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": 0.35021189938906927,
+  "epoch": 0.4002421707303649,
   "eval_steps": 500,
-  "global_step": 6363,
+  "global_step": 7272,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -44548,6 +44548,6369 @@
       "learning_rate": 9.280914683069837e-06,
       "loss": 0.8314,
       "step": 6363
+    },
+    {
+      "epoch": 0.35026693819142496,
+      "grad_norm": 0.8059762716293335,
+      "learning_rate": 9.280690706573633e-06,
+      "loss": 0.7695,
+      "step": 6364
+    },
+    {
+      "epoch": 0.3503219769937806,
+      "grad_norm": 0.8053386807441711,
+      "learning_rate": 9.280466697904902e-06,
+      "loss": 0.8941,
+      "step": 6365
+    },
+    {
+      "epoch": 0.3503770157961363,
+      "grad_norm": 0.6703817248344421,
+      "learning_rate": 9.280242657065329e-06,
+      "loss": 0.5978,
+      "step": 6366
+    },
+    {
+      "epoch": 0.3504320545984919,
+      "grad_norm": 0.9359784722328186,
+      "learning_rate": 9.280018584056598e-06,
+      "loss": 0.8479,
+      "step": 6367
+    },
+    {
+      "epoch": 0.3504870934008476,
+      "grad_norm": 0.7692418098449707,
+      "learning_rate": 9.279794478880393e-06,
+      "loss": 0.7254,
+      "step": 6368
+    },
+    {
+      "epoch": 0.35054213220320324,
+      "grad_norm": 0.7992031574249268,
+      "learning_rate": 9.279570341538397e-06,
+      "loss": 0.6749,
+      "step": 6369
+    },
+    {
+      "epoch": 0.3505971710055589,
+      "grad_norm": 0.7735288739204407,
+      "learning_rate": 9.279346172032297e-06,
+      "loss": 0.8545,
+      "step": 6370
+    },
+    {
+      "epoch": 0.35065220980791456,
+      "grad_norm": 0.7124339938163757,
+      "learning_rate": 9.279121970363778e-06,
+      "loss": 0.8066,
+      "step": 6371
+    },
+    {
+      "epoch": 0.35070724861027025,
+      "grad_norm": 0.8116535544395447,
+      "learning_rate": 9.278897736534521e-06,
+      "loss": 0.8197,
+      "step": 6372
+    },
+    {
+      "epoch": 0.3507622874126259,
+      "grad_norm": 0.9377869963645935,
+      "learning_rate": 9.278673470546217e-06,
+      "loss": 0.74,
+      "step": 6373
+    },
+    {
+      "epoch": 0.3508173262149816,
+      "grad_norm": 0.6726253628730774,
+      "learning_rate": 9.278449172400548e-06,
+      "loss": 0.6389,
+      "step": 6374
+    },
+    {
+      "epoch": 0.3508723650173372,
+      "grad_norm": 0.8470593094825745,
+      "learning_rate": 9.278224842099198e-06,
+      "loss": 0.8059,
+      "step": 6375
+    },
+    {
+      "epoch": 0.3509274038196929,
+      "grad_norm": 0.7041867971420288,
+      "learning_rate": 9.278000479643857e-06,
+      "loss": 0.7409,
+      "step": 6376
+    },
+    {
+      "epoch": 0.35098244262204853,
+      "grad_norm": 0.7467322945594788,
+      "learning_rate": 9.27777608503621e-06,
+      "loss": 0.823,
+      "step": 6377
+    },
+    {
+      "epoch": 0.3510374814244042,
+      "grad_norm": 0.7211065888404846,
+      "learning_rate": 9.277551658277942e-06,
+      "loss": 0.7655,
+      "step": 6378
+    },
+    {
+      "epoch": 0.35109252022675985,
+      "grad_norm": 0.7709450125694275,
+      "learning_rate": 9.27732719937074e-06,
+      "loss": 0.8938,
+      "step": 6379
+    },
+    {
+      "epoch": 0.35114755902911554,
+      "grad_norm": 0.7672929167747498,
+      "learning_rate": 9.277102708316293e-06,
+      "loss": 0.6814,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3512025978314712,
+      "grad_norm": 0.7334907650947571,
+      "learning_rate": 9.276878185116287e-06,
+      "loss": 0.6608,
+      "step": 6381
+    },
+    {
+      "epoch": 0.35125763663382686,
+      "grad_norm": 0.7011460065841675,
+      "learning_rate": 9.27665362977241e-06,
+      "loss": 0.8196,
+      "step": 6382
+    },
+    {
+      "epoch": 0.3513126754361825,
+      "grad_norm": 0.7388820052146912,
+      "learning_rate": 9.276429042286349e-06,
+      "loss": 0.8793,
+      "step": 6383
+    },
+    {
+      "epoch": 0.3513677142385382,
+      "grad_norm": 0.809725821018219,
+      "learning_rate": 9.27620442265979e-06,
+      "loss": 0.6976,
+      "step": 6384
+    },
+    {
+      "epoch": 0.3514227530408938,
+      "grad_norm": 0.6933012008666992,
+      "learning_rate": 9.275979770894424e-06,
+      "loss": 0.759,
+      "step": 6385
+    },
+    {
+      "epoch": 0.3514777918432495,
+      "grad_norm": 0.7928480505943298,
+      "learning_rate": 9.27575508699194e-06,
+      "loss": 0.7462,
+      "step": 6386
+    },
+    {
+      "epoch": 0.35153283064560514,
+      "grad_norm": 0.8461304903030396,
+      "learning_rate": 9.275530370954024e-06,
+      "loss": 0.8184,
+      "step": 6387
+    },
+    {
+      "epoch": 0.35158786944796083,
+      "grad_norm": 0.7624425292015076,
+      "learning_rate": 9.275305622782366e-06,
+      "loss": 0.7913,
+      "step": 6388
+    },
+    {
+      "epoch": 0.35164290825031647,
+      "grad_norm": 0.7103675007820129,
+      "learning_rate": 9.275080842478657e-06,
+      "loss": 0.7633,
+      "step": 6389
+    },
+    {
+      "epoch": 0.35169794705267216,
+      "grad_norm": 0.9002664089202881,
+      "learning_rate": 9.274856030044583e-06,
+      "loss": 0.7643,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3517529858550278,
+      "grad_norm": 0.7658692002296448,
+      "learning_rate": 9.274631185481836e-06,
+      "loss": 0.8028,
+      "step": 6391
+    },
+    {
+      "epoch": 0.3518080246573835,
+      "grad_norm": 0.6747875809669495,
+      "learning_rate": 9.274406308792106e-06,
+      "loss": 0.695,
+      "step": 6392
+    },
+    {
+      "epoch": 0.3518630634597391,
+      "grad_norm": 0.8197165131568909,
+      "learning_rate": 9.27418139997708e-06,
+      "loss": 0.7218,
+      "step": 6393
+    },
+    {
+      "epoch": 0.3519181022620948,
+      "grad_norm": 0.7597750425338745,
+      "learning_rate": 9.273956459038453e-06,
+      "loss": 0.7738,
+      "step": 6394
+    },
+    {
+      "epoch": 0.35197314106445043,
+      "grad_norm": 0.7365928888320923,
+      "learning_rate": 9.273731485977912e-06,
+      "loss": 0.7906,
+      "step": 6395
+    },
+    {
+      "epoch": 0.3520281798668061,
+      "grad_norm": 0.7313928604125977,
+      "learning_rate": 9.273506480797151e-06,
+      "loss": 0.834,
+      "step": 6396
+    },
+    {
+      "epoch": 0.35208321866916176,
+      "grad_norm": 0.758886456489563,
+      "learning_rate": 9.273281443497858e-06,
+      "loss": 0.8883,
+      "step": 6397
+    },
+    {
+      "epoch": 0.35213825747151745,
+      "grad_norm": 0.7318256497383118,
+      "learning_rate": 9.273056374081726e-06,
+      "loss": 0.7463,
+      "step": 6398
+    },
+    {
+      "epoch": 0.3521932962738731,
+      "grad_norm": 0.778448224067688,
+      "learning_rate": 9.272831272550446e-06,
+      "loss": 0.6838,
+      "step": 6399
+    },
+    {
+      "epoch": 0.3522483350762287,
+      "grad_norm": 0.7392274141311646,
+      "learning_rate": 9.272606138905709e-06,
+      "loss": 0.7237,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3523033738785844,
+      "grad_norm": 0.8803032040596008,
+      "learning_rate": 9.272380973149209e-06,
+      "loss": 0.7839,
+      "step": 6401
+    },
+    {
+      "epoch": 0.35235841268094004,
+      "grad_norm": 0.7506754994392395,
+      "learning_rate": 9.272155775282636e-06,
+      "loss": 0.7665,
+      "step": 6402
+    },
+    {
+      "epoch": 0.3524134514832957,
+      "grad_norm": 0.8136595487594604,
+      "learning_rate": 9.271930545307686e-06,
+      "loss": 0.9111,
+      "step": 6403
+    },
+    {
+      "epoch": 0.35246849028565136,
+      "grad_norm": 0.7976880073547363,
+      "learning_rate": 9.271705283226047e-06,
+      "loss": 0.735,
+      "step": 6404
+    },
+    {
+      "epoch": 0.35252352908800705,
+      "grad_norm": 0.89708411693573,
+      "learning_rate": 9.271479989039415e-06,
+      "loss": 0.7698,
+      "step": 6405
+    },
+    {
+      "epoch": 0.3525785678903627,
+      "grad_norm": 0.8618703484535217,
+      "learning_rate": 9.271254662749484e-06,
+      "loss": 0.9001,
+      "step": 6406
+    },
+    {
+      "epoch": 0.35263360669271837,
+      "grad_norm": 0.7143027186393738,
+      "learning_rate": 9.271029304357946e-06,
+      "loss": 0.8188,
+      "step": 6407
+    },
+    {
+      "epoch": 0.352688645495074,
+      "grad_norm": 0.795365571975708,
+      "learning_rate": 9.270803913866496e-06,
+      "loss": 0.7389,
+      "step": 6408
+    },
+    {
+      "epoch": 0.3527436842974297,
+      "grad_norm": 0.6947643756866455,
+      "learning_rate": 9.270578491276825e-06,
+      "loss": 0.7278,
+      "step": 6409
+    },
+    {
+      "epoch": 0.3527987230997853,
+      "grad_norm": 0.7806137204170227,
+      "learning_rate": 9.27035303659063e-06,
+      "loss": 0.808,
+      "step": 6410
+    },
+    {
+      "epoch": 0.352853761902141,
+      "grad_norm": 0.8908704519271851,
+      "learning_rate": 9.270127549809606e-06,
+      "loss": 0.8659,
+      "step": 6411
+    },
+    {
+      "epoch": 0.35290880070449665,
+      "grad_norm": 0.8171417713165283,
+      "learning_rate": 9.269902030935445e-06,
+      "loss": 0.7918,
+      "step": 6412
+    },
+    {
+      "epoch": 0.35296383950685234,
+      "grad_norm": 0.7556712627410889,
+      "learning_rate": 9.269676479969842e-06,
+      "loss": 0.7121,
+      "step": 6413
+    },
+    {
+      "epoch": 0.353018878309208,
+      "grad_norm": 0.8080483675003052,
+      "learning_rate": 9.269450896914495e-06,
+      "loss": 0.8185,
+      "step": 6414
+    },
+    {
+      "epoch": 0.35307391711156366,
+      "grad_norm": 0.8514583706855774,
+      "learning_rate": 9.2692252817711e-06,
+      "loss": 0.8055,
+      "step": 6415
+    },
+    {
+      "epoch": 0.3531289559139193,
+      "grad_norm": 0.7914162278175354,
+      "learning_rate": 9.268999634541347e-06,
+      "loss": 0.759,
+      "step": 6416
+    },
+    {
+      "epoch": 0.353183994716275,
+      "grad_norm": 0.6452118754386902,
+      "learning_rate": 9.268773955226937e-06,
+      "loss": 0.6797,
+      "step": 6417
+    },
+    {
+      "epoch": 0.3532390335186306,
+      "grad_norm": 0.6876220107078552,
+      "learning_rate": 9.268548243829565e-06,
+      "loss": 0.7365,
+      "step": 6418
+    },
+    {
+      "epoch": 0.3532940723209863,
+      "grad_norm": 0.758550226688385,
+      "learning_rate": 9.268322500350926e-06,
+      "loss": 0.7069,
+      "step": 6419
+    },
+    {
+      "epoch": 0.35334911112334194,
+      "grad_norm": 0.7905879020690918,
+      "learning_rate": 9.268096724792718e-06,
+      "loss": 0.8024,
+      "step": 6420
+    },
+    {
+      "epoch": 0.35340414992569763,
+      "grad_norm": 0.755253255367279,
+      "learning_rate": 9.267870917156638e-06,
+      "loss": 0.8018,
+      "step": 6421
+    },
+    {
+      "epoch": 0.35345918872805326,
+      "grad_norm": 0.6879923343658447,
+      "learning_rate": 9.267645077444382e-06,
+      "loss": 0.7267,
+      "step": 6422
+    },
+    {
+      "epoch": 0.35351422753040895,
+      "grad_norm": 0.766214907169342,
+      "learning_rate": 9.267419205657649e-06,
+      "loss": 0.7801,
+      "step": 6423
+    },
+    {
+      "epoch": 0.3535692663327646,
+      "grad_norm": 0.868776798248291,
+      "learning_rate": 9.267193301798135e-06,
+      "loss": 0.9234,
+      "step": 6424
+    },
+    {
+      "epoch": 0.3536243051351203,
+      "grad_norm": 1.2007492780685425,
+      "learning_rate": 9.266967365867536e-06,
+      "loss": 0.7743,
+      "step": 6425
+    },
+    {
+      "epoch": 0.3536793439374759,
+      "grad_norm": 0.7445551156997681,
+      "learning_rate": 9.266741397867556e-06,
+      "loss": 0.6755,
+      "step": 6426
+    },
+    {
+      "epoch": 0.3537343827398316,
+      "grad_norm": 0.7493785619735718,
+      "learning_rate": 9.266515397799889e-06,
+      "loss": 0.7891,
+      "step": 6427
+    },
+    {
+      "epoch": 0.35378942154218723,
+      "grad_norm": 0.6718230843544006,
+      "learning_rate": 9.266289365666234e-06,
+      "loss": 0.6908,
+      "step": 6428
+    },
+    {
+      "epoch": 0.3538444603445429,
+      "grad_norm": 0.7783547639846802,
+      "learning_rate": 9.266063301468289e-06,
+      "loss": 0.7115,
+      "step": 6429
+    },
+    {
+      "epoch": 0.35389949914689856,
+      "grad_norm": 0.745627224445343,
+      "learning_rate": 9.265837205207755e-06,
+      "loss": 0.8421,
+      "step": 6430
+    },
+    {
+      "epoch": 0.35395453794925424,
+      "grad_norm": 0.7314152717590332,
+      "learning_rate": 9.26561107688633e-06,
+      "loss": 0.807,
+      "step": 6431
+    },
+    {
+      "epoch": 0.3540095767516099,
+      "grad_norm": 0.6975863575935364,
+      "learning_rate": 9.265384916505714e-06,
+      "loss": 0.7787,
+      "step": 6432
+    },
+    {
+      "epoch": 0.35406461555396557,
+      "grad_norm": 0.9758319854736328,
+      "learning_rate": 9.265158724067608e-06,
+      "loss": 0.8668,
+      "step": 6433
+    },
+    {
+      "epoch": 0.3541196543563212,
+      "grad_norm": 0.7686764001846313,
+      "learning_rate": 9.264932499573711e-06,
+      "loss": 0.7428,
+      "step": 6434
+    },
+    {
+      "epoch": 0.3541746931586769,
+      "grad_norm": 0.8761935830116272,
+      "learning_rate": 9.26470624302572e-06,
+      "loss": 0.8022,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3542297319610325,
+      "grad_norm": 0.9145118594169617,
+      "learning_rate": 9.264479954425341e-06,
+      "loss": 0.7994,
+      "step": 6436
+    },
+    {
+      "epoch": 0.3542847707633882,
+      "grad_norm": 0.8217951655387878,
+      "learning_rate": 9.264253633774271e-06,
+      "loss": 0.7235,
+      "step": 6437
+    },
+    {
+      "epoch": 0.35433980956574385,
+      "grad_norm": 0.7624716758728027,
+      "learning_rate": 9.264027281074214e-06,
+      "loss": 0.8238,
+      "step": 6438
+    },
+    {
+      "epoch": 0.35439484836809954,
+      "grad_norm": 0.7772085070610046,
+      "learning_rate": 9.26380089632687e-06,
+      "loss": 0.7941,
+      "step": 6439
+    },
+    {
+      "epoch": 0.35444988717045517,
+      "grad_norm": 1.0462371110916138,
+      "learning_rate": 9.263574479533937e-06,
+      "loss": 0.8255,
+      "step": 6440
+    },
+    {
+      "epoch": 0.35450492597281086,
+      "grad_norm": 0.8523101210594177,
+      "learning_rate": 9.263348030697119e-06,
+      "loss": 0.8489,
+      "step": 6441
+    },
+    {
+      "epoch": 0.3545599647751665,
+      "grad_norm": 1.0292255878448486,
+      "learning_rate": 9.26312154981812e-06,
+      "loss": 0.7989,
+      "step": 6442
+    },
+    {
+      "epoch": 0.3546150035775221,
+      "grad_norm": 0.7621143460273743,
+      "learning_rate": 9.262895036898641e-06,
+      "loss": 0.8154,
+      "step": 6443
+    },
+    {
+      "epoch": 0.3546700423798778,
+      "grad_norm": 0.7158074378967285,
+      "learning_rate": 9.262668491940382e-06,
+      "loss": 0.7821,
+      "step": 6444
+    },
+    {
+      "epoch": 0.35472508118223345,
+      "grad_norm": 0.7969478964805603,
+      "learning_rate": 9.26244191494505e-06,
+      "loss": 0.8535,
+      "step": 6445
+    },
+    {
+      "epoch": 0.35478011998458914,
+      "grad_norm": 0.9244762063026428,
+      "learning_rate": 9.262215305914345e-06,
+      "loss": 0.7585,
+      "step": 6446
+    },
+    {
+      "epoch": 0.35483515878694477,
+      "grad_norm": 0.6862454414367676,
+      "learning_rate": 9.26198866484997e-06,
+      "loss": 0.7294,
+      "step": 6447
+    },
+    {
+      "epoch": 0.35489019758930046,
+      "grad_norm": 0.6816834211349487,
+      "learning_rate": 9.261761991753629e-06,
+      "loss": 0.7763,
+      "step": 6448
+    },
+    {
+      "epoch": 0.3549452363916561,
+      "grad_norm": 0.792539119720459,
+      "learning_rate": 9.261535286627025e-06,
+      "loss": 0.7829,
+      "step": 6449
+    },
+    {
+      "epoch": 0.3550002751940118,
+      "grad_norm": 0.8563211560249329,
+      "learning_rate": 9.261308549471866e-06,
+      "loss": 0.8945,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3550553139963674,
+      "grad_norm": 0.7241078019142151,
+      "learning_rate": 9.26108178028985e-06,
+      "loss": 0.6936,
+      "step": 6451
+    },
+    {
+      "epoch": 0.3551103527987231,
+      "grad_norm": 0.7150034308433533,
+      "learning_rate": 9.260854979082682e-06,
+      "loss": 0.7689,
+      "step": 6452
+    },
+    {
+      "epoch": 0.35516539160107874,
+      "grad_norm": 0.8630193471908569,
+      "learning_rate": 9.260628145852073e-06,
+      "loss": 0.8506,
+      "step": 6453
+    },
+    {
+      "epoch": 0.35522043040343443,
+      "grad_norm": 0.7133893370628357,
+      "learning_rate": 9.26040128059972e-06,
+      "loss": 0.7976,
+      "step": 6454
+    },
+    {
+      "epoch": 0.35527546920579006,
+      "grad_norm": 0.6984630823135376,
+      "learning_rate": 9.260174383327332e-06,
+      "loss": 0.7442,
+      "step": 6455
+    },
+    {
+      "epoch": 0.35533050800814575,
+      "grad_norm": 0.7166933417320251,
+      "learning_rate": 9.259947454036613e-06,
+      "loss": 0.813,
+      "step": 6456
+    },
+    {
+      "epoch": 0.3553855468105014,
+      "grad_norm": 0.7353581190109253,
+      "learning_rate": 9.259720492729272e-06,
+      "loss": 0.8157,
+      "step": 6457
+    },
+    {
+      "epoch": 0.3554405856128571,
+      "grad_norm": 0.6810038089752197,
+      "learning_rate": 9.259493499407011e-06,
+      "loss": 0.7423,
+      "step": 6458
+    },
+    {
+      "epoch": 0.3554956244152127,
+      "grad_norm": 1.1599586009979248,
+      "learning_rate": 9.259266474071535e-06,
+      "loss": 0.7159,
+      "step": 6459
+    },
+    {
+      "epoch": 0.3555506632175684,
+      "grad_norm": 0.7857629060745239,
+      "learning_rate": 9.259039416724554e-06,
+      "loss": 0.7846,
+      "step": 6460
+    },
+    {
+      "epoch": 0.35560570201992403,
+      "grad_norm": 0.705333948135376,
+      "learning_rate": 9.258812327367773e-06,
+      "loss": 0.751,
+      "step": 6461
+    },
+    {
+      "epoch": 0.3556607408222797,
+      "grad_norm": 0.6899998188018799,
+      "learning_rate": 9.258585206002897e-06,
+      "loss": 0.7303,
+      "step": 6462
+    },
+    {
+      "epoch": 0.35571577962463535,
+      "grad_norm": 0.8007912039756775,
+      "learning_rate": 9.258358052631637e-06,
+      "loss": 0.7363,
+      "step": 6463
+    },
+    {
+      "epoch": 0.35577081842699104,
+      "grad_norm": 0.9403146505355835,
+      "learning_rate": 9.258130867255695e-06,
+      "loss": 0.9096,
+      "step": 6464
+    },
+    {
+      "epoch": 0.3558258572293467,
+      "grad_norm": 0.7069174647331238,
+      "learning_rate": 9.257903649876782e-06,
+      "loss": 0.7362,
+      "step": 6465
+    },
+    {
+      "epoch": 0.35588089603170237,
+      "grad_norm": 0.770807683467865,
+      "learning_rate": 9.257676400496607e-06,
+      "loss": 0.7904,
+      "step": 6466
+    },
+    {
+      "epoch": 0.355935934834058,
+      "grad_norm": 0.8586871027946472,
+      "learning_rate": 9.257449119116874e-06,
+      "loss": 0.7596,
+      "step": 6467
+    },
+    {
+      "epoch": 0.3559909736364137,
+      "grad_norm": 0.6934101581573486,
+      "learning_rate": 9.257221805739294e-06,
+      "loss": 0.6655,
+      "step": 6468
+    },
+    {
+      "epoch": 0.3560460124387693,
+      "grad_norm": 0.9494497179985046,
+      "learning_rate": 9.256994460365573e-06,
+      "loss": 0.7923,
+      "step": 6469
+    },
+    {
+      "epoch": 0.356101051241125,
+      "grad_norm": 0.7131130695343018,
+      "learning_rate": 9.256767082997422e-06,
+      "loss": 0.819,
+      "step": 6470
+    },
+    {
+      "epoch": 0.35615609004348064,
+      "grad_norm": 0.8641398549079895,
+      "learning_rate": 9.25653967363655e-06,
+      "loss": 0.8275,
+      "step": 6471
+    },
+    {
+      "epoch": 0.35621112884583633,
+      "grad_norm": 0.7350367307662964,
+      "learning_rate": 9.256312232284665e-06,
+      "loss": 0.7991,
+      "step": 6472
+    },
+    {
+      "epoch": 0.35626616764819197,
+      "grad_norm": 0.8174671530723572,
+      "learning_rate": 9.256084758943476e-06,
+      "loss": 0.7147,
+      "step": 6473
+    },
+    {
+      "epoch": 0.35632120645054766,
+      "grad_norm": 0.7560263872146606,
+      "learning_rate": 9.255857253614693e-06,
+      "loss": 0.7435,
+      "step": 6474
+    },
+    {
+      "epoch": 0.3563762452529033,
+      "grad_norm": 0.7465197443962097,
+      "learning_rate": 9.255629716300025e-06,
+      "loss": 0.8228,
+      "step": 6475
+    },
+    {
