm": 1.8228788375854492,
+      "learning_rate": 9.24623742509624e-05,
+      "loss": 0.8661,
+      "step": 27235
+    },
+    {
+      "epoch": 1.0487006737247353,
+      "grad_norm": 1.9597275257110596,
+      "learning_rate": 9.243222101692617e-05,
+      "loss": 0.8972,
+      "step": 27240
+    },
+    {
+      "epoch": 1.048893166506256,
+      "grad_norm": 0.8910306692123413,
+      "learning_rate": 9.24020684749132e-05,
+      "loss": 0.7583,
+      "step": 27245
+    },
+    {
+      "epoch": 1.0490856592877766,
+      "grad_norm": 1.1714658737182617,
+      "learning_rate": 9.237191662768071e-05,
+      "loss": 0.7962,
+      "step": 27250
+    },
+    {
+      "epoch": 1.0492781520692973,
+      "grad_norm": 1.0180914402008057,
+      "learning_rate": 9.234176547798595e-05,
+      "loss": 0.7615,
+      "step": 27255
+    },
+    {
+      "epoch": 1.0494706448508182,
+      "grad_norm": 1.5986478328704834,
+      "learning_rate": 9.231161502858599e-05,
+      "loss": 0.6797,
+      "step": 27260
+    },
+    {
+      "epoch": 1.0496631376323389,
+      "grad_norm": 1.1098225116729736,
+      "learning_rate": 9.228146528223787e-05,
+      "loss": 0.867,
+      "step": 27265
+    },
+    {
+      "epoch": 1.0498556304138595,
+      "grad_norm": 1.7311850786209106,
+      "learning_rate": 9.225131624169868e-05,
+      "loss": 0.8742,
+      "step": 27270
+    },
+    {
+      "epoch": 1.0500481231953802,
+      "grad_norm": 1.0296027660369873,
+      "learning_rate": 9.222116790972526e-05,
+      "loss": 0.9372,
+      "step": 27275
+    },
+    {
+      "epoch": 1.0502406159769009,
+      "grad_norm": 1.6226547956466675,
+      "learning_rate": 9.219102028907448e-05,
+      "loss": 0.7823,
+      "step": 27280
+    },
+    {
+      "epoch": 1.0504331087584216,
+      "grad_norm": 1.745197057723999,
+      "learning_rate": 9.216087338250321e-05,
+      "loss": 0.7434,
+      "step": 27285
+    },
+    {
+      "epoch": 1.0506256015399422,
+      "grad_norm": 1.3191584348678589,
+      "learning_rate": 9.21307271927681e-05,
+      "loss": 1.0177,
+      "step": 27290
+    },
+    {
+      "epoch": 1.050818094321463,
+      "grad_norm": 1.4447453022003174,
+      "learning_rate": 9.21005817226259e-05,
+      "loss": 0.7972,
+      "step": 27295
+    },
+    {
+      "epoch": 1.0510105871029836,
+      "grad_norm": 1.4396947622299194,
+      "learning_rate": 9.207043697483315e-05,
+      "loss": 0.901,
+      "step": 27300
+    },
+    {
+      "epoch": 1.0512030798845042,
+      "grad_norm": 2.1123616695404053,
+      "learning_rate": 9.204029295214641e-05,
+      "loss": 1.0325,
+      "step": 27305
+    },
+    {
+      "epoch": 1.0513955726660251,
+      "grad_norm": 1.1420029401779175,
+      "learning_rate": 9.201014965732216e-05,
+      "loss": 0.8349,
+      "step": 27310
+    },
+    {
+      "epoch": 1.0515880654475458,
+      "grad_norm": 1.5591840744018555,
+      "learning_rate": 9.198000709311681e-05,
+      "loss": 0.8653,
+      "step": 27315
+    },
+    {
+      "epoch": 1.0517805582290665,
+      "grad_norm": 1.5964603424072266,
+      "learning_rate": 9.194986526228662e-05,
+      "loss": 0.9203,
+      "step": 27320
+    },
+    {
+      "epoch": 1.0519730510105871,
+      "grad_norm": 1.5072814226150513,
+      "learning_rate": 9.191972416758798e-05,
+      "loss": 0.8917,
+      "step": 27325
+    },
+    {
+      "epoch": 1.0521655437921078,
+      "grad_norm": 1.7254325151443481,
+      "learning_rate": 9.188958381177705e-05,
+      "loss": 0.7569,
+      "step": 27330
+    },
+    {
+      "epoch": 1.0523580365736285,
+      "grad_norm": 1.856126070022583,
+      "learning_rate": 9.18594441976099e-05,
+      "loss": 0.8379,
+      "step": 27335
+    },
+    {
+      "epoch": 1.0525505293551491,
+      "grad_norm": 1.1711968183517456,
+      "learning_rate": 9.18293053278427e-05,
+      "loss": 0.9595,
+      "step": 27340
+    },
+    {
+      "epoch": 1.0527430221366698,
+      "grad_norm": 0.8440997004508972,
+      "learning_rate": 9.179916720523138e-05,
+      "loss": 0.8221,
+      "step": 27345
+    },
+    {
+      "epoch": 1.0529355149181905,
+      "grad_norm": 1.532020092010498,
+      "learning_rate": 9.176902983253187e-05,
+      "loss": 0.8155,
+      "step": 27350
+    },
+    {
+      "epoch": 1.0531280076997112,
+      "grad_norm": 1.3105454444885254,
+      "learning_rate": 9.173889321250009e-05,
+      "loss": 1.0119,
+      "step": 27355
+    },
+    {
+      "epoch": 1.053320500481232,
+      "grad_norm": 1.6066848039627075,
+      "learning_rate": 9.170875734789182e-05,
+      "loss": 0.8037,
+      "step": 27360
+    },
+    {
+      "epoch": 1.0535129932627527,
+      "grad_norm": 0.9645727872848511,
+      "learning_rate": 9.167862224146271e-05,
+      "loss": 0.8052,
+      "step": 27365
+    },
+    {
+      "epoch": 1.0537054860442734,
+      "grad_norm": 1.2492599487304688,
+      "learning_rate": 9.164848789596851e-05,
+      "loss": 0.8299,
+      "step": 27370
+    },
+    {
+      "epoch": 1.053897978825794,
+      "grad_norm": 1.861069679260254,
+      "learning_rate": 9.161835431416478e-05,
+      "loss": 0.8103,
+      "step": 27375
+    },
+    {
+      "epoch": 1.0540904716073147,
+      "grad_norm": 1.3065990209579468,
+      "learning_rate": 9.158822149880699e-05,
+      "loss": 0.7432,
+      "step": 27380
+    },
+    {
+      "epoch": 1.0542829643888354,
+      "grad_norm": 1.0527342557907104,
+      "learning_rate": 9.155808945265063e-05,
+      "loss": 0.9543,
+      "step": 27385
+    },
+    {
+      "epoch": 1.054475457170356,
+      "grad_norm": 1.0329846143722534,
+      "learning_rate": 9.152795817845109e-05,
+      "loss": 0.7798,
+      "step": 27390
+    },
+    {
+      "epoch": 1.0546679499518767,
+      "grad_norm": 1.2137219905853271,
+      "learning_rate": 9.149782767896364e-05,
+      "loss": 0.8731,
+      "step": 27395
+    },
+    {
+      "epoch": 1.0548604427333974,
+      "grad_norm": 1.7884348630905151,
+      "learning_rate": 9.146769795694351e-05,
+      "loss": 0.7893,
+      "step": 27400
+    },
+    {
+      "epoch": 1.0550529355149183,
+      "grad_norm": 1.842133641242981,
+      "learning_rate": 9.143756901514591e-05,
+      "loss": 1.0068,
+      "step": 27405
+    },
+    {
+      "epoch": 1.055245428296439,
+      "grad_norm": 1.8429181575775146,
+      "learning_rate": 9.140744085632587e-05,
+      "loss": 0.8311,
+      "step": 27410
+    },
+    {
+      "epoch": 1.0554379210779596,
+      "grad_norm": 1.6385420560836792,
+      "learning_rate": 9.137731348323848e-05,
+      "loss": 0.9761,
+      "step": 27415
+    },
+    {
+      "epoch": 1.0556304138594803,
+      "grad_norm": 1.289536714553833,
+      "learning_rate": 9.134718689863863e-05,
+      "loss": 0.9112,
+      "step": 27420
+    },
+    {
+      "epoch": 1.055822906641001,
+      "grad_norm": 1.3178956508636475,
+      "learning_rate": 9.13170611052812e-05,
+      "loss": 0.9767,
+      "step": 27425
+    },
+    {
+      "epoch": 1.0560153994225217,
+      "grad_norm": 1.588117003440857,
+      "learning_rate": 9.128693610592104e-05,
+      "loss": 0.9646,
+      "step": 27430
+    },
+    {
+      "epoch": 1.0562078922040423,
+      "grad_norm": 2.409613609313965,
+      "learning_rate": 9.125681190331287e-05,
+      "loss": 0.9672,
+      "step": 27435
+    },
+    {
+      "epoch": 1.056400384985563,
+      "grad_norm": 2.1366968154907227,
+      "learning_rate": 9.123271311673885e-05,
+      "loss": 1.0157,
+      "step": 27440
+    },
+    {
+      "epoch": 1.0565928777670837,
+      "grad_norm": 1.3184314966201782,
+      "learning_rate": 9.12025903552259e-05,
+      "loss": 0.7433,
+      "step": 27445
+    },
+    {
+      "epoch": 1.0567853705486043,
+      "grad_norm": 1.710648775100708,
+      "learning_rate": 9.117246839817776e-05,
+      "loss": 1.145,
+      "step": 27450
+    },
+    {
+      "epoch": 1.0569778633301252,
+      "grad_norm": 1.0037586688995361,
+      "learning_rate": 9.114234724834895e-05,
+      "loss": 0.8956,
+      "step": 27455
+    },
+    {
+      "epoch": 1.057170356111646,
+      "grad_norm": 1.9379615783691406,
+      "learning_rate": 9.11122269084938e-05,
+      "loss": 0.8227,
+      "step": 27460
+    },
+    {
+      "epoch": 1.0573628488931666,
+      "grad_norm": 1.1451090574264526,
+      "learning_rate": 9.108210738136661e-05,
+      "loss": 1.0623,
+      "step": 27465
+    },
+    {
+      "epoch": 1.0575553416746872,
+      "grad_norm": 1.3829764127731323,
+      "learning_rate": 9.105198866972162e-05,
+      "loss": 0.6903,
+      "step": 27470
+    },
+    {
+      "epoch": 1.057747834456208,
+      "grad_norm": 0.7273270487785339,
+      "learning_rate": 9.102187077631302e-05,
+      "loss": 0.7338,
+      "step": 27475
+    },
+    {
+      "epoch": 1.0579403272377286,
+      "grad_norm": 1.3922228813171387,
+      "learning_rate": 9.099175370389482e-05,
+      "loss": 0.9767,
+      "step": 27480
+    },
+    {
+      "epoch": 1.0581328200192492,
+      "grad_norm": 1.5572834014892578,
+      "learning_rate": 9.096163745522112e-05,
+      "loss": 0.7942,
+      "step": 27485
+    },
+    {
+      "epoch": 1.05832531280077,
+      "grad_norm": 1.2158164978027344,
+      "learning_rate": 9.093152203304578e-05,
+      "loss": 0.8304,
+      "step": 27490
+    },
+    {
+      "epoch": 1.0585178055822906,
+      "grad_norm": 1.2715733051300049,
+      "learning_rate": 9.090140744012265e-05,
+      "loss": 0.7802,
+      "step": 27495
+    },
+    {
+      "epoch": 1.0587102983638113,
+      "grad_norm": 1.2508288621902466,
+      "learning_rate": 9.087129367920557e-05,
+      "loss": 0.8579,
+      "step": 27500
+    },
+    {
+      "epoch": 1.0589027911453321,
+      "grad_norm": 1.9734135866165161,
+      "learning_rate": 9.084118075304821e-05,
+      "loss": 0.8703,
+      "step": 27505
+    },
+    {
+      "epoch": 1.0590952839268528,
+      "grad_norm": 1.4569380283355713,
+      "learning_rate": 9.081106866440415e-05,
+      "loss": 0.9245,
+      "step": 27510
+    },
+    {
+      "epoch": 1.0592877767083735,
+      "grad_norm": 1.2966618537902832,
+      "learning_rate": 9.078095741602704e-05,
+      "loss": 0.8588,
+      "step": 27515
+    },
+    {
+      "epoch": 1.0594802694898942,
+      "grad_norm": 1.5596932172775269,
+      "learning_rate": 9.07508470106703e-05,
+      "loss": 0.8499,
+      "step": 27520
+    },
+    {
+      "epoch": 1.0596727622714148,
+      "grad_norm": 1.3186569213867188,
+      "learning_rate": 9.07207374510873e-05,
+      "loss": 0.825,
+      "step": 27525
+    },
+    {
+      "epoch": 1.0598652550529355,
+      "grad_norm": 1.393056869506836,
+      "learning_rate": 9.069062874003141e-05,
+      "loss": 0.7853,
+      "step": 27530
+    },
+    {
+      "epoch": 1.0600577478344562,
+      "grad_norm": 1.091132402420044,
+      "learning_rate": 9.066052088025587e-05,
+      "loss": 0.9176,
+      "step": 27535
+    },
+    {
+      "epoch": 1.0602502406159768,
+      "grad_norm": 2.1310064792633057,
+      "learning_rate": 9.063041387451378e-05,
+      "loss": 0.6352,
+      "step": 27540
+    },
+    {
+      "epoch": 1.0604427333974975,
+      "grad_norm": 1.153936505317688,
+      "learning_rate": 9.060030772555833e-05,
+      "loss": 0.9243,
+      "step": 27545
+    },
+    {
+      "epoch": 1.0606352261790182,
+      "grad_norm": 1.3241852521896362,
+      "learning_rate": 9.05702024361424e-05,
+      "loss": 0.7682,
+      "step": 27550
+    },
+    {
+      "epoch": 1.060827718960539,
+      "grad_norm": 1.4337151050567627,
+      "learning_rate": 9.054009800901906e-05,
+      "loss": 0.8711,
+      "step": 27555
+    },
+    {
+      "epoch": 1.0610202117420597,
+      "grad_norm": 1.3560181856155396,
+      "learning_rate": 9.050999444694108e-05,
+      "loss": 0.8178,
+      "step": 27560
+    },
+    {
+      "epoch": 1.0612127045235804,
+      "grad_norm": 1.4069291353225708,
+      "learning_rate": 9.047989175266123e-05,
+      "loss": 0.7403,
+      "step": 27565
+    },
+    {
+      "epoch": 1.061405197305101,
+      "grad_norm": 2.0509390830993652,
+      "learning_rate": 9.044978992893219e-05,
+      "loss": 1.0252,
+      "step": 27570
+    },
+    {
+      "epoch": 1.0615976900866217,
+      "grad_norm": 0.6099883913993835,
+      "learning_rate": 9.041968897850664e-05,
+      "loss": 0.8478,
+      "step": 27575
+    },
+    {
+      "epoch": 1.0617901828681424,
+      "grad_norm": 1.1845272779464722,
+      "learning_rate": 9.038958890413705e-05,