+      "epoch": 0.356431284055259,
+      "grad_norm": 0.7130733728408813,
+      "learning_rate": 9.255402147001184e-06,
+      "loss": 0.8361,
+      "step": 6476
+    },
+    {
+      "epoch": 0.3564863228576146,
+      "grad_norm": 0.7200759053230286,
+      "learning_rate": 9.255174545719882e-06,
+      "loss": 0.7387,
+      "step": 6477
+    },
+    {
+      "epoch": 0.3565413616599703,
+      "grad_norm": 0.8387622237205505,
+      "learning_rate": 9.254946912457826e-06,
+      "loss": 0.8427,
+      "step": 6478
+    },
+    {
+      "epoch": 0.35659640046232594,
+      "grad_norm": 0.7263510823249817,
+      "learning_rate": 9.254719247216725e-06,
+      "loss": 0.712,
+      "step": 6479
+    },
+    {
+      "epoch": 0.3566514392646816,
+      "grad_norm": 0.7393862009048462,
+      "learning_rate": 9.254491549998296e-06,
+      "loss": 0.6916,
+      "step": 6480
+    },
+    {
+      "epoch": 0.35670647806703726,
+      "grad_norm": 0.7289569973945618,
+      "learning_rate": 9.254263820804246e-06,
+      "loss": 0.7561,
+      "step": 6481
+    },
+    {
+      "epoch": 0.35676151686939295,
+      "grad_norm": 0.7597448825836182,
+      "learning_rate": 9.254036059636288e-06,
+      "loss": 0.853,
+      "step": 6482
+    },
+    {
+      "epoch": 0.3568165556717486,
+      "grad_norm": 0.7652063369750977,
+      "learning_rate": 9.253808266496136e-06,
+      "loss": 0.7652,
+      "step": 6483
+    },
+    {
+      "epoch": 0.35687159447410427,
+      "grad_norm": 1.193938136100769,
+      "learning_rate": 9.253580441385497e-06,
+      "loss": 0.8288,
+      "step": 6484
+    },
+    {
+      "epoch": 0.3569266332764599,
+      "grad_norm": 0.9258719086647034,
+      "learning_rate": 9.253352584306087e-06,
+      "loss": 0.807,
+      "step": 6485
+    },
+    {
+      "epoch": 0.35698167207881554,
+      "grad_norm": 0.78384929895401,
+      "learning_rate": 9.253124695259617e-06,
+      "loss": 0.7785,
+      "step": 6486
+    },
+    {
+      "epoch": 0.3570367108811712,
+      "grad_norm": 0.801403284072876,
+      "learning_rate": 9.252896774247802e-06,
+      "loss": 0.8382,
+      "step": 6487
+    },
+    {
+      "epoch": 0.35709174968352686,
+      "grad_norm": 0.9472376108169556,
+      "learning_rate": 9.25266882127235e-06,
+      "loss": 0.8661,
+      "step": 6488
+    },
+    {
+      "epoch": 0.35714678848588255,
+      "grad_norm": 0.7575686573982239,
+      "learning_rate": 9.252440836334981e-06,
+      "loss": 0.8428,
+      "step": 6489
+    },
+    {
+      "epoch": 0.3572018272882382,
+      "grad_norm": 0.736282467842102,
+      "learning_rate": 9.252212819437402e-06,
+      "loss": 0.801,
+      "step": 6490
+    },
+    {
+      "epoch": 0.35725686609059387,
+      "grad_norm": 0.7420864701271057,
+      "learning_rate": 9.251984770581332e-06,
+      "loss": 0.8849,
+      "step": 6491
+    },
+    {
+      "epoch": 0.3573119048929495,
+      "grad_norm": 0.7129189372062683,
+      "learning_rate": 9.251756689768482e-06,
+      "loss": 0.7716,
+      "step": 6492
+    },
+    {
+      "epoch": 0.3573669436953052,
+      "grad_norm": 0.7777297496795654,
+      "learning_rate": 9.251528577000566e-06,
+      "loss": 0.8183,
+      "step": 6493
+    },
+    {
+      "epoch": 0.35742198249766083,
+      "grad_norm": 0.7644590139389038,
+      "learning_rate": 9.2513004322793e-06,
+      "loss": 0.6319,
+      "step": 6494
+    },
+    {
+      "epoch": 0.3574770213000165,
+      "grad_norm": 0.7112484574317932,
+      "learning_rate": 9.251072255606399e-06,
+      "loss": 0.8012,
+      "step": 6495
+    },
+    {
+      "epoch": 0.35753206010237215,
+      "grad_norm": 0.7772265076637268,
+      "learning_rate": 9.250844046983576e-06,
+      "loss": 0.8372,
+      "step": 6496
+    },
+    {
+      "epoch": 0.35758709890472784,
+      "grad_norm": 0.9530157446861267,
+      "learning_rate": 9.250615806412546e-06,
+      "loss": 0.8683,
+      "step": 6497
+    },
+    {
+      "epoch": 0.3576421377070835,
+      "grad_norm": 0.7249575257301331,
+      "learning_rate": 9.250387533895026e-06,
+      "loss": 0.7091,
+      "step": 6498
+    },
+    {
+      "epoch": 0.35769717650943916,
+      "grad_norm": 0.8549422025680542,
+      "learning_rate": 9.25015922943273e-06,
+      "loss": 0.8376,
+      "step": 6499
+    },
+    {
+      "epoch": 0.3577522153117948,
+      "grad_norm": 0.74477618932724,
+      "learning_rate": 9.249930893027376e-06,
+      "loss": 0.7594,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3578072541141505,
+      "grad_norm": 0.8269739151000977,
+      "learning_rate": 9.24970252468068e-06,
+      "loss": 0.6473,
+      "step": 6501
+    },
+    {
+      "epoch": 0.3578622929165061,
+      "grad_norm": 0.8375437259674072,
+      "learning_rate": 9.249474124394358e-06,
+      "loss": 0.7631,
+      "step": 6502
+    },
+    {
+      "epoch": 0.3579173317188618,
+      "grad_norm": 0.8680340051651001,
+      "learning_rate": 9.249245692170123e-06,
+      "loss": 0.7863,
+      "step": 6503
+    },
+    {
+      "epoch": 0.35797237052121744,
+      "grad_norm": 0.7179692983627319,
+      "learning_rate": 9.249017228009696e-06,
+      "loss": 0.8022,
+      "step": 6504
+    },
+    {
+      "epoch": 0.35802740932357313,
+      "grad_norm": 0.7797464728355408,
+      "learning_rate": 9.248788731914794e-06,
+      "loss": 0.8067,
+      "step": 6505
+    },
+    {
+      "epoch": 0.35808244812592877,
+      "grad_norm": 0.8032993674278259,
+      "learning_rate": 9.248560203887133e-06,
+      "loss": 0.7383,
+      "step": 6506
+    },
+    {
+      "epoch": 0.35813748692828445,
+      "grad_norm": 0.7714722156524658,
+      "learning_rate": 9.24833164392843e-06,
+      "loss": 0.7149,
+      "step": 6507
+    },
+    {
+      "epoch": 0.3581925257306401,
+      "grad_norm": 0.7492430210113525,
+      "learning_rate": 9.248103052040404e-06,
+      "loss": 0.7645,
+      "step": 6508
+    },
+    {
+      "epoch": 0.3582475645329958,
+      "grad_norm": 0.6843901872634888,
+      "learning_rate": 9.247874428224773e-06,
+      "loss": 0.7183,
+      "step": 6509
+    },
+    {
+      "epoch": 0.3583026033353514,
+      "grad_norm": 0.8370186686515808,
+      "learning_rate": 9.247645772483254e-06,
+      "loss": 0.7832,
+      "step": 6510
+    },
+    {
+      "epoch": 0.3583576421377071,
+      "grad_norm": 0.7907791137695312,
+      "learning_rate": 9.247417084817567e-06,
+      "loss": 0.8742,
+      "step": 6511
+    },
+    {
+      "epoch": 0.35841268094006273,
+      "grad_norm": 0.7950869798660278,
+      "learning_rate": 9.247188365229428e-06,
+      "loss": 0.8705,
+      "step": 6512
+    },
+    {
+      "epoch": 0.3584677197424184,
+      "grad_norm": 0.7276936173439026,
+      "learning_rate": 9.24695961372056e-06,
+      "loss": 0.7629,
+      "step": 6513
+    },
+    {
+      "epoch": 0.35852275854477406,
+      "grad_norm": 0.7761141657829285,
+      "learning_rate": 9.24673083029268e-06,
+      "loss": 0.8813,
+      "step": 6514
+    },
+    {
+      "epoch": 0.35857779734712975,
+      "grad_norm": 0.7528283596038818,
+      "learning_rate": 9.24650201494751e-06,
+      "loss": 0.7885,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3586328361494854,
+      "grad_norm": 0.8972534537315369,
+      "learning_rate": 9.246273167686765e-06,
+      "loss": 0.9081,
+      "step": 6516
+    },
+    {
+      "epoch": 0.35868787495184107,
+      "grad_norm": 0.7658557891845703,
+      "learning_rate": 9.246044288512168e-06,
+      "loss": 0.8451,
+      "step": 6517
+    },
+    {
+      "epoch": 0.3587429137541967,
+      "grad_norm": 0.8013193607330322,
+      "learning_rate": 9.245815377425438e-06,
+      "loss": 0.7236,
+      "step": 6518
+    },
+    {
+      "epoch": 0.3587979525565524,
+      "grad_norm": 0.8134163022041321,
+      "learning_rate": 9.245586434428298e-06,
+      "loss": 0.908,
+      "step": 6519
+    },
+    {
+      "epoch": 0.358852991358908,
+      "grad_norm": 0.6479801535606384,
+      "learning_rate": 9.245357459522466e-06,
+      "loss": 0.7397,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3589080301612637,
+      "grad_norm": 0.70014488697052,
+      "learning_rate": 9.245128452709665e-06,
+      "loss": 0.6898,
+      "step": 6521
+    },
+    {
+      "epoch": 0.35896306896361935,
+      "grad_norm": 0.7645437717437744,
+      "learning_rate": 9.244899413991613e-06,
+      "loss": 0.8319,
+      "step": 6522
+    },
+    {
+      "epoch": 0.35901810776597504,
+      "grad_norm": 0.6812799572944641,
+      "learning_rate": 9.244670343370033e-06,
+      "loss": 0.7359,
+      "step": 6523
+    },
+    {
+      "epoch": 0.35907314656833067,
+      "grad_norm": 0.6573774218559265,
+      "learning_rate": 9.244441240846647e-06,
+      "loss": 0.742,
+      "step": 6524
+    },
+    {
+      "epoch": 0.35912818537068636,
+      "grad_norm": 0.7870661020278931,
+      "learning_rate": 9.244212106423178e-06,
+      "loss": 0.7307,
+      "step": 6525
+    },
+    {
+      "epoch": 0.359183224173042,
+      "grad_norm": 0.9163166284561157,
+      "learning_rate": 9.243982940101347e-06,
+      "loss": 0.8584,
+      "step": 6526
+    },
+    {
+      "epoch": 0.3592382629753977,
+      "grad_norm": 0.766888439655304,
+      "learning_rate": 9.243753741882874e-06,
+      "loss": 0.8093,
+      "step": 6527
+    },
+    {
+      "epoch": 0.3592933017777533,
+      "grad_norm": 0.7831236124038696,
+      "learning_rate": 9.243524511769486e-06,
+      "loss": 0.8665,
+      "step": 6528
+    },
+    {
+      "epoch": 0.35934834058010895,
+      "grad_norm": 0.7485133409500122,
+      "learning_rate": 9.243295249762904e-06,
+      "loss": 0.7336,
+      "step": 6529
+    },
+    {
+      "epoch": 0.35940337938246464,
+      "grad_norm": 0.7231502532958984,
+      "learning_rate": 9.24306595586485e-06,
+      "loss": 0.8095,
+      "step": 6530
+    },
+    {
+      "epoch": 0.35945841818482027,
+      "grad_norm": 0.821898877620697,
+      "learning_rate": 9.242836630077048e-06,
+      "loss": 0.831,
+      "step": 6531
+    },
+    {
+      "epoch": 0.35951345698717596,
+      "grad_norm": 0.6792737245559692,
+      "learning_rate": 9.242607272401223e-06,
+      "loss": 0.7183,
+      "step": 6532
+    },
+    {
+      "epoch": 0.3595684957895316,
+      "grad_norm": 0.7200430631637573,
+      "learning_rate": 9.242377882839095e-06,
+      "loss": 0.7256,
+      "step": 6533
+    },
+    {
+      "epoch": 0.3596235345918873,
+      "grad_norm": 0.6713700890541077,
+      "learning_rate": 9.242148461392393e-06,
+      "loss": 0.7416,
+      "step": 6534
+    },
+    {
+      "epoch": 0.3596785733942429,
+      "grad_norm": 0.7054564356803894,
+      "learning_rate": 9.241919008062836e-06,
+      "loss": 0.6856,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3597336121965986,
+      "grad_norm": 0.7516196966171265,
+      "learning_rate": 9.241689522852152e-06,
+      "loss": 0.7149,
+      "step": 6536
+    },
+    {
+      "epoch": 0.35978865099895424,
+      "grad_norm": 0.8547651767730713,
+      "learning_rate": 9.241460005762067e-06,
+      "loss": 0.7075,
+      "step": 6537
+    },
+    {
+      "epoch": 0.35984368980130993,
+      "grad_norm": 0.6791819334030151,
+      "learning_rate": 9.241230456794302e-06,
+      "loss": 0.6449,
+      "step": 6538
+    },
+    {
+      "epoch": 0.35989872860366556,
+      "grad_norm": 0.8365122079849243,
+      "learning_rate": 9.241000875950583e-06,
+      "loss": 0.7619,
+      "step": 6539
+    },
+    {
+      "epoch": 0.35995376740602125,
+      "grad_norm": 0.763829231262207,
+      "learning_rate": 9.24077126323264e-06,
+      "loss": 0.71,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3600088062083769,
+      "grad_norm": 0.7698483467102051,
+      "learning_rate": 9.240541618642193e-06,
+      "loss": 0.7949,
+      "step": 6541
+    },
+    {
+      "epoch": 0.3600638450107326,
+      "grad_norm": 0.7331508994102478,
+      "learning_rate": 9.24031194218097e-06,
+      "loss": 0.8292,
+      "step": 6542
+    },
+    {
+      "epoch": 0.3601188838130882,
+      "grad_norm": 0.7507451772689819,
+      "learning_rate": 9.2400822338507e-06,
+      "loss": 0.8651,
+      "step": 6543
+    },
+    {
+      "epoch": 0.3601739226154439,
+      "grad_norm": 0.8537001609802246,
+      "learning_rate": 9.239852493653104e-06,
+      "loss": 0.848,
+      "step": 6544
+    },
+    {
+      "epoch": 0.36022896141779953,
+      "grad_norm": 0.683311939239502,
+      "learning_rate": 9.239622721589913e-06,
+      "loss": 0.803,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3602840002201552,
+      "grad_norm": 0.6916974186897278,
+      "learning_rate": 9.239392917662852e-06,
+      "loss": 0.8037,
+      "step": 6546
+    },
+    {
+      "epoch": 0.36033903902251085,
+      "grad_norm": 0.798795223236084,
+      "learning_rate": 9.23916308187365e-06,
+      "loss": 0.8037,
+      "step": 6547
+    },
+    {
+      "epoch": 0.36039407782486654,
+      "grad_norm": 0.7284069657325745,
+      "learning_rate": 9.238933214224032e-06,
+      "loss": 0.7365,
+      "step": 6548
+    },
+    {
+      "epoch": 0.3604491166272222,
+      "grad_norm": 0.7789250016212463,
+      "learning_rate": 9.238703314715727e-06,
+      "loss": 0.788,
+      "step": 6549
+    },
+    {
+      "epoch": 0.36050415542957787,
+      "grad_norm": 0.7029675841331482,
+      "learning_rate": 9.238473383350462e-06,
+      "loss": 0.7796,
+      "step": 6550
+    },
+    {
+      "epoch": 0.3605591942319335,
+      "grad_norm": 0.9094457626342773,
+      "learning_rate": 9.238243420129965e-06,
+      "loss": 0.7884,
+      "step": 6551
+    },
+    {
+      "epoch": 0.3606142330342892,
+      "grad_norm": 0.8253848552703857,
+      "learning_rate": 9.238013425055965e-06,
+      "loss": 0.7671,
+      "step": 6552
+    },
+    {
+      "epoch": 0.3606692718366448,
+      "grad_norm": 0.7052987813949585,
+      "learning_rate": 9.237783398130193e-06,
+      "loss": 0.7511,
+      "step": 6553
+    },
+    {
+      "epoch": 0.3607243106390005,
+      "grad_norm": 0.7506607174873352,
+      "learning_rate": 9.237553339354373e-06,
+      "loss": 0.6804,
+      "step": 6554
+    },
+    {
+      "epoch": 0.36077934944135615,
+      "grad_norm": 0.725106418132782,
+      "learning_rate": 9.237323248730237e-06,
+      "loss": 0.7658,
+      "step": 6555
+    },
+    {
+      "epoch": 0.36083438824371183,
+      "grad_norm": 0.8164945244789124,
+      "learning_rate": 9.237093126259515e-06,
+      "loss": 0.7857,
+      "step": 6556
+    },
+    {
+      "epoch": 0.36088942704606747,
+      "grad_norm": 0.6937377452850342,
+      "learning_rate": 9.236862971943934e-06,
+      "loss": 0.6985,
+      "step": 6557
+    },
+    {
+      "epoch": 0.36094446584842316,
+      "grad_norm": 0.7511105537414551,
+      "learning_rate": 9.236632785785225e-06,
+      "loss": 0.7891,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3609995046507788,
+      "grad_norm": 0.7217637896537781,
+      "learning_rate": 9.236402567785118e-06,
+      "loss": 0.7942,
+      "step": 6559
+    },
+    {
+      "epoch": 0.3610545434531345,
+      "grad_norm": 1.1438478231430054,
+      "learning_rate": 9.236172317945343e-06,
+      "loss": 0.8311,
+      "step": 6560
+    },
+    {
+      "epoch": 0.3611095822554901,
+      "grad_norm": 0.7414245009422302,
+      "learning_rate": 9.23594203626763e-06,
+      "loss": 0.7726,
+      "step": 6561
+    },
+    {
+      "epoch": 0.3611646210578458,
+      "grad_norm": 0.7762154340744019,
+      "learning_rate": 9.235711722753712e-06,
+      "loss": 0.7891,
+      "step": 6562
+    },
+    {
+      "epoch": 0.36121965986020144,
+      "grad_norm": 0.7368801832199097,
+      "learning_rate": 9.23548137740532e-06,
+      "loss": 0.7656,
+      "step": 6563
+    },
+    {
+      "epoch": 0.3612746986625571,
+      "grad_norm": 0.7571502923965454,
+      "learning_rate": 9.235251000224181e-06,
+      "loss": 0.7845,
+      "step": 6564
+    },
+    {
+      "epoch": 0.36132973746491276,
+      "grad_norm": 0.8078309297561646,
+      "learning_rate": 9.235020591212031e-06,
+      "loss": 0.7969,
+      "step": 6565
+    },
+    {
+      "epoch": 0.36138477626726845,
+      "grad_norm": 0.6897913813591003,
+      "learning_rate": 9.234790150370599e-06,
+      "loss": 0.6922,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3614398150696241,
+      "grad_norm": 0.8053449988365173,
+      "learning_rate": 9.234559677701618e-06,
+      "loss": 0.9126,
+      "step": 6567
+    },
+    {
+      "epoch": 0.36149485387197977,
+      "grad_norm": 0.8400903940200806,
+      "learning_rate": 9.23432917320682e-06,
+      "loss": 0.8144,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3615498926743354,
+      "grad_norm": 0.7753110527992249,
+      "learning_rate": 9.234098636887935e-06,
+      "loss": 0.7025,
+      "step": 6569
+    },
+    {
+      "epoch": 0.3616049314766911,
+      "grad_norm": 0.7901243567466736,
+      "learning_rate": 9.233868068746702e-06,
+      "loss": 0.783,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3616599702790467,
+      "grad_norm": 1.2297497987747192,
+      "learning_rate": 9.233637468784849e-06,
+      "loss": 0.8541,
+      "step": 6571
+    },
+    {
+      "epoch": 0.36171500908140236,
+      "grad_norm": 0.7590478658676147,
+      "learning_rate": 9.233406837004108e-06,
+      "loss": 0.7856,
+      "step": 6572
+    },
+    {
+      "epoch": 0.36177004788375805,
+      "grad_norm": 0.6651493310928345,
+      "learning_rate": 9.233176173406216e-06,
+      "loss": 0.6822,
+      "step": 6573
+    },
+    {
+      "epoch": 0.3618250866861137,
+      "grad_norm": 0.7760787010192871,
+      "learning_rate": 9.232945477992905e-06,
+      "loss": 0.8017,
+      "step": 6574
+    },
+    {
+      "epoch": 0.3618801254884694,
+      "grad_norm": 0.8788009285926819,
+      "learning_rate": 9.232714750765908e-06,
+      "loss": 0.7812,
+      "step": 6575
+    },
+    {
+      "epoch": 0.361935164290825,
+      "grad_norm": 0.7014517188072205,
+      "learning_rate": 9.232483991726961e-06,
+      "loss": 0.7293,
+      "step": 6576
+    },
+    {
+      "epoch": 0.3619902030931807,
+      "grad_norm": 0.7586061954498291,
+      "learning_rate": 9.232253200877797e-06,
+      "loss": 0.7953,
+      "step": 6577
+    },
+    {
+      "epoch": 0.36204524189553633,
+      "grad_norm": 0.8202564120292664,
+      "learning_rate": 9.232022378220151e-06,
+      "loss": 0.8545,
+      "step": 6578
+    },
+    {
+      "epoch": 0.362100280697892,
+      "grad_norm": 0.7816846966743469,
+      "learning_rate": 9.231791523755758e-06,
+      "loss": 0.8573,
+      "step": 6579
+    },
+    {
+      "epoch": 0.36215531950024765,
+      "grad_norm": 0.883222222328186,
+      "learning_rate": 9.23156063748635e-06,
+      "loss": 0.7733,
+      "step": 6580
+    },
+    {
+      "epoch": 0.36221035830260334,
+      "grad_norm": 0.8472830057144165,
+      "learning_rate": 9.231329719413668e-06,
+      "loss": 0.8931,
+      "step": 6581
+    },
+    {
+      "epoch": 0.362265397104959,
+      "grad_norm": 0.7916087508201599,
+      "learning_rate": 9.231098769539443e-06,
+      "loss": 0.8806,
+      "step": 6582
+    },
+    {
+      "epoch": 0.36232043590731466,
+      "grad_norm": 0.815339982509613,
+      "learning_rate": 9.230867787865414e-06,
+      "loss": 0.9081,
+      "step": 6583
+    },
+    {
+      "epoch": 0.3623754747096703,
+      "grad_norm": 1.2352560758590698,
+      "learning_rate": 9.230636774393312e-06,
+      "loss": 0.726,
+      "step": 6584
+    },
+    {
+      "epoch": 0.362430513512026,
+      "grad_norm": 0.759308397769928,
+      "learning_rate": 9.230405729124878e-06,
+      "loss": 0.7648,
+      "step": 6585
+    },
+    {
+      "epoch": 0.3624855523143816,
+      "grad_norm": 0.8285754323005676,
+      "learning_rate": 9.230174652061847e-06,
+      "loss": 0.7972,
+      "step": 6586
+    },
+    {
+      "epoch": 0.3625405911167373,
+      "grad_norm": 0.7393043041229248,
+      "learning_rate": 9.229943543205956e-06,
+      "loss": 0.7859,
+      "step": 6587
+    },
+    {
+      "epoch": 0.36259562991909294,
+      "grad_norm": 0.7354594469070435,
+      "learning_rate": 9.229712402558942e-06,
+      "loss": 0.6683,
+      "step": 6588
+    },
+    {
+      "epoch": 0.36265066872144863,
+      "grad_norm": 0.8244406580924988,
+      "learning_rate": 9.229481230122543e-06,
+      "loss": 0.6977,
+      "step": 6589
+    },
+    {
+      "epoch": 0.36270570752380427,
+      "grad_norm": 0.810565173625946,
+      "learning_rate": 9.229250025898493e-06,
+      "loss": 0.7278,
+      "step": 6590
+    },
+    {
+      "epoch": 0.36276074632615996,
+      "grad_norm": 0.7443352937698364,
+      "learning_rate": 9.229018789888532e-06,
+      "loss": 0.7821,
+      "step": 6591
+    },
+    {
+      "epoch": 0.3628157851285156,
+      "grad_norm": 0.9211748838424683,
+      "learning_rate": 9.228787522094398e-06,
+      "loss": 0.9174,
+      "step": 6592
+    },
+    {
+      "epoch": 0.3628708239308713,
+      "grad_norm": 0.7099255919456482,
+      "learning_rate": 9.22855622251783e-06,
+      "loss": 0.74,
+      "step": 6593
+    },
+    {
+      "epoch": 0.3629258627332269,
+      "grad_norm": 0.7373029589653015,
+      "learning_rate": 9.228324891160564e-06,
+      "loss": 0.7909,
+      "step": 6594
+    },
+    {
+      "epoch": 0.3629809015355826,
+      "grad_norm": 0.8774755001068115,
+      "learning_rate": 9.22809352802434e-06,
+      "loss": 0.8354,
+      "step": 6595
+    },
+    {
+      "epoch": 0.36303594033793823,
+      "grad_norm": 0.7547696232795715,
+      "learning_rate": 9.227862133110899e-06,
+      "loss": 0.6942,
+      "step": 6596
+    },
+    {
+      "epoch": 0.3630909791402939,
+      "grad_norm": 0.7868191003799438,
+      "learning_rate": 9.227630706421975e-06,
+      "loss": 0.7575,
+      "step": 6597
+    },
+    {
+      "epoch": 0.36314601794264956,
+      "grad_norm": 0.6753721237182617,
+      "learning_rate": 9.227399247959312e-06,
+      "loss": 0.7092,
+      "step": 6598
+    },
+    {
+      "epoch": 0.36320105674500525,
+      "grad_norm": 0.7317304611206055,
+      "learning_rate": 9.227167757724646e-06,
+      "loss": 0.8372,
+      "step": 6599
+    },
+    {
+      "epoch": 0.3632560955473609,
+      "grad_norm": 0.8928040266036987,
+      "learning_rate": 9.226936235719721e-06,
+      "loss": 0.8536,
+      "step": 6600
+    },
+    {
+      "epoch": 0.36331113434971657,
+      "grad_norm": 0.7178280353546143,
+      "learning_rate": 9.226704681946275e-06,
+      "loss": 0.7648,
+      "step": 6601
+    },
+    {
+      "epoch": 0.3633661731520722,
+      "grad_norm": 0.7439851760864258,
+      "learning_rate": 9.226473096406046e-06,
+      "loss": 0.8284,
+      "step": 6602
+    },
+    {
+      "epoch": 0.3634212119544279,
+      "grad_norm": 0.7000887989997864,
+      "learning_rate": 9.226241479100777e-06,
+      "loss": 0.7797,
+      "step": 6603
+    },
+    {
+      "epoch": 0.3634762507567835,
+      "grad_norm": 0.7882626056671143,
+      "learning_rate": 9.226009830032209e-06,
+      "loss": 0.72,
+      "step": 6604
+    },
+    {
+      "epoch": 0.3635312895591392,
+      "grad_norm": 0.6445927619934082,
+      "learning_rate": 9.225778149202081e-06,
+      "loss": 0.6785,
+      "step": 6605
+    },
+    {
+      "epoch": 0.36358632836149485,
+      "grad_norm": 0.7348469495773315,
+      "learning_rate": 9.225546436612137e-06,
+      "loss": 0.8117,
+      "step": 6606
+    },
+    {
+      "epoch": 0.36364136716385054,
+      "grad_norm": 0.7455001473426819,
+      "learning_rate": 9.225314692264118e-06,
+      "loss": 0.8196,
+      "step": 6607
+    },
+    {
+      "epoch": 0.36369640596620617,
+      "grad_norm": 0.7149390578269958,
+      "learning_rate": 9.225082916159762e-06,
+      "loss": 0.8841,
+      "step": 6608
+    },
+    {
+      "epoch": 0.36375144476856186,
+      "grad_norm": 0.7095748782157898,
+      "learning_rate": 9.224851108300816e-06,
+      "loss": 0.7336,
+      "step": 6609
+    },
+    {
+      "epoch": 0.3638064835709175,
+      "grad_norm": 0.7112231850624084,
+      "learning_rate": 9.224619268689019e-06,
+      "loss": 0.8606,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3638615223732732,
+      "grad_norm": 0.8052846789360046,
+      "learning_rate": 9.224387397326115e-06,
+      "loss": 0.7838,
+      "step": 6611
+    },
+    {
+      "epoch": 0.3639165611756288,
+      "grad_norm": 0.7538836002349854,
+      "learning_rate": 9.224155494213846e-06,
+      "loss": 0.8252,
+      "step": 6612
+    },
+    {
+      "epoch": 0.3639715999779845,
+      "grad_norm": 0.6968722343444824,
+      "learning_rate": 9.223923559353956e-06,
+      "loss": 0.759,
+      "step": 6613
+    },
+    {
+      "epoch": 0.36402663878034014,
+      "grad_norm": 0.7797368168830872,
+      "learning_rate": 9.223691592748185e-06,
+      "loss": 0.8452,
+      "step": 6614
+    },
+    {
+      "epoch": 0.3640816775826958,
+      "grad_norm": 0.7738572955131531,
+      "learning_rate": 9.223459594398278e-06,
+      "loss": 0.806,
+      "step": 6615
+    },
+    {
+      "epoch": 0.36413671638505146,
+      "grad_norm": 0.7998547554016113,
+      "learning_rate": 9.223227564305983e-06,
+      "loss": 0.748,
+      "step": 6616
+    },
+    {
+      "epoch": 0.3641917551874071,
+      "grad_norm": 0.838666558265686,
+      "learning_rate": 9.222995502473037e-06,
+      "loss": 0.8252,
+      "step": 6617
+    },
+    {
+      "epoch": 0.3642467939897628,
+      "grad_norm": 1.1672697067260742,
+      "learning_rate": 9.222763408901189e-06,
+      "loss": 0.806,
+      "step": 6618
+    },
+    {
+      "epoch": 0.3643018327921184,
+      "grad_norm": 0.6721193194389343,
+      "learning_rate": 9.22253128359218e-06,
+      "loss": 0.6897,
+      "step": 6619
+    },
+    {
+      "epoch": 0.3643568715944741,
+      "grad_norm": 0.8152795433998108,
+      "learning_rate": 9.222299126547758e-06,
+      "loss": 0.8377,
+      "step": 6620
+    },
+    {
+      "epoch": 0.36441191039682974,
+      "grad_norm": 0.7959492206573486,
+      "learning_rate": 9.222066937769664e-06,
+      "loss": 0.8496,
+      "step": 6621
+    },
+    {
+      "epoch": 0.36446694919918543,
+      "grad_norm": 0.7759784460067749,
+      "learning_rate": 9.221834717259646e-06,
+      "loss": 0.7736,
+      "step": 6622
+    },
+    {
+      "epoch": 0.36452198800154106,
+      "grad_norm": 0.6929076313972473,
+      "learning_rate": 9.221602465019449e-06,
+      "loss": 0.7759,
+      "step": 6623
+    },
+    {
+      "epoch": 0.36457702680389675,
+      "grad_norm": 0.7323315143585205,
+      "learning_rate": 9.221370181050817e-06,
+      "loss": 0.7958,
+      "step": 6624
+    },
+    {
+      "epoch": 0.3646320656062524,
+      "grad_norm": 0.7177294492721558,
+      "learning_rate": 9.221137865355496e-06,
+      "loss": 0.8405,
+      "step": 6625
+    },
+    {
+      "epoch": 0.3646871044086081,
+      "grad_norm": 0.7425093650817871,
+      "learning_rate": 9.220905517935235e-06,
+      "loss": 0.7722,
+      "step": 6626
+    },
+    {
+      "epoch": 0.3647421432109637,
+      "grad_norm": 0.8761040568351746,
+      "learning_rate": 9.220673138791775e-06,
+      "loss": 0.8617,
+      "step": 6627
+    },
+    {
+      "epoch": 0.3647971820133194,
+      "grad_norm": 0.927509069442749,
+      "learning_rate": 9.220440727926869e-06,
+      "loss": 0.7839,
+      "step": 6628
+    },
+    {
+      "epoch": 0.36485222081567503,
+      "grad_norm": 0.874399721622467,
+      "learning_rate": 9.220208285342258e-06,
+      "loss": 0.9697,
+      "step": 6629
+    },
+    {
+      "epoch": 0.3649072596180307,
+      "grad_norm": 0.931384801864624,
+      "learning_rate": 9.219975811039691e-06,
+      "loss": 0.8142,
+      "step": 6630
+    },
+    {
+      "epoch": 0.36496229842038636,
+      "grad_norm": 0.8567885160446167,
+      "learning_rate": 9.219743305020916e-06,
+      "loss": 0.7623,
+      "step": 6631
+    },
+    {
+      "epoch": 0.36501733722274204,
+      "grad_norm": 0.7287514209747314,
+      "learning_rate": 9.21951076728768e-06,
+      "loss": 0.8044,
+      "step": 6632
+    },
+    {
+      "epoch": 0.3650723760250977,
+      "grad_norm": 0.7234703302383423,
+      "learning_rate": 9.21927819784173e-06,
+      "loss": 0.7736,
+      "step": 6633
+    },
+    {
+      "epoch": 0.36512741482745337,
+      "grad_norm": 0.7174978256225586,
+      "learning_rate": 9.219045596684815e-06,
+      "loss": 0.7658,
+      "step": 6634
+    },
+    {
+      "epoch": 0.365182453629809,
+      "grad_norm": 0.751075804233551,
+      "learning_rate": 9.218812963818682e-06,
+      "loss": 0.7586,
+      "step": 6635
+    },
+    {
+      "epoch": 0.3652374924321647,
+      "grad_norm": 0.755283534526825,
+      "learning_rate": 9.21858029924508e-06,
+      "loss": 0.8904,
+      "step": 6636
+    },
+    {
+      "epoch": 0.3652925312345203,
+      "grad_norm": 0.6439716815948486,
+      "learning_rate": 9.21834760296576e-06,
+      "loss": 0.7335,
+      "step": 6637
+    },
+    {
+      "epoch": 0.365347570036876,
+      "grad_norm": 0.735285758972168,
+      "learning_rate": 9.218114874982467e-06,
+      "loss": 0.7193,
+      "step": 6638
+    },
+    {
+      "epoch": 0.36540260883923165,
+      "grad_norm": 0.7724307775497437,
+      "learning_rate": 9.217882115296952e-06,
+      "loss": 0.8322,
+      "step": 6639
+    },
+    {
+      "epoch": 0.36545764764158734,
+      "grad_norm": 0.7771303653717041,
+      "learning_rate": 9.217649323910964e-06,
+      "loss": 0.7952,
+      "step": 6640
+    },
+    {
+      "epoch": 0.36551268644394297,
+      "grad_norm": 0.7753337621688843,
+      "learning_rate": 9.217416500826251e-06,
+      "loss": 0.8501,
+      "step": 6641
+    },
+    {
+      "epoch": 0.36556772524629866,
+      "grad_norm": 0.8104514479637146,
+      "learning_rate": 9.217183646044567e-06,
+      "loss": 0.8503,
+      "step": 6642
+    },
+    {
+      "epoch": 0.3656227640486543,
+      "grad_norm": 0.7191929221153259,
+      "learning_rate": 9.21695075956766e-06,
+      "loss": 0.7578,
+      "step": 6643
+    },
+    {
+      "epoch": 0.36567780285101,
+      "grad_norm": 0.745837926864624,
+      "learning_rate": 9.216717841397277e-06,
+      "loss": 0.819,
+      "step": 6644
+    },
+    {
+      "epoch": 0.3657328416533656,
+      "grad_norm": 0.7019662261009216,
+      "learning_rate": 9.216484891535174e-06,
+      "loss": 0.8024,
+      "step": 6645
+    },
+    {
+      "epoch": 0.3657878804557213,
+      "grad_norm": 0.9709738492965698,
+      "learning_rate": 9.216251909983095e-06,
+      "loss": 0.7653,
+      "step": 6646
+    },
+    {
+      "epoch": 0.36584291925807694,
+      "grad_norm": 0.7973032593727112,
+      "learning_rate": 9.2160188967428e-06,
+      "loss": 0.8071,
+      "step": 6647
+    },
+    {
+      "epoch": 0.3658979580604326,
+      "grad_norm": 0.6945796012878418,
+      "learning_rate": 9.215785851816034e-06,
+      "loss": 0.6831,
+      "step": 6648
+    },
+    {
+      "epoch": 0.36595299686278826,
+      "grad_norm": 0.8685100674629211,
+      "learning_rate": 9.21555277520455e-06,
+      "loss": 0.821,
+      "step": 6649
+    },
+    {
+      "epoch": 0.36600803566514395,
+      "grad_norm": 1.0164310932159424,
+      "learning_rate": 9.2153196669101e-06,
+      "loss": 0.7861,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3660630744674996,
+      "grad_norm": 0.8572850227355957,
+      "learning_rate": 9.215086526934435e-06,
+      "loss": 0.7982,
+      "step": 6651
+    },
+    {
+      "epoch": 0.36611811326985527,
+      "grad_norm": 0.7481987476348877,
+      "learning_rate": 9.214853355279307e-06,
+      "loss": 0.8258,
+      "step": 6652
+    },
+    {
+      "epoch": 0.3661731520722109,
+      "grad_norm": 0.750344455242157,
+      "learning_rate": 9.214620151946472e-06,
+      "loss": 0.7842,
+      "step": 6653
+    },
+    {
+      "epoch": 0.3662281908745666,
+      "grad_norm": 1.0266414880752563,
+      "learning_rate": 9.214386916937678e-06,
+      "loss": 0.7313,
+      "step": 6654
+    },
+    {
+      "epoch": 0.36628322967692223,
+      "grad_norm": 0.7913589477539062,
+      "learning_rate": 9.214153650254682e-06,
+      "loss": 0.8251,
+      "step": 6655
+    },
+    {
+      "epoch": 0.3663382684792779,
+      "grad_norm": 0.7185465693473816,
+      "learning_rate": 9.213920351899235e-06,
+      "loss": 0.7145,
+      "step": 6656
+    },
+    {
+      "epoch": 0.36639330728163355,
+      "grad_norm": 0.7185063362121582,
+      "learning_rate": 9.213687021873088e-06,
+      "loss": 0.8321,
+      "step": 6657
+    },
+    {
+      "epoch": 0.3664483460839892,
+      "grad_norm": 0.8380091190338135,
+      "learning_rate": 9.213453660178e-06,
+      "loss": 0.8293,
+      "step": 6658
+    },
+    {
+      "epoch": 0.3665033848863449,
+      "grad_norm": 0.7569485306739807,
+      "learning_rate": 9.21322026681572e-06,
+      "loss": 0.7201,
+      "step": 6659
+    },
+    {
+      "epoch": 0.3665584236887005,
+      "grad_norm": 0.7212445735931396,
+      "learning_rate": 9.212986841788005e-06,
+      "loss": 0.7869,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3666134624910562,
+      "grad_norm": 0.9435489773750305,
+      "learning_rate": 9.212753385096612e-06,
+      "loss": 0.8469,
+      "step": 6661
+    },
+    {
+      "epoch": 0.36666850129341183,
+      "grad_norm": 0.6609265208244324,
+      "learning_rate": 9.212519896743289e-06,
+      "loss": 0.6446,
+      "step": 6662
+    },
+    {
+      "epoch": 0.3667235400957675,
+      "grad_norm": 0.7232604026794434,
+      "learning_rate": 9.212286376729794e-06,
+      "loss": 0.7138,
+      "step": 6663
+    },
+    {
+      "epoch": 0.36677857889812315,
+      "grad_norm": 0.7276197075843811,
+      "learning_rate": 9.212052825057882e-06,
+      "loss": 0.725,
+      "step": 6664
+    },
+    {
+      "epoch": 0.36683361770047884,
+      "grad_norm": 0.7029727101325989,
+      "learning_rate": 9.21181924172931e-06,
+      "loss": 0.6973,
+      "step": 6665
+    },
+    {
+      "epoch": 0.3668886565028345,
+      "grad_norm": 0.7292968034744263,
+      "learning_rate": 9.21158562674583e-06,
+      "loss": 0.6984,
+      "step": 6666
+    },
+    {
+      "epoch": 0.36694369530519017,
+      "grad_norm": 0.6977009177207947,
+      "learning_rate": 9.2113519801092e-06,
+      "loss": 0.7752,
+      "step": 6667
+    },
+    {
+      "epoch": 0.3669987341075458,
+      "grad_norm": 0.8019471764564514,
+      "learning_rate": 9.211118301821176e-06,
+      "loss": 0.7481,
+      "step": 6668
+    },
+    {
+      "epoch": 0.3670537729099015,
+      "grad_norm": 0.8097867965698242,
+      "learning_rate": 9.210884591883516e-06,
+      "loss": 0.8077,
+      "step": 6669
+    },
+    {
+      "epoch": 0.3671088117122571,
+      "grad_norm": 1.1622828245162964,
+      "learning_rate": 9.210650850297973e-06,
+      "loss": 0.8053,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3671638505146128,
+      "grad_norm": 0.8188957571983337,
+      "learning_rate": 9.210417077066304e-06,
+      "loss": 0.7731,
+      "step": 6671
+    },
+    {
+      "epoch": 0.36721888931696844,
+      "grad_norm": 0.8531584739685059,
+      "learning_rate": 9.210183272190269e-06,
+      "loss": 0.8183,
+      "step": 6672
+    },
+    {
+      "epoch": 0.36727392811932413,
+      "grad_norm": 0.8007203936576843,
+      "learning_rate": 9.209949435671624e-06,
+      "loss": 0.7906,
+      "step": 6673
+    },
+    {
+      "epoch": 0.36732896692167977,
+      "grad_norm": 0.8284860253334045,
+      "learning_rate": 9.209715567512126e-06,
+      "loss": 0.7845,
+      "step": 6674
+    },
+    {
+      "epoch": 0.36738400572403546,
+      "grad_norm": 0.7735304236412048,
+      "learning_rate": 9.209481667713533e-06,
+      "loss": 0.7333,
+      "step": 6675
+    },
+    {
+      "epoch": 0.3674390445263911,
+      "grad_norm": 0.7390912771224976,
+      "learning_rate": 9.209247736277601e-06,
+      "loss": 0.7992,
+      "step": 6676
+    },
+    {
+      "epoch": 0.3674940833287468,
+      "grad_norm": 0.6871926784515381,
+      "learning_rate": 9.209013773206091e-06,
+      "loss": 0.7765,
+      "step": 6677
+    },
+    {
+      "epoch": 0.3675491221311024,
+      "grad_norm": 0.7241746187210083,
+      "learning_rate": 9.208779778500758e-06,
+      "loss": 0.7124,
+      "step": 6678
+    },
+    {
+      "epoch": 0.3676041609334581,
+      "grad_norm": 0.7362630367279053,
+      "learning_rate": 9.208545752163365e-06,
+      "loss": 0.7695,
+      "step": 6679
+    },
+    {
+      "epoch": 0.36765919973581374,
+      "grad_norm": 0.7577944993972778,
+      "learning_rate": 9.208311694195669e-06,
+      "loss": 0.8302,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3677142385381694,
+      "grad_norm": 0.7182355523109436,
+      "learning_rate": 9.208077604599427e-06,
+      "loss": 0.8182,
+      "step": 6681
+    },
+    {
+      "epoch": 0.36776927734052506,
+      "grad_norm": 0.7636679410934448,
+      "learning_rate": 9.207843483376402e-06,
+      "loss": 0.7266,
+      "step": 6682
+    },
+    {
+      "epoch": 0.36782431614288075,
+      "grad_norm": 0.7325936555862427,
+      "learning_rate": 9.207609330528349e-06,
+      "loss": 0.735,
+      "step": 6683
+    },
+    {
+      "epoch": 0.3678793549452364,
+      "grad_norm": 1.1119143962860107,
+      "learning_rate": 9.207375146057033e-06,
+      "loss": 1.0124,
+      "step": 6684
+    },
+    {
+      "epoch": 0.36793439374759207,
+      "grad_norm": 0.7694228291511536,
+      "learning_rate": 9.207140929964212e-06,
+      "loss": 0.7803,
+      "step": 6685
+    },
+    {
+      "epoch": 0.3679894325499477,
+      "grad_norm": 0.7628658413887024,
+      "learning_rate": 9.206906682251644e-06,
+      "loss": 0.8057,
+      "step": 6686
+    },
+    {
+      "epoch": 0.3680444713523034,
+      "grad_norm": 0.766266942024231,
+      "learning_rate": 9.206672402921092e-06,
+      "loss": 0.7827,
+      "step": 6687
+    },
+    {
+      "epoch": 0.368099510154659,
+      "grad_norm": 0.7355746626853943,
+      "learning_rate": 9.206438091974316e-06,
+      "loss": 0.8146,
+      "step": 6688
+    },
+    {
+      "epoch": 0.3681545489570147,
+      "grad_norm": 0.8464547395706177,
+      "learning_rate": 9.20620374941308e-06,
+      "loss": 0.8296,
+      "step": 6689
+    },
+    {
+      "epoch": 0.36820958775937035,
+      "grad_norm": 0.7113955616950989,
+      "learning_rate": 9.20596937523914e-06,
+      "loss": 0.7621,
+      "step": 6690
+    },
+    {
+      "epoch": 0.36826462656172604,
+      "grad_norm": 0.7141324877738953,
+      "learning_rate": 9.205734969454259e-06,
+      "loss": 0.738,
+      "step": 6691
+    },
+    {
+      "epoch": 0.36831966536408167,
+      "grad_norm": 0.7576237320899963,
+      "learning_rate": 9.2055005320602e-06,
+      "loss": 0.7727,
+      "step": 6692
+    },
+    {
+      "epoch": 0.36837470416643736,
+      "grad_norm": 0.7448444962501526,
+      "learning_rate": 9.205266063058727e-06,
+      "loss": 0.8238,
+      "step": 6693
+    },
+    {
+      "epoch": 0.368429742968793,
+      "grad_norm": 0.7441811561584473,
+      "learning_rate": 9.205031562451599e-06,
+      "loss": 0.7518,
+      "step": 6694
+    },
+    {
+      "epoch": 0.3684847817711487,
+      "grad_norm": 0.9284115433692932,
+      "learning_rate": 9.20479703024058e-06,
+      "loss": 0.817,
+      "step": 6695
+    },
+    {
+      "epoch": 0.3685398205735043,
+      "grad_norm": 0.7019243836402893,
+      "learning_rate": 9.204562466427431e-06,
+      "loss": 0.7403,
+      "step": 6696
+    },
+    {
+      "epoch": 0.36859485937586,
+      "grad_norm": 0.6345306634902954,
+      "learning_rate": 9.204327871013917e-06,
+      "loss": 0.7058,
+      "step": 6697
+    },
+    {
+      "epoch": 0.36864989817821564,
+      "grad_norm": 0.7375063300132751,
+      "learning_rate": 9.2040932440018e-06,
+      "loss": 0.831,
+      "step": 6698
+    },
+    {
+      "epoch": 0.36870493698057133,
+      "grad_norm": 0.8213731050491333,
+      "learning_rate": 9.203858585392842e-06,
+      "loss": 0.7677,
+      "step": 6699
+    },
+    {
+      "epoch": 0.36875997578292696,
+      "grad_norm": 0.7114601731300354,
+      "learning_rate": 9.203623895188809e-06,
+      "loss": 0.8015,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3688150145852826,
+      "grad_norm": 0.7707667350769043,
+      "learning_rate": 9.203389173391463e-06,
+      "loss": 0.7758,
+      "step": 6701
+    },
+    {
+      "epoch": 0.3688700533876383,
+      "grad_norm": 0.7374396920204163,
+      "learning_rate": 9.203154420002572e-06,
+      "loss": 0.7583,
+      "step": 6702
+    },
+    {
+      "epoch": 0.3689250921899939,
+      "grad_norm": 0.7156866192817688,
+      "learning_rate": 9.202919635023895e-06,
+      "loss": 0.8173,
+      "step": 6703
+    },
+    {