+      "loss": 0.9869,
+      "step": 27580
+    },
+    {
+      "epoch": 1.061982675649663,
+      "grad_norm": 1.2324453592300415,
+      "learning_rate": 9.035948970857589e-05,
+      "loss": 0.8353,
+      "step": 27585
+    },
+    {
+      "epoch": 1.0621751684311838,
+      "grad_norm": 2.6329379081726074,
+      "learning_rate": 9.032939139457556e-05,
+      "loss": 0.8846,
+      "step": 27590
+    },
+    {
+      "epoch": 1.0623676612127044,
+      "grad_norm": 1.4631036520004272,
+      "learning_rate": 9.029929396488826e-05,
+      "loss": 0.8486,
+      "step": 27595
+    },
+    {
+      "epoch": 1.0625601539942253,
+      "grad_norm": 2.0869596004486084,
+      "learning_rate": 9.026919742226633e-05,
+      "loss": 0.9341,
+      "step": 27600
+    },
+    {
+      "epoch": 1.062752646775746,
+      "grad_norm": 1.2778350114822388,
+      "learning_rate": 9.023910176946182e-05,
+      "loss": 0.7416,
+      "step": 27605
+    },
+    {
+      "epoch": 1.0629451395572667,
+      "grad_norm": 1.2002792358398438,
+      "learning_rate": 9.020900700922675e-05,
+      "loss": 0.8239,
+      "step": 27610
+    },
+    {
+      "epoch": 1.0631376323387873,
+      "grad_norm": 1.811063528060913,
+      "learning_rate": 9.017891314431318e-05,
+      "loss": 0.961,
+      "step": 27615
+    },
+    {
+      "epoch": 1.063330125120308,
+      "grad_norm": 1.2313002347946167,
+      "learning_rate": 9.014882017747294e-05,
+      "loss": 0.987,
+      "step": 27620
+    },
+    {
+      "epoch": 1.0635226179018287,
+      "grad_norm": 1.319166898727417,
+      "learning_rate": 9.01187281114578e-05,
+      "loss": 0.8669,
+      "step": 27625
+    },
+    {
+      "epoch": 1.0637151106833493,
+      "grad_norm": 1.418129801750183,
+      "learning_rate": 9.008863694901955e-05,
+      "loss": 0.8651,
+      "step": 27630
+    },
+    {
+      "epoch": 1.06390760346487,
+      "grad_norm": 1.349688172340393,
+      "learning_rate": 9.00585466929098e-05,
+      "loss": 0.7572,
+      "step": 27635
+    },
+    {
+      "epoch": 1.0641000962463907,
+      "grad_norm": 1.4476646184921265,
+      "learning_rate": 9.002845734588005e-05,
+      "loss": 0.8944,
+      "step": 27640
+    },
+    {
+      "epoch": 1.0642925890279114,
+      "grad_norm": 2.580551862716675,
+      "learning_rate": 8.999836891068187e-05,
+      "loss": 0.6778,
+      "step": 27645
+    },
+    {
+      "epoch": 1.0644850818094322,
+      "grad_norm": 1.5499801635742188,
+      "learning_rate": 8.99682813900666e-05,
+      "loss": 0.7813,
+      "step": 27650
+    },
+    {
+      "epoch": 1.064677574590953,
+      "grad_norm": 3.0829336643218994,
+      "learning_rate": 8.993819478678549e-05,
+      "loss": 0.9905,
+      "step": 27655
+    },
+    {
+      "epoch": 1.0648700673724736,
+      "grad_norm": 1.489709734916687,
+      "learning_rate": 8.990810910358987e-05,
+      "loss": 0.9231,
+      "step": 27660
+    },
+    {
+      "epoch": 1.0650625601539943,
+      "grad_norm": 1.199367642402649,
+      "learning_rate": 8.987802434323081e-05,
+      "loss": 0.971,
+      "step": 27665
+    },
+    {
+      "epoch": 1.065255052935515,
+      "grad_norm": 1.2478957176208496,
+      "learning_rate": 8.984794050845933e-05,
+      "loss": 0.9586,
+      "step": 27670
+    },
+    {
+      "epoch": 1.0654475457170356,
+      "grad_norm": 1.5783430337905884,
+      "learning_rate": 8.981785760202647e-05,
+      "loss": 0.7706,
+      "step": 27675
+    },
+    {
+      "epoch": 1.0656400384985563,
+      "grad_norm": 1.6337023973464966,
+      "learning_rate": 8.97877756266831e-05,
+      "loss": 0.7538,
+      "step": 27680
+    },
+    {
+      "epoch": 1.065832531280077,
+      "grad_norm": 1.195513367652893,
+      "learning_rate": 8.975769458517997e-05,
+      "loss": 0.9013,
+      "step": 27685
+    },
+    {
+      "epoch": 1.0660250240615976,
+      "grad_norm": 1.2982577085494995,
+      "learning_rate": 8.972761448026785e-05,
+      "loss": 0.9172,
+      "step": 27690
+    },
+    {
+      "epoch": 1.0662175168431185,
+      "grad_norm": 1.3517059087753296,
+      "learning_rate": 8.96975353146973e-05,
+      "loss": 0.8085,
+      "step": 27695
+    },
+    {
+      "epoch": 1.0664100096246392,
+      "grad_norm": 0.9628300666809082,
+      "learning_rate": 8.966745709121893e-05,
+      "loss": 0.9518,
+      "step": 27700
+    },
+    {
+      "epoch": 1.0666025024061598,
+      "grad_norm": 1.872689127922058,
+      "learning_rate": 8.963737981258315e-05,
+      "loss": 0.8118,
+      "step": 27705
+    },
+    {
+      "epoch": 1.0667949951876805,
+      "grad_norm": 1.307530403137207,
+      "learning_rate": 8.960730348154037e-05,
+      "loss": 0.7289,
+      "step": 27710
+    },
+    {
+      "epoch": 1.0669874879692012,
+      "grad_norm": 1.127726435661316,
+      "learning_rate": 8.957722810084077e-05,
+      "loss": 0.8415,
+      "step": 27715
+    },
+    {
+      "epoch": 1.0671799807507218,
+      "grad_norm": 1.2474530935287476,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 0.9341,
+      "step": 27720
+    },
+    {
+      "epoch": 1.0673724735322425,
+      "grad_norm": 1.6957672834396362,
+      "learning_rate": 8.951708020147212e-05,
+      "loss": 0.9228,
+      "step": 27725
+    },
+    {
+      "epoch": 1.0675649663137632,
+      "grad_norm": 2.220019817352295,
+      "learning_rate": 8.94870076883031e-05,
+      "loss": 0.9219,
+      "step": 27730
+    },
+    {
+      "epoch": 1.0677574590952839,
+      "grad_norm": 0.9045952558517456,
+      "learning_rate": 8.945693613647763e-05,
+      "loss": 0.6171,
+      "step": 27735
+    },
+    {
+      "epoch": 1.0679499518768045,
+      "grad_norm": 1.5750988721847534,
+      "learning_rate": 8.94268655487455e-05,
+      "loss": 0.9442,
+      "step": 27740
+    },
+    {
+      "epoch": 1.0681424446583252,
+      "grad_norm": 1.328931450843811,
+      "learning_rate": 8.939679592785646e-05,
+      "loss": 0.8479,
+      "step": 27745
+    },
+    {
+      "epoch": 1.068334937439846,
+      "grad_norm": 1.809605598449707,
+      "learning_rate": 8.936672727656021e-05,
+      "loss": 0.8478,
+      "step": 27750
+    },
+    {
+      "epoch": 1.0685274302213668,
+      "grad_norm": 1.630718469619751,
+      "learning_rate": 8.933665959760632e-05,
+      "loss": 0.7898,
+      "step": 27755
+    },
+    {
+      "epoch": 1.0687199230028874,
+      "grad_norm": 1.7636911869049072,
+      "learning_rate": 8.930659289374422e-05,
+      "loss": 0.8724,
+      "step": 27760
+    },
+    {
+      "epoch": 1.068912415784408,
+      "grad_norm": 1.659645915031433,
+      "learning_rate": 8.927652716772342e-05,
+      "loss": 1.0174,
+      "step": 27765
+    },
+    {
+      "epoch": 1.0691049085659288,
+      "grad_norm": 2.0495598316192627,
+      "learning_rate": 8.924646242229317e-05,
+      "loss": 0.8524,
+      "step": 27770
+    },
+    {
+      "epoch": 1.0692974013474494,
+      "grad_norm": 0.9576983451843262,
+      "learning_rate": 8.921639866020264e-05,
+      "loss": 0.8567,
+      "step": 27775
+    },
+    {
+      "epoch": 1.06948989412897,
+      "grad_norm": 1.3908922672271729,
+      "learning_rate": 8.918633588420106e-05,
+      "loss": 0.9384,
+      "step": 27780
+    },
+    {
+      "epoch": 1.0696823869104908,
+      "grad_norm": 1.0620615482330322,
+      "learning_rate": 8.915627409703745e-05,
+      "loss": 0.9836,
+      "step": 27785
+    },
+    {
+      "epoch": 1.0698748796920114,
+      "grad_norm": 1.0709110498428345,
+      "learning_rate": 8.912621330146071e-05,
+      "loss": 0.9535,
+      "step": 27790
+    },
+    {
+      "epoch": 1.0700673724735323,
+      "grad_norm": 1.297035813331604,
+      "learning_rate": 8.909615350021973e-05,
+      "loss": 0.8007,
+      "step": 27795
+    },
+    {
+      "epoch": 1.070259865255053,
+      "grad_norm": 0.9443962574005127,
+      "learning_rate": 8.906609469606329e-05,
+      "loss": 0.9376,
+      "step": 27800
+    },
+    {
+      "epoch": 1.0704523580365737,
+      "grad_norm": 2.0467529296875,
+      "learning_rate": 8.903603689174006e-05,
+      "loss": 0.9165,
+      "step": 27805
+    },
+    {
+      "epoch": 1.0706448508180944,
+      "grad_norm": 1.3403770923614502,
+      "learning_rate": 8.900598008999864e-05,
+      "loss": 0.8009,
+      "step": 27810
+    },
+    {
+      "epoch": 1.070837343599615,
+      "grad_norm": 1.8648827075958252,
+      "learning_rate": 8.897592429358746e-05,
+      "loss": 0.8834,
+      "step": 27815
+    },
+    {
+      "epoch": 1.0710298363811357,
+      "grad_norm": 1.1583093404769897,
+      "learning_rate": 8.894586950525502e-05,
+      "loss": 0.8895,
+      "step": 27820
+    },
+    {
+      "epoch": 1.0712223291626564,
+      "grad_norm": 1.1083035469055176,
+      "learning_rate": 8.89158157277496e-05,
+      "loss": 0.9152,
+      "step": 27825
+    },
+    {
+      "epoch": 1.071414821944177,
+      "grad_norm": 0.8488582372665405,
+      "learning_rate": 8.888576296381936e-05,
+      "loss": 0.6838,
+      "step": 27830
+    },
+    {
+      "epoch": 1.0716073147256977,
+      "grad_norm": 2.6488595008850098,
+      "learning_rate": 8.885571121621251e-05,
+      "loss": 0.9074,
+      "step": 27835
+    },
+    {
+      "epoch": 1.0717998075072184,
+      "grad_norm": 1.0296202898025513,
+      "learning_rate": 8.882566048767705e-05,
+      "loss": 0.8427,
+      "step": 27840
+    },
+    {
+      "epoch": 1.0719923002887393,
+      "grad_norm": 1.2265279293060303,
+      "learning_rate": 8.879561078096088e-05,
+      "loss": 0.9545,
+      "step": 27845
+    },
+    {
+      "epoch": 1.07218479307026,
+      "grad_norm": 1.3146073818206787,
+      "learning_rate": 8.87655620988119e-05,
+      "loss": 0.8112,
+      "step": 27850
+    },
+    {
+      "epoch": 1.0723772858517806,
+      "grad_norm": 1.1017197370529175,
+      "learning_rate": 8.873551444397785e-05,
+      "loss": 0.739,
+      "step": 27855
+    },
+    {
+      "epoch": 1.0725697786333013,
+      "grad_norm": 1.3772103786468506,
+      "learning_rate": 8.870546781920636e-05,
+      "loss": 0.86,
+      "step": 27860
+    },
+    {
+      "epoch": 1.072762271414822,
+      "grad_norm": 1.4853140115737915,
+      "learning_rate": 8.867542222724504e-05,
+      "loss": 0.9195,
+      "step": 27865
+    },
+    {
+      "epoch": 1.0729547641963426,
+      "grad_norm": 1.3116743564605713,
+      "learning_rate": 8.864537767084135e-05,
+      "loss": 0.8492,
+      "step": 27870
+    },
+    {
+      "epoch": 1.0731472569778633,
+      "grad_norm": 1.78131103515625,
+      "learning_rate": 8.861533415274262e-05,
+      "loss": 0.8794,
+      "step": 27875
+    },
+    {
+      "epoch": 1.073339749759384,
+      "grad_norm": 1.1533085107803345,
+      "learning_rate": 8.858529167569619e-05,
+      "loss": 0.952,
+      "step": 27880
+    },
+    {
+      "epoch": 1.0735322425409046,
+      "grad_norm": 2.657715320587158,
+      "learning_rate": 8.855525024244923e-05,
+      "loss": 0.885,
+      "step": 27885
+    },
+    {
+      "epoch": 1.0737247353224255,
+      "grad_norm": 1.6451362371444702,
+      "learning_rate": 8.85252098557488e-05,
+      "loss": 0.823,
+      "step": 27890
+    },
+    {
+      "epoch": 1.0739172281039462,
+      "grad_norm": 1.7528831958770752,
+      "learning_rate": 8.84951705183419e-05,
+      "loss": 0.7666,
+      "step": 27895
+    },
+    {
+      "epoch": 1.0741097208854669,
+      "grad_norm": 1.086206078529358,
+      "learning_rate": 8.846513223297549e-05,
+      "loss": 0.8837,
+      "step": 27900
+    },
+    {
+      "epoch": 1.0743022136669875,
+      "grad_norm": 1.413942575454712,
+      "learning_rate": 8.843509500239628e-05,
+      "loss": 0.737,
+      "step": 27905
+    },
+    {
+      "epoch": 1.0744947064485082,
+      "grad_norm": 1.2918885946273804,
+      "learning_rate": 8.840505882935106e-05,
+      "loss": 0.7301,
+      "step": 27910
+    },
+    {
+      "epoch": 1.0746871992300289,
+      "grad_norm": 1.7935266494750977,
+      "learning_rate": 8.837502371658639e-05,
+      "loss": 0.9754,
+      "step": 27915
+    },
+    {
+      "epoch": 1.0748796920115495,
+      "grad_norm": 1.8554896116256714,
+      "learning_rate": 8.834498966684876e-05,
+      "loss": 1.0135,
+      "step": 27920
+    },
+    {
+      "epoch": 1.0750721847930702,
+      "grad_norm": 1.3628082275390625,
+      "learning_rate": 8.831495668288467e-05,
+      "loss": 0.836,
+      "step": 27925
+    },
+    {
+      "epoch": 1.0752646775745909,