+      "epoch": 0.3689801309923496,
+      "grad_norm": 0.6811904311180115,
+      "learning_rate": 9.2026848184572e-06,
+      "loss": 0.7441,
+      "step": 6704
+    },
+    {
+      "epoch": 0.36903516979470524,
+      "grad_norm": 0.7515163421630859,
+      "learning_rate": 9.20244997030425e-06,
+      "loss": 0.7927,
+      "step": 6705
+    },
+    {
+      "epoch": 0.36909020859706093,
+      "grad_norm": 0.761116087436676,
+      "learning_rate": 9.202215090566813e-06,
+      "loss": 0.7686,
+      "step": 6706
+    },
+    {
+      "epoch": 0.36914524739941657,
+      "grad_norm": 0.8726711869239807,
+      "learning_rate": 9.20198017924665e-06,
+      "loss": 0.7831,
+      "step": 6707
+    },
+    {
+      "epoch": 0.36920028620177225,
+      "grad_norm": 0.6868153810501099,
+      "learning_rate": 9.20174523634553e-06,
+      "loss": 0.7855,
+      "step": 6708
+    },
+    {
+      "epoch": 0.3692553250041279,
+      "grad_norm": 0.7140498757362366,
+      "learning_rate": 9.201510261865218e-06,
+      "loss": 0.8144,
+      "step": 6709
+    },
+    {
+      "epoch": 0.3693103638064836,
+      "grad_norm": 0.8745181560516357,
+      "learning_rate": 9.201275255807478e-06,
+      "loss": 0.9204,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3693654026088392,
+      "grad_norm": 0.6535945534706116,
+      "learning_rate": 9.20104021817408e-06,
+      "loss": 0.7729,
+      "step": 6711
+    },
+    {
+      "epoch": 0.3694204414111949,
+      "grad_norm": 0.655857503414154,
+      "learning_rate": 9.200805148966785e-06,
+      "loss": 0.8373,
+      "step": 6712
+    },
+    {
+      "epoch": 0.36947548021355053,
+      "grad_norm": 0.8393271565437317,
+      "learning_rate": 9.200570048187365e-06,
+      "loss": 0.8532,
+      "step": 6713
+    },
+    {
+      "epoch": 0.3695305190159062,
+      "grad_norm": 0.7484574913978577,
+      "learning_rate": 9.200334915837585e-06,
+      "loss": 0.8411,
+      "step": 6714
+    },
+    {
+      "epoch": 0.36958555781826186,
+      "grad_norm": 0.9913665652275085,
+      "learning_rate": 9.200099751919212e-06,
+      "loss": 0.9011,
+      "step": 6715
+    },
+    {
+      "epoch": 0.36964059662061755,
+      "grad_norm": 0.7314063310623169,
+      "learning_rate": 9.199864556434013e-06,
+      "loss": 0.7184,
+      "step": 6716
+    },
+    {
+      "epoch": 0.3696956354229732,
+      "grad_norm": 0.7881553173065186,
+      "learning_rate": 9.199629329383758e-06,
+      "loss": 0.796,
+      "step": 6717
+    },
+    {
+      "epoch": 0.36975067422532887,
+      "grad_norm": 0.7440283298492432,
+      "learning_rate": 9.199394070770212e-06,
+      "loss": 0.7472,
+      "step": 6718
+    },
+    {
+      "epoch": 0.3698057130276845,
+      "grad_norm": 0.6916326880455017,
+      "learning_rate": 9.199158780595144e-06,
+      "loss": 0.6808,
+      "step": 6719
+    },
+    {
+      "epoch": 0.3698607518300402,
+      "grad_norm": 0.8482714295387268,
+      "learning_rate": 9.198923458860323e-06,
+      "loss": 0.7795,
+      "step": 6720
+    },
+    {
+      "epoch": 0.3699157906323958,
+      "grad_norm": 0.7541999816894531,
+      "learning_rate": 9.198688105567516e-06,
+      "loss": 0.7917,
+      "step": 6721
+    },
+    {
+      "epoch": 0.3699708294347515,
+      "grad_norm": 0.794335126876831,
+      "learning_rate": 9.198452720718494e-06,
+      "loss": 0.8463,
+      "step": 6722
+    },
+    {
+      "epoch": 0.37002586823710715,
+      "grad_norm": 0.7866827845573425,
+      "learning_rate": 9.198217304315025e-06,
+      "loss": 0.7938,
+      "step": 6723
+    },
+    {
+      "epoch": 0.37008090703946284,
+      "grad_norm": 0.7393556833267212,
+      "learning_rate": 9.19798185635888e-06,
+      "loss": 0.7825,
+      "step": 6724
+    },
+    {
+      "epoch": 0.37013594584181847,
+      "grad_norm": 0.7131090760231018,
+      "learning_rate": 9.197746376851825e-06,
+      "loss": 0.7184,
+      "step": 6725
+    },
+    {
+      "epoch": 0.37019098464417416,
+      "grad_norm": 0.7054039239883423,
+      "learning_rate": 9.197510865795634e-06,
+      "loss": 0.7458,
+      "step": 6726
+    },
+    {
+      "epoch": 0.3702460234465298,
+      "grad_norm": 0.7437009811401367,
+      "learning_rate": 9.197275323192073e-06,
+      "loss": 0.7921,
+      "step": 6727
+    },
+    {
+      "epoch": 0.3703010622488855,
+      "grad_norm": 1.0703076124191284,
+      "learning_rate": 9.197039749042916e-06,
+      "loss": 0.771,
+      "step": 6728
+    },
+    {
+      "epoch": 0.3703561010512411,
+      "grad_norm": 0.8278045654296875,
+      "learning_rate": 9.196804143349929e-06,
+      "loss": 0.8984,
+      "step": 6729
+    },
+    {
+      "epoch": 0.3704111398535968,
+      "grad_norm": 0.7713067531585693,
+      "learning_rate": 9.196568506114887e-06,
+      "loss": 0.7702,
+      "step": 6730
+    },
+    {
+      "epoch": 0.37046617865595244,
+      "grad_norm": 0.9040505290031433,
+      "learning_rate": 9.19633283733956e-06,
+      "loss": 0.7113,
+      "step": 6731
+    },
+    {
+      "epoch": 0.3705212174583081,
+      "grad_norm": 0.8853700757026672,
+      "learning_rate": 9.196097137025718e-06,
+      "loss": 0.8445,
+      "step": 6732
+    },
+    {
+      "epoch": 0.37057625626066376,
+      "grad_norm": 0.6870817542076111,
+      "learning_rate": 9.195861405175133e-06,
+      "loss": 0.7613,
+      "step": 6733
+    },
+    {
+      "epoch": 0.37063129506301945,
+      "grad_norm": 0.7539152503013611,
+      "learning_rate": 9.195625641789579e-06,
+      "loss": 0.7478,
+      "step": 6734
+    },
+    {
+      "epoch": 0.3706863338653751,
+      "grad_norm": 0.7084356546401978,
+      "learning_rate": 9.195389846870822e-06,
+      "loss": 0.7803,
+      "step": 6735
+    },
+    {
+      "epoch": 0.3707413726677308,
+      "grad_norm": 0.7883948087692261,
+      "learning_rate": 9.19515402042064e-06,
+      "loss": 0.8606,
+      "step": 6736
+    },
+    {
+      "epoch": 0.3707964114700864,
+      "grad_norm": 0.714948296546936,
+      "learning_rate": 9.194918162440804e-06,
+      "loss": 0.8066,
+      "step": 6737
+    },
+    {
+      "epoch": 0.3708514502724421,
+      "grad_norm": 0.7110786437988281,
+      "learning_rate": 9.194682272933085e-06,
+      "loss": 0.7439,
+      "step": 6738
+    },
+    {
+      "epoch": 0.37090648907479773,
+      "grad_norm": 0.7281045317649841,
+      "learning_rate": 9.194446351899257e-06,
+      "loss": 0.7772,
+      "step": 6739
+    },
+    {
+      "epoch": 0.3709615278771534,
+      "grad_norm": 0.7351245880126953,
+      "learning_rate": 9.194210399341093e-06,
+      "loss": 0.8777,
+      "step": 6740
+    },
+    {
+      "epoch": 0.37101656667950905,
+      "grad_norm": 0.8028532266616821,
+      "learning_rate": 9.193974415260367e-06,
+      "loss": 0.7461,
+      "step": 6741
+    },
+    {
+      "epoch": 0.37107160548186474,
+      "grad_norm": 0.8015451431274414,
+      "learning_rate": 9.19373839965885e-06,
+      "loss": 0.8006,
+      "step": 6742
+    },
+    {
+      "epoch": 0.3711266442842204,
+      "grad_norm": 0.9567442536354065,
+      "learning_rate": 9.193502352538321e-06,
+      "loss": 0.8636,
+      "step": 6743
+    },
+    {
+      "epoch": 0.371181683086576,
+      "grad_norm": 1.1413114070892334,
+      "learning_rate": 9.193266273900547e-06,
+      "loss": 0.8976,
+      "step": 6744
+    },
+    {
+      "epoch": 0.3712367218889317,
+      "grad_norm": 0.6971789002418518,
+      "learning_rate": 9.19303016374731e-06,
+      "loss": 0.7419,
+      "step": 6745
+    },
+    {
+      "epoch": 0.37129176069128733,
+      "grad_norm": 0.8117435574531555,
+      "learning_rate": 9.192794022080378e-06,
+      "loss": 0.8166,
+      "step": 6746
+    },
+    {
+      "epoch": 0.371346799493643,
+      "grad_norm": 0.7748119831085205,
+      "learning_rate": 9.19255784890153e-06,
+      "loss": 0.8073,
+      "step": 6747
+    },
+    {
+      "epoch": 0.37140183829599865,
+      "grad_norm": 0.6550068259239197,
+      "learning_rate": 9.192321644212539e-06,
+      "loss": 0.6976,
+      "step": 6748
+    },
+    {
+      "epoch": 0.37145687709835434,
+      "grad_norm": 0.7931404709815979,
+      "learning_rate": 9.19208540801518e-06,
+      "loss": 0.7153,
+      "step": 6749
+    },
+    {
+      "epoch": 0.37151191590071,
+      "grad_norm": 0.7107539176940918,
+      "learning_rate": 9.19184914031123e-06,
+      "loss": 0.7616,
+      "step": 6750
+    },
+    {
+      "epoch": 0.37156695470306567,
+      "grad_norm": 0.6983848810195923,
+      "learning_rate": 9.191612841102463e-06,
+      "loss": 0.6507,
+      "step": 6751
+    },
+    {
+      "epoch": 0.3716219935054213,
+      "grad_norm": 0.7653477787971497,
+      "learning_rate": 9.191376510390657e-06,
+      "loss": 0.708,
+      "step": 6752
+    },
+    {
+      "epoch": 0.371677032307777,
+      "grad_norm": 0.8903954029083252,
+      "learning_rate": 9.191140148177586e-06,
+      "loss": 0.8131,
+      "step": 6753
+    },
+    {
+      "epoch": 0.3717320711101326,
+      "grad_norm": 0.7584933042526245,
+      "learning_rate": 9.190903754465028e-06,
+      "loss": 0.8178,
+      "step": 6754
+    },
+    {
+      "epoch": 0.3717871099124883,
+      "grad_norm": 0.7338405847549438,
+      "learning_rate": 9.19066732925476e-06,
+      "loss": 0.7717,
+      "step": 6755
+    },
+    {
+      "epoch": 0.37184214871484395,
+      "grad_norm": 0.764944851398468,
+      "learning_rate": 9.190430872548557e-06,
+      "loss": 0.7762,
+      "step": 6756
+    },
+    {
+      "epoch": 0.37189718751719963,
+      "grad_norm": 0.7362231612205505,
+      "learning_rate": 9.190194384348199e-06,
+      "loss": 0.8277,
+      "step": 6757
+    },
+    {
+      "epoch": 0.37195222631955527,
+      "grad_norm": 0.7462226748466492,
+      "learning_rate": 9.18995786465546e-06,
+      "loss": 0.7362,
+      "step": 6758
+    },
+    {
+      "epoch": 0.37200726512191096,
+      "grad_norm": 0.7769725322723389,
+      "learning_rate": 9.18972131347212e-06,
+      "loss": 0.8217,
+      "step": 6759
+    },
+    {
+      "epoch": 0.3720623039242666,
+      "grad_norm": 0.7263969779014587,
+      "learning_rate": 9.189484730799956e-06,
+      "loss": 0.7719,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3721173427266223,
+      "grad_norm": 0.7612473964691162,
+      "learning_rate": 9.189248116640746e-06,
+      "loss": 0.7149,
+      "step": 6761
+    },
+    {
+      "epoch": 0.3721723815289779,
+      "grad_norm": 0.6813042759895325,
+      "learning_rate": 9.189011470996268e-06,
+      "loss": 0.7119,
+      "step": 6762
+    },
+    {
+      "epoch": 0.3722274203313336,
+      "grad_norm": 0.7376571297645569,
+      "learning_rate": 9.188774793868302e-06,
+      "loss": 0.7998,
+      "step": 6763
+    },
+    {
+      "epoch": 0.37228245913368924,
+      "grad_norm": 0.8592102527618408,
+      "learning_rate": 9.188538085258626e-06,
+      "loss": 0.8026,
+      "step": 6764
+    },
+    {
+      "epoch": 0.3723374979360449,
+      "grad_norm": 0.7666613459587097,
+      "learning_rate": 9.188301345169017e-06,
+      "loss": 0.8571,
+      "step": 6765
+    },
+    {
+      "epoch": 0.37239253673840056,
+      "grad_norm": 0.7118985652923584,
+      "learning_rate": 9.188064573601258e-06,
+      "loss": 0.7637,
+      "step": 6766
+    },
+    {
+      "epoch": 0.37244757554075625,
+      "grad_norm": 0.8247082233428955,
+      "learning_rate": 9.187827770557127e-06,
+      "loss": 0.8209,
+      "step": 6767
+    },
+    {
+      "epoch": 0.3725026143431119,
+      "grad_norm": 0.7259567975997925,
+      "learning_rate": 9.187590936038403e-06,
+      "loss": 0.7918,
+      "step": 6768
+    },
+    {
+      "epoch": 0.37255765314546757,
+      "grad_norm": 0.7409893274307251,
+      "learning_rate": 9.187354070046867e-06,
+      "loss": 0.8004,
+      "step": 6769
+    },
+    {
+      "epoch": 0.3726126919478232,
+      "grad_norm": 0.8163084387779236,
+      "learning_rate": 9.187117172584298e-06,
+      "loss": 0.8452,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3726677307501789,
+      "grad_norm": 0.9241586923599243,
+      "learning_rate": 9.186880243652477e-06,
+      "loss": 0.8939,
+      "step": 6771
+    },
+    {
+      "epoch": 0.3727227695525345,
+      "grad_norm": 0.710434079170227,
+      "learning_rate": 9.186643283253185e-06,
+      "loss": 0.7337,
+      "step": 6772
+    },
+    {
+      "epoch": 0.3727778083548902,
+      "grad_norm": 0.7850505709648132,
+      "learning_rate": 9.186406291388203e-06,
+      "loss": 0.7892,
+      "step": 6773
+    },
+    {
+      "epoch": 0.37283284715724585,
+      "grad_norm": 0.813979983329773,
+      "learning_rate": 9.186169268059311e-06,
+      "loss": 0.7993,
+      "step": 6774
+    },
+    {
+      "epoch": 0.37288788595960154,
+      "grad_norm": 0.7923213243484497,
+      "learning_rate": 9.185932213268292e-06,
+      "loss": 0.7501,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3729429247619572,
+      "grad_norm": 0.7923155426979065,
+      "learning_rate": 9.185695127016928e-06,
+      "loss": 0.8435,
+      "step": 6776
+    },
+    {
+      "epoch": 0.37299796356431286,
+      "grad_norm": 0.69893479347229,
+      "learning_rate": 9.185458009306999e-06,
+      "loss": 0.7155,
+      "step": 6777
+    },
+    {
+      "epoch": 0.3730530023666685,
+      "grad_norm": 0.7848305106163025,
+      "learning_rate": 9.185220860140289e-06,
+      "loss": 0.7971,
+      "step": 6778
+    },
+    {
+      "epoch": 0.3731080411690242,
+      "grad_norm": 0.6707655787467957,
+      "learning_rate": 9.184983679518578e-06,
+      "loss": 0.6939,
+      "step": 6779
+    },
+    {
+      "epoch": 0.3731630799713798,
+      "grad_norm": 0.6612532734870911,
+      "learning_rate": 9.18474646744365e-06,
+      "loss": 0.7361,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3732181187737355,
+      "grad_norm": 0.7753985524177551,
+      "learning_rate": 9.184509223917288e-06,
+      "loss": 0.7263,
+      "step": 6781
+    },
+    {
+      "epoch": 0.37327315757609114,
+      "grad_norm": 0.6856646537780762,
+      "learning_rate": 9.184271948941275e-06,
+      "loss": 0.6923,
+      "step": 6782
+    },
+    {
+      "epoch": 0.37332819637844683,
+      "grad_norm": 0.7223647832870483,
+      "learning_rate": 9.184034642517393e-06,
+      "loss": 0.793,
+      "step": 6783
+    },
+    {
+      "epoch": 0.37338323518080246,
+      "grad_norm": 0.7428838014602661,
+      "learning_rate": 9.183797304647428e-06,
+      "loss": 0.7781,
+      "step": 6784
+    },
+    {
+      "epoch": 0.37343827398315815,
+      "grad_norm": 0.7301773428916931,
+      "learning_rate": 9.183559935333161e-06,
+      "loss": 0.7964,
+      "step": 6785
+    },
+    {
+      "epoch": 0.3734933127855138,
+      "grad_norm": 0.7883384823799133,
+      "learning_rate": 9.183322534576378e-06,
+      "loss": 0.8904,
+      "step": 6786
+    },
+    {
+      "epoch": 0.3735483515878694,
+      "grad_norm": 0.7943564653396606,
+      "learning_rate": 9.183085102378864e-06,
+      "loss": 0.7229,
+      "step": 6787
+    },
+    {
+      "epoch": 0.3736033903902251,
+      "grad_norm": 0.7385129928588867,
+      "learning_rate": 9.1828476387424e-06,
+      "loss": 0.7967,
+      "step": 6788
+    },
+    {
+      "epoch": 0.37365842919258074,
+      "grad_norm": 0.7968102097511292,
+      "learning_rate": 9.182610143668775e-06,
+      "loss": 0.8016,
+      "step": 6789
+    },
+    {
+      "epoch": 0.37371346799493643,
+      "grad_norm": 0.7810283303260803,
+      "learning_rate": 9.18237261715977e-06,
+      "loss": 0.8956,
+      "step": 6790
+    },
+    {
+      "epoch": 0.37376850679729207,
+      "grad_norm": 0.7110065221786499,
+      "learning_rate": 9.182135059217172e-06,
+      "loss": 0.7808,
+      "step": 6791
+    },
+    {
+      "epoch": 0.37382354559964776,
+      "grad_norm": 0.7513633370399475,
+      "learning_rate": 9.181897469842767e-06,
+      "loss": 0.8236,
+      "step": 6792
+    },
+    {
+      "epoch": 0.3738785844020034,
+      "grad_norm": 0.7850426435470581,
+      "learning_rate": 9.18165984903834e-06,
+      "loss": 0.8642,
+      "step": 6793
+    },
+    {
+      "epoch": 0.3739336232043591,
+      "grad_norm": 1.4948225021362305,
+      "learning_rate": 9.181422196805676e-06,
+      "loss": 0.8765,
+      "step": 6794
+    },
+    {
+      "epoch": 0.3739886620067147,
+      "grad_norm": 0.8242343068122864,
+      "learning_rate": 9.181184513146563e-06,
+      "loss": 0.7213,
+      "step": 6795
+    },
+    {
+      "epoch": 0.3740437008090704,
+      "grad_norm": 0.8017476797103882,
+      "learning_rate": 9.180946798062786e-06,
+      "loss": 0.655,
+      "step": 6796
+    },
+    {
+      "epoch": 0.37409873961142603,
+      "grad_norm": 0.9573387503623962,
+      "learning_rate": 9.180709051556132e-06,
+      "loss": 0.8674,
+      "step": 6797
+    },
+    {
+      "epoch": 0.3741537784137817,
+      "grad_norm": 0.7575511932373047,
+      "learning_rate": 9.180471273628388e-06,
+      "loss": 0.8672,
+      "step": 6798
+    },
+    {
+      "epoch": 0.37420881721613736,
+      "grad_norm": 0.7723323702812195,
+      "learning_rate": 9.180233464281343e-06,
+      "loss": 0.7698,
+      "step": 6799
+    },
+    {
+      "epoch": 0.37426385601849305,
+      "grad_norm": 0.8352731466293335,
+      "learning_rate": 9.17999562351678e-06,
+      "loss": 0.9248,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3743188948208487,
+      "grad_norm": 0.7459322214126587,
+      "learning_rate": 9.179757751336488e-06,
+      "loss": 0.7561,
+      "step": 6801
+    },
+    {
+      "epoch": 0.37437393362320437,
+      "grad_norm": 0.8053051829338074,
+      "learning_rate": 9.179519847742257e-06,
+      "loss": 0.8743,
+      "step": 6802
+    },
+    {
+      "epoch": 0.37442897242556,
+      "grad_norm": 0.7781768441200256,
+      "learning_rate": 9.179281912735873e-06,
+      "loss": 0.7426,
+      "step": 6803
+    },
+    {
+      "epoch": 0.3744840112279157,
+      "grad_norm": 0.6812007427215576,
+      "learning_rate": 9.179043946319126e-06,
+      "loss": 0.761,
+      "step": 6804
+    },
+    {
+      "epoch": 0.3745390500302713,
+      "grad_norm": 0.8327108025550842,
+      "learning_rate": 9.178805948493803e-06,
+      "loss": 0.7633,
+      "step": 6805
+    },
+    {
+      "epoch": 0.374594088832627,
+      "grad_norm": 0.7519007921218872,
+      "learning_rate": 9.178567919261692e-06,
+      "loss": 0.8268,
+      "step": 6806
+    },
+    {
+      "epoch": 0.37464912763498265,
+      "grad_norm": 0.7507897019386292,
+      "learning_rate": 9.178329858624584e-06,
+      "loss": 0.8734,
+      "step": 6807
+    },
+    {
+      "epoch": 0.37470416643733834,
+      "grad_norm": 0.6874666213989258,
+      "learning_rate": 9.178091766584267e-06,
+      "loss": 0.6669,
+      "step": 6808
+    },
+    {
+      "epoch": 0.37475920523969397,
+      "grad_norm": 0.6987403631210327,
+      "learning_rate": 9.17785364314253e-06,
+      "loss": 0.7627,
+      "step": 6809
+    },
+    {
+      "epoch": 0.37481424404204966,
+      "grad_norm": 0.7777343392372131,
+      "learning_rate": 9.177615488301163e-06,
+      "loss": 0.7637,
+      "step": 6810
+    },
+    {
+      "epoch": 0.3748692828444053,
+      "grad_norm": 0.71980881690979,
+      "learning_rate": 9.177377302061958e-06,
+      "loss": 0.7964,
+      "step": 6811
+    },
+    {
+      "epoch": 0.374924321646761,
+      "grad_norm": 0.627328634262085,
+      "learning_rate": 9.177139084426704e-06,
+      "loss": 0.6862,
+      "step": 6812
+    },
+    {
+      "epoch": 0.3749793604491166,
+      "grad_norm": 0.7099852561950684,
+      "learning_rate": 9.176900835397188e-06,
+      "loss": 0.7592,
+      "step": 6813
+    },
+    {
+      "epoch": 0.3750343992514723,
+      "grad_norm": 0.7880212664604187,
+      "learning_rate": 9.176662554975205e-06,
+      "loss": 0.756,
+      "step": 6814
+    },
+    {
+      "epoch": 0.37508943805382794,
+      "grad_norm": 0.7347460389137268,
+      "learning_rate": 9.176424243162546e-06,
+      "loss": 0.8537,
+      "step": 6815
+    },
+    {
+      "epoch": 0.37514447685618363,
+      "grad_norm": 0.7020999789237976,
+      "learning_rate": 9.176185899960996e-06,
+      "loss": 0.7844,
+      "step": 6816
+    },
+    {
+      "epoch": 0.37519951565853926,
+      "grad_norm": 0.6857696175575256,
+      "learning_rate": 9.175947525372355e-06,
+      "loss": 0.8491,
+      "step": 6817
+    },
+    {