+      "grad_norm": 1.181231141090393,
+      "learning_rate": 8.828492476744037e-05,
+      "loss": 1.0394,
+      "step": 27930
+    },
+    {
+      "epoch": 1.0754571703561115,
+      "grad_norm": 1.8281370401382446,
+      "learning_rate": 8.825489392326205e-05,
+      "loss": 0.8396,
+      "step": 27935
+    },
+    {
+      "epoch": 1.0756496631376324,
+      "grad_norm": 1.069940209388733,
+      "learning_rate": 8.822486415309592e-05,
+      "loss": 0.9147,
+      "step": 27940
+    },
+    {
+      "epoch": 1.075842155919153,
+      "grad_norm": 1.711498737335205,
+      "learning_rate": 8.819483545968797e-05,
+      "loss": 0.7523,
+      "step": 27945
+    },
+    {
+      "epoch": 1.0760346487006738,
+      "grad_norm": 1.5465821027755737,
+      "learning_rate": 8.816480784578406e-05,
+      "loss": 0.8782,
+      "step": 27950
+    },
+    {
+      "epoch": 1.0762271414821944,
+      "grad_norm": 1.0662082433700562,
+      "learning_rate": 8.813478131413013e-05,
+      "loss": 0.6925,
+      "step": 27955
+    },
+    {
+      "epoch": 1.0764196342637151,
+      "grad_norm": 0.8745738863945007,
+      "learning_rate": 8.810475586747183e-05,
+      "loss": 0.7974,
+      "step": 27960
+    },
+    {
+      "epoch": 1.0766121270452358,
+      "grad_norm": 1.5703550577163696,
+      "learning_rate": 8.807473150855476e-05,
+      "loss": 0.7965,
+      "step": 27965
+    },
+    {
+      "epoch": 1.0768046198267565,
+      "grad_norm": 2.1191911697387695,
+      "learning_rate": 8.804470824012456e-05,
+      "loss": 0.8466,
+      "step": 27970
+    },
+    {
+      "epoch": 1.0769971126082771,
+      "grad_norm": 2.149782419204712,
+      "learning_rate": 8.801468606492659e-05,
+      "loss": 1.009,
+      "step": 27975
+    },
+    {
+      "epoch": 1.0771896053897978,
+      "grad_norm": 1.0900071859359741,
+      "learning_rate": 8.798466498570614e-05,
+      "loss": 0.8092,
+      "step": 27980
+    },
+    {
+      "epoch": 1.0773820981713187,
+      "grad_norm": 1.022477149963379,
+      "learning_rate": 8.795464500520851e-05,
+      "loss": 0.9038,
+      "step": 27985
+    },
+    {
+      "epoch": 1.0775745909528394,
+      "grad_norm": 1.4904948472976685,
+      "learning_rate": 8.792462612617882e-05,
+      "loss": 0.8276,
+      "step": 27990
+    },
+    {
+      "epoch": 1.07776708373436,
+      "grad_norm": 1.8655019998550415,
+      "learning_rate": 8.789460835136203e-05,
+      "loss": 0.9897,
+      "step": 27995
+    },
+    {
+      "epoch": 1.0779595765158807,
+      "grad_norm": 1.3288123607635498,
+      "learning_rate": 8.786459168350316e-05,
+      "loss": 0.7634,
+      "step": 28000
+    },
+    {
+      "epoch": 1.0781520692974014,
+      "grad_norm": 1.4979158639907837,
+      "learning_rate": 8.783457612534699e-05,
+      "loss": 1.0058,
+      "step": 28005
+    },
+    {
+      "epoch": 1.078344562078922,
+      "grad_norm": 0.987000048160553,
+      "learning_rate": 8.780456167963821e-05,
+      "loss": 0.7555,
+      "step": 28010
+    },
+    {
+      "epoch": 1.0785370548604427,
+      "grad_norm": 1.124505639076233,
+      "learning_rate": 8.77745483491215e-05,
+      "loss": 0.8965,
+      "step": 28015
+    },
+    {
+      "epoch": 1.0787295476419634,
+      "grad_norm": 1.8975436687469482,
+      "learning_rate": 8.774453613654136e-05,
+      "loss": 0.7971,
+      "step": 28020
+    },
+    {
+      "epoch": 1.078922040423484,
+      "grad_norm": 2.273435115814209,
+      "learning_rate": 8.77145250446422e-05,
+      "loss": 0.9065,
+      "step": 28025
+    },
+    {
+      "epoch": 1.0791145332050047,
+      "grad_norm": 2.141667366027832,
+      "learning_rate": 8.768451507616834e-05,
+      "loss": 0.937,
+      "step": 28030
+    },
+    {
+      "epoch": 1.0793070259865254,
+      "grad_norm": 1.4163758754730225,
+      "learning_rate": 8.7654506233864e-05,
+      "loss": 0.8959,
+      "step": 28035
+    },
+    {
+      "epoch": 1.0794995187680463,
+      "grad_norm": 2.3507914543151855,
+      "learning_rate": 8.762449852047326e-05,
+      "loss": 0.9258,
+      "step": 28040
+    },
+    {
+      "epoch": 1.079692011549567,
+      "grad_norm": 1.0547107458114624,
+      "learning_rate": 8.759449193874018e-05,
+      "loss": 0.9966,
+      "step": 28045
+    },
+    {
+      "epoch": 1.0798845043310876,
+      "grad_norm": 1.1531494855880737,
+      "learning_rate": 8.756448649140864e-05,
+      "loss": 0.8557,
+      "step": 28050
+    },
+    {
+      "epoch": 1.0800769971126083,
+      "grad_norm": 0.93568354845047,
+      "learning_rate": 8.753448218122239e-05,
+      "loss": 0.9067,
+      "step": 28055
+    },
+    {
+      "epoch": 1.080269489894129,
+      "grad_norm": 1.3804124593734741,
+      "learning_rate": 8.75044790109252e-05,
+      "loss": 0.9757,
+      "step": 28060
+    },
+    {
+      "epoch": 1.0804619826756496,
+      "grad_norm": 1.065398097038269,
+      "learning_rate": 8.747447698326058e-05,
+      "loss": 0.8644,
+      "step": 28065
+    },
+    {
+      "epoch": 1.0806544754571703,
+      "grad_norm": 2.9915854930877686,
+      "learning_rate": 8.744447610097213e-05,
+      "loss": 0.8151,
+      "step": 28070
+    },
+    {
+      "epoch": 1.080846968238691,
+      "grad_norm": 1.506426453590393,
+      "learning_rate": 8.741447636680317e-05,
+      "loss": 1.007,
+      "step": 28075
+    },
+    {
+      "epoch": 1.0810394610202116,
+      "grad_norm": 0.8236234784126282,
+      "learning_rate": 8.738447778349692e-05,
+      "loss": 0.8457,
+      "step": 28080
+    },
+    {
+      "epoch": 1.0812319538017325,
+      "grad_norm": 1.224062442779541,
+      "learning_rate": 8.735448035379665e-05,
+      "loss": 0.8969,
+      "step": 28085
+    },
+    {
+      "epoch": 1.0814244465832532,
+      "grad_norm": 1.5686476230621338,
+      "learning_rate": 8.732448408044539e-05,
+      "loss": 0.8932,
+      "step": 28090
+    },
+    {
+      "epoch": 1.0816169393647739,
+      "grad_norm": 1.104203462600708,
+      "learning_rate": 8.729448896618603e-05,
+      "loss": 0.8526,
+      "step": 28095
+    },
+    {
+      "epoch": 1.0818094321462945,
+      "grad_norm": 0.723469614982605,
+      "learning_rate": 8.726449501376153e-05,
+      "loss": 0.761,
+      "step": 28100
+    },
+    {
+      "epoch": 1.0820019249278152,
+      "grad_norm": 1.4393385648727417,
+      "learning_rate": 8.723450222591462e-05,
+      "loss": 0.876,
+      "step": 28105
+    },
+    {
+      "epoch": 1.0821944177093359,
+      "grad_norm": 1.0453604459762573,
+      "learning_rate": 8.720451060538789e-05,
+      "loss": 0.7178,
+      "step": 28110
+    },
+    {
+      "epoch": 1.0823869104908566,
+      "grad_norm": 1.7050689458847046,
+      "learning_rate": 8.717452015492388e-05,
+      "loss": 0.8199,
+      "step": 28115
+    },
+    {
+      "epoch": 1.0825794032723772,
+      "grad_norm": 1.400423526763916,
+      "learning_rate": 8.71445308772651e-05,
+      "loss": 0.8114,
+      "step": 28120
+    },
+    {
+      "epoch": 1.082771896053898,
+      "grad_norm": 1.93320631980896,
+      "learning_rate": 8.711454277515377e-05,
+      "loss": 0.7295,
+      "step": 28125
+    },
+    {
+      "epoch": 1.0829643888354186,
+      "grad_norm": 1.5179084539413452,
+      "learning_rate": 8.708455585133216e-05,
+      "loss": 1.083,
+      "step": 28130
+    },
+    {
+      "epoch": 1.0831568816169395,
+      "grad_norm": 2.2808666229248047,
+      "learning_rate": 8.705457010854235e-05,
+      "loss": 0.9567,
+      "step": 28135
+    },
+    {
+      "epoch": 1.0833493743984601,
+      "grad_norm": 1.0833265781402588,
+      "learning_rate": 8.702458554952635e-05,
+      "loss": 0.8316,
+      "step": 28140
+    },
+    {
+      "epoch": 1.0835418671799808,
+      "grad_norm": 1.4321517944335938,
+      "learning_rate": 8.699460217702606e-05,
+      "loss": 0.7606,
+      "step": 28145
+    },
+    {
+      "epoch": 1.0837343599615015,
+      "grad_norm": 0.9563835263252258,
+      "learning_rate": 8.696461999378324e-05,
+      "loss": 0.8625,
+      "step": 28150
+    },
+    {
+      "epoch": 1.0839268527430221,
+      "grad_norm": 1.0265910625457764,
+      "learning_rate": 8.693463900253954e-05,
+      "loss": 0.842,
+      "step": 28155
+    },
+    {
+      "epoch": 1.0841193455245428,
+      "grad_norm": 1.2975974082946777,
+      "learning_rate": 8.690465920603659e-05,
+      "loss": 0.9416,
+      "step": 28160
+    },
+    {
+      "epoch": 1.0843118383060635,
+      "grad_norm": 0.9792843461036682,
+      "learning_rate": 8.687468060701582e-05,
+      "loss": 0.786,
+      "step": 28165
+    },
+    {
+      "epoch": 1.0845043310875842,
+      "grad_norm": 1.5455747842788696,
+      "learning_rate": 8.684470320821851e-05,
+      "loss": 0.9023,
+      "step": 28170
+    },
+    {
+      "epoch": 1.0846968238691048,
+      "grad_norm": 1.3106191158294678,
+      "learning_rate": 8.681472701238599e-05,
+      "loss": 0.793,
+      "step": 28175
+    },
+    {
+      "epoch": 1.0848893166506257,
+      "grad_norm": 1.0658930540084839,
+      "learning_rate": 8.678475202225933e-05,
+      "loss": 0.8117,
+      "step": 28180
+    },
+    {
+      "epoch": 1.0850818094321464,
+      "grad_norm": 1.1267075538635254,
+      "learning_rate": 8.675477824057953e-05,
+      "loss": 0.9166,
+      "step": 28185
+    },
+    {
+      "epoch": 1.085274302213667,
+      "grad_norm": 2.638794183731079,
+      "learning_rate": 8.672480567008753e-05,
+      "loss": 0.9889,
+      "step": 28190
+    },
+    {
+      "epoch": 1.0854667949951877,
+      "grad_norm": 1.0422507524490356,
+      "learning_rate": 8.669483431352414e-05,
+      "loss": 0.8901,
+      "step": 28195
+    },
+    {
+      "epoch": 1.0856592877767084,
+      "grad_norm": 1.2292778491973877,
+      "learning_rate": 8.666486417362996e-05,
+      "loss": 0.917,
+      "step": 28200
+    },
+    {
+      "epoch": 1.085851780558229,
+      "grad_norm": 1.1497111320495605,
+      "learning_rate": 8.663489525314568e-05,
+      "loss": 0.7443,
+      "step": 28205
+    },
+    {
+      "epoch": 1.0860442733397497,
+      "grad_norm": 1.712801456451416,
+      "learning_rate": 8.660492755481167e-05,
+      "loss": 1.0358,
+      "step": 28210
+    },
+    {
+      "epoch": 1.0862367661212704,
+      "grad_norm": 1.057096004486084,
+      "learning_rate": 8.657496108136826e-05,
+      "loss": 0.829,
+      "step": 28215
+    },
+    {
+      "epoch": 1.086429258902791,
+      "grad_norm": 0.9071182012557983,
+      "learning_rate": 8.654499583555576e-05,
+      "loss": 0.8827,
+      "step": 28220
+    },
+    {
+      "epoch": 1.0866217516843117,
+      "grad_norm": 1.586111068725586,
+      "learning_rate": 8.651503182011427e-05,
+      "loss": 0.8244,
+      "step": 28225
+    },
+    {
+      "epoch": 1.0868142444658326,
+      "grad_norm": 0.845116913318634,
+      "learning_rate": 8.648506903778377e-05,
+      "loss": 0.9803,
+      "step": 28230
+    },
+    {
+      "epoch": 1.0870067372473533,
+      "grad_norm": 1.4213730096817017,
+      "learning_rate": 8.64551074913042e-05,
+      "loss": 0.7278,
+      "step": 28235
+    },
+    {
+      "epoch": 1.087199230028874,
+      "grad_norm": 1.0875608921051025,
+      "learning_rate": 8.64251471834153e-05,
+      "loss": 0.8259,
+      "step": 28240
+    },
+    {
+      "epoch": 1.0873917228103946,
+      "grad_norm": 0.9991816878318787,
+      "learning_rate": 8.639518811685676e-05,
+      "loss": 0.847,
+      "step": 28245
+    },
+    {
+      "epoch": 1.0875842155919153,
+      "grad_norm": 1.6519125699996948,
+      "learning_rate": 8.636523029436819e-05,
+      "loss": 0.8359,
+      "step": 28250
+    },
+    {
+      "epoch": 1.087776708373436,
+      "grad_norm": 1.6119956970214844,
+      "learning_rate": 8.633527371868896e-05,
+      "loss": 0.9016,
+      "step": 28255
+    },
+    {
+      "epoch": 1.0879692011549567,
+      "grad_norm": 0.7314580678939819,
+      "learning_rate": 8.630531839255839e-05,
+      "loss": 0.6791,
+      "step": 28260
+    },
+    {
+      "epoch": 1.0881616939364773,
+      "grad_norm": 0.9393051266670227,
+      "learning_rate": 8.627536431871578e-05,
+      "loss": 0.8801,
+      "step": 28265
+    },
+    {
+      "epoch": 1.088354186717998,
+      "grad_norm": 2.3234903812408447,
+      "learning_rate": 8.624541149990018e-05,
+      "loss": 0.8656,
+      "step": 28270
+    },
+    {
+      "epoch": 1.0885466794995187,
+      "grad_norm": 0.9072073101997375,