+      "epoch": 0.37525455446089495,
+      "grad_norm": 0.6882391571998596,
+      "learning_rate": 9.175709119398409e-06,
+      "loss": 0.7797,
+      "step": 6818
+    },
+    {
+      "epoch": 0.3753095932632506,
+      "grad_norm": 0.7788485288619995,
+      "learning_rate": 9.17547068204095e-06,
+      "loss": 0.6898,
+      "step": 6819
+    },
+    {
+      "epoch": 0.3753646320656063,
+      "grad_norm": 0.8529300093650818,
+      "learning_rate": 9.17523221330177e-06,
+      "loss": 0.8113,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3754196708679619,
+      "grad_norm": 0.6297540068626404,
+      "learning_rate": 9.174993713182663e-06,
+      "loss": 0.7133,
+      "step": 6821
+    },
+    {
+      "epoch": 0.3754747096703176,
+      "grad_norm": 0.8225051760673523,
+      "learning_rate": 9.174755181685422e-06,
+      "loss": 0.83,
+      "step": 6822
+    },
+    {
+      "epoch": 0.37552974847267323,
+      "grad_norm": 0.7445290684700012,
+      "learning_rate": 9.174516618811838e-06,
+      "loss": 0.8597,
+      "step": 6823
+    },
+    {
+      "epoch": 0.3755847872750289,
+      "grad_norm": 0.7890744209289551,
+      "learning_rate": 9.174278024563706e-06,
+      "loss": 0.8021,
+      "step": 6824
+    },
+    {
+      "epoch": 0.37563982607738455,
+      "grad_norm": 0.644434928894043,
+      "learning_rate": 9.174039398942815e-06,
+      "loss": 0.7154,
+      "step": 6825
+    },
+    {
+      "epoch": 0.37569486487974024,
+      "grad_norm": 0.7664980292320251,
+      "learning_rate": 9.173800741950962e-06,
+      "loss": 0.8496,
+      "step": 6826
+    },
+    {
+      "epoch": 0.3757499036820959,
+      "grad_norm": 0.8062339425086975,
+      "learning_rate": 9.173562053589942e-06,
+      "loss": 0.7736,
+      "step": 6827
+    },
+    {
+      "epoch": 0.37580494248445157,
+      "grad_norm": 0.6334213018417358,
+      "learning_rate": 9.173323333861543e-06,
+      "loss": 0.6513,
+      "step": 6828
+    },
+    {
+      "epoch": 0.3758599812868072,
+      "grad_norm": 0.6825501322746277,
+      "learning_rate": 9.173084582767567e-06,
+      "loss": 0.755,
+      "step": 6829
+    },
+    {
+      "epoch": 0.37591502008916283,
+      "grad_norm": 0.7353835105895996,
+      "learning_rate": 9.172845800309801e-06,
+      "loss": 0.7783,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3759700588915185,
+      "grad_norm": 0.7830193638801575,
+      "learning_rate": 9.172606986490046e-06,
+      "loss": 0.7352,
+      "step": 6831
+    },
+    {
+      "epoch": 0.37602509769387416,
+      "grad_norm": 0.7464943528175354,
+      "learning_rate": 9.172368141310091e-06,
+      "loss": 0.6454,
+      "step": 6832
+    },
+    {
+      "epoch": 0.37608013649622984,
+      "grad_norm": 0.7171493172645569,
+      "learning_rate": 9.172129264771736e-06,
+      "loss": 0.7978,
+      "step": 6833
+    },
+    {
+      "epoch": 0.3761351752985855,
+      "grad_norm": 0.6929624676704407,
+      "learning_rate": 9.171890356876774e-06,
+      "loss": 0.8026,
+      "step": 6834
+    },
+    {
+      "epoch": 0.37619021410094117,
+      "grad_norm": 0.7240758538246155,
+      "learning_rate": 9.171651417627e-06,
+      "loss": 0.8469,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3762452529032968,
+      "grad_norm": 0.7713736891746521,
+      "learning_rate": 9.17141244702421e-06,
+      "loss": 0.8307,
+      "step": 6836
+    },
+    {
+      "epoch": 0.3763002917056525,
+      "grad_norm": 0.7417639493942261,
+      "learning_rate": 9.171173445070203e-06,
+      "loss": 0.8165,
+      "step": 6837
+    },
+    {
+      "epoch": 0.3763553305080081,
+      "grad_norm": 0.811005711555481,
+      "learning_rate": 9.17093441176677e-06,
+      "loss": 0.8418,
+      "step": 6838
+    },
+    {
+      "epoch": 0.3764103693103638,
+      "grad_norm": 0.9996818900108337,
+      "learning_rate": 9.170695347115713e-06,
+      "loss": 0.851,
+      "step": 6839
+    },
+    {
+      "epoch": 0.37646540811271945,
+      "grad_norm": 0.7703381776809692,
+      "learning_rate": 9.170456251118824e-06,
+      "loss": 0.8308,
+      "step": 6840
+    },
+    {
+      "epoch": 0.37652044691507514,
+      "grad_norm": 0.7194466590881348,
+      "learning_rate": 9.170217123777904e-06,
+      "loss": 0.699,
+      "step": 6841
+    },
+    {
+      "epoch": 0.37657548571743077,
+      "grad_norm": 0.7146462202072144,
+      "learning_rate": 9.169977965094748e-06,
+      "loss": 0.8247,
+      "step": 6842
+    },
+    {
+      "epoch": 0.37663052451978646,
+      "grad_norm": 0.7490555047988892,
+      "learning_rate": 9.169738775071153e-06,
+      "loss": 0.8627,
+      "step": 6843
+    },
+    {
+      "epoch": 0.3766855633221421,
+      "grad_norm": 0.827996015548706,
+      "learning_rate": 9.169499553708919e-06,
+      "loss": 0.7454,
+      "step": 6844
+    },
+    {
+      "epoch": 0.3767406021244978,
+      "grad_norm": 0.7185913324356079,
+      "learning_rate": 9.16926030100984e-06,
+      "loss": 0.7018,
+      "step": 6845
+    },
+    {
+      "epoch": 0.3767956409268534,
+      "grad_norm": 0.7879654169082642,
+      "learning_rate": 9.169021016975718e-06,
+      "loss": 0.8144,
+      "step": 6846
+    },
+    {
+      "epoch": 0.3768506797292091,
+      "grad_norm": 0.7072417736053467,
+      "learning_rate": 9.168781701608352e-06,
+      "loss": 0.7572,
+      "step": 6847
+    },
+    {
+      "epoch": 0.37690571853156474,
+      "grad_norm": 0.7359803915023804,
+      "learning_rate": 9.168542354909536e-06,
+      "loss": 0.7712,
+      "step": 6848
+    },
+    {
+      "epoch": 0.3769607573339204,
+      "grad_norm": 0.7672479748725891,
+      "learning_rate": 9.168302976881072e-06,
+      "loss": 0.7696,
+      "step": 6849
+    },
+    {
+      "epoch": 0.37701579613627606,
+      "grad_norm": 0.7276006937026978,
+      "learning_rate": 9.168063567524758e-06,
+      "loss": 0.8235,
+      "step": 6850
+    },
+    {
+      "epoch": 0.37707083493863175,
+      "grad_norm": 0.673577606678009,
+      "learning_rate": 9.167824126842396e-06,
+      "loss": 0.6515,
+      "step": 6851
+    },
+    {
+      "epoch": 0.3771258737409874,
+      "grad_norm": 0.7257997989654541,
+      "learning_rate": 9.167584654835782e-06,
+      "loss": 0.729,
+      "step": 6852
+    },
+    {
+      "epoch": 0.37718091254334307,
+      "grad_norm": 0.6655071377754211,
+      "learning_rate": 9.167345151506717e-06,
+      "loss": 0.7917,
+      "step": 6853
+    },
+    {
+      "epoch": 0.3772359513456987,
+      "grad_norm": 0.7603726983070374,
+      "learning_rate": 9.167105616857002e-06,
+      "loss": 0.8383,
+      "step": 6854
+    },
+    {
+      "epoch": 0.3772909901480544,
+      "grad_norm": 0.7066939473152161,
+      "learning_rate": 9.166866050888437e-06,
+      "loss": 0.7589,
+      "step": 6855
+    },
+    {
+      "epoch": 0.37734602895041003,
+      "grad_norm": 0.7002355456352234,
+      "learning_rate": 9.16662645360282e-06,
+      "loss": 0.8305,
+      "step": 6856
+    },
+    {
+      "epoch": 0.3774010677527657,
+      "grad_norm": 0.9499780535697937,
+      "learning_rate": 9.166386825001957e-06,
+      "loss": 0.78,
+      "step": 6857
+    },
+    {
+      "epoch": 0.37745610655512135,
+      "grad_norm": 0.7136938571929932,
+      "learning_rate": 9.166147165087645e-06,
+      "loss": 0.7449,
+      "step": 6858
+    },
+    {
+      "epoch": 0.37751114535747704,
+      "grad_norm": 0.740443766117096,
+      "learning_rate": 9.165907473861687e-06,
+      "loss": 0.8228,
+      "step": 6859
+    },
+    {
+      "epoch": 0.3775661841598327,
+      "grad_norm": 0.7649856209754944,
+      "learning_rate": 9.165667751325879e-06,
+      "loss": 0.7762,
+      "step": 6860
+    },
+    {
+      "epoch": 0.37762122296218836,
+      "grad_norm": 0.743251383304596,
+      "learning_rate": 9.165427997482032e-06,
+      "loss": 0.7536,
+      "step": 6861
+    },
+    {
+      "epoch": 0.377676261764544,
+      "grad_norm": 0.7023851871490479,
+      "learning_rate": 9.165188212331941e-06,
+      "loss": 0.7327,
+      "step": 6862
+    },
+    {
+      "epoch": 0.3777313005668997,
+      "grad_norm": 0.7304333448410034,
+      "learning_rate": 9.164948395877411e-06,
+      "loss": 0.8816,
+      "step": 6863
+    },
+    {
+      "epoch": 0.3777863393692553,
+      "grad_norm": 0.6666659116744995,
+      "learning_rate": 9.164708548120244e-06,
+      "loss": 0.7821,
+      "step": 6864
+    },
+    {
+      "epoch": 0.377841378171611,
+      "grad_norm": 0.6542865037918091,
+      "learning_rate": 9.164468669062242e-06,
+      "loss": 0.7044,
+      "step": 6865
+    },
+    {
+      "epoch": 0.37789641697396664,
+      "grad_norm": 0.7436043620109558,
+      "learning_rate": 9.16422875870521e-06,
+      "loss": 0.8492,
+      "step": 6866
+    },
+    {
+      "epoch": 0.37795145577632233,
+      "grad_norm": 0.7660424709320068,
+      "learning_rate": 9.163988817050947e-06,
+      "loss": 0.7236,
+      "step": 6867
+    },
+    {
+      "epoch": 0.37800649457867797,
+      "grad_norm": 0.7288914918899536,
+      "learning_rate": 9.16374884410126e-06,
+      "loss": 0.6361,
+      "step": 6868
+    },
+    {
+      "epoch": 0.37806153338103365,
+      "grad_norm": 0.884832501411438,
+      "learning_rate": 9.163508839857948e-06,
+      "loss": 0.8112,
+      "step": 6869
+    },
+    {
+      "epoch": 0.3781165721833893,
+      "grad_norm": 0.937660813331604,
+      "learning_rate": 9.163268804322822e-06,
+      "loss": 0.6405,
+      "step": 6870
+    },
+    {
+      "epoch": 0.378171610985745,
+      "grad_norm": 0.8295212388038635,
+      "learning_rate": 9.16302873749768e-06,
+      "loss": 0.8107,
+      "step": 6871
+    },
+    {
+      "epoch": 0.3782266497881006,
+      "grad_norm": 1.0573647022247314,
+      "learning_rate": 9.16278863938433e-06,
+      "loss": 0.7792,
+      "step": 6872
+    },
+    {
+      "epoch": 0.37828168859045624,
+      "grad_norm": 0.8450027108192444,
+      "learning_rate": 9.162548509984574e-06,
+      "loss": 0.8103,
+      "step": 6873
+    },
+    {
+      "epoch": 0.37833672739281193,
+      "grad_norm": 0.7372947931289673,
+      "learning_rate": 9.162308349300218e-06,
+      "loss": 0.8232,
+      "step": 6874
+    },
+    {
+      "epoch": 0.37839176619516757,
+      "grad_norm": 0.7573776841163635,
+      "learning_rate": 9.162068157333066e-06,
+      "loss": 0.773,
+      "step": 6875
+    },
+    {
+      "epoch": 0.37844680499752326,
+      "grad_norm": 0.7883201241493225,
+      "learning_rate": 9.161827934084924e-06,
+      "loss": 0.7561,
+      "step": 6876
+    },
+    {
+      "epoch": 0.3785018437998789,
+      "grad_norm": 0.7195025086402893,
+      "learning_rate": 9.161587679557598e-06,
+      "loss": 0.798,
+      "step": 6877
+    },
+    {
+      "epoch": 0.3785568826022346,
+      "grad_norm": 0.7047843337059021,
+      "learning_rate": 9.161347393752891e-06,
+      "loss": 0.8122,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3786119214045902,
+      "grad_norm": 0.7354363203048706,
+      "learning_rate": 9.161107076672613e-06,
+      "loss": 0.7296,
+      "step": 6879
+    },
+    {
+      "epoch": 0.3786669602069459,
+      "grad_norm": 0.7748313546180725,
+      "learning_rate": 9.160866728318567e-06,
+      "loss": 0.9576,
+      "step": 6880
+    },
+    {
+      "epoch": 0.37872199900930154,
+      "grad_norm": 0.7197638750076294,
+      "learning_rate": 9.16062634869256e-06,
+      "loss": 0.8054,
+      "step": 6881
+    },
+    {
+      "epoch": 0.3787770378116572,
+      "grad_norm": 0.7086492776870728,
+      "learning_rate": 9.1603859377964e-06,
+      "loss": 0.8938,
+      "step": 6882
+    },
+    {
+      "epoch": 0.37883207661401286,
+      "grad_norm": 0.7764425873756409,
+      "learning_rate": 9.160145495631894e-06,
+      "loss": 0.7562,
+      "step": 6883
+    },
+    {
+      "epoch": 0.37888711541636855,
+      "grad_norm": 0.7673479914665222,
+      "learning_rate": 9.159905022200846e-06,
+      "loss": 0.6783,
+      "step": 6884
+    },
+    {
+      "epoch": 0.3789421542187242,
+      "grad_norm": 0.7323669195175171,
+      "learning_rate": 9.159664517505067e-06,
+      "loss": 0.8274,
+      "step": 6885
+    },
+    {
+      "epoch": 0.37899719302107987,
+      "grad_norm": 0.8283136487007141,
+      "learning_rate": 9.159423981546362e-06,
+      "loss": 0.7184,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3790522318234355,
+      "grad_norm": 0.6949145793914795,
+      "learning_rate": 9.15918341432654e-06,
+      "loss": 0.7843,
+      "step": 6887
+    },
+    {
+      "epoch": 0.3791072706257912,
+      "grad_norm": 0.8584639430046082,
+      "learning_rate": 9.158942815847408e-06,
+      "loss": 0.71,
+      "step": 6888
+    },
+    {
+      "epoch": 0.3791623094281468,
+      "grad_norm": 0.7125271558761597,
+      "learning_rate": 9.158702186110777e-06,
+      "loss": 0.7432,
+      "step": 6889
+    },
+    {
+      "epoch": 0.3792173482305025,
+      "grad_norm": 0.6657430529594421,
+      "learning_rate": 9.158461525118452e-06,
+      "loss": 0.6715,
+      "step": 6890
+    },
+    {
+      "epoch": 0.37927238703285815,
+      "grad_norm": 0.770226240158081,
+      "learning_rate": 9.158220832872243e-06,
+      "loss": 0.7029,
+      "step": 6891
+    },
+    {
+      "epoch": 0.37932742583521384,
+      "grad_norm": 0.7697272300720215,
+      "learning_rate": 9.15798010937396e-06,
+      "loss": 0.686,
+      "step": 6892
+    },
+    {
+      "epoch": 0.37938246463756947,
+      "grad_norm": 0.7693290710449219,
+      "learning_rate": 9.157739354625413e-06,
+      "loss": 0.7669,
+      "step": 6893
+    },
+    {
+      "epoch": 0.37943750343992516,
+      "grad_norm": 0.8365996479988098,
+      "learning_rate": 9.157498568628406e-06,
+      "loss": 0.8254,
+      "step": 6894
+    },
+    {
+      "epoch": 0.3794925422422808,
+      "grad_norm": 0.8075883388519287,
+      "learning_rate": 9.157257751384756e-06,
+      "loss": 0.8311,
+      "step": 6895
+    },
+    {
+      "epoch": 0.3795475810446365,
+      "grad_norm": 0.8422812819480896,
+      "learning_rate": 9.15701690289627e-06,
+      "loss": 0.9173,
+      "step": 6896
+    },
+    {
+      "epoch": 0.3796026198469921,
+      "grad_norm": 0.7930355072021484,
+      "learning_rate": 9.156776023164755e-06,
+      "loss": 0.9376,
+      "step": 6897
+    },
+    {
+      "epoch": 0.3796576586493478,
+      "grad_norm": 0.7877563238143921,
+      "learning_rate": 9.156535112192026e-06,
+      "loss": 0.8358,
+      "step": 6898
+    },
+    {
+      "epoch": 0.37971269745170344,
+      "grad_norm": 0.7712885141372681,
+      "learning_rate": 9.156294169979891e-06,
+      "loss": 0.8781,
+      "step": 6899
+    },
+    {
+      "epoch": 0.37976773625405913,
+      "grad_norm": 0.6953728199005127,
+      "learning_rate": 9.156053196530162e-06,
+      "loss": 0.7861,
+      "step": 6900
+    },
+    {
+      "epoch": 0.37982277505641476,
+      "grad_norm": 0.9581564664840698,
+      "learning_rate": 9.155812191844649e-06,
+      "loss": 0.8294,
+      "step": 6901
+    },
+    {
+      "epoch": 0.37987781385877045,
+      "grad_norm": 0.738571286201477,
+      "learning_rate": 9.155571155925166e-06,
+      "loss": 0.7998,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3799328526611261,
+      "grad_norm": 0.7059765458106995,
+      "learning_rate": 9.155330088773519e-06,
+      "loss": 0.7877,
+      "step": 6903
+    },
+    {
+      "epoch": 0.3799878914634818,
+      "grad_norm": 0.8572642207145691,
+      "learning_rate": 9.155088990391527e-06,
+      "loss": 0.7333,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3800429302658374,
+      "grad_norm": 0.7442637085914612,
+      "learning_rate": 9.154847860780996e-06,
+      "loss": 0.685,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3800979690681931,
+      "grad_norm": 0.7787682414054871,
+      "learning_rate": 9.154606699943741e-06,
+      "loss": 0.7893,
+      "step": 6906
+    },
+    {
+      "epoch": 0.38015300787054873,
+      "grad_norm": 0.8973822593688965,
+      "learning_rate": 9.154365507881574e-06,
+      "loss": 0.8297,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3802080466729044,
+      "grad_norm": 0.7759919166564941,
+      "learning_rate": 9.154124284596311e-06,
+      "loss": 0.8257,
+      "step": 6908
+    },
+    {
+      "epoch": 0.38026308547526005,
+      "grad_norm": 0.8042850494384766,
+      "learning_rate": 9.153883030089759e-06,
+      "loss": 0.8024,
+      "step": 6909
+    },
+    {
+      "epoch": 0.38031812427761574,
+      "grad_norm": 0.8285790085792542,
+      "learning_rate": 9.153641744363733e-06,
+      "loss": 0.7824,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3803731630799714,
+      "grad_norm": 0.7225445508956909,
+      "learning_rate": 9.15340042742005e-06,
+      "loss": 0.8065,
+      "step": 6911
+    },
+    {
+      "epoch": 0.38042820188232707,
+      "grad_norm": 0.7685298919677734,
+      "learning_rate": 9.15315907926052e-06,
+      "loss": 0.8151,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3804832406846827,
+      "grad_norm": 0.9005589485168457,
+      "learning_rate": 9.152917699886958e-06,
+      "loss": 0.8413,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3805382794870384,
+      "grad_norm": 0.8715279698371887,
+      "learning_rate": 9.152676289301178e-06,
+      "loss": 0.7233,
+      "step": 6914
+    },
+    {
+      "epoch": 0.380593318289394,
+      "grad_norm": 0.8764133453369141,
+      "learning_rate": 9.152434847504996e-06,
+      "loss": 0.783,
+      "step": 6915
+    },
+    {
+      "epoch": 0.38064835709174966,
+      "grad_norm": 0.6847019195556641,
+      "learning_rate": 9.152193374500225e-06,
+      "loss": 0.7133,
+      "step": 6916
+    },
+    {
+      "epoch": 0.38070339589410535,
+      "grad_norm": 0.7562721371650696,
+      "learning_rate": 9.151951870288678e-06,
+      "loss": 0.8155,
+      "step": 6917
+    },
+    {
+      "epoch": 0.380758434696461,
+      "grad_norm": 0.6888439059257507,
+      "learning_rate": 9.151710334872173e-06,
+      "loss": 0.6395,
+      "step": 6918
+    },
+    {
+      "epoch": 0.38081347349881667,
+      "grad_norm": 1.0951511859893799,
+      "learning_rate": 9.151468768252525e-06,
+      "loss": 0.8936,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3808685123011723,
+      "grad_norm": 0.7261115908622742,
+      "learning_rate": 9.151227170431549e-06,
+      "loss": 0.7864,
+      "step": 6920
+    },
+    {
+      "epoch": 0.380923551103528,
+      "grad_norm": 1.2851859331130981,
+      "learning_rate": 9.150985541411061e-06,
+      "loss": 0.9419,
+      "step": 6921
+    },
+    {
+      "epoch": 0.3809785899058836,
+      "grad_norm": 0.7621721625328064,
+      "learning_rate": 9.150743881192876e-06,
+      "loss": 0.7773,
+      "step": 6922
+    },
+    {
+      "epoch": 0.3810336287082393,
+      "grad_norm": 0.7605605721473694,
+      "learning_rate": 9.150502189778811e-06,
+      "loss": 0.8752,
+      "step": 6923
+    },
+    {
+      "epoch": 0.38108866751059495,
+      "grad_norm": 0.8422327041625977,
+      "learning_rate": 9.150260467170683e-06,
+      "loss": 0.8555,
+      "step": 6924
+    },
+    {
+      "epoch": 0.38114370631295064,
+      "grad_norm": 0.7227829098701477,
+      "learning_rate": 9.15001871337031e-06,
+      "loss": 0.7637,
+      "step": 6925
+    },
+    {
+      "epoch": 0.38119874511530627,
+      "grad_norm": 0.6568942666053772,
+      "learning_rate": 9.149776928379506e-06,
+      "loss": 0.6944,
+      "step": 6926
+    },
+    {
+      "epoch": 0.38125378391766196,
+      "grad_norm": 0.9317567944526672,
+      "learning_rate": 9.149535112200087e-06,
+      "loss": 0.8098,
+      "step": 6927
+    },
+    {
+      "epoch": 0.3813088227200176,
+      "grad_norm": 0.6374759674072266,
+      "learning_rate": 9.149293264833877e-06,
+      "loss": 0.6654,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3813638615223733,
+      "grad_norm": 0.7276837825775146,
+      "learning_rate": 9.149051386282685e-06,
+      "loss": 0.7728,
+      "step": 6929
+    },
+    {
+      "epoch": 0.3814189003247289,
+      "grad_norm": 0.7573683261871338,
+      "learning_rate": 9.148809476548337e-06,
+      "loss": 0.7681,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3814739391270846,
+      "grad_norm": 0.7535703778266907,
+      "learning_rate": 9.148567535632647e-06,
+      "loss": 0.8498,
+      "step": 6931
+    },
+    {