+      "learning_rate": 8.621545993885054e-05,
+      "loss": 0.8702,
+      "step": 28275
+    },
+    {
+      "epoch": 1.0887391722810396,
+      "grad_norm": 1.1024999618530273,
+      "learning_rate": 8.618550963830583e-05,
+      "loss": 0.9067,
+      "step": 28280
+    },
+    {
+      "epoch": 1.0889316650625602,
+      "grad_norm": 1.7317997217178345,
+      "learning_rate": 8.615556060100474e-05,
+      "loss": 0.975,
+      "step": 28285
+    },
+    {
+      "epoch": 1.089124157844081,
+      "grad_norm": 1.9279747009277344,
+      "learning_rate": 8.612561282968588e-05,
+      "loss": 0.9101,
+      "step": 28290
+    },
+    {
+      "epoch": 1.0893166506256016,
+      "grad_norm": 1.848134994506836,
+      "learning_rate": 8.609566632708786e-05,
+      "loss": 0.7642,
+      "step": 28295
+    },
+    {
+      "epoch": 1.0895091434071222,
+      "grad_norm": 1.3254327774047852,
+      "learning_rate": 8.606572109594906e-05,
+      "loss": 1.0743,
+      "step": 28300
+    },
+    {
+      "epoch": 1.089701636188643,
+      "grad_norm": 1.483266830444336,
+      "learning_rate": 8.60357771390077e-05,
+      "loss": 0.7679,
+      "step": 28305
+    },
+    {
+      "epoch": 1.0898941289701636,
+      "grad_norm": 1.0175541639328003,
+      "learning_rate": 8.600583445900206e-05,
+      "loss": 0.8647,
+      "step": 28310
+    },
+    {
+      "epoch": 1.0900866217516842,
+      "grad_norm": 1.3284434080123901,
+      "learning_rate": 8.597589305867012e-05,
+      "loss": 0.8266,
+      "step": 28315
+    },
+    {
+      "epoch": 1.090279114533205,
+      "grad_norm": 1.2873848676681519,
+      "learning_rate": 8.59459529407499e-05,
+      "loss": 0.9412,
+      "step": 28320
+    },
+    {
+      "epoch": 1.0904716073147256,
+      "grad_norm": 1.6508309841156006,
+      "learning_rate": 8.591601410797918e-05,
+      "loss": 0.8731,
+      "step": 28325
+    },
+    {
+      "epoch": 1.0906641000962465,
+      "grad_norm": 1.7631590366363525,
+      "learning_rate": 8.588607656309565e-05,
+      "loss": 0.8443,
+      "step": 28330
+    },
+    {
+      "epoch": 1.0908565928777672,
+      "grad_norm": 1.6212012767791748,
+      "learning_rate": 8.585614030883695e-05,
+      "loss": 0.9399,
+      "step": 28335
+    },
+    {
+      "epoch": 1.0910490856592878,
+      "grad_norm": 1.2359917163848877,
+      "learning_rate": 8.58262053479405e-05,
+      "loss": 0.8504,
+      "step": 28340
+    },
+    {
+      "epoch": 1.0912415784408085,
+      "grad_norm": 1.0911660194396973,
+      "learning_rate": 8.579627168314368e-05,
+      "loss": 0.7694,
+      "step": 28345
+    },
+    {
+      "epoch": 1.0914340712223292,
+      "grad_norm": 1.2516300678253174,
+      "learning_rate": 8.576633931718374e-05,
+      "loss": 0.7347,
+      "step": 28350
+    },
+    {
+      "epoch": 1.0916265640038498,
+      "grad_norm": 1.3555577993392944,
+      "learning_rate": 8.573640825279779e-05,
+      "loss": 0.9815,
+      "step": 28355
+    },
+    {
+      "epoch": 1.0918190567853705,
+      "grad_norm": 1.2513625621795654,
+      "learning_rate": 8.570647849272276e-05,
+      "loss": 0.856,
+      "step": 28360
+    },
+    {
+      "epoch": 1.0920115495668912,
+      "grad_norm": 1.3024554252624512,
+      "learning_rate": 8.567655003969563e-05,
+      "loss": 0.8517,
+      "step": 28365
+    },
+    {
+      "epoch": 1.0922040423484118,
+      "grad_norm": 0.9453557729721069,
+      "learning_rate": 8.564662289645313e-05,
+      "loss": 0.9392,
+      "step": 28370
+    },
+    {
+      "epoch": 1.0923965351299327,
+      "grad_norm": 1.9402847290039062,
+      "learning_rate": 8.561669706573184e-05,
+      "loss": 0.9265,
+      "step": 28375
+    },
+    {
+      "epoch": 1.0925890279114534,
+      "grad_norm": 0.8357999324798584,
+      "learning_rate": 8.558677255026835e-05,
+      "loss": 0.8744,
+      "step": 28380
+    },
+    {
+      "epoch": 1.092781520692974,
+      "grad_norm": 1.287455439567566,
+      "learning_rate": 8.555684935279904e-05,
+      "loss": 0.7733,
+      "step": 28385
+    },
+    {
+      "epoch": 1.0929740134744947,
+      "grad_norm": 1.2665926218032837,
+      "learning_rate": 8.552692747606014e-05,
+      "loss": 0.8229,
+      "step": 28390
+    },
+    {
+      "epoch": 1.0931665062560154,
+      "grad_norm": 1.5380029678344727,
+      "learning_rate": 8.549700692278788e-05,
+      "loss": 0.9208,
+      "step": 28395
+    },
+    {
+      "epoch": 1.093358999037536,
+      "grad_norm": 1.7858625650405884,
+      "learning_rate": 8.546708769571827e-05,
+      "loss": 0.8756,
+      "step": 28400
+    },
+    {
+      "epoch": 1.0935514918190568,
+      "grad_norm": 2.022096633911133,
+      "learning_rate": 8.543716979758717e-05,
+      "loss": 0.828,
+      "step": 28405
+    },
+    {
+      "epoch": 1.0937439846005774,
+      "grad_norm": 1.1538034677505493,
+      "learning_rate": 8.540725323113047e-05,
+      "loss": 0.8699,
+      "step": 28410
+    },
+    {
+      "epoch": 1.093936477382098,
+      "grad_norm": 2.0683727264404297,
+      "learning_rate": 8.537733799908379e-05,
+      "loss": 0.9554,
+      "step": 28415
+    },
+    {
+      "epoch": 1.0941289701636188,
+      "grad_norm": 1.3001796007156372,
+      "learning_rate": 8.534742410418265e-05,
+      "loss": 0.9327,
+      "step": 28420
+    },
+    {
+      "epoch": 1.0943214629451397,
+      "grad_norm": 0.7399016618728638,
+      "learning_rate": 8.531751154916254e-05,
+      "loss": 0.904,
+      "step": 28425
+    },
+    {
+      "epoch": 1.0945139557266603,
+      "grad_norm": 1.1065526008605957,
+      "learning_rate": 8.528760033675875e-05,
+      "loss": 0.7723,
+      "step": 28430
+    },
+    {
+      "epoch": 1.094706448508181,
+      "grad_norm": 1.9489480257034302,
+      "learning_rate": 8.525769046970642e-05,
+      "loss": 0.8233,
+      "step": 28435
+    },
+    {
+      "epoch": 1.0948989412897017,
+      "grad_norm": 2.0606977939605713,
+      "learning_rate": 8.522778195074064e-05,
+      "loss": 0.9149,
+      "step": 28440
+    },
+    {
+      "epoch": 1.0950914340712223,
+      "grad_norm": 1.784341812133789,
+      "learning_rate": 8.519787478259637e-05,
+      "loss": 0.8731,
+      "step": 28445
+    },
+    {
+      "epoch": 1.095283926852743,
+      "grad_norm": 0.9500823020935059,
+      "learning_rate": 8.516796896800836e-05,
+      "loss": 0.8144,
+      "step": 28450
+    },
+    {
+      "epoch": 1.0954764196342637,
+      "grad_norm": 1.5665167570114136,
+      "learning_rate": 8.513806450971137e-05,
+      "loss": 0.7198,
+      "step": 28455
+    },
+    {
+      "epoch": 1.0956689124157843,
+      "grad_norm": 0.8766529560089111,
+      "learning_rate": 8.51081614104399e-05,
+      "loss": 0.8412,
+      "step": 28460
+    },
+    {
+      "epoch": 1.095861405197305,
+      "grad_norm": 1.1104719638824463,
+      "learning_rate": 8.507825967292842e-05,
+      "loss": 0.7384,
+      "step": 28465
+    },
+    {
+      "epoch": 1.096053897978826,
+      "grad_norm": 1.2092657089233398,
+      "learning_rate": 8.504835929991128e-05,
+      "loss": 0.9372,
+      "step": 28470
+    },
+    {
+      "epoch": 1.0962463907603466,
+      "grad_norm": 1.244201898574829,
+      "learning_rate": 8.501846029412262e-05,
+      "loss": 1.0071,
+      "step": 28475
+    },
+    {
+      "epoch": 1.0964388835418672,
+      "grad_norm": 1.6996936798095703,
+      "learning_rate": 8.49885626582965e-05,
+      "loss": 0.8313,
+      "step": 28480
+    },
+    {
+      "epoch": 1.096631376323388,
+      "grad_norm": 1.2669320106506348,
+      "learning_rate": 8.49586663951669e-05,
+      "loss": 0.8763,
+      "step": 28485
+    },
+    {
+      "epoch": 1.0968238691049086,
+      "grad_norm": 1.1832565069198608,
+      "learning_rate": 8.492877150746764e-05,
+      "loss": 0.9931,
+      "step": 28490
+    },
+    {
+      "epoch": 1.0970163618864293,
+      "grad_norm": 0.8875114321708679,
+      "learning_rate": 8.489887799793232e-05,
+      "loss": 0.9127,
+      "step": 28495
+    },
+    {
+      "epoch": 1.09720885466795,
+      "grad_norm": 0.9826710224151611,
+      "learning_rate": 8.486898586929464e-05,
+      "loss": 0.8378,
+      "step": 28500
+    },
+    {
+      "epoch": 1.0974013474494706,
+      "grad_norm": 1.0573769807815552,
+      "learning_rate": 8.483909512428796e-05,
+      "loss": 0.8276,
+      "step": 28505
+    },
+    {
+      "epoch": 1.0975938402309913,
+      "grad_norm": 0.9481573104858398,
+      "learning_rate": 8.480920576564555e-05,
+      "loss": 0.8358,
+      "step": 28510
+    },
+    {
+      "epoch": 1.097786333012512,
+      "grad_norm": 1.1130627393722534,
+      "learning_rate": 8.477931779610068e-05,
+      "loss": 0.9346,
+      "step": 28515
+    },
+    {
+      "epoch": 1.0979788257940326,
+      "grad_norm": 2.038252115249634,
+      "learning_rate": 8.474943121838638e-05,
+      "loss": 0.787,
+      "step": 28520
+    },
+    {
+      "epoch": 1.0981713185755535,
+      "grad_norm": 1.3713279962539673,
+      "learning_rate": 8.471954603523553e-05,
+      "loss": 0.805,
+      "step": 28525
+    },
+    {
+      "epoch": 1.0983638113570742,
+      "grad_norm": 1.3389123678207397,
+      "learning_rate": 8.468966224938101e-05,
+      "loss": 0.8322,
+      "step": 28530
+    },
+    {
+      "epoch": 1.0985563041385948,
+      "grad_norm": 2.199444532394409,
+      "learning_rate": 8.465977986355545e-05,
+      "loss": 1.0354,
+      "step": 28535
+    },
+    {
+      "epoch": 1.0987487969201155,
+      "grad_norm": 1.5610674619674683,
+      "learning_rate": 8.462989888049137e-05,
+      "loss": 0.9338,
+      "step": 28540
+    },
+    {
+      "epoch": 1.0989412897016362,
+      "grad_norm": 1.1883718967437744,
+      "learning_rate": 8.460001930292125e-05,
+      "loss": 0.7661,
+      "step": 28545
+    },
+    {
+      "epoch": 1.0991337824831569,
+      "grad_norm": 1.4481984376907349,
+      "learning_rate": 8.457014113357736e-05,
+      "loss": 0.8787,
+      "step": 28550
+    },
+    {
+      "epoch": 1.0993262752646775,
+      "grad_norm": 1.3184305429458618,
+      "learning_rate": 8.454026437519182e-05,
+      "loss": 0.789,
+      "step": 28555
+    },
+    {
+      "epoch": 1.0995187680461982,
+      "grad_norm": 1.6092246770858765,
+      "learning_rate": 8.45103890304967e-05,
+      "loss": 0.9798,
+      "step": 28560
+    },
+    {
+      "epoch": 1.0997112608277189,
+      "grad_norm": 1.512444019317627,
+      "learning_rate": 8.448051510222391e-05,
+      "loss": 0.7606,
+      "step": 28565
+    },
+    {
+      "epoch": 1.0999037536092398,
+      "grad_norm": 0.880029022693634,
+      "learning_rate": 8.445064259310518e-05,
+      "loss": 0.7627,
+      "step": 28570
+    },
+    {
+      "epoch": 1.1000962463907604,
+      "grad_norm": 1.796959638595581,
+      "learning_rate": 8.442077150587222e-05,
+      "loss": 0.8827,
+      "step": 28575
+    },
+    {
+      "epoch": 1.100288739172281,
+      "grad_norm": 1.1779733896255493,
+      "learning_rate": 8.439090184325644e-05,
+      "loss": 0.9346,
+      "step": 28580
+    },
+    {
+      "epoch": 1.1004812319538018,
+      "grad_norm": 0.7998082041740417,
+      "learning_rate": 8.436103360798934e-05,
+      "loss": 0.799,
+      "step": 28585
+    },
+    {
+      "epoch": 1.1006737247353224,
+      "grad_norm": 1.9549007415771484,
+      "learning_rate": 8.433116680280212e-05,
+      "loss": 1.0649,
+      "step": 28590
+    },
+    {
+      "epoch": 1.100866217516843,
+      "grad_norm": 1.2117373943328857,
+      "learning_rate": 8.430130143042584e-05,
+      "loss": 0.8325,
+      "step": 28595
+    },
+    {
+      "epoch": 1.1010587102983638,
+      "grad_norm": 1.4036140441894531,
+      "learning_rate": 8.427143749359161e-05,
+      "loss": 0.7584,
+      "step": 28600
+    },
+    {
+      "epoch": 1.1012512030798844,
+      "grad_norm": 0.979111909866333,
+      "learning_rate": 8.424157499503022e-05,
+      "loss": 0.8124,
+      "step": 28605
+    },
+    {
+      "epoch": 1.1014436958614051,
+      "grad_norm": 1.6154838800430298,
+      "learning_rate": 8.421171393747237e-05,
+      "loss": 1.0092,
+      "step": 28610
+    },
+    {
+      "epoch": 1.1016361886429258,
+      "grad_norm": 1.6151224374771118,
+      "learning_rate": 8.418185432364873e-05,
+      "loss": 0.8584,
+      "step": 28615
+    },
+    {
+      "epoch": 1.1018286814244467,
+      "grad_norm": 0.9842272996902466,
+      "learning_rate": 8.415199615628974e-05,
+      "loss": 0.819,