+      "epoch": 0.38152897792944024,
+      "grad_norm": 0.7510126233100891,
+      "learning_rate": 9.148325563537432e-06,
+      "loss": 0.7874,
+      "step": 6932
+    },
+    {
+      "epoch": 0.3815840167317959,
+      "grad_norm": 0.7809224724769592,
+      "learning_rate": 9.148083560264515e-06,
+      "loss": 0.7223,
+      "step": 6933
+    },
+    {
+      "epoch": 0.38163905553415156,
+      "grad_norm": 0.7433155179023743,
+      "learning_rate": 9.14784152581571e-06,
+      "loss": 0.7914,
+      "step": 6934
+    },
+    {
+      "epoch": 0.38169409433650725,
+      "grad_norm": 0.7142858505249023,
+      "learning_rate": 9.14759946019284e-06,
+      "loss": 0.781,
+      "step": 6935
+    },
+    {
+      "epoch": 0.3817491331388629,
+      "grad_norm": 0.7910202741622925,
+      "learning_rate": 9.147357363397721e-06,
+      "loss": 0.755,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3818041719412186,
+      "grad_norm": 1.007727026939392,
+      "learning_rate": 9.147115235432176e-06,
+      "loss": 0.7809,
+      "step": 6937
+    },
+    {
+      "epoch": 0.3818592107435742,
+      "grad_norm": 0.7227005362510681,
+      "learning_rate": 9.146873076298024e-06,
+      "loss": 0.7276,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3819142495459299,
+      "grad_norm": 0.6945967674255371,
+      "learning_rate": 9.146630885997081e-06,
+      "loss": 0.825,
+      "step": 6939
+    },
+    {
+      "epoch": 0.38196928834828553,
+      "grad_norm": 0.6719669103622437,
+      "learning_rate": 9.146388664531172e-06,
+      "loss": 0.6486,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3820243271506412,
+      "grad_norm": 0.7528467178344727,
+      "learning_rate": 9.146146411902115e-06,
+      "loss": 0.8143,
+      "step": 6941
+    },
+    {
+      "epoch": 0.38207936595299685,
+      "grad_norm": 0.6835548877716064,
+      "learning_rate": 9.145904128111732e-06,
+      "loss": 0.7742,
+      "step": 6942
+    },
+    {
+      "epoch": 0.38213440475535254,
+      "grad_norm": 0.7829870581626892,
+      "learning_rate": 9.145661813161844e-06,
+      "loss": 0.8147,
+      "step": 6943
+    },
+    {
+      "epoch": 0.3821894435577082,
+      "grad_norm": 0.6833155155181885,
+      "learning_rate": 9.145419467054271e-06,
+      "loss": 0.7615,
+      "step": 6944
+    },
+    {
+      "epoch": 0.38224448236006386,
+      "grad_norm": 0.7577275037765503,
+      "learning_rate": 9.145177089790833e-06,
+      "loss": 0.8611,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3822995211624195,
+      "grad_norm": 0.7102984189987183,
+      "learning_rate": 9.144934681373356e-06,
+      "loss": 0.8373,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3823545599647752,
+      "grad_norm": 0.6906121373176575,
+      "learning_rate": 9.144692241803658e-06,
+      "loss": 0.8314,
+      "step": 6947
+    },
+    {
+      "epoch": 0.3824095987671308,
+      "grad_norm": 0.7790967226028442,
+      "learning_rate": 9.144449771083563e-06,
+      "loss": 0.8285,
+      "step": 6948
+    },
+    {
+      "epoch": 0.3824646375694865,
+      "grad_norm": 0.8420237898826599,
+      "learning_rate": 9.144207269214893e-06,
+      "loss": 0.8159,
+      "step": 6949
+    },
+    {
+      "epoch": 0.38251967637184214,
+      "grad_norm": 0.7944310307502747,
+      "learning_rate": 9.143964736199471e-06,
+      "loss": 0.7981,
+      "step": 6950
+    },
+    {
+      "epoch": 0.38257471517419783,
+      "grad_norm": 0.7610076069831848,
+      "learning_rate": 9.14372217203912e-06,
+      "loss": 0.8011,
+      "step": 6951
+    },
+    {
+      "epoch": 0.38262975397655347,
+      "grad_norm": 0.7183333039283752,
+      "learning_rate": 9.143479576735661e-06,
+      "loss": 0.7504,
+      "step": 6952
+    },
+    {
+      "epoch": 0.38268479277890916,
+      "grad_norm": 0.7363573312759399,
+      "learning_rate": 9.14323695029092e-06,
+      "loss": 0.7561,
+      "step": 6953
+    },
+    {
+      "epoch": 0.3827398315812648,
+      "grad_norm": 0.7330427765846252,
+      "learning_rate": 9.142994292706716e-06,
+      "loss": 0.754,
+      "step": 6954
+    },
+    {
+      "epoch": 0.3827948703836205,
+      "grad_norm": 0.8307509422302246,
+      "learning_rate": 9.142751603984879e-06,
+      "loss": 0.8059,
+      "step": 6955
+    },
+    {
+      "epoch": 0.3828499091859761,
+      "grad_norm": 0.7340347766876221,
+      "learning_rate": 9.142508884127228e-06,
+      "loss": 0.8636,
+      "step": 6956
+    },
+    {
+      "epoch": 0.3829049479883318,
+      "grad_norm": 0.7032678127288818,
+      "learning_rate": 9.14226613313559e-06,
+      "loss": 0.8237,
+      "step": 6957
+    },
+    {
+      "epoch": 0.38295998679068743,
+      "grad_norm": 0.769809365272522,
+      "learning_rate": 9.142023351011788e-06,
+      "loss": 0.7523,
+      "step": 6958
+    },
+    {
+      "epoch": 0.38301502559304307,
+      "grad_norm": 0.7446833252906799,
+      "learning_rate": 9.141780537757647e-06,
+      "loss": 0.8382,
+      "step": 6959
+    },
+    {
+      "epoch": 0.38307006439539876,
+      "grad_norm": 0.6926285028457642,
+      "learning_rate": 9.141537693374994e-06,
+      "loss": 0.7997,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3831251031977544,
+      "grad_norm": 0.7303034067153931,
+      "learning_rate": 9.141294817865651e-06,
+      "loss": 0.794,
+      "step": 6961
+    },
+    {
+      "epoch": 0.3831801420001101,
+      "grad_norm": 0.7453297972679138,
+      "learning_rate": 9.141051911231445e-06,
+      "loss": 0.7031,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3832351808024657,
+      "grad_norm": 0.8503912091255188,
+      "learning_rate": 9.140808973474201e-06,
+      "loss": 0.7855,
+      "step": 6963
+    },
+    {
+      "epoch": 0.3832902196048214,
+      "grad_norm": 0.7304036617279053,
+      "learning_rate": 9.140566004595746e-06,
+      "loss": 0.7062,
+      "step": 6964
+    },
+    {
+      "epoch": 0.38334525840717704,
+      "grad_norm": 0.7534968852996826,
+      "learning_rate": 9.140323004597904e-06,
+      "loss": 0.8138,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3834002972095327,
+      "grad_norm": 0.8122013807296753,
+      "learning_rate": 9.140079973482503e-06,
+      "loss": 0.7769,
+      "step": 6966
+    },
+    {
+      "epoch": 0.38345533601188836,
+      "grad_norm": 0.7345744967460632,
+      "learning_rate": 9.13983691125137e-06,
+      "loss": 0.7588,
+      "step": 6967
+    },
+    {
+      "epoch": 0.38351037481424405,
+      "grad_norm": 0.7251620292663574,
+      "learning_rate": 9.13959381790633e-06,
+      "loss": 0.8027,
+      "step": 6968
+    },
+    {
+      "epoch": 0.3835654136165997,
+      "grad_norm": 0.7157594561576843,
+      "learning_rate": 9.139350693449212e-06,
+      "loss": 0.7233,
+      "step": 6969
+    },
+    {
+      "epoch": 0.38362045241895537,
+      "grad_norm": 0.8076621890068054,
+      "learning_rate": 9.139107537881842e-06,
+      "loss": 0.7256,
+      "step": 6970
+    },
+    {
+      "epoch": 0.383675491221311,
+      "grad_norm": 0.717182993888855,
+      "learning_rate": 9.138864351206047e-06,
+      "loss": 0.7003,
+      "step": 6971
+    },
+    {
+      "epoch": 0.3837305300236667,
+      "grad_norm": 0.7534194588661194,
+      "learning_rate": 9.138621133423656e-06,
+      "loss": 0.7315,
+      "step": 6972
+    },
+    {
+      "epoch": 0.3837855688260223,
+      "grad_norm": 0.6400160193443298,
+      "learning_rate": 9.138377884536494e-06,
+      "loss": 0.6814,
+      "step": 6973
+    },
+    {
+      "epoch": 0.383840607628378,
+      "grad_norm": 0.7319507002830505,
+      "learning_rate": 9.138134604546394e-06,
+      "loss": 0.7942,
+      "step": 6974
+    },
+    {
+      "epoch": 0.38389564643073365,
+      "grad_norm": 0.7109829783439636,
+      "learning_rate": 9.137891293455181e-06,
+      "loss": 0.7528,
+      "step": 6975
+    },
+    {
+      "epoch": 0.38395068523308934,
+      "grad_norm": 1.006724238395691,
+      "learning_rate": 9.137647951264685e-06,
+      "loss": 0.7652,
+      "step": 6976
+    },
+    {
+      "epoch": 0.384005724035445,
+      "grad_norm": 0.7080540060997009,
+      "learning_rate": 9.137404577976736e-06,
+      "loss": 0.7706,
+      "step": 6977
+    },
+    {
+      "epoch": 0.38406076283780066,
+      "grad_norm": 0.7551368474960327,
+      "learning_rate": 9.137161173593161e-06,
+      "loss": 0.8202,
+      "step": 6978
+    },
+    {
+      "epoch": 0.3841158016401563,
+      "grad_norm": 0.6624314785003662,
+      "learning_rate": 9.13691773811579e-06,
+      "loss": 0.7258,
+      "step": 6979
+    },
+    {
+      "epoch": 0.384170840442512,
+      "grad_norm": 0.9603848457336426,
+      "learning_rate": 9.136674271546451e-06,
+      "loss": 0.9415,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3842258792448676,
+      "grad_norm": 0.6964829564094543,
+      "learning_rate": 9.136430773886977e-06,
+      "loss": 0.7604,
+      "step": 6981
+    },
+    {
+      "epoch": 0.3842809180472233,
+      "grad_norm": 0.6503588557243347,
+      "learning_rate": 9.136187245139197e-06,
+      "loss": 0.7141,
+      "step": 6982
+    },
+    {
+      "epoch": 0.38433595684957894,
+      "grad_norm": 0.9179829359054565,
+      "learning_rate": 9.13594368530494e-06,
+      "loss": 0.7619,
+      "step": 6983
+    },
+    {
+      "epoch": 0.38439099565193463,
+      "grad_norm": 0.7993278503417969,
+      "learning_rate": 9.135700094386038e-06,
+      "loss": 0.832,
+      "step": 6984
+    },
+    {
+      "epoch": 0.38444603445429026,
+      "grad_norm": 0.8136988282203674,
+      "learning_rate": 9.13545647238432e-06,
+      "loss": 0.8127,
+      "step": 6985
+    },
+    {
+      "epoch": 0.38450107325664595,
+      "grad_norm": 0.9918104410171509,
+      "learning_rate": 9.135212819301619e-06,
+      "loss": 0.836,
+      "step": 6986
+    },
+    {
+      "epoch": 0.3845561120590016,
+      "grad_norm": 0.7767511010169983,
+      "learning_rate": 9.134969135139765e-06,
+      "loss": 0.8391,
+      "step": 6987
+    },
+    {
+      "epoch": 0.3846111508613573,
+      "grad_norm": 0.6889285445213318,
+      "learning_rate": 9.134725419900589e-06,
+      "loss": 0.7639,
+      "step": 6988
+    },
+    {
+      "epoch": 0.3846661896637129,
+      "grad_norm": 1.803467035293579,
+      "learning_rate": 9.134481673585924e-06,
+      "loss": 0.7629,
+      "step": 6989
+    },
+    {
+      "epoch": 0.3847212284660686,
+      "grad_norm": 0.721581757068634,
+      "learning_rate": 9.134237896197603e-06,
+      "loss": 0.8194,
+      "step": 6990
+    },
+    {
+      "epoch": 0.38477626726842423,
+      "grad_norm": 0.8163189888000488,
+      "learning_rate": 9.133994087737456e-06,
+      "loss": 0.7789,
+      "step": 6991
+    },
+    {
+      "epoch": 0.3848313060707799,
+      "grad_norm": 0.7518420815467834,
+      "learning_rate": 9.133750248207315e-06,
+      "loss": 0.7529,
+      "step": 6992
+    },
+    {
+      "epoch": 0.38488634487313556,
+      "grad_norm": 0.7318000197410583,
+      "learning_rate": 9.133506377609015e-06,
+      "loss": 0.7829,
+      "step": 6993
+    },
+    {
+      "epoch": 0.38494138367549124,
+      "grad_norm": 0.7765058875083923,
+      "learning_rate": 9.133262475944386e-06,
+      "loss": 0.7902,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3849964224778469,
+      "grad_norm": 0.845567524433136,
+      "learning_rate": 9.133018543215265e-06,
+      "loss": 0.8117,
+      "step": 6995
+    },
+    {
+      "epoch": 0.38505146128020257,
+      "grad_norm": 0.7081887125968933,
+      "learning_rate": 9.13277457942348e-06,
+      "loss": 0.8131,
+      "step": 6996
+    },
+    {
+      "epoch": 0.3851065000825582,
+      "grad_norm": 0.7447869777679443,
+      "learning_rate": 9.132530584570869e-06,
+      "loss": 0.7765,
+      "step": 6997
+    },
+    {
+      "epoch": 0.3851615388849139,
+      "grad_norm": 0.8554795384407043,
+      "learning_rate": 9.132286558659265e-06,
+      "loss": 0.8966,
+      "step": 6998
+    },
+    {
+      "epoch": 0.3852165776872695,
+      "grad_norm": 0.7117023468017578,
+      "learning_rate": 9.1320425016905e-06,
+      "loss": 0.7461,
+      "step": 6999
+    },
+    {
+      "epoch": 0.3852716164896252,
+      "grad_norm": 0.6965934038162231,
+      "learning_rate": 9.131798413666411e-06,
+      "loss": 0.6827,
+      "step": 7000
+    },
+    {
+      "epoch": 0.38532665529198085,
+      "grad_norm": 0.7449018359184265,
+      "learning_rate": 9.13155429458883e-06,
+      "loss": 0.7562,
+      "step": 7001
+    },
+    {
+      "epoch": 0.3853816940943365,
+      "grad_norm": 0.7764221429824829,
+      "learning_rate": 9.131310144459593e-06,
+      "loss": 0.7842,
+      "step": 7002
+    },
+    {
+      "epoch": 0.38543673289669217,
+      "grad_norm": 0.9788658618927002,
+      "learning_rate": 9.131065963280536e-06,
+      "loss": 0.7857,
+      "step": 7003
+    },
+    {
+      "epoch": 0.3854917716990478,
+      "grad_norm": 0.7900908589363098,
+      "learning_rate": 9.13082175105349e-06,
+      "loss": 0.8733,
+      "step": 7004
+    },
+    {
+      "epoch": 0.3855468105014035,
+      "grad_norm": 0.814822793006897,
+      "learning_rate": 9.130577507780298e-06,
+      "loss": 0.8032,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3856018493037591,
+      "grad_norm": 1.0648475885391235,
+      "learning_rate": 9.130333233462789e-06,
+      "loss": 0.8078,
+      "step": 7006
+    },
+    {
+      "epoch": 0.3856568881061148,
+      "grad_norm": 0.7359917163848877,
+      "learning_rate": 9.130088928102799e-06,
+      "loss": 0.6491,
+      "step": 7007
+    },
+    {
+      "epoch": 0.38571192690847045,
+      "grad_norm": 0.7321771383285522,
+      "learning_rate": 9.129844591702169e-06,
+      "loss": 0.7663,
+      "step": 7008
+    },
+    {
+      "epoch": 0.38576696571082614,
+      "grad_norm": 0.6937146186828613,
+      "learning_rate": 9.129600224262732e-06,
+      "loss": 0.7835,
+      "step": 7009
+    },
+    {
+      "epoch": 0.38582200451318177,
+      "grad_norm": 0.7330107688903809,
+      "learning_rate": 9.129355825786323e-06,
+      "loss": 0.7626,
+      "step": 7010
+    },
+    {
+      "epoch": 0.38587704331553746,
+      "grad_norm": 0.7021715044975281,
+      "learning_rate": 9.129111396274783e-06,
+      "loss": 0.7115,
+      "step": 7011
+    },
+    {
+      "epoch": 0.3859320821178931,
+      "grad_norm": 0.6599563360214233,
+      "learning_rate": 9.128866935729947e-06,
+      "loss": 0.6554,
+      "step": 7012
+    },
+    {
+      "epoch": 0.3859871209202488,
+      "grad_norm": 0.7323513031005859,
+      "learning_rate": 9.128622444153652e-06,
+      "loss": 0.7392,
+      "step": 7013
+    },
+    {
+      "epoch": 0.3860421597226044,
+      "grad_norm": 0.681888222694397,
+      "learning_rate": 9.128377921547736e-06,
+      "loss": 0.7474,
+      "step": 7014
+    },
+    {
+      "epoch": 0.3860971985249601,
+      "grad_norm": 0.8454889059066772,
+      "learning_rate": 9.128133367914036e-06,
+      "loss": 0.8355,
+      "step": 7015
+    },
+    {
+      "epoch": 0.38615223732731574,
+      "grad_norm": 0.7514123916625977,
+      "learning_rate": 9.12788878325439e-06,
+      "loss": 0.7683,
+      "step": 7016
+    },
+    {
+      "epoch": 0.38620727612967143,
+      "grad_norm": 0.7317092418670654,
+      "learning_rate": 9.12764416757064e-06,
+      "loss": 0.7201,
+      "step": 7017
+    },
+    {
+      "epoch": 0.38626231493202706,
+      "grad_norm": 0.7626729011535645,
+      "learning_rate": 9.127399520864619e-06,
+      "loss": 0.7701,
+      "step": 7018
+    },
+    {
+      "epoch": 0.38631735373438275,
+      "grad_norm": 0.9790363311767578,
+      "learning_rate": 9.127154843138168e-06,
+      "loss": 0.8034,
+      "step": 7019
+    },
+    {
+      "epoch": 0.3863723925367384,
+      "grad_norm": 0.663593590259552,
+      "learning_rate": 9.126910134393125e-06,
+      "loss": 0.661,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3864274313390941,
+      "grad_norm": 0.6599924564361572,
+      "learning_rate": 9.126665394631332e-06,
+      "loss": 0.7395,
+      "step": 7021
+    },
+    {
+      "epoch": 0.3864824701414497,
+      "grad_norm": 0.8493411540985107,
+      "learning_rate": 9.126420623854625e-06,
+      "loss": 0.8008,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3865375089438054,
+      "grad_norm": 0.7587194442749023,
+      "learning_rate": 9.126175822064846e-06,
+      "loss": 0.7533,
+      "step": 7023
+    },
+    {
+      "epoch": 0.38659254774616103,
+      "grad_norm": 0.773764431476593,
+      "learning_rate": 9.125930989263835e-06,
+      "loss": 0.75,
+      "step": 7024
+    },
+    {
+      "epoch": 0.3866475865485167,
+      "grad_norm": 0.7126749753952026,
+      "learning_rate": 9.12568612545343e-06,
+      "loss": 0.7794,
+      "step": 7025
+    },
+    {
+      "epoch": 0.38670262535087235,
+      "grad_norm": 0.7404584884643555,
+      "learning_rate": 9.125441230635472e-06,
+      "loss": 0.7264,
+      "step": 7026
+    },
+    {
+      "epoch": 0.38675766415322804,
+      "grad_norm": 0.8057644367218018,
+      "learning_rate": 9.125196304811804e-06,
+      "loss": 0.8058,
+      "step": 7027
+    },
+    {
+      "epoch": 0.3868127029555837,
+      "grad_norm": 0.9586995840072632,
+      "learning_rate": 9.124951347984263e-06,
+      "loss": 0.7659,
+      "step": 7028
+    },
+    {
+      "epoch": 0.38686774175793937,
+      "grad_norm": 0.7567793726921082,
+      "learning_rate": 9.124706360154693e-06,
+      "loss": 0.8961,
+      "step": 7029
+    },
+    {
+      "epoch": 0.386922780560295,
+      "grad_norm": 0.8523182272911072,
+      "learning_rate": 9.124461341324934e-06,
+      "loss": 0.8815,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3869778193626507,
+      "grad_norm": 0.7466379404067993,
+      "learning_rate": 9.124216291496826e-06,
+      "loss": 0.7817,
+      "step": 7031
+    },
+    {
+      "epoch": 0.3870328581650063,
+      "grad_norm": 0.6721325516700745,
+      "learning_rate": 9.123971210672214e-06,
+      "loss": 0.7637,
+      "step": 7032
+    },
+    {
+      "epoch": 0.387087896967362,
+      "grad_norm": 0.6620928049087524,
+      "learning_rate": 9.123726098852936e-06,
+      "loss": 0.6956,
+      "step": 7033
+    },
+    {
+      "epoch": 0.38714293576971764,
+      "grad_norm": 0.6784290671348572,
+      "learning_rate": 9.12348095604084e-06,
+      "loss": 0.7034,
+      "step": 7034
+    },
+    {
+      "epoch": 0.38719797457207333,
+      "grad_norm": 0.7138848304748535,
+      "learning_rate": 9.123235782237763e-06,
+      "loss": 0.6037,
+      "step": 7035
+    },
+    {
+      "epoch": 0.38725301337442897,
+      "grad_norm": 0.8473613858222961,
+      "learning_rate": 9.122990577445548e-06,
+      "loss": 0.8157,
+      "step": 7036
+    },
+    {
+      "epoch": 0.38730805217678466,
+      "grad_norm": 0.835381031036377,
+      "learning_rate": 9.122745341666041e-06,
+      "loss": 0.8736,
+      "step": 7037
+    },
+    {
+      "epoch": 0.3873630909791403,
+      "grad_norm": 0.8823271989822388,
+      "learning_rate": 9.122500074901083e-06,
+      "loss": 0.7448,
+      "step": 7038
+    },
+    {
+      "epoch": 0.387418129781496,
+      "grad_norm": 0.6494244933128357,
+      "learning_rate": 9.122254777152519e-06,
+      "loss": 0.7423,
+      "step": 7039
+    },
+    {
+      "epoch": 0.3874731685838516,
+      "grad_norm": 0.7232181429862976,
+      "learning_rate": 9.122009448422191e-06,
+      "loss": 0.8489,
+      "step": 7040
+    },
+    {
+      "epoch": 0.3875282073862073,
+      "grad_norm": 0.7357699275016785,
+      "learning_rate": 9.121764088711945e-06,
+      "loss": 0.8799,
+      "step": 7041
+    },
+    {
+      "epoch": 0.38758324618856294,
+      "grad_norm": 0.7638574838638306,
+      "learning_rate": 9.121518698023621e-06,
+      "loss": 0.8539,
+      "step": 7042
+    },
+    {
+      "epoch": 0.3876382849909186,
+      "grad_norm": 0.7407062649726868,
+      "learning_rate": 9.121273276359068e-06,
+      "loss": 0.7152,
+      "step": 7043
+    },
+    {
+      "epoch": 0.38769332379327426,
+      "grad_norm": 0.6945983171463013,
+      "learning_rate": 9.121027823720126e-06,
+      "loss": 0.8224,
+      "step": 7044
+    },
+    {
+      "epoch": 0.3877483625956299,
+      "grad_norm": 0.7163639068603516,
+      "learning_rate": 9.120782340108643e-06,
+      "loss": 0.808,
+      "step": 7045
+    },
+    {
+      "epoch": 0.3878034013979856,
+      "grad_norm": 0.7062035799026489,
+      "learning_rate": 9.120536825526463e-06,
+      "loss": 0.783,
+      "step": 7046
+    },
+    {
+      "epoch": 0.3878584402003412,
+      "grad_norm": 0.7459971308708191,
+      "learning_rate": 9.120291279975431e-06,
+      "loss": 0.8219,
+      "step": 7047
+    },
+    {
+      "epoch": 0.3879134790026969,
+      "grad_norm": 0.9016150236129761,
+      "learning_rate": 9.120045703457394e-06,
+      "loss": 0.8605,
+      "step": 7048
+    },
+    {
+      "epoch": 0.38796851780505254,
+      "grad_norm": 0.78440922498703,
+      "learning_rate": 9.119800095974193e-06,
+      "loss": 0.8424,
+      "step": 7049
+    },
+    {
+      "epoch": 0.3880235566074082,
+      "grad_norm": 0.751504123210907,
+      "learning_rate": 9.119554457527681e-06,
+      "loss": 0.701,
+      "step": 7050
+    },
+    {
+      "epoch": 0.38807859540976386,
+      "grad_norm": 0.7540284991264343,
+      "learning_rate": 9.119308788119698e-06,
+      "loss": 0.7912,
+      "step": 7051
+    },
+    {
+      "epoch": 0.38813363421211955,
+      "grad_norm": 0.7977007627487183,
+      "learning_rate": 9.119063087752094e-06,
+      "loss": 0.9297,
+      "step": 7052
+    },
+    {
+      "epoch": 0.3881886730144752,
+      "grad_norm": 0.6923508644104004,
+      "learning_rate": 9.118817356426715e-06,
+      "loss": 0.7458,
+      "step": 7053
+    },
+    {
+      "epoch": 0.38824371181683087,
+      "grad_norm": 0.7170272469520569,
+      "learning_rate": 9.118571594145406e-06,
+      "loss": 0.733,
+      "step": 7054
+    },
+    {
+      "epoch": 0.3882987506191865,
+      "grad_norm": 0.7547701001167297,
+      "learning_rate": 9.118325800910015e-06,
+      "loss": 0.7758,
+      "step": 7055
+    },
+    {
+      "epoch": 0.3883537894215422,
+      "grad_norm": 0.7921421527862549,
+      "learning_rate": 9.118079976722391e-06,
+      "loss": 0.8262,
+      "step": 7056
+    },
+    {
+      "epoch": 0.38840882822389783,
+      "grad_norm": 0.734470784664154,
+      "learning_rate": 9.117834121584379e-06,
+      "loss": 0.817,
+      "step": 7057
+    },
+    {
+      "epoch": 0.3884638670262535,
+      "grad_norm": 0.8106420040130615,
+      "learning_rate": 9.117588235497829e-06,
+      "loss": 0.8203,
+      "step": 7058
+    },
+    {
+      "epoch": 0.38851890582860915,
+      "grad_norm": 0.7355543375015259,
+      "learning_rate": 9.117342318464588e-06,
+      "loss": 0.8076,
+      "step": 7059
+    },
+    {
+      "epoch": 0.38857394463096484,
+      "grad_norm": 0.7665252685546875,
+      "learning_rate": 9.117096370486504e-06,
+      "loss": 0.7611,
+      "step": 7060
+    },
+    {
+      "epoch": 0.3886289834333205,
+      "grad_norm": 0.7968598008155823,
+      "learning_rate": 9.116850391565426e-06,
+      "loss": 0.6461,
+      "step": 7061
+    },
+    {
+      "epoch": 0.38868402223567616,
+      "grad_norm": 0.7187741994857788,
+      "learning_rate": 9.116604381703203e-06,
+      "loss": 0.7982,
+      "step": 7062
+    },
+    {
+      "epoch": 0.3887390610380318,
+      "grad_norm": 0.8566913604736328,
+      "learning_rate": 9.11635834090168e-06,
+      "loss": 0.9072,
+      "step": 7063
+    },
+    {
+      "epoch": 0.3887940998403875,
+      "grad_norm": 0.7120797038078308,
+      "learning_rate": 9.116112269162714e-06,
+      "loss": 0.7353,
+      "step": 7064
+    },
+    {
+      "epoch": 0.3888491386427431,
+      "grad_norm": 0.7230019569396973,
+      "learning_rate": 9.115866166488148e-06,
+      "loss": 0.7717,
+      "step": 7065
+    },
+    {
+      "epoch": 0.3889041774450988,
+      "grad_norm": 0.6650584936141968,
+      "learning_rate": 9.115620032879833e-06,
+      "loss": 0.7384,
+      "step": 7066
+    },
+    {
+      "epoch": 0.38895921624745444,
+      "grad_norm": 0.970750629901886,
+      "learning_rate": 9.115373868339621e-06,
+      "loss": 0.8478,
+      "step": 7067
+    },
+    {
+      "epoch": 0.38901425504981013,
+      "grad_norm": 0.7066280245780945,
+      "learning_rate": 9.115127672869359e-06,
+      "loss": 0.7638,
+      "step": 7068
+    },
+    {
+      "epoch": 0.38906929385216577,
+      "grad_norm": 0.6952232718467712,
+      "learning_rate": 9.1148814464709e-06,
+      "loss": 0.7869,
+      "step": 7069
+    },
+    {
+      "epoch": 0.38912433265452145,
+      "grad_norm": 0.804489254951477,
+      "learning_rate": 9.114635189146094e-06,
+      "loss": 0.7905,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3891793714568771,
+      "grad_norm": 0.6988457441329956,
+      "learning_rate": 9.114388900896791e-06,
+      "loss": 0.7107,
+      "step": 7071
+    },
+    {
+      "epoch": 0.3892344102592328,
+      "grad_norm": 0.6379980444908142,
+      "learning_rate": 9.114142581724842e-06,
+      "loss": 0.733,
+      "step": 7072
+    },
+    {
+      "epoch": 0.3892894490615884,
+      "grad_norm": 0.7238649129867554,
+      "learning_rate": 9.113896231632098e-06,
+      "loss": 0.8252,
+      "step": 7073
+    },
+    {
+      "epoch": 0.3893444878639441,
+      "grad_norm": 0.7168585062026978,
+      "learning_rate": 9.113649850620412e-06,
+      "loss": 0.6459,
+      "step": 7074
+    },
+    {
+      "epoch": 0.38939952666629973,
+      "grad_norm": 0.7315915822982788,
+      "learning_rate": 9.113403438691634e-06,
+      "loss": 0.7557,
+      "step": 7075
+    },
+    {
+      "epoch": 0.3894545654686554,
+      "grad_norm": 0.7438754439353943,
+      "learning_rate": 9.11315699584762e-06,
+      "loss": 0.7938,
+      "step": 7076
+    },
+    {
+      "epoch": 0.38950960427101106,
+      "grad_norm": 0.7497848272323608,
+      "learning_rate": 9.112910522090215e-06,
+      "loss": 0.8232,
+      "step": 7077
+    },
+    {
+      "epoch": 0.38956464307336675,
+      "grad_norm": 0.8072896003723145,
+      "learning_rate": 9.112664017421277e-06,
+      "loss": 0.7974,
+      "step": 7078
+    },
+    {
+      "epoch": 0.3896196818757224,
+      "grad_norm": 0.7255920767784119,
+      "learning_rate": 9.112417481842657e-06,
+      "loss": 0.7658,
+      "step": 7079
+    },
+    {
+      "epoch": 0.38967472067807807,
+      "grad_norm": 0.6263132095336914,
+      "learning_rate": 9.112170915356209e-06,
+      "loss": 0.7188,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3897297594804337,
+      "grad_norm": 0.6817660927772522,
+      "learning_rate": 9.111924317963785e-06,
+      "loss": 0.7406,
+      "step": 7081
+    },
+    {
+      "epoch": 0.3897847982827894,
+      "grad_norm": 0.7829134464263916,
+      "learning_rate": 9.111677689667238e-06,
+      "loss": 0.8406,
+      "step": 7082
+    },
+    {
+      "epoch": 0.389839837085145,
+      "grad_norm": 0.7122843861579895,
+      "learning_rate": 9.111431030468421e-06,
+      "loss": 0.7722,
+      "step": 7083
+    },
+    {
+      "epoch": 0.3898948758875007,
+      "grad_norm": 0.7041764259338379,
+      "learning_rate": 9.11118434036919e-06,
+      "loss": 0.8307,
+      "step": 7084
+    },
+    {
+      "epoch": 0.38994991468985635,
+      "grad_norm": 0.7582009434700012,
+      "learning_rate": 9.110937619371398e-06,
+      "loss": 0.7461,
+      "step": 7085
+    },
+    {
+      "epoch": 0.39000495349221204,
+      "grad_norm": 0.7156100273132324,
+      "learning_rate": 9.110690867476899e-06,
+      "loss": 0.7294,
+      "step": 7086
+    },
+    {
+      "epoch": 0.39005999229456767,
+      "grad_norm": 0.79449063539505,
+      "learning_rate": 9.110444084687549e-06,
+      "loss": 0.8652,
+      "step": 7087
+    },
+    {
+      "epoch": 0.3901150310969233,
+      "grad_norm": 0.7692831754684448,
+      "learning_rate": 9.1101972710052e-06,
+      "loss": 0.7899,
+      "step": 7088
+    },
+    {
+      "epoch": 0.390170069899279,
+      "grad_norm": 0.7189639806747437,
+      "learning_rate": 9.109950426431708e-06,
+      "loss": 0.726,
+      "step": 7089
+    },
+    {
+      "epoch": 0.3902251087016346,
+      "grad_norm": 0.7491177916526794,
+      "learning_rate": 9.10970355096893e-06,
+      "loss": 0.8881,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3902801475039903,
+      "grad_norm": 0.783027172088623,
+      "learning_rate": 9.10945664461872e-06,
+      "loss": 0.7728,
+      "step": 7091
+    },
+    {
+      "epoch": 0.39033518630634595,
+      "grad_norm": 1.0871556997299194,
+      "learning_rate": 9.109209707382934e-06,
+      "loss": 0.8059,
+      "step": 7092
+    },
+    {
+      "epoch": 0.39039022510870164,
+      "grad_norm": 0.7287113666534424,
+      "learning_rate": 9.108962739263429e-06,
+      "loss": 0.7896,
+      "step": 7093
+    },
+    {
+      "epoch": 0.39044526391105727,
+      "grad_norm": 0.7801700234413147,
+      "learning_rate": 9.108715740262058e-06,
+      "loss": 0.8012,
+      "step": 7094
+    },
+    {
+      "epoch": 0.39050030271341296,
+      "grad_norm": 0.846709132194519,
+      "learning_rate": 9.10846871038068e-06,
+      "loss": 0.8392,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3905553415157686,
+      "grad_norm": 0.7408092617988586,
+      "learning_rate": 9.10822164962115e-06,
+      "loss": 0.8657,
+      "step": 7096
+    },
+    {
+      "epoch": 0.3906103803181243,
+      "grad_norm": 0.6748743057250977,
+      "learning_rate": 9.107974557985328e-06,
+      "loss": 0.7659,
+      "step": 7097
+    },
+    {
+      "epoch": 0.3906654191204799,
+      "grad_norm": 0.7512170672416687,
+      "learning_rate": 9.107727435475067e-06,
+      "loss": 0.7704,
+      "step": 7098
+    },
+    {
+      "epoch": 0.3907204579228356,
+      "grad_norm": 0.9039596319198608,
+      "learning_rate": 9.107480282092227e-06,
+      "loss": 0.8412,
+      "step": 7099
+    },
+    {
+      "epoch": 0.39077549672519124,
+      "grad_norm": 0.829785943031311,
+      "learning_rate": 9.107233097838663e-06,
+      "loss": 0.8229,
+      "step": 7100
+    },
+    {
+      "epoch": 0.39083053552754693,
+      "grad_norm": 0.7597842812538147,
+      "learning_rate": 9.106985882716238e-06,
+      "loss": 0.7798,
+      "step": 7101
+    },
+    {
+      "epoch": 0.39088557432990256,
+      "grad_norm": 0.7619945406913757,
+      "learning_rate": 9.106738636726802e-06,
+      "loss": 0.7504,
+      "step": 7102
+    },
+    {
+      "epoch": 0.39094061313225825,
+      "grad_norm": 0.6791092157363892,
+      "learning_rate": 9.10649135987222e-06,
+      "loss": 0.8167,
+      "step": 7103
+    },
+    {
+      "epoch": 0.3909956519346139,
+      "grad_norm": 0.7977412343025208,
+      "learning_rate": 9.10624405215435e-06,
+      "loss": 0.8252,
+      "step": 7104
+    },
+    {
+      "epoch": 0.3910506907369696,
+      "grad_norm": 0.7329283356666565,
+      "learning_rate": 9.105996713575047e-06,
+      "loss": 0.7084,
+      "step": 7105
+    },
+    {
+      "epoch": 0.3911057295393252,
+      "grad_norm": 0.7125133872032166,
+      "learning_rate": 9.105749344136172e-06,
+      "loss": 0.6672,
+      "step": 7106
+    },
+    {
+      "epoch": 0.3911607683416809,
+      "grad_norm": 0.6974679827690125,
+      "learning_rate": 9.105501943839583e-06,
+      "loss": 0.7354,
+      "step": 7107
+    },
+    {
+      "epoch": 0.39121580714403653,
+      "grad_norm": 0.7191265225410461,
+      "learning_rate": 9.10525451268714e-06,
+      "loss": 0.8133,
+      "step": 7108
+    },
+    {
+      "epoch": 0.3912708459463922,
+      "grad_norm": 0.7188206911087036,
+      "learning_rate": 9.105007050680704e-06,
+      "loss": 0.7947,
+      "step": 7109
+    },
+    {
+      "epoch": 0.39132588474874785,
+      "grad_norm": 0.9017364382743835,
+      "learning_rate": 9.104759557822135e-06,
+      "loss": 0.7848,
+      "step": 7110
+    },
+    {
+      "epoch": 0.39138092355110354,
+      "grad_norm": 0.7551164031028748,
+      "learning_rate": 9.104512034113292e-06,
+      "loss": 0.8266,
+      "step": 7111
+    },
+    {
+      "epoch": 0.3914359623534592,
+      "grad_norm": 0.7810001969337463,
+      "learning_rate": 9.104264479556033e-06,
+      "loss": 0.7731,
+      "step": 7112
+    },
+    {
+      "epoch": 0.39149100115581487,
+      "grad_norm": 0.787723183631897,
+      "learning_rate": 9.104016894152223e-06,
+      "loss": 0.8008,
+      "step": 7113
+    },
+    {
+      "epoch": 0.3915460399581705,
+      "grad_norm": 0.7303524017333984,
+      "learning_rate": 9.103769277903718e-06,
+      "loss": 0.826,
+      "step": 7114
+    },
+    {
+      "epoch": 0.3916010787605262,
+      "grad_norm": 0.707759439945221,
+      "learning_rate": 9.103521630812384e-06,
+      "loss": 0.6303,
+      "step": 7115
+    },
+    {
+      "epoch": 0.3916561175628818,
+      "grad_norm": 0.6929940581321716,
+      "learning_rate": 9.10327395288008e-06,
+      "loss": 0.733,
+      "step": 7116
+    },
+    {
+      "epoch": 0.3917111563652375,
+      "grad_norm": 0.7133205533027649,
+      "learning_rate": 9.103026244108667e-06,
+      "loss": 0.8421,
+      "step": 7117
+    },
+    {
+      "epoch": 0.39176619516759315,
+      "grad_norm": 1.2049434185028076,
+      "learning_rate": 9.102778504500005e-06,
+      "loss": 0.8618,
+      "step": 7118
+    },
+    {
+      "epoch": 0.39182123396994883,
+      "grad_norm": 0.7792720198631287,
+      "learning_rate": 9.10253073405596e-06,
+      "loss": 0.717,
+      "step": 7119
+    },
+    {
+      "epoch": 0.39187627277230447,
+      "grad_norm": 0.7234412431716919,
+      "learning_rate": 9.10228293277839e-06,
+      "loss": 0.7547,
+      "step": 7120
+    },
+    {
+      "epoch": 0.39193131157466016,
+      "grad_norm": 0.6845420002937317,
+      "learning_rate": 9.102035100669162e-06,
+      "loss": 0.7255,
+      "step": 7121
+    },
+    {
+      "epoch": 0.3919863503770158,
+      "grad_norm": 0.7446799874305725,
+      "learning_rate": 9.101787237730135e-06,
+      "loss": 0.7947,
+      "step": 7122
+    },
+    {
+      "epoch": 0.3920413891793715,
+      "grad_norm": 0.812924325466156,
+      "learning_rate": 9.101539343963176e-06,
+      "loss": 0.843,
+      "step": 7123
+    },
+    {
+      "epoch": 0.3920964279817271,
+      "grad_norm": 0.7373847365379333,
+      "learning_rate": 9.101291419370141e-06,
+      "loss": 0.7703,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3921514667840828,
+      "grad_norm": 0.8305120468139648,
+      "learning_rate": 9.101043463952899e-06,
+      "loss": 0.8904,
+      "step": 7125
+    },
+    {
+      "epoch": 0.39220650558643844,
+      "grad_norm": 0.7263030409812927,
+      "learning_rate": 9.100795477713313e-06,
+      "loss": 0.8319,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3922615443887941,
+      "grad_norm": 0.8358581066131592,
+      "learning_rate": 9.100547460653245e-06,
+      "loss": 0.8305,
+      "step": 7127
+    },
+    {
+      "epoch": 0.39231658319114976,
+      "grad_norm": 0.6608800292015076,
+      "learning_rate": 9.10029941277456e-06,
+      "loss": 0.7815,
+      "step": 7128
+    },
+    {
+      "epoch": 0.39237162199350545,
+      "grad_norm": 0.8590257167816162,
+      "learning_rate": 9.100051334079122e-06,
+      "loss": 0.8292,
+      "step": 7129
+    },
+    {
+      "epoch": 0.3924266607958611,
+      "grad_norm": 0.6241755485534668,
+      "learning_rate": 9.099803224568797e-06,
+      "loss": 0.6568,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3924816995982167,
+      "grad_norm": 0.7298059463500977,
+      "learning_rate": 9.099555084245447e-06,
+      "loss": 0.727,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3925367384005724,
+      "grad_norm": 0.7741055488586426,
+      "learning_rate": 9.099306913110939e-06,
+      "loss": 0.8481,
+      "step": 7132
+    },
+    {
+      "epoch": 0.39259177720292804,
+      "grad_norm": 0.9674170613288879,
+      "learning_rate": 9.099058711167137e-06,
+      "loss": 0.8507,
+      "step": 7133
+    },
+    {
+      "epoch": 0.3926468160052837,
+      "grad_norm": 0.7285159826278687,
+      "learning_rate": 9.098810478415907e-06,
+      "loss": 0.766,
+      "step": 7134
+    },
+    {
+      "epoch": 0.39270185480763936,
+      "grad_norm": 0.7215660810470581,
+      "learning_rate": 9.098562214859115e-06,
+      "loss": 0.794,
+      "step": 7135
+    },
+    {
+      "epoch": 0.39275689360999505,
+      "grad_norm": 0.764437735080719,
+      "learning_rate": 9.098313920498627e-06,
+      "loss": 0.8228,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3928119324123507,
+      "grad_norm": 0.7222796082496643,
+      "learning_rate": 9.098065595336309e-06,
+      "loss": 0.8064,
+      "step": 7137
+    },
+    {
+      "epoch": 0.3928669712147064,
+      "grad_norm": 0.7044625878334045,
+      "learning_rate": 9.097817239374024e-06,
+      "loss": 0.8017,
+      "step": 7138
+    },
+    {
+      "epoch": 0.392922010017062,
+      "grad_norm": 0.7929979562759399,
+      "learning_rate": 9.097568852613646e-06,
+      "loss": 0.7527,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3929770488194177,
+      "grad_norm": 0.7833721041679382,
+      "learning_rate": 9.097320435057033e-06,
+      "loss": 0.8335,
+      "step": 7140
+    },
+    {
+      "epoch": 0.39303208762177333,
+      "grad_norm": 0.8365728259086609,
+      "learning_rate": 9.097071986706058e-06,
+      "loss": 0.6439,
+      "step": 7141
+    },
+    {
+      "epoch": 0.393087126424129,
+      "grad_norm": 0.7547842264175415,
+      "learning_rate": 9.096823507562588e-06,
+      "loss": 0.8316,
+      "step": 7142
+    },
+    {
+      "epoch": 0.39314216522648465,
+      "grad_norm": 0.6598891019821167,
+      "learning_rate": 9.09657499762849e-06,
+      "loss": 0.6547,
+      "step": 7143
+    },
+    {
+      "epoch": 0.39319720402884034,
+      "grad_norm": 0.7913638949394226,
+      "learning_rate": 9.096326456905627e-06,
+      "loss": 0.7964,
+      "step": 7144
+    },
+    {
+      "epoch": 0.393252242831196,
+      "grad_norm": 0.6927905082702637,
+      "learning_rate": 9.096077885395874e-06,
+      "loss": 0.7836,
+      "step": 7145
+    },
+    {
+      "epoch": 0.39330728163355166,
+      "grad_norm": 0.7505417466163635,
+      "learning_rate": 9.095829283101094e-06,
+      "loss": 0.7707,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3933623204359073,
+      "grad_norm": 0.8797083497047424,
+      "learning_rate": 9.095580650023158e-06,
+      "loss": 0.866,
+      "step": 7147
+    },
+    {
+      "epoch": 0.393417359238263,
+      "grad_norm": 0.7023645639419556,
+      "learning_rate": 9.095331986163935e-06,
+      "loss": 0.7013,
+      "step": 7148
+    },
+    {
+      "epoch": 0.3934723980406186,
+      "grad_norm": 0.697354793548584,
+      "learning_rate": 9.095083291525293e-06,
+      "loss": 0.7691,
+      "step": 7149
+    },
+    {
+      "epoch": 0.3935274368429743,
+      "grad_norm": 0.7211105227470398,
+      "learning_rate": 9.094834566109101e-06,
+      "loss": 0.6816,
+      "step": 7150
+    },
+    {
+      "epoch": 0.39358247564532994,
+      "grad_norm": 0.8593278527259827,
+      "learning_rate": 9.094585809917227e-06,
+      "loss": 0.915,
+      "step": 7151
+    },
+    {
+      "epoch": 0.39363751444768563,
+      "grad_norm": 0.7406070828437805,
+      "learning_rate": 9.094337022951545e-06,
+      "loss": 0.7825,
+      "step": 7152
+    },
+    {
+      "epoch": 0.39369255325004127,
+      "grad_norm": 0.7644504308700562,
+      "learning_rate": 9.09408820521392e-06,
+      "loss": 0.6796,
+      "step": 7153
+    },
+    {
+      "epoch": 0.39374759205239696,
+      "grad_norm": 0.8239033222198486,
+      "learning_rate": 9.093839356706224e-06,
+      "loss": 0.8396,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3938026308547526,
+      "grad_norm": 0.6433991193771362,
+      "learning_rate": 9.093590477430327e-06,
+      "loss": 0.6941,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3938576696571083,
+      "grad_norm": 0.6979972124099731,
+      "learning_rate": 9.093341567388102e-06,
+      "loss": 0.8142,
+      "step": 7156
+    },
+    {
+      "epoch": 0.3939127084594639,
+      "grad_norm": 0.7062026262283325,
+      "learning_rate": 9.093092626581414e-06,
+      "loss": 0.804,
+      "step": 7157
+    },
+    {
+      "epoch": 0.3939677472618196,
+      "grad_norm": 0.7070814967155457,
+      "learning_rate": 9.09284365501214e-06,
+      "loss": 0.765,
+      "step": 7158
+    },
+    {
+      "epoch": 0.39402278606417523,
+      "grad_norm": 0.8577908873558044,
+      "learning_rate": 9.092594652682147e-06,
+      "loss": 0.7074,
+      "step": 7159
+    },
+    {
+      "epoch": 0.3940778248665309,