+      "step": 28620
+    },
+    {
+      "epoch": 1.1020211742059673,
+      "grad_norm": 1.6019426584243774,
+      "learning_rate": 8.412213943812566e-05,
+      "loss": 0.7806,
+      "step": 28625
+    },
+    {
+      "epoch": 1.102213666987488,
+      "grad_norm": 1.4778478145599365,
+      "learning_rate": 8.40922841718868e-05,
+      "loss": 0.878,
+      "step": 28630
+    },
+    {
+      "epoch": 1.1024061597690087,
+      "grad_norm": 1.3348811864852905,
+      "learning_rate": 8.406243036030316e-05,
+      "loss": 0.8364,
+      "step": 28635
+    },
+    {
+      "epoch": 1.1025986525505294,
+      "grad_norm": 2.31394100189209,
+      "learning_rate": 8.403257800610465e-05,
+      "loss": 0.8119,
+      "step": 28640
+    },
+    {
+      "epoch": 1.10279114533205,
+      "grad_norm": 2.239318370819092,
+      "learning_rate": 8.400272711202115e-05,
+      "loss": 0.865,
+      "step": 28645
+    },
+    {
+      "epoch": 1.1029836381135707,
+      "grad_norm": 2.2287495136260986,
+      "learning_rate": 8.397287768078227e-05,
+      "loss": 0.9477,
+      "step": 28650
+    },
+    {
+      "epoch": 1.1031761308950914,
+      "grad_norm": 1.1342699527740479,
+      "learning_rate": 8.394302971511757e-05,
+      "loss": 0.8375,
+      "step": 28655
+    },
+    {
+      "epoch": 1.103368623676612,
+      "grad_norm": 1.326915979385376,
+      "learning_rate": 8.39131832177564e-05,
+      "loss": 0.8521,
+      "step": 28660
+    },
+    {
+      "epoch": 1.103561116458133,
+      "grad_norm": 1.2437760829925537,
+      "learning_rate": 8.38833381914281e-05,
+      "loss": 0.8906,
+      "step": 28665
+    },
+    {
+      "epoch": 1.1037536092396536,
+      "grad_norm": 1.5128203630447388,
+      "learning_rate": 8.385349463886174e-05,
+      "loss": 0.9937,
+      "step": 28670
+    },
+    {
+      "epoch": 1.1039461020211743,
+      "grad_norm": 1.7990190982818604,
+      "learning_rate": 8.382365256278635e-05,
+      "loss": 0.8797,
+      "step": 28675
+    },
+    {
+      "epoch": 1.104138594802695,
+      "grad_norm": 0.9128474593162537,
+      "learning_rate": 8.379381196593075e-05,
+      "loss": 0.8568,
+      "step": 28680
+    },
+    {
+      "epoch": 1.1043310875842156,
+      "grad_norm": 0.8986829519271851,
+      "learning_rate": 8.376397285102369e-05,
+      "loss": 0.6961,
+      "step": 28685
+    },
+    {
+      "epoch": 1.1045235803657363,
+      "grad_norm": 1.1345797777175903,
+      "learning_rate": 8.373413522079377e-05,
+      "loss": 0.8966,
+      "step": 28690
+    },
+    {
+      "epoch": 1.104716073147257,
+      "grad_norm": 0.927941083908081,
+      "learning_rate": 8.370429907796945e-05,
+      "loss": 0.955,
+      "step": 28695
+    },
+    {
+      "epoch": 1.1049085659287776,
+      "grad_norm": 1.2426468133926392,
+      "learning_rate": 8.367446442527898e-05,
+      "loss": 0.6989,
+      "step": 28700
+    },
+    {
+      "epoch": 1.1051010587102983,
+      "grad_norm": 1.2944527864456177,
+      "learning_rate": 8.364463126545064e-05,
+      "loss": 0.8288,
+      "step": 28705
+    },
+    {
+      "epoch": 1.105293551491819,
+      "grad_norm": 0.8285412788391113,
+      "learning_rate": 8.361479960121242e-05,
+      "loss": 0.7919,
+      "step": 28710
+    },
+    {
+      "epoch": 1.1054860442733399,
+      "grad_norm": 1.3904072046279907,
+      "learning_rate": 8.358496943529219e-05,
+      "loss": 0.8751,
+      "step": 28715
+    },
+    {
+      "epoch": 1.1056785370548605,
+      "grad_norm": 1.4166022539138794,
+      "learning_rate": 8.355514077041782e-05,
+      "loss": 0.8239,
+      "step": 28720
+    },
+    {
+      "epoch": 1.1058710298363812,
+      "grad_norm": 1.4481518268585205,
+      "learning_rate": 8.352531360931688e-05,
+      "loss": 0.7997,
+      "step": 28725
+    },
+    {
+      "epoch": 1.1060635226179019,
+      "grad_norm": 1.293131709098816,
+      "learning_rate": 8.349548795471683e-05,
+      "loss": 0.8535,
+      "step": 28730
+    },
+    {
+      "epoch": 1.1062560153994225,
+      "grad_norm": 1.2604767084121704,
+      "learning_rate": 8.346566380934511e-05,
+      "loss": 0.892,
+      "step": 28735
+    },
+    {
+      "epoch": 1.1064485081809432,
+      "grad_norm": 1.0578190088272095,
+      "learning_rate": 8.343584117592894e-05,
+      "loss": 0.8556,
+      "step": 28740
+    },
+    {
+      "epoch": 1.1066410009624639,
+      "grad_norm": 0.7822481989860535,
+      "learning_rate": 8.340602005719529e-05,
+      "loss": 0.8676,
+      "step": 28745
+    },
+    {
+      "epoch": 1.1068334937439845,
+      "grad_norm": 1.1127487421035767,
+      "learning_rate": 8.337620045587123e-05,
+      "loss": 0.8864,
+      "step": 28750
+    },
+    {
+      "epoch": 1.1070259865255052,
+      "grad_norm": 1.4301708936691284,
+      "learning_rate": 8.334638237468352e-05,
+      "loss": 0.7653,
+      "step": 28755
+    },
+    {
+      "epoch": 1.107218479307026,
+      "grad_norm": 1.5563634634017944,
+      "learning_rate": 8.331656581635879e-05,
+      "loss": 0.8272,
+      "step": 28760
+    },
+    {
+      "epoch": 1.1074109720885468,
+      "grad_norm": 1.9292504787445068,
+      "learning_rate": 8.328675078362363e-05,
+      "loss": 0.9858,
+      "step": 28765
+    },
+    {
+      "epoch": 1.1076034648700674,
+      "grad_norm": 0.9912068247795105,
+      "learning_rate": 8.325693727920438e-05,
+      "loss": 0.805,
+      "step": 28770
+    },
+    {
+      "epoch": 1.1077959576515881,
+      "grad_norm": 1.0305575132369995,
+      "learning_rate": 8.32271253058273e-05,
+      "loss": 0.9662,
+      "step": 28775
+    },
+    {
+      "epoch": 1.1079884504331088,
+      "grad_norm": 1.2367445230484009,
+      "learning_rate": 8.31973148662185e-05,
+      "loss": 0.9497,
+      "step": 28780
+    },
+    {
+      "epoch": 1.1081809432146295,
+      "grad_norm": 1.2530933618545532,
+      "learning_rate": 8.316750596310396e-05,
+      "loss": 0.8991,
+      "step": 28785
+    },
+    {
+      "epoch": 1.1083734359961501,
+      "grad_norm": 0.8720499873161316,
+      "learning_rate": 8.313769859920947e-05,
+      "loss": 0.7811,
+      "step": 28790
+    },
+    {
+      "epoch": 1.1085659287776708,
+      "grad_norm": 1.3998817205429077,
+      "learning_rate": 8.310789277726079e-05,
+      "loss": 0.8827,
+      "step": 28795
+    },
+    {
+      "epoch": 1.1087584215591915,
+      "grad_norm": 2.0500388145446777,
+      "learning_rate": 8.30780884999834e-05,
+      "loss": 0.8259,
+      "step": 28800
+    },
+    {
+      "epoch": 1.1089509143407121,
+      "grad_norm": 1.5865284204483032,
+      "learning_rate": 8.304828577010266e-05,
+      "loss": 0.7836,
+      "step": 28805
+    },
+    {
+      "epoch": 1.1091434071222328,
+      "grad_norm": 1.2251750230789185,
+      "learning_rate": 8.301848459034396e-05,
+      "loss": 0.773,
+      "step": 28810
+    },
+    {
+      "epoch": 1.1093358999037537,
+      "grad_norm": 1.2303904294967651,
+      "learning_rate": 8.298868496343234e-05,
+      "loss": 0.9684,
+      "step": 28815
+    },
+    {
+      "epoch": 1.1095283926852744,
+      "grad_norm": 1.153835415840149,
+      "learning_rate": 8.295888689209274e-05,
+      "loss": 0.834,
+      "step": 28820
+    },
+    {
+      "epoch": 1.109720885466795,
+      "grad_norm": 1.5492242574691772,
+      "learning_rate": 8.29290903790501e-05,
+      "loss": 0.7547,
+      "step": 28825
+    },
+    {
+      "epoch": 1.1099133782483157,
+      "grad_norm": 1.5805435180664062,
+      "learning_rate": 8.289929542702905e-05,
+      "loss": 0.8278,
+      "step": 28830
+    },
+    {
+      "epoch": 1.1101058710298364,
+      "grad_norm": 1.0030450820922852,
+      "learning_rate": 8.286950203875412e-05,
+      "loss": 0.8725,
+      "step": 28835
+    },
+    {
+      "epoch": 1.110298363811357,
+      "grad_norm": 1.6949737071990967,
+      "learning_rate": 8.28397102169498e-05,
+      "loss": 0.8892,
+      "step": 28840
+    },
+    {
+      "epoch": 1.1104908565928777,
+      "grad_norm": 1.040964126586914,
+      "learning_rate": 8.280991996434025e-05,
+      "loss": 0.9165,
+      "step": 28845
+    },
+    {
+      "epoch": 1.1106833493743984,
+      "grad_norm": 1.1363362073898315,
+      "learning_rate": 8.278013128364967e-05,
+      "loss": 0.8674,
+      "step": 28850
+    },
+    {
+      "epoch": 1.110875842155919,
+      "grad_norm": 0.8567915558815002,
+      "learning_rate": 8.275034417760205e-05,
+      "loss": 0.8689,
+      "step": 28855
+    },
+    {
+      "epoch": 1.11106833493744,
+      "grad_norm": 1.597185492515564,
+      "learning_rate": 8.272055864892112e-05,
+      "loss": 0.7718,
+      "step": 28860
+    },
+    {
+      "epoch": 1.1112608277189606,
+      "grad_norm": 0.7887008786201477,
+      "learning_rate": 8.269077470033068e-05,
+      "loss": 0.8011,
+      "step": 28865
+    },
+    {
+      "epoch": 1.1114533205004813,
+      "grad_norm": 1.3229820728302002,
+      "learning_rate": 8.266099233455423e-05,
+      "loss": 0.8735,
+      "step": 28870
+    },
+    {
+      "epoch": 1.111645813282002,
+      "grad_norm": 1.3275078535079956,
+      "learning_rate": 8.263121155431515e-05,
+      "loss": 0.9547,
+      "step": 28875
+    },
+    {
+      "epoch": 1.1118383060635226,
+      "grad_norm": 1.2389519214630127,
+      "learning_rate": 8.260143236233673e-05,
+      "loss": 1.079,
+      "step": 28880
+    },
+    {
+      "epoch": 1.1120307988450433,
+      "grad_norm": 1.1708885431289673,
+      "learning_rate": 8.257165476134207e-05,
+      "loss": 0.9086,
+      "step": 28885
+    },
+    {
+      "epoch": 1.112223291626564,
+      "grad_norm": 1.1817975044250488,
+      "learning_rate": 8.254187875405411e-05,
+      "loss": 0.914,
+      "step": 28890
+    },
+    {
+      "epoch": 1.1124157844080846,
+      "grad_norm": 1.4715931415557861,
+      "learning_rate": 8.251210434319572e-05,
+      "loss": 0.9382,
+      "step": 28895
+    },
+    {
+      "epoch": 1.1126082771896053,
+      "grad_norm": 1.0971523523330688,
+      "learning_rate": 8.248233153148953e-05,
+      "loss": 0.6749,
+      "step": 28900
+    },
+    {
+      "epoch": 1.112800769971126,
+      "grad_norm": 0.9213184714317322,
+      "learning_rate": 8.245256032165804e-05,
+      "loss": 0.8894,
+      "step": 28905
+    },
+    {
+      "epoch": 1.1129932627526469,
+      "grad_norm": 1.3410868644714355,
+      "learning_rate": 8.242279071642371e-05,
+      "loss": 0.8031,
+      "step": 28910
+    },
+    {
+      "epoch": 1.1131857555341675,
+      "grad_norm": 0.9468138813972473,
+      "learning_rate": 8.239302271850874e-05,
+      "loss": 0.9081,
+      "step": 28915
+    },
+    {
+      "epoch": 1.1133782483156882,
+      "grad_norm": 0.9489649534225464,
+      "learning_rate": 8.236325633063515e-05,
+      "loss": 0.9047,
+      "step": 28920
+    },
+    {
+      "epoch": 1.1135707410972089,
+      "grad_norm": 1.7119163274765015,
+      "learning_rate": 8.233349155552499e-05,
+      "loss": 0.7558,
+      "step": 28925
+    },
+    {
+      "epoch": 1.1137632338787296,
+      "grad_norm": 1.610944151878357,
+      "learning_rate": 8.230372839589998e-05,
+      "loss": 0.868,
+      "step": 28930
+    },
+    {
+      "epoch": 1.1139557266602502,
+      "grad_norm": 1.2964144945144653,
+      "learning_rate": 8.227396685448176e-05,
+      "loss": 0.7636,
+      "step": 28935
+    },
+    {
+      "epoch": 1.114148219441771,
+      "grad_norm": 1.2506333589553833,
+      "learning_rate": 8.224420693399187e-05,
+      "loss": 0.8933,
+      "step": 28940
+    },
+    {
+      "epoch": 1.1143407122232916,
+      "grad_norm": 1.260901689529419,
+      "learning_rate": 8.221444863715165e-05,
+      "loss": 0.8687,
+      "step": 28945
+    },
+    {
+      "epoch": 1.1145332050048122,
+      "grad_norm": 0.6778833866119385,
+      "learning_rate": 8.218469196668222e-05,
+      "loss": 0.7826,
+      "step": 28950
+    },
+    {
+      "epoch": 1.1147256977863331,
+      "grad_norm": 2.43918514251709,
+      "learning_rate": 8.215493692530475e-05,
+      "loss": 1.0159,
+      "step": 28955
+    },
+    {
+      "epoch": 1.1149181905678538,
+      "grad_norm": 1.1964274644851685,
+      "learning_rate": 8.21251835157401e-05,
+      "loss": 0.8936,
+      "step": 28960
+    },
+    {
+      "epoch": 1.1151106833493745,
+      "grad_norm": 1.0071650743484497,
+      "learning_rate": 8.209543174070894e-05,
+      "loss": 0.894,
+      "step": 28965
+    },
+    {
+      "epoch": 1.1153031761308951,
+      "grad_norm": 1.0765222311019897,
+      "learning_rate": 8.2065681602932e-05,
+      "loss": 0.943,
+      "step": 28970
+    },
+    {
+      "epoch": 1.1154956689124158,
+      "grad_norm": 1.2114105224609375,
+      "learning_rate": 8.203593310512967e-05,
+      "loss": 0.9682,
+      "step": 28975
+    },
+    {
+      "epoch": 1.1156881616939365,
+      "grad_norm": 0.9356218576431274,
+      "learning_rate": 8.200618625002223e-05,
+      "loss": 0.7903,
+      "step": 28980
+    },
+    {
+      "epoch": 1.1158806544754571,
+      "grad_norm": 1.1432327032089233,
+      "learning_rate": 8.197644104032987e-05,
+      "loss": 0.79,
+      "step": 28985
+    },
+    {
+      "epoch": 1.1160731472569778,
+      "grad_norm": 1.7446037530899048,
+      "learning_rate": 8.194669747877259e-05,
+      "loss": 0.8792,
+      "step": 28990
+    },
+    {
+      "epoch": 1.1162656400384985,
+      "grad_norm": 2.04817533493042,
+      "learning_rate": 8.191695556807022e-05,
+      "loss": 0.9906,
+      "step": 28995
+    },
+    {
+      "epoch": 1.1164581328200192,
+      "grad_norm": 1.7263697385787964,
+      "learning_rate": 8.18872153109425e-05,
+      "loss": 0.7607,
+      "step": 29000
+    },
+    {
+      "epoch": 1.1166506256015398,
+      "grad_norm": 1.177629828453064,
+      "learning_rate": 8.185747671010891e-05,
+      "loss": 0.7068,
+      "step": 29005
+    },
+    {
+      "epoch": 1.1168431183830607,
+      "grad_norm": 0.9758390188217163,
+      "learning_rate": 8.18277397682889e-05,
+      "loss": 0.9517,
+      "step": 29010
+    },
+    {
+      "epoch": 1.1170356111645814,
+      "grad_norm": 1.2482577562332153,
+      "learning_rate": 8.179800448820174e-05,
+      "loss": 0.839,
+      "step": 29015
+    },
+    {
+      "epoch": 1.117228103946102,
+      "grad_norm": 1.2977570295333862,
+      "learning_rate": 8.176827087256649e-05,
+      "loss": 0.8067,
+      "step": 29020
+    },
+    {
+      "epoch": 1.1174205967276227,
+      "grad_norm": 1.2474614381790161,
+      "learning_rate": 8.173853892410206e-05,
+      "loss": 0.947,
+      "step": 29025
+    },
+    {
+      "epoch": 1.1176130895091434,
+      "grad_norm": 1.5093929767608643,
+      "learning_rate": 8.17088086455273e-05,
+      "loss": 0.67,
+      "step": 29030
+    },
+    {
+      "epoch": 1.117805582290664,
+      "grad_norm": 1.944604754447937,
+      "learning_rate": 8.167908003956082e-05,
+      "loss": 0.8075,
+      "step": 29035
+    },
+    {
+      "epoch": 1.1179980750721847,
+      "grad_norm": 1.2519726753234863,
+      "learning_rate": 8.164935310892108e-05,
+      "loss": 0.9158,
+      "step": 29040
+    },
+    {
+      "epoch": 1.1181905678537054,
+      "grad_norm": 1.3411145210266113,
+      "learning_rate": 8.161962785632646e-05,
+      "loss": 0.9831,
+      "step": 29045
+    },
+    {
+      "epoch": 1.118383060635226,
+      "grad_norm": 1.7142786979675293,
+      "learning_rate": 8.158990428449514e-05,
+      "loss": 0.8387,
+      "step": 29050
+    },
+    {
+      "epoch": 1.118575553416747,
+      "grad_norm": 1.1233665943145752,
+      "learning_rate": 8.156018239614505e-05,
+      "loss": 0.8295,
+      "step": 29055
+    },
+    {
+      "epoch": 1.1187680461982676,
+      "grad_norm": 1.210729718208313,
+      "learning_rate": 8.153046219399418e-05,
+      "loss": 0.8221,
+      "step": 29060
+    },
+    {
+      "epoch": 1.1189605389797883,
+      "grad_norm": 1.134706735610962,
+      "learning_rate": 8.150074368076019e-05,
+      "loss": 0.8517,
+      "step": 29065
+    },
+    {
+      "epoch": 1.119153031761309,
+      "grad_norm": 1.5711884498596191,
+      "learning_rate": 8.14710268591606e-05,
+      "loss": 0.8291,
+      "step": 29070
+    },
+    {
+      "epoch": 1.1193455245428297,
+      "grad_norm": 1.2242764234542847,
+      "learning_rate": 8.144131173191292e-05,
+      "loss": 0.8554,
+      "step": 29075
+    },
+    {
+      "epoch": 1.1195380173243503,
+      "grad_norm": 0.9937723875045776,
+      "learning_rate": 8.141159830173432e-05,
+      "loss": 0.9085,
+      "step": 29080
+    },
+    {
+      "epoch": 1.119730510105871,
+      "grad_norm": 1.4553319215774536,
+      "learning_rate": 8.138188657134188e-05,
+      "loss": 0.7981,
+      "step": 29085
+    },
+    {
+      "epoch": 1.1199230028873917,
+      "grad_norm": 1.1583729982376099,
+      "learning_rate": 8.135217654345263e-05,
+      "loss": 0.8145,
+      "step": 29090
+    },
+    {
+      "epoch": 1.1201154956689123,
+      "grad_norm": 1.2025479078292847,
+      "learning_rate": 8.132246822078328e-05,
+      "loss": 0.925,
+      "step": 29095
+    },
+    {
+      "epoch": 1.120307988450433,
+      "grad_norm": 1.4865069389343262,
+      "learning_rate": 8.12927616060505e-05,
+      "loss": 0.7075,
+      "step": 29100
+    },
+    {
+      "epoch": 1.120500481231954,
+      "grad_norm": 2.3236899375915527,
+      "learning_rate": 8.126305670197073e-05,
+      "loss": 0.9911,
+      "step": 29105
+    },
+    {
+      "epoch": 1.1206929740134746,
+      "grad_norm": 0.9563505053520203,
+      "learning_rate": 8.12333535112603e-05,
+      "loss": 0.7803,
+      "step": 29110
+    },
+    {
+      "epoch": 1.1208854667949952,
+      "grad_norm": 1.071954607963562,
+      "learning_rate": 8.12036520366354e-05,
+      "loss": 0.808,
+      "step": 29115
+    },
+    {
+      "epoch": 1.121077959576516,
+      "grad_norm": 0.5228383541107178,
+      "learning_rate": 8.117395228081202e-05,
+      "loss": 0.8721,
+      "step": 29120
+    },
+    {
+      "epoch": 1.1212704523580366,
+      "grad_norm": 1.238578200340271,
+      "learning_rate": 8.114425424650592e-05,
+      "loss": 0.8783,
+      "step": 29125
+    },
+    {
+      "epoch": 1.1214629451395572,
+      "grad_norm": 0.9781158566474915,
+      "learning_rate": 8.111455793643292e-05,
+      "loss": 0.8673,
+      "step": 29130
+    },
+    {
+      "epoch": 1.121655437921078,
+      "grad_norm": 1.529806137084961,
+      "learning_rate": 8.10848633533085e-05,
+      "loss": 0.9377,
+      "step": 29135
+    },
+    {
+      "epoch": 1.1218479307025986,
+      "grad_norm": 1.0390987396240234,
+      "learning_rate": 8.105517049984797e-05,
+      "loss": 0.8467,
+      "step": 29140
+    },
+    {
+      "epoch": 1.1220404234841193,
+      "grad_norm": 1.066085696220398,
+      "learning_rate": 8.102547937876664e-05,
+      "loss": 0.7405,
+      "step": 29145
+    },
+    {
+      "epoch": 1.1222329162656401,
+      "grad_norm": 0.9266438484191895,
+      "learning_rate": 8.099578999277953e-05,
+      "loss": 0.8576,
+      "step": 29150
+    },
+    {
+      "epoch": 1.1224254090471608,
+      "grad_norm": 1.52492356300354,
+      "learning_rate": 8.09661023446015e-05,
+      "loss": 0.86,
+      "step": 29155
+    },
+    {
+      "epoch": 1.1226179018286815,
+      "grad_norm": 1.8862872123718262,
+      "learning_rate": 8.093641643694736e-05,
+      "loss": 0.9879,
+      "step": 29160
+    },
+    {
+      "epoch": 1.1228103946102022,
+      "grad_norm": 1.223181128501892,
+      "learning_rate": 8.090673227253165e-05,
+      "loss": 0.8792,
+      "step": 29165
+    },
+    {
+      "epoch": 1.1230028873917228,
+      "grad_norm": 1.4018183946609497,
+      "learning_rate": 8.087704985406875e-05,
+      "loss": 0.9087,
+      "step": 29170
+    },
+    {
+      "epoch": 1.1231953801732435,
+      "grad_norm": 0.8999775052070618,
+      "learning_rate": 8.084736918427302e-05,
+      "loss": 0.9368,
+      "step": 29175
+    },
+    {
+      "epoch": 1.1233878729547642,
+      "grad_norm": 1.3538589477539062,
+      "learning_rate": 8.081769026585848e-05,
+      "loss": 0.7531,
+      "step": 29180
+    },
+    {
+      "epoch": 1.1235803657362848,
+      "grad_norm": 1.039942741394043,
+      "learning_rate": 8.078801310153906e-05,
+      "loss": 0.89,
+      "step": 29185
+    },
+    {
+      "epoch": 1.1237728585178055,
+      "grad_norm": 1.8016685247421265,
+      "learning_rate": 8.075833769402861e-05,
+      "loss": 0.8751,
+      "step": 29190
+    },
+    {
+      "epoch": 1.1239653512993262,
+      "grad_norm": 1.5245671272277832,
+      "learning_rate": 8.072866404604073e-05,
+      "loss": 0.7489,
+      "step": 29195
+    },
+    {
+      "epoch": 1.124157844080847,
+      "grad_norm": 1.9634850025177002,
+      "learning_rate": 8.069899216028883e-05,
+      "loss": 0.7756,
+      "step": 29200
+    },
+    {
+      "epoch": 1.1243503368623677,
+      "grad_norm": 1.0487682819366455,
+      "learning_rate": 8.066932203948625e-05,
+      "loss": 0.697,
+      "step": 29205
+    },
+    {
+      "epoch": 1.1245428296438884,
+      "grad_norm": 1.292322039604187,
+      "learning_rate": 8.063965368634611e-05,
+      "loss": 0.8285,
+      "step": 29210
+    },
+    {
+      "epoch": 1.124735322425409,
+      "grad_norm": 1.1624624729156494,
+      "learning_rate": 8.06099871035814e-05,
+      "loss": 0.7489,
+      "step": 29215
+    },
+    {
+      "epoch": 1.1249278152069297,
+      "grad_norm": 0.9536521434783936,
+      "learning_rate": 8.058032229390493e-05,
+      "loss": 0.8356,
+      "step": 29220
+    },
+    {
+      "epoch": 1.1251203079884504,
+      "grad_norm": 1.499011516571045,
+      "learning_rate": 8.055065926002933e-05,
+      "loss": 0.964,
+      "step": 29225
+    },
+    {
+      "epoch": 1.125312800769971,
+      "grad_norm": 2.0605900287628174,
+      "learning_rate": 8.052099800466708e-05,
+      "loss": 0.8371,
+      "step": 29230
+    },
+    {
+      "epoch": 1.1255052935514918,
+      "grad_norm": 1.7930225133895874,
+      "learning_rate": 8.049133853053057e-05,
+      "loss": 0.7997,
+      "step": 29235
+    },
+    {
+      "epoch": 1.1256977863330124,
+      "grad_norm": 2.043825626373291,
+      "learning_rate": 8.04616808403319e-05,
+      "loss": 0.8135,
+      "step": 29240
+    },
+    {
+      "epoch": 1.1258902791145333,
+      "grad_norm": 1.0916283130645752,
+      "learning_rate": 8.043202493678306e-05,
+      "loss": 0.8819,
+      "step": 29245
+    },
+    {
+      "epoch": 1.126082771896054,
+      "grad_norm": 1.748055100440979,
+      "learning_rate": 8.040237082259594e-05,
+      "loss": 0.8833,
+      "step": 29250
+    },
+    {
+      "epoch": 1.1262752646775747,
+      "grad_norm": 1.4773916006088257,
+      "learning_rate": 8.03727185004822e-05,
+      "loss": 0.7417,
+      "step": 29255
+    },
+    {
+      "epoch": 1.1264677574590953,
+      "grad_norm": 0.9700600504875183,
+      "learning_rate": 8.034306797315328e-05,
+      "loss": 0.7682,
+      "step": 29260
+    },
+    {
+      "epoch": 1.126660250240616,
+      "grad_norm": 1.6670438051223755,
+      "learning_rate": 8.031341924332063e-05,
+      "loss": 0.8024,
+      "step": 29265
+    },
+    {
+      "epoch": 1.1268527430221367,
+      "grad_norm": 1.4361203908920288,
+      "learning_rate": 8.028377231369538e-05,
+      "loss": 0.9223,
+      "step": 29270
+    },
+    {
+      "epoch": 1.1270452358036573,
+      "grad_norm": 2.1577651500701904,
+      "learning_rate": 8.02541271869885e-05,
+      "loss": 0.7592,
+      "step": 29275
+    },
+    {
+      "epoch": 1.127237728585178,
+      "grad_norm": 1.7753071784973145,
+      "learning_rate": 8.022448386591093e-05,
+      "loss": 0.7609,
+      "step": 29280
+    },
+    {
+      "epoch": 1.1274302213666987,
+      "grad_norm": 1.2456077337265015,
+      "learning_rate": 8.01948423531733e-05,
+      "loss": 0.9491,
+      "step": 29285
+    },
+    {
+      "epoch": 1.1276227141482194,
+      "grad_norm": 1.634460687637329,
+      "learning_rate": 8.01652026514861e-05,
+      "loss": 0.9136,
+      "step": 29290
+    },
+    {
+      "epoch": 1.12781520692974,
+      "grad_norm": 1.8183555603027344,
+      "learning_rate": 8.013556476355976e-05,
+      "loss": 0.8587,
+      "step": 29295
+    },
+    {
+      "epoch": 1.128007699711261,
+      "grad_norm": 1.5357935428619385,
+      "learning_rate": 8.010592869210445e-05,
+      "loss": 0.8325,
+      "step": 29300
+    },
+    {
+      "epoch": 1.1282001924927816,
+      "grad_norm": 1.0693864822387695,
+      "learning_rate": 8.007629443983013e-05,
+      "loss": 0.7546,
+      "step": 29305
+    },
+    {
+      "epoch": 1.1283926852743023,
+      "grad_norm": 1.100164771080017,
+      "learning_rate": 8.004666200944673e-05,
+      "loss": 0.9358,
+      "step": 29310
+    },
+    {
+      "epoch": 1.128585178055823,
+      "grad_norm": 1.2228190898895264,
+      "learning_rate": 8.00170314036639e-05,