+      "grad_norm": 0.7386197447776794,
+      "learning_rate": 9.092345619593309e-06,
+      "loss": 0.7629,
+      "step": 7160
+    },
+    {
+      "epoch": 0.39413286366888656,
+      "grad_norm": 0.8048123121261597,
+      "learning_rate": 9.092096555747496e-06,
+      "loss": 0.9225,
+      "step": 7161
+    },
+    {
+      "epoch": 0.39418790247124225,
+      "grad_norm": 0.7479888200759888,
+      "learning_rate": 9.091847461146582e-06,
+      "loss": 0.7284,
+      "step": 7162
+    },
+    {
+      "epoch": 0.3942429412735979,
+      "grad_norm": 0.7448734045028687,
+      "learning_rate": 9.091598335792438e-06,
+      "loss": 0.8694,
+      "step": 7163
+    },
+    {
+      "epoch": 0.39429798007595357,
+      "grad_norm": 0.7511261701583862,
+      "learning_rate": 9.091349179686935e-06,
+      "loss": 0.7822,
+      "step": 7164
+    },
+    {
+      "epoch": 0.3943530188783092,
+      "grad_norm": 0.7079344391822815,
+      "learning_rate": 9.091099992831946e-06,
+      "loss": 0.7238,
+      "step": 7165
+    },
+    {
+      "epoch": 0.3944080576806649,
+      "grad_norm": 0.7007229328155518,
+      "learning_rate": 9.090850775229347e-06,
+      "loss": 0.7269,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3944630964830205,
+      "grad_norm": 0.769800066947937,
+      "learning_rate": 9.090601526881007e-06,
+      "loss": 0.7894,
+      "step": 7167
+    },
+    {
+      "epoch": 0.3945181352853762,
+      "grad_norm": 0.7211676836013794,
+      "learning_rate": 9.090352247788801e-06,
+      "loss": 0.7998,
+      "step": 7168
+    },
+    {
+      "epoch": 0.39457317408773185,
+      "grad_norm": 0.6784254312515259,
+      "learning_rate": 9.090102937954602e-06,
+      "loss": 0.7576,
+      "step": 7169
+    },
+    {
+      "epoch": 0.39462821289008754,
+      "grad_norm": 0.7696946859359741,
+      "learning_rate": 9.089853597380285e-06,
+      "loss": 0.8395,
+      "step": 7170
+    },
+    {
+      "epoch": 0.39468325169244317,
+      "grad_norm": 0.8720405697822571,
+      "learning_rate": 9.089604226067723e-06,
+      "loss": 0.8971,
+      "step": 7171
+    },
+    {
+      "epoch": 0.39473829049479886,
+      "grad_norm": 0.8457947373390198,
+      "learning_rate": 9.08935482401879e-06,
+      "loss": 0.7002,
+      "step": 7172
+    },
+    {
+      "epoch": 0.3947933292971545,
+      "grad_norm": 0.8181997537612915,
+      "learning_rate": 9.089105391235361e-06,
+      "loss": 0.8949,
+      "step": 7173
+    },
+    {
+      "epoch": 0.3948483680995101,
+      "grad_norm": 0.7717136144638062,
+      "learning_rate": 9.08885592771931e-06,
+      "loss": 0.829,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3949034069018658,
+      "grad_norm": 0.6941567063331604,
+      "learning_rate": 9.088606433472514e-06,
+      "loss": 0.7592,
+      "step": 7175
+    },
+    {
+      "epoch": 0.39495844570422145,
+      "grad_norm": 0.7358599901199341,
+      "learning_rate": 9.088356908496845e-06,
+      "loss": 0.8657,
+      "step": 7176
+    },
+    {
+      "epoch": 0.39501348450657714,
+      "grad_norm": 1.1329307556152344,
+      "learning_rate": 9.08810735279418e-06,
+      "loss": 0.8307,
+      "step": 7177
+    },
+    {
+      "epoch": 0.3950685233089328,
+      "grad_norm": 0.7011532187461853,
+      "learning_rate": 9.087857766366395e-06,
+      "loss": 0.7487,
+      "step": 7178
+    },
+    {
+      "epoch": 0.39512356211128846,
+      "grad_norm": 0.7390572428703308,
+      "learning_rate": 9.087608149215366e-06,
+      "loss": 0.8244,
+      "step": 7179
+    },
+    {
+      "epoch": 0.3951786009136441,
+      "grad_norm": 0.6907634735107422,
+      "learning_rate": 9.087358501342966e-06,
+      "loss": 0.751,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3952336397159998,
+      "grad_norm": 0.7467379570007324,
+      "learning_rate": 9.087108822751076e-06,
+      "loss": 0.8549,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3952886785183554,
+      "grad_norm": 0.7493302226066589,
+      "learning_rate": 9.086859113441568e-06,
+      "loss": 0.8332,
+      "step": 7182
+    },
+    {
+      "epoch": 0.3953437173207111,
+      "grad_norm": 0.8364959955215454,
+      "learning_rate": 9.086609373416321e-06,
+      "loss": 0.7873,
+      "step": 7183
+    },
+    {
+      "epoch": 0.39539875612306674,
+      "grad_norm": 0.7330418825149536,
+      "learning_rate": 9.086359602677214e-06,
+      "loss": 0.7861,
+      "step": 7184
+    },
+    {
+      "epoch": 0.39545379492542243,
+      "grad_norm": 0.7296311855316162,
+      "learning_rate": 9.086109801226121e-06,
+      "loss": 0.7946,
+      "step": 7185
+    },
+    {
+      "epoch": 0.39550883372777806,
+      "grad_norm": 0.7884660363197327,
+      "learning_rate": 9.085859969064921e-06,
+      "loss": 0.7851,
+      "step": 7186
+    },
+    {
+      "epoch": 0.39556387253013375,
+      "grad_norm": 0.7311955690383911,
+      "learning_rate": 9.08561010619549e-06,
+      "loss": 0.7645,
+      "step": 7187
+    },
+    {
+      "epoch": 0.3956189113324894,
+      "grad_norm": 0.7447296977043152,
+      "learning_rate": 9.085360212619707e-06,
+      "loss": 0.7446,
+      "step": 7188
+    },
+    {
+      "epoch": 0.3956739501348451,
+      "grad_norm": 0.755628228187561,
+      "learning_rate": 9.08511028833945e-06,
+      "loss": 0.8107,
+      "step": 7189
+    },
+    {
+      "epoch": 0.3957289889372007,
+      "grad_norm": 0.6800833940505981,
+      "learning_rate": 9.0848603333566e-06,
+      "loss": 0.7471,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3957840277395564,
+      "grad_norm": 0.6396341919898987,
+      "learning_rate": 9.08461034767303e-06,
+      "loss": 0.6797,
+      "step": 7191
+    },
+    {
+      "epoch": 0.39583906654191203,
+      "grad_norm": 0.729680597782135,
+      "learning_rate": 9.084360331290625e-06,
+      "loss": 0.7224,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3958941053442677,
+      "grad_norm": 0.7630584239959717,
+      "learning_rate": 9.084110284211259e-06,
+      "loss": 0.8203,
+      "step": 7193
+    },
+    {
+      "epoch": 0.39594914414662336,
+      "grad_norm": 0.8799235820770264,
+      "learning_rate": 9.083860206436813e-06,
+      "loss": 0.8312,
+      "step": 7194
+    },
+    {
+      "epoch": 0.39600418294897904,
+      "grad_norm": 0.797081708908081,
+      "learning_rate": 9.083610097969169e-06,
+      "loss": 0.7561,
+      "step": 7195
+    },
+    {
+      "epoch": 0.3960592217513347,
+      "grad_norm": 0.7408759593963623,
+      "learning_rate": 9.083359958810203e-06,
+      "loss": 0.7854,
+      "step": 7196
+    },
+    {
+      "epoch": 0.39611426055369037,
+      "grad_norm": 0.7552130222320557,
+      "learning_rate": 9.083109788961797e-06,
+      "loss": 0.8145,
+      "step": 7197
+    },
+    {
+      "epoch": 0.396169299356046,
+      "grad_norm": 0.7147447466850281,
+      "learning_rate": 9.08285958842583e-06,
+      "loss": 0.792,
+      "step": 7198
+    },
+    {
+      "epoch": 0.3962243381584017,
+      "grad_norm": 0.7416259050369263,
+      "learning_rate": 9.082609357204183e-06,
+      "loss": 0.7801,
+      "step": 7199
+    },
+    {
+      "epoch": 0.3962793769607573,
+      "grad_norm": 0.7551109194755554,
+      "learning_rate": 9.082359095298741e-06,
+      "loss": 0.8841,
+      "step": 7200
+    },
+    {
+      "epoch": 0.396334415763113,
+      "grad_norm": 0.761472225189209,
+      "learning_rate": 9.082108802711377e-06,
+      "loss": 0.9061,
+      "step": 7201
+    },
+    {
+      "epoch": 0.39638945456546865,
+      "grad_norm": 0.7234126329421997,
+      "learning_rate": 9.081858479443977e-06,
+      "loss": 0.8308,
+      "step": 7202
+    },
+    {
+      "epoch": 0.39644449336782434,
+      "grad_norm": 0.7204816341400146,
+      "learning_rate": 9.08160812549842e-06,
+      "loss": 0.7481,
+      "step": 7203
+    },
+    {
+      "epoch": 0.39649953217017997,
+      "grad_norm": 0.7207956910133362,
+      "learning_rate": 9.081357740876591e-06,
+      "loss": 0.762,
+      "step": 7204
+    },
+    {
+      "epoch": 0.39655457097253566,
+      "grad_norm": 0.7967123985290527,
+      "learning_rate": 9.081107325580367e-06,
+      "loss": 0.7931,
+      "step": 7205
+    },
+    {
+      "epoch": 0.3966096097748913,
+      "grad_norm": 0.9839354753494263,
+      "learning_rate": 9.080856879611635e-06,
+      "loss": 0.8182,
+      "step": 7206
+    },
+    {
+      "epoch": 0.396664648577247,
+      "grad_norm": 0.8468357920646667,
+      "learning_rate": 9.080606402972274e-06,
+      "loss": 0.7056,
+      "step": 7207
+    },
+    {
+      "epoch": 0.3967196873796026,
+      "grad_norm": 0.6549574136734009,
+      "learning_rate": 9.080355895664169e-06,
+      "loss": 0.7604,
+      "step": 7208
+    },
+    {
+      "epoch": 0.3967747261819583,
+      "grad_norm": 0.7475417256355286,
+      "learning_rate": 9.080105357689201e-06,
+      "loss": 0.7107,
+      "step": 7209
+    },
+    {
+      "epoch": 0.39682976498431394,
+      "grad_norm": 0.7464179992675781,
+      "learning_rate": 9.079854789049251e-06,
+      "loss": 0.793,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3968848037866696,
+      "grad_norm": 0.8332071900367737,
+      "learning_rate": 9.079604189746207e-06,
+      "loss": 0.8383,
+      "step": 7211
+    },
+    {
+      "epoch": 0.39693984258902526,
+      "grad_norm": 0.722055196762085,
+      "learning_rate": 9.07935355978195e-06,
+      "loss": 0.8569,
+      "step": 7212
+    },
+    {
+      "epoch": 0.39699488139138095,
+      "grad_norm": 0.7442018389701843,
+      "learning_rate": 9.079102899158363e-06,
+      "loss": 0.8165,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3970499201937366,
+      "grad_norm": 0.6986141204833984,
+      "learning_rate": 9.07885220787733e-06,
+      "loss": 0.7562,
+      "step": 7214
+    },
+    {
+      "epoch": 0.39710495899609227,
+      "grad_norm": 0.7718464732170105,
+      "learning_rate": 9.078601485940736e-06,
+      "loss": 0.8529,
+      "step": 7215
+    },
+    {
+      "epoch": 0.3971599977984479,
+      "grad_norm": 0.7583653330802917,
+      "learning_rate": 9.078350733350464e-06,
+      "loss": 0.7855,
+      "step": 7216
+    },
+    {
+      "epoch": 0.39721503660080354,
+      "grad_norm": 0.7699223160743713,
+      "learning_rate": 9.078099950108401e-06,
+      "loss": 0.8061,
+      "step": 7217
+    },
+    {
+      "epoch": 0.39727007540315923,
+      "grad_norm": 0.7374141812324524,
+      "learning_rate": 9.07784913621643e-06,
+      "loss": 0.789,
+      "step": 7218
+    },
+    {
+      "epoch": 0.39732511420551486,
+      "grad_norm": 0.7446104884147644,
+      "learning_rate": 9.077598291676436e-06,
+      "loss": 0.8381,
+      "step": 7219
+    },
+    {
+      "epoch": 0.39738015300787055,
+      "grad_norm": 0.7017301917076111,
+      "learning_rate": 9.077347416490305e-06,
+      "loss": 0.7153,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3974351918102262,
+      "grad_norm": 0.7676172852516174,
+      "learning_rate": 9.077096510659922e-06,
+      "loss": 0.8029,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3974902306125819,
+      "grad_norm": 0.9340602159500122,
+      "learning_rate": 9.076845574187174e-06,
+      "loss": 0.7865,
+      "step": 7222
+    },
+    {
+      "epoch": 0.3975452694149375,
+      "grad_norm": 0.8634235262870789,
+      "learning_rate": 9.076594607073945e-06,
+      "loss": 0.7606,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3976003082172932,
+      "grad_norm": 0.8967369198799133,
+      "learning_rate": 9.076343609322123e-06,
+      "loss": 0.7011,
+      "step": 7224
+    },
+    {
+      "epoch": 0.39765534701964883,
+      "grad_norm": 0.7269352078437805,
+      "learning_rate": 9.076092580933594e-06,
+      "loss": 0.8043,
+      "step": 7225
+    },
+    {
+      "epoch": 0.3977103858220045,
+      "grad_norm": 0.7550628781318665,
+      "learning_rate": 9.075841521910243e-06,
+      "loss": 0.7344,
+      "step": 7226
+    },
+    {
+      "epoch": 0.39776542462436015,
+      "grad_norm": 0.6973844766616821,
+      "learning_rate": 9.075590432253958e-06,
+      "loss": 0.6995,
+      "step": 7227
+    },
+    {
+      "epoch": 0.39782046342671584,
+      "grad_norm": 0.648560643196106,
+      "learning_rate": 9.075339311966627e-06,
+      "loss": 0.6997,
+      "step": 7228
+    },
+    {
+      "epoch": 0.3978755022290715,
+      "grad_norm": 0.8457548022270203,
+      "learning_rate": 9.075088161050134e-06,
+      "loss": 0.8548,
+      "step": 7229
+    },
+    {
+      "epoch": 0.39793054103142717,
+      "grad_norm": 0.7644637823104858,
+      "learning_rate": 9.074836979506373e-06,
+      "loss": 0.6966,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3979855798337828,
+      "grad_norm": 0.7146210670471191,
+      "learning_rate": 9.074585767337227e-06,
+      "loss": 0.7673,
+      "step": 7231
+    },
+    {
+      "epoch": 0.3980406186361385,
+      "grad_norm": 0.8570694327354431,
+      "learning_rate": 9.074334524544585e-06,
+      "loss": 0.8233,
+      "step": 7232
+    },
+    {
+      "epoch": 0.3980956574384941,
+      "grad_norm": 0.7257633805274963,
+      "learning_rate": 9.074083251130334e-06,
+      "loss": 0.7464,
+      "step": 7233
+    },
+    {
+      "epoch": 0.3981506962408498,
+      "grad_norm": 0.9377032518386841,
+      "learning_rate": 9.073831947096365e-06,
+      "loss": 0.7814,
+      "step": 7234
+    },
+    {
+      "epoch": 0.39820573504320544,
+      "grad_norm": 0.8105629086494446,
+      "learning_rate": 9.073580612444566e-06,
+      "loss": 0.8069,
+      "step": 7235
+    },
+    {
+      "epoch": 0.39826077384556113,
+      "grad_norm": 0.7874456644058228,
+      "learning_rate": 9.073329247176824e-06,
+      "loss": 0.8414,
+      "step": 7236
+    },
+    {
+      "epoch": 0.39831581264791677,
+      "grad_norm": 0.6829617023468018,
+      "learning_rate": 9.07307785129503e-06,
+      "loss": 0.7633,
+      "step": 7237
+    },
+    {
+      "epoch": 0.39837085145027246,
+      "grad_norm": 0.6838501691818237,
+      "learning_rate": 9.072826424801075e-06,
+      "loss": 0.6972,
+      "step": 7238
+    },
+    {
+      "epoch": 0.3984258902526281,
+      "grad_norm": 0.7054216861724854,
+      "learning_rate": 9.072574967696845e-06,
+      "loss": 0.8049,
+      "step": 7239
+    },
+    {
+      "epoch": 0.3984809290549838,
+      "grad_norm": 0.9462615847587585,
+      "learning_rate": 9.072323479984232e-06,
+      "loss": 0.7988,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3985359678573394,
+      "grad_norm": 0.7334465980529785,
+      "learning_rate": 9.072071961665128e-06,
+      "loss": 0.7538,
+      "step": 7241
+    },
+    {
+      "epoch": 0.3985910066596951,
+      "grad_norm": 0.7506609559059143,
+      "learning_rate": 9.071820412741418e-06,
+      "loss": 0.7991,
+      "step": 7242
+    },
+    {
+      "epoch": 0.39864604546205074,
+      "grad_norm": 0.6858688592910767,
+      "learning_rate": 9.071568833214998e-06,
+      "loss": 0.7258,
+      "step": 7243
+    },
+    {
+      "epoch": 0.3987010842644064,
+      "grad_norm": 0.8117396235466003,
+      "learning_rate": 9.071317223087754e-06,
+      "loss": 0.752,
+      "step": 7244
+    },
+    {
+      "epoch": 0.39875612306676206,
+      "grad_norm": 0.7772389054298401,
+      "learning_rate": 9.071065582361582e-06,
+      "loss": 0.7444,
+      "step": 7245
+    },
+    {
+      "epoch": 0.39881116186911775,
+      "grad_norm": 0.7221882939338684,
+      "learning_rate": 9.07081391103837e-06,
+      "loss": 0.8035,
+      "step": 7246
+    },
+    {
+      "epoch": 0.3988662006714734,
+      "grad_norm": 0.8113289475440979,
+      "learning_rate": 9.07056220912001e-06,
+      "loss": 0.7623,
+      "step": 7247
+    },
+    {
+      "epoch": 0.39892123947382907,
+      "grad_norm": 0.730823278427124,
+      "learning_rate": 9.070310476608395e-06,
+      "loss": 0.7872,
+      "step": 7248
+    },
+    {
+      "epoch": 0.3989762782761847,
+      "grad_norm": 0.7690893411636353,
+      "learning_rate": 9.070058713505415e-06,
+      "loss": 0.7402,
+      "step": 7249
+    },
+    {
+      "epoch": 0.3990313170785404,
+      "grad_norm": 0.6768597364425659,
+      "learning_rate": 9.069806919812963e-06,
+      "loss": 0.7283,
+      "step": 7250
+    },
+    {
+      "epoch": 0.399086355880896,
+      "grad_norm": 0.6938686370849609,
+      "learning_rate": 9.069555095532932e-06,
+      "loss": 0.7209,
+      "step": 7251
+    },
+    {
+      "epoch": 0.3991413946832517,
+      "grad_norm": 0.7162025570869446,
+      "learning_rate": 9.069303240667215e-06,
+      "loss": 0.7915,
+      "step": 7252
+    },
+    {
+      "epoch": 0.39919643348560735,
+      "grad_norm": 0.9170399308204651,
+      "learning_rate": 9.069051355217704e-06,
+      "loss": 0.8399,
+      "step": 7253
+    },
+    {
+      "epoch": 0.39925147228796304,
+      "grad_norm": 0.7080186009407043,
+      "learning_rate": 9.068799439186291e-06,
+      "loss": 0.8678,
+      "step": 7254
+    },
+    {
+      "epoch": 0.39930651109031867,
+      "grad_norm": 1.013613224029541,
+      "learning_rate": 9.068547492574872e-06,
+      "loss": 0.817,
+      "step": 7255
+    },
+    {
+      "epoch": 0.39936154989267436,
+      "grad_norm": 0.6911013722419739,
+      "learning_rate": 9.068295515385337e-06,
+      "loss": 0.7048,
+      "step": 7256
+    },
+    {
+      "epoch": 0.39941658869503,
+      "grad_norm": 0.748219907283783,
+      "learning_rate": 9.068043507619584e-06,
+      "loss": 0.8115,
+      "step": 7257
+    },
+    {
+      "epoch": 0.3994716274973857,
+      "grad_norm": 0.6763347387313843,
+      "learning_rate": 9.067791469279504e-06,
+      "loss": 0.763,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3995266662997413,
+      "grad_norm": 0.7291030287742615,
+      "learning_rate": 9.067539400366993e-06,
+      "loss": 0.7319,
+      "step": 7259
+    },
+    {
+      "epoch": 0.39958170510209695,
+      "grad_norm": 0.6515628695487976,
+      "learning_rate": 9.067287300883945e-06,
+      "loss": 0.7903,
+      "step": 7260
+    },
+    {
+      "epoch": 0.39963674390445264,
+      "grad_norm": 0.7815985679626465,
+      "learning_rate": 9.067035170832253e-06,
+      "loss": 0.8241,
+      "step": 7261
+    },
+    {
+      "epoch": 0.3996917827068083,
+      "grad_norm": 0.6747417449951172,
+      "learning_rate": 9.066783010213812e-06,
+      "loss": 0.7544,
+      "step": 7262
+    },
+    {
+      "epoch": 0.39974682150916396,
+      "grad_norm": 0.6568340063095093,
+      "learning_rate": 9.066530819030522e-06,
+      "loss": 0.7754,
+      "step": 7263
+    },
+    {
+      "epoch": 0.3998018603115196,
+      "grad_norm": 0.6703339219093323,
+      "learning_rate": 9.066278597284273e-06,
+      "loss": 0.7581,
+      "step": 7264
+    },
+    {
+      "epoch": 0.3998568991138753,
+      "grad_norm": 0.7421279549598694,
+      "learning_rate": 9.066026344976962e-06,
+      "loss": 0.7974,
+      "step": 7265
+    },
+    {
+      "epoch": 0.3999119379162309,
+      "grad_norm": 0.7226015329360962,
+      "learning_rate": 9.065774062110486e-06,
+      "loss": 0.7777,
+      "step": 7266
+    },
+    {
+      "epoch": 0.3999669767185866,
+      "grad_norm": 0.7092894911766052,
+      "learning_rate": 9.06552174868674e-06,
+      "loss": 0.7885,
+      "step": 7267
+    },
+    {
+      "epoch": 0.40002201552094224,
+      "grad_norm": 0.837902307510376,
+      "learning_rate": 9.065269404707622e-06,
+      "loss": 0.7425,
+      "step": 7268
+    },
+    {
+      "epoch": 0.40007705432329793,
+      "grad_norm": 0.803811252117157,
+      "learning_rate": 9.065017030175027e-06,
+      "loss": 0.8418,
+      "step": 7269
+    },
+    {
+      "epoch": 0.40013209312565357,
+      "grad_norm": 0.8110278248786926,
+      "learning_rate": 9.064764625090854e-06,
+      "loss": 0.7724,
+      "step": 7270
+    },
+    {
+      "epoch": 0.40018713192800925,
+      "grad_norm": 0.7305173277854919,
+      "learning_rate": 9.064512189456995e-06,
+      "loss": 0.7465,
+      "step": 7271
+    },
+    {
+      "epoch": 0.4002421707303649,
+      "grad_norm": 0.7312467694282532,
+      "learning_rate": 9.06425972327535e-06,
+      "loss": 0.8406,
+      "step": 7272
     }
   ],
   "logging_steps": 1,
@@ -44567,7 +50930,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.877760299885396e+19,
+  "total_flos": 2.1460117712975954e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null