+      "loss": 0.9255,
+      "step": 29315
+    },
+    {
+      "epoch": 1.1287776708373436,
+      "grad_norm": 1.408209204673767,
+      "learning_rate": 7.998740262519118e-05,
+      "loss": 0.7362,
+      "step": 29320
+    },
+    {
+      "epoch": 1.1289701636188643,
+      "grad_norm": 2.5063939094543457,
+      "learning_rate": 7.995777567673791e-05,
+      "loss": 0.7319,
+      "step": 29325
+    },
+    {
+      "epoch": 1.129162656400385,
+      "grad_norm": 1.9858168363571167,
+      "learning_rate": 7.992815056101331e-05,
+      "loss": 0.9686,
+      "step": 29330
+    },
+    {
+      "epoch": 1.1293551491819056,
+      "grad_norm": 1.276520013809204,
+      "learning_rate": 7.989852728072635e-05,
+      "loss": 0.8679,
+      "step": 29335
+    },
+    {
+      "epoch": 1.1295476419634265,
+      "grad_norm": 1.9740073680877686,
+      "learning_rate": 7.986890583858593e-05,
+      "loss": 0.9984,
+      "step": 29340
+    },
+    {
+      "epoch": 1.1297401347449472,
+      "grad_norm": 0.7693697214126587,
+      "learning_rate": 7.983928623730073e-05,
+      "loss": 0.7382,
+      "step": 29345
+    },
+    {
+      "epoch": 1.1299326275264678,
+      "grad_norm": 1.349514365196228,
+      "learning_rate": 7.980966847957918e-05,
+      "loss": 0.9728,
+      "step": 29350
+    },
+    {
+      "epoch": 1.1301251203079885,
+      "grad_norm": 1.5095326900482178,
+      "learning_rate": 7.978005256812977e-05,
+      "loss": 0.9931,
+      "step": 29355
+    },
+    {
+      "epoch": 1.1303176130895092,
+      "grad_norm": 1.1477704048156738,
+      "learning_rate": 7.975043850566053e-05,
+      "loss": 0.7388,
+      "step": 29360
+    },
+    {
+      "epoch": 1.1305101058710298,
+      "grad_norm": 1.9314098358154297,
+      "learning_rate": 7.972082629487962e-05,
+      "loss": 0.8511,
+      "step": 29365
+    },
+    {
+      "epoch": 1.1307025986525505,
+      "grad_norm": 2.503371477127075,
+      "learning_rate": 7.969121593849481e-05,
+      "loss": 0.8943,
+      "step": 29370
+    },
+    {
+      "epoch": 1.1308950914340712,
+      "grad_norm": 1.379692792892456,
+      "learning_rate": 7.966160743921371e-05,
+      "loss": 0.8356,
+      "step": 29375
+    },
+    {
+      "epoch": 1.1310875842155919,
+      "grad_norm": 1.3109577894210815,
+      "learning_rate": 7.963200079974394e-05,
+      "loss": 0.9496,
+      "step": 29380
+    },
+    {
+      "epoch": 1.1312800769971125,
+      "grad_norm": 1.2980083227157593,
+      "learning_rate": 7.960239602279275e-05,
+      "loss": 0.9289,
+      "step": 29385
+    },
+    {
+      "epoch": 1.1314725697786332,
+      "grad_norm": 1.131455659866333,
+      "learning_rate": 7.95727931110673e-05,
+      "loss": 0.8283,
+      "step": 29390
+    },
+    {
+      "epoch": 1.131665062560154,
+      "grad_norm": 2.513030529022217,
+      "learning_rate": 7.954319206727464e-05,
+      "loss": 0.8418,
+      "step": 29395
+    },
+    {
+      "epoch": 1.1318575553416748,
+      "grad_norm": 1.2375028133392334,
+      "learning_rate": 7.951359289412154e-05,
+      "loss": 0.9583,
+      "step": 29400
+    },
+    {
+      "epoch": 1.1320500481231954,
+      "grad_norm": 1.336858868598938,
+      "learning_rate": 7.948399559431463e-05,
+      "loss": 0.8757,
+      "step": 29405
+    },
+    {
+      "epoch": 1.132242540904716,
+      "grad_norm": 1.1518652439117432,
+      "learning_rate": 7.945440017056045e-05,
+      "loss": 0.8316,
+      "step": 29410
+    },
+    {
+      "epoch": 1.1324350336862368,
+      "grad_norm": 1.9742542505264282,
+      "learning_rate": 7.942480662556527e-05,
+      "loss": 0.7784,
+      "step": 29415
+    },
+    {
+      "epoch": 1.1326275264677574,
+      "grad_norm": 1.131545066833496,
+      "learning_rate": 7.939521496203521e-05,
+      "loss": 0.974,
+      "step": 29420
+    },
+    {
+      "epoch": 1.132820019249278,
+      "grad_norm": 1.558993935585022,
+      "learning_rate": 7.936562518267625e-05,
+      "loss": 1.0205,
+      "step": 29425
+    },
+    {
+      "epoch": 1.1330125120307988,
+      "grad_norm": 1.082209825515747,
+      "learning_rate": 7.93360372901942e-05,
+      "loss": 1.0551,
+      "step": 29430
+    },
+    {
+      "epoch": 1.1332050048123194,
+      "grad_norm": 1.8491493463516235,
+      "learning_rate": 7.930645128729462e-05,
+      "loss": 0.8227,
+      "step": 29435
+    },
+    {
+      "epoch": 1.1333974975938403,
+      "grad_norm": 1.647204875946045,
+      "learning_rate": 7.927686717668302e-05,
+      "loss": 0.9909,
+      "step": 29440
+    },
+    {
+      "epoch": 1.133589990375361,
+      "grad_norm": 1.1486252546310425,
+      "learning_rate": 7.924728496106464e-05,
+      "loss": 0.8855,
+      "step": 29445
+    },
+    {
+      "epoch": 1.1337824831568817,
+      "grad_norm": 1.3618193864822388,
+      "learning_rate": 7.921770464314453e-05,
+      "loss": 0.8427,
+      "step": 29450
+    },
+    {
+      "epoch": 1.1339749759384024,
+      "grad_norm": 1.4775711297988892,
+      "learning_rate": 7.91881262256277e-05,
+      "loss": 0.8925,
+      "step": 29455
+    },
+    {
+      "epoch": 1.134167468719923,
+      "grad_norm": 1.0450371503829956,
+      "learning_rate": 7.915854971121886e-05,
+      "loss": 0.6872,
+      "step": 29460
+    },
+    {
+      "epoch": 1.1343599615014437,
+      "grad_norm": 1.4951140880584717,
+      "learning_rate": 7.912897510262255e-05,
+      "loss": 0.7814,
+      "step": 29465
+    },
+    {
+      "epoch": 1.1345524542829644,
+      "grad_norm": 0.9859576225280762,
+      "learning_rate": 7.909940240254325e-05,
+      "loss": 0.7146,
+      "step": 29470
+    },
+    {
+      "epoch": 1.134744947064485,
+      "grad_norm": 2.4736711978912354,
+      "learning_rate": 7.906983161368515e-05,
+      "loss": 1.0436,
+      "step": 29475
+    },
+    {
+      "epoch": 1.1349374398460057,
+      "grad_norm": 0.912487268447876,
+      "learning_rate": 7.904026273875225e-05,
+      "loss": 0.8881,
+      "step": 29480
+    },
+    {
+      "epoch": 1.1351299326275264,
+      "grad_norm": 1.1811585426330566,
+      "learning_rate": 7.901069578044852e-05,
+      "loss": 0.8878,
+      "step": 29485
+    },
+    {
+      "epoch": 1.135322425409047,
+      "grad_norm": 1.4744383096694946,
+      "learning_rate": 7.898113074147762e-05,
+      "loss": 0.8272,
+      "step": 29490
+    },
+    {
+      "epoch": 1.135514918190568,
+      "grad_norm": 1.061719536781311,
+      "learning_rate": 7.895748009403729e-05,
+      "loss": 1.3432,
+      "step": 29495
+    },
+    {
+      "epoch": 1.1357074109720886,
+      "grad_norm": 1.2878315448760986,
+      "learning_rate": 7.892791851667824e-05,
+      "loss": 1.045,
+      "step": 29500
+    },
+    {
+      "epoch": 1.1358999037536093,
+      "grad_norm": 1.3372122049331665,
+      "learning_rate": 7.889835886622145e-05,
+      "loss": 0.8942,
+      "step": 29505
+    },
+    {
+      "epoch": 1.13609239653513,
+      "grad_norm": 1.7610641717910767,
+      "learning_rate": 7.886880114536997e-05,
+      "loss": 0.9568,
+      "step": 29510
+    },
+    {
+      "epoch": 1.1362848893166506,
+      "grad_norm": 1.9298161268234253,
+      "learning_rate": 7.883924535682665e-05,
+      "loss": 0.8845,
+      "step": 29515
+    },
+    {
+      "epoch": 1.1364773820981713,
+      "grad_norm": 1.6231752634048462,
+      "learning_rate": 7.880969150329412e-05,
+      "loss": 0.7979,
+      "step": 29520
+    },
+    {
+      "epoch": 1.136669874879692,
+      "grad_norm": 2.3804054260253906,
+      "learning_rate": 7.878013958747496e-05,
+      "loss": 1.0056,
+      "step": 29525
+    },
+    {
+      "epoch": 1.1368623676612126,
+      "grad_norm": 1.7176682949066162,
+      "learning_rate": 7.875058961207148e-05,
+      "loss": 0.8073,
+      "step": 29530
+    },
+    {
+      "epoch": 1.1370548604427335,
+      "grad_norm": 1.6357464790344238,
+      "learning_rate": 7.872104157978576e-05,
+      "loss": 1.0359,
+      "step": 29535
+    },
+    {
+      "epoch": 1.1372473532242542,
+      "grad_norm": 0.9791454076766968,
+      "learning_rate": 7.869149549331989e-05,
+      "loss": 0.5608,
+      "step": 29540
+    },
+    {
+      "epoch": 1.1374398460057749,
+      "grad_norm": 1.282942533493042,
+      "learning_rate": 7.866195135537558e-05,
+      "loss": 0.8483,
+      "step": 29545
+    },
+    {
+      "epoch": 1.1376323387872955,
+      "grad_norm": 1.7176309823989868,
+      "learning_rate": 7.863240916865445e-05,
+      "loss": 0.8454,
+      "step": 29550
+    },
+    {
+      "epoch": 1.1378248315688162,
+      "grad_norm": 2.1249489784240723,
+      "learning_rate": 7.860286893585798e-05,
+      "loss": 1.0242,
+      "step": 29555
+    },
+    {
+      "epoch": 1.1380173243503369,
+      "grad_norm": 1.37616765499115,
+      "learning_rate": 7.85733306596874e-05,
+      "loss": 0.7752,
+      "step": 29560
+    },
+    {
+      "epoch": 1.1382098171318575,
+      "grad_norm": 1.312436580657959,
+      "learning_rate": 7.854379434284375e-05,
+      "loss": 0.7942,
+      "step": 29565
+    },
+    {
+      "epoch": 1.1384023099133782,
+      "grad_norm": 1.2645299434661865,
+      "learning_rate": 7.851425998802801e-05,
+      "loss": 0.8355,
+      "step": 29570
+    },
+    {
+      "epoch": 1.1385948026948989,
+      "grad_norm": 1.6107802391052246,
+      "learning_rate": 7.848472759794087e-05,
+      "loss": 0.7958,
+      "step": 29575
+    },
+    {
+      "epoch": 1.1387872954764195,
+      "grad_norm": 1.874605655670166,
+      "learning_rate": 7.845519717528284e-05,
+      "loss": 1.0369,
+      "step": 29580
+    },
+    {
+      "epoch": 1.1389797882579402,
+      "grad_norm": 0.8068329691886902,
+      "learning_rate": 7.842566872275433e-05,
+      "loss": 0.7825,
+      "step": 29585
+    },
+    {
+      "epoch": 1.139172281039461,
+      "grad_norm": 0.9913658499717712,
+      "learning_rate": 7.839614224305551e-05,
+      "loss": 0.8762,
+      "step": 29590
+    },
+    {
+      "epoch": 1.1393647738209818,
+      "grad_norm": 1.0012116432189941,
+      "learning_rate": 7.836661773888632e-05,
+      "loss": 0.7111,
+      "step": 29595
+    },
+    {
+      "epoch": 1.1395572666025024,
+      "grad_norm": 1.5434329509735107,
+      "learning_rate": 7.833709521294668e-05,
+      "loss": 0.7117,
+      "step": 29600
+    },
+    {
+      "epoch": 1.1397497593840231,
+      "grad_norm": 1.6689229011535645,
+      "learning_rate": 7.830757466793617e-05,
+      "loss": 0.878,
+      "step": 29605
+    },
+    {
+      "epoch": 1.1399422521655438,
+      "grad_norm": 1.267446756362915,
+      "learning_rate": 7.827805610655423e-05,
+      "loss": 0.8429,
+      "step": 29610
+    },
+    {
+      "epoch": 1.1401347449470645,
+      "grad_norm": 1.4829270839691162,
+      "learning_rate": 7.824853953150019e-05,
+      "loss": 0.9683,
+      "step": 29615
+    },
+    {
+      "epoch": 1.1403272377285851,
+      "grad_norm": 1.028507113456726,
+      "learning_rate": 7.821902494547309e-05,
+      "loss": 0.7345,
+      "step": 29620
+    },
+    {
+      "epoch": 1.1405197305101058,
+      "grad_norm": 1.30324125289917,
+      "learning_rate": 7.818951235117191e-05,
+      "loss": 0.8416,
+      "step": 29625
+    },
+    {
+      "epoch": 1.1407122232916265,
+      "grad_norm": 1.0241384506225586,
+      "learning_rate": 7.816000175129534e-05,
+      "loss": 0.9229,
+      "step": 29630
+    },
+    {
+      "epoch": 1.1409047160731474,
+      "grad_norm": 1.6417129039764404,
+      "learning_rate": 7.813049314854189e-05,
+      "loss": 1.075,
+      "step": 29635
+    },
+    {
+      "epoch": 1.141097208854668,
+      "grad_norm": 0.99880450963974,
+      "learning_rate": 7.810098654561002e-05,
+      "loss": 0.6836,
+      "step": 29640
+    },
+    {
+      "epoch": 1.1412897016361887,
+      "grad_norm": 1.3350822925567627,
+      "learning_rate": 7.807148194519784e-05,
+      "loss": 0.8302,
+      "step": 29645
+    },
+    {
+      "epoch": 1.1414821944177094,
+      "grad_norm": 1.4701656103134155,
+      "learning_rate": 7.804197935000336e-05,
+      "loss": 0.8758,
+      "step": 29650
+    },
+    {
+      "epoch": 1.14167468719923,
+      "grad_norm": 1.867395281791687,
+      "learning_rate": 7.801247876272444e-05,
+      "loss": 0.8234,
+      "step": 29655
+    },
+    {
+      "epoch": 1.1418671799807507,
+      "grad_norm": 1.0658506155014038,
+      "learning_rate": 7.798298018605868e-05,
+      "loss": 0.854,
+      "step": 29660
+    },
+    {
+      "epoch": 1.1420596727622714,
+      "grad_norm": 1.4265475273132324,
+      "learning_rate": 7.795348362270349e-05,
+      "loss": 0.8507,
+      "step": 29665
+    },
+    {
+      "epoch": 1.142252165543792,
+      "grad_norm": 1.2551696300506592,
+      "learning_rate": 7.792398907535622e-05,
+      "loss": 0.9168,
+      "step": 29670
+    },
+    {
+      "epoch": 1.1424446583253127,
+      "grad_norm": 1.5808075666427612,
+      "learning_rate": 7.789449654671391e-05,
+      "loss": 0.9197,
+      "step": 29675
+    },
+    {
+      "epoch": 1.1426371511068334,
+      "grad_norm": 1.2171685695648193,
+      "learning_rate": 7.786500603947342e-05,
+      "loss": 0.8965,
+      "step": 29680
+    },
+    {
+      "epoch": 1.142829643888354,
+      "grad_norm": 1.4671955108642578,
+      "learning_rate": 7.783551755633152e-05,
+      "loss": 0.84,
+      "step": 29685
+    },
+    {
+      "epoch": 1.143022136669875,
+      "grad_norm": 0.9525561332702637,
+      "learning_rate": 7.780603109998475e-05,
+      "loss": 0.9211,
+      "step": 29690
+    },
+    {
+      "epoch": 1.1432146294513956,
+      "grad_norm": 2.0476224422454834,
+      "learning_rate": 7.777654667312934e-05,
+      "loss": 0.8695,
+      "step": 29695
+    },
+    {
+      "epoch": 1.1434071222329163,
+      "grad_norm": 0.7733902335166931,
+      "learning_rate": 7.774706427846157e-05,
+      "loss": 0.8636,
+      "step": 29700
+    },
+    {
+      "epoch": 1.143599615014437,
+      "grad_norm": 2.125204086303711,
+      "learning_rate": 7.771758391867736e-05,
+      "loss": 1.0092,
+      "step": 29705
+    },
+    {
+      "epoch": 1.1437921077959576,
+      "grad_norm": 1.576104998588562,
+      "learning_rate": 7.768810559647248e-05,
+      "loss": 0.7684,
+      "step": 29710
+    },
+    {
+      "epoch": 1.1439846005774783,
+      "grad_norm": 1.8463459014892578,
+      "learning_rate": 7.765862931454256e-05,
+      "loss": 0.9924,
+      "step": 29715
+    },
+    {
+      "epoch": 1.144177093358999,
+      "grad_norm": 1.1519972085952759,
+      "learning_rate": 7.762915507558295e-05,
+      "loss": 0.8446,
+      "step": 29720
+    },
+    {
+      "epoch": 1.1443695861405196,
+      "grad_norm": 1.6794402599334717,
+      "learning_rate": 7.759968288228892e-05,
+      "loss": 0.8016,
+      "step": 29725
+    },
+    {
+      "epoch": 1.1445620789220405,
+      "grad_norm": 0.8587988018989563,
+      "learning_rate": 7.757021273735554e-05,
+      "loss": 0.7753,
+      "step": 29730
+    },
+    {
+      "epoch": 1.1447545717035612,
+      "grad_norm": 1.1401095390319824,
+      "learning_rate": 7.754074464347761e-05,
+      "loss": 0.7876,
+      "step": 29735
+    },
+    {
+      "epoch": 1.1449470644850819,
+      "grad_norm": 0.9994805455207825,
+      "learning_rate": 7.751127860334977e-05,
+      "loss": 0.9192,
+      "step": 29740
+    },
+    {
+      "epoch": 1.1451395572666025,
+      "grad_norm": 1.1678563356399536,
+      "learning_rate": 7.748181461966658e-05,
+      "loss": 0.8725,
+      "step": 29745
+    },
+    {
+      "epoch": 1.1453320500481232,
+      "grad_norm": 0.8279195427894592,
+      "learning_rate": 7.745235269512224e-05,
+      "loss": 0.7797,
+      "step": 29750
+    },
+    {
+      "epoch": 1.1455245428296439,
+      "grad_norm": 1.5446690320968628,
+      "learning_rate": 7.742289283241086e-05,
+      "loss": 0.7591,
+      "step": 29755
+    },
+    {
+      "epoch": 1.1457170356111646,
+      "grad_norm": 1.552101969718933,
+      "learning_rate": 7.739343503422641e-05,
+      "loss": 0.958,
+      "step": 29760
+    },
+    {
+      "epoch": 1.1459095283926852,
+      "grad_norm": 1.6127655506134033,
+      "learning_rate": 7.736397930326259e-05,
+      "loss": 0.8331,
+      "step": 29765
+    },
+    {
+      "epoch": 1.146102021174206,
+      "grad_norm": 1.63381028175354,
+      "learning_rate": 7.733452564221284e-05,
+      "loss": 1.0676,
+      "step": 29770
+    },
+    {
+      "epoch": 1.1462945139557266,
+      "grad_norm": 2.0071661472320557,
+      "learning_rate": 7.730507405377062e-05,
+      "loss": 0.8955,
+      "step": 29775
+    },
+    {
+      "epoch": 1.1464870067372472,
+      "grad_norm": 1.577379822731018,
+      "learning_rate": 7.727562454062906e-05,
+      "loss": 0.9404,
+      "step": 29780
+    },
+    {
+      "epoch": 1.1466794995187681,
+      "grad_norm": 1.0836290121078491,
+      "learning_rate": 7.724617710548104e-05,
+      "loss": 0.9939,
+      "step": 29785
+    },
+    {
+      "epoch": 1.1468719923002888,
+      "grad_norm": 1.5348353385925293,
+      "learning_rate": 7.721673175101944e-05,
+      "loss": 0.8655,
+      "step": 29790
+    },
+    {
+      "epoch": 1.1470644850818095,
+      "grad_norm": 1.7705544233322144,
+      "learning_rate": 7.718728847993679e-05,
+      "loss": 0.9213,
+      "step": 29795
+    },
+    {
+      "epoch": 1.1472569778633301,
+      "grad_norm": 1.228938341140747,
+      "learning_rate": 7.715784729492546e-05,
+      "loss": 0.8888,
+      "step": 29800
+    },
+    {
+      "epoch": 1.1474494706448508,
+      "grad_norm": 1.3519344329833984,
+      "learning_rate": 7.712840819867771e-05,
+      "loss": 0.7919,
+      "step": 29805
+    },
+    {
+      "epoch": 1.1476419634263715,
+      "grad_norm": 2.427093267440796,
+      "learning_rate": 7.709897119388553e-05,
+      "loss": 0.9861,
+      "step": 29810
+    },
+    {
+      "epoch": 1.1478344562078922,
+      "grad_norm": 1.0887788534164429,
+      "learning_rate": 7.706953628324069e-05,
+      "loss": 0.9012,
+      "step": 29815
+    },
+    {
+      "epoch": 1.1480269489894128,
+      "grad_norm": 1.7134822607040405,
+      "learning_rate": 7.704010346943488e-05,
+      "loss": 0.849,
+      "step": 29820
+    },
+    {
+      "epoch": 1.1482194417709337,
+      "grad_norm": 0.9797286987304688,
+      "learning_rate": 7.701067275515952e-05,
+      "loss": 0.8742,
+      "step": 29825
+    },
+    {
+      "epoch": 1.1484119345524544,
+      "grad_norm": 1.1243619918823242,
+      "learning_rate": 7.698124414310584e-05,
+      "loss": 0.7453,
+      "step": 29830
+    },
+    {
+      "epoch": 1.148604427333975,
+      "grad_norm": 0.9771775603294373,
+      "learning_rate": 7.695181763596489e-05,
+      "loss": 0.8668,
+      "step": 29835
+    },
+    {
+      "epoch": 1.1487969201154957,
+      "grad_norm": 1.8043612241744995,
+      "learning_rate": 7.692239323642753e-05,
+      "loss": 0.9907,
+      "step": 29840
+    },
+    {
+      "epoch": 1.1489894128970164,
+      "grad_norm": 2.071789503097534,
+      "learning_rate": 7.689297094718444e-05,
+      "loss": 0.8891,
+      "step": 29845
+    },
+    {
+      "epoch": 1.149181905678537,
+      "grad_norm": 1.8060542345046997,
+      "learning_rate": 7.686355077092611e-05,
+      "loss": 1.0377,
+      "step": 29850
+    },
+    {
+      "epoch": 1.1493743984600577,
+      "grad_norm": 1.0479457378387451,
+      "learning_rate": 7.683413271034279e-05,
+      "loss": 0.7771,
+      "step": 29855
+    },
+    {
+      "epoch": 1.1495668912415784,
+      "grad_norm": 1.4982253313064575,
+      "learning_rate": 7.680471676812453e-05,
+      "loss": 0.8573,
+      "step": 29860
+    },
+    {
+      "epoch": 1.149759384023099,
+      "grad_norm": 1.1916199922561646,
+      "learning_rate": 7.677530294696131e-05,
+      "loss": 0.726,
+      "step": 29865
+    },
+    {
+      "epoch": 1.1499518768046197,
+      "grad_norm": 1.0968197584152222,
+      "learning_rate": 7.674589124954274e-05,
+      "loss": 0.7354,
+      "step": 29870
+    },
+    {
+      "epoch": 1.1501443695861404,
+      "grad_norm": 1.3688973188400269,
+      "learning_rate": 7.671648167855842e-05,
+      "loss": 0.7691,
+      "step": 29875
+    },
+    {
+      "epoch": 1.1503368623676613,
+      "grad_norm": 2.174705743789673,
+      "learning_rate": 7.66870742366976e-05,
+      "loss": 0.8993,
+      "step": 29880
+    },
+    {
+      "epoch": 1.150529355149182,
+      "grad_norm": 1.718237280845642,
+      "learning_rate": 7.665766892664938e-05,
+      "loss": 0.8576,
+      "step": 29885
+    },
+    {
+      "epoch": 1.1507218479307026,
+      "grad_norm": 1.2384010553359985,
+      "learning_rate": 7.662826575110276e-05,
+      "loss": 0.8853,
+      "step": 29890
+    },
+    {
+      "epoch": 1.1509143407122233,
+      "grad_norm": 1.3564085960388184,
+      "learning_rate": 7.65988647127464e-05,
+      "loss": 0.7828,
+      "step": 29895
+    },
+    {
+      "epoch": 1.151106833493744,
+      "grad_norm": 1.820794701576233,
+      "learning_rate": 7.65694658142688e-05,
+      "loss": 0.893,
+      "step": 29900
+    },
+    {
+      "epoch": 1.1512993262752647,
+      "grad_norm": 2.1679587364196777,
+      "learning_rate": 7.654006905835838e-05,
+      "loss": 0.8594,
+      "step": 29905
+    },
+    {
+      "epoch": 1.1514918190567853,
+      "grad_norm": 2.045226573944092,
+      "learning_rate": 7.651067444770324e-05,
+      "loss": 1.0744,
+      "step": 29910
+    },
+    {
+      "epoch": 1.151684311838306,
+      "grad_norm": 1.4716484546661377,
+      "learning_rate": 7.648128198499128e-05,
+      "loss": 0.6354,
+      "step": 29915
+    },
+    {
+      "epoch": 1.1518768046198267,
+      "grad_norm": 2.3998184204101562,
+      "learning_rate": 7.64518916729103e-05,
+      "loss": 0.7672,
+      "step": 29920
+    },
+    {
+      "epoch": 1.1520692974013476,
+      "grad_norm": 1.5186283588409424,
+      "learning_rate": 7.642250351414786e-05,
+      "loss": 0.8835,
+      "step": 29925
+    },
+    {
+      "epoch": 1.1522617901828682,
+      "grad_norm": 1.9617729187011719,
+      "learning_rate": 7.639311751139127e-05,
+      "loss": 0.828,
+      "step": 29930
+    },
+    {
+      "epoch": 1.152454282964389,
+      "grad_norm": 0.981577455997467,
+      "learning_rate": 7.636373366732771e-05,
+      "loss": 0.8872,
+      "step": 29935
+    },
+    {
+      "epoch": 1.1526467757459096,
+      "grad_norm": 1.1515179872512817,
+      "learning_rate": 7.633435198464411e-05,
+      "loss": 0.9887,
+      "step": 29940
+    },
+    {
+      "epoch": 1.1528392685274302,
+      "grad_norm": 1.667133092880249,
+      "learning_rate": 7.630497246602725e-05,
+      "loss": 0.8211,
+      "step": 29945
+    },
+    {
+      "epoch": 1.153031761308951,
+      "grad_norm": 0.8391001224517822,
+      "learning_rate": 7.627559511416373e-05,
+      "loss": 0.8981,
+      "step": 29950
+    },
+    {
+      "epoch": 1.1532242540904716,
+      "grad_norm": 1.4484329223632812,
+      "learning_rate": 7.624621993173985e-05,
+      "loss": 0.865,
+      "step": 29955
+    },
+    {
+      "epoch": 1.1534167468719922,
+      "grad_norm": 1.7156023979187012,
+      "learning_rate": 7.621684692144178e-05,
+      "loss": 0.8699,
+      "step": 29960
+    },
+    {
+      "epoch": 1.153609239653513,
+      "grad_norm": 1.1260522603988647,
+      "learning_rate": 7.618747608595555e-05,
+      "loss": 0.9687,
+      "step": 29965
+    },
+    {
+      "epoch": 1.1538017324350336,
+      "grad_norm": 0.9241002202033997,
+      "learning_rate": 7.615810742796688e-05,
+      "loss": 0.7649,
+      "step": 29970
+    },
+    {
+      "epoch": 1.1539942252165543,
+      "grad_norm": 1.7365343570709229,
+      "learning_rate": 7.612874095016132e-05,
+      "loss": 0.9088,
+      "step": 29975
+    },
+    {
+      "epoch": 1.1541867179980752,
+      "grad_norm": 1.0506408214569092,
+      "learning_rate": 7.60993766552243e-05,
+      "loss": 0.9671,
+      "step": 29980
+    },
+    {
+      "epoch": 1.1543792107795958,
+      "grad_norm": 1.6173326969146729,
+      "learning_rate": 7.607001454584097e-05,
+      "loss": 0.9204,
+      "step": 29985
+    },
+    {
+      "epoch": 1.1545717035611165,
+      "grad_norm": 0.9955803155899048,
+      "learning_rate": 7.604065462469623e-05,
+      "loss": 0.923,
+      "step": 29990
+    },
+    {
+      "epoch": 1.1547641963426372,
+      "grad_norm": 1.820084571838379,
+      "learning_rate": 7.601129689447495e-05,
+      "loss": 0.9913,
+      "step": 29995
+    },
+    {
+      "epoch": 1.1549566891241578,
+      "grad_norm": 1.0505925416946411,
+      "learning_rate": 7.598194135786166e-05,
+      "loss": 0.8086,
+      "step": 30000
     }
   ],
   "logging_steps": 5,
@@ -28014,7 +42014,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 2,
   "save_steps": 10000,
-  "total_flos": 6.241026539658117e+17,
+  "total_flos": 9.370503974649938e+17,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null