diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15609 @@
+{
+  "best_metric": 0.19376881420612335,
+  "best_model_checkpoint": "d:\\\\whisper-large-v3-pt-3000h-4\\checkpoint-5529",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 55290,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004521613311629589,
+      "grad_norm": 2.3342323303222656,
+      "learning_rate": 2.5e-08,
+      "loss": 0.0451,
+      "step": 25
+    },
+    {
+      "epoch": 0.009043226623259178,
+      "grad_norm": 2.0830228328704834,
+      "learning_rate": 5e-08,
+      "loss": 0.0393,
+      "step": 50
+    },
+    {
+      "epoch": 0.013564839934888768,
+      "grad_norm": 3.438735008239746,
+      "learning_rate": 7.5e-08,
+      "loss": 0.0355,
+      "step": 75
+    },
+    {
+      "epoch": 0.018086453246518357,
+      "grad_norm": 14.74426555633545,
+      "learning_rate": 1e-07,
+      "loss": 0.0359,
+      "step": 100
+    },
+    {
+      "epoch": 0.022608066558147948,
+      "grad_norm": 2.5502216815948486,
+      "learning_rate": 1.25e-07,
+      "loss": 0.028,
+      "step": 125
+    },
+    {
+      "epoch": 0.027129679869777535,
+      "grad_norm": 6.838224411010742,
+      "learning_rate": 1.5e-07,
+      "loss": 0.0214,
+      "step": 150
+    },
+    {
+      "epoch": 0.03165129318140713,
+      "grad_norm": 6.352050304412842,
+      "learning_rate": 1.75e-07,
+      "loss": 0.0336,
+      "step": 175
+    },
+    {
+      "epoch": 0.03617290649303671,
+      "grad_norm": 8.222951889038086,
+      "learning_rate": 2e-07,
+      "loss": 0.028,
+      "step": 200
+    },
+    {
+      "epoch": 0.040694519804666304,
+      "grad_norm": 5.666439533233643,
+      "learning_rate": 2.25e-07,
+      "loss": 0.0295,
+      "step": 225
+    },
+    {
+      "epoch": 0.045216133116295895,
+      "grad_norm": 0.2596105635166168,
+      "learning_rate": 2.5e-07,
+      "loss": 0.0338,
+      "step": 250
+    },
+    {
+      "epoch": 0.049737746427925486,
+      "grad_norm": 4.480187892913818,
+      "learning_rate": 2.75e-07,
+      "loss": 0.0365,
+      "step": 275
+    },
+    {
+      "epoch": 0.05425935973955507,
+      "grad_norm": 0.934718132019043,
+      "learning_rate": 3e-07,
+      "loss": 0.0284,
+      "step": 300
+    },
+    {
+      "epoch": 0.05878097305118466,
+      "grad_norm": 1.4461249113082886,
+      "learning_rate": 3.25e-07,
+      "loss": 0.0199,
+      "step": 325
+    },
+    {
+      "epoch": 0.06330258636281426,
+      "grad_norm": 3.453277587890625,
+      "learning_rate": 3.5e-07,
+      "loss": 0.0401,
+      "step": 350
+    },
+    {
+      "epoch": 0.06782419967444384,
+      "grad_norm": 6.596461772918701,
+      "learning_rate": 3.75e-07,
+      "loss": 0.0522,
+      "step": 375
+    },
+    {
+      "epoch": 0.07234581298607343,
+      "grad_norm": 0.7677063345909119,
+      "learning_rate": 3.99e-07,
+      "loss": 0.1004,
+      "step": 400
+    },
+    {
+      "epoch": 0.07686742629770302,
+      "grad_norm": 7.2375264167785645,
+      "learning_rate": 4.24e-07,
+      "loss": 0.0313,
+      "step": 425
+    },
+    {
+      "epoch": 0.08138903960933261,
+      "grad_norm": 4.461888790130615,
+      "learning_rate": 4.49e-07,
+      "loss": 0.0358,
+      "step": 450
+    },
+    {
+      "epoch": 0.0859106529209622,
+      "grad_norm": 2.9905307292938232,
+      "learning_rate": 4.7399999999999993e-07,
+      "loss": 0.035,
+      "step": 475
+    },
+    {
+      "epoch": 0.09043226623259179,
+      "grad_norm": 2.32279896736145,
+      "learning_rate": 4.99e-07,
+      "loss": 0.0296,
+      "step": 500
+    },
+    {
+      "epoch": 0.09495387954422138,
+      "grad_norm": 8.56678581237793,
+      "learning_rate": 5.24e-07,
+      "loss": 0.0357,
+      "step": 525
+    },
+    {
+      "epoch": 0.09947549285585097,
+      "grad_norm": 2.7547428607940674,
+      "learning_rate": 5.490000000000001e-07,
+      "loss": 0.0378,
+      "step": 550
+    },
+    {
+      "epoch": 0.10399710616748056,
+      "grad_norm": 1.2645742893218994,
+      "learning_rate": 5.739999999999999e-07,
+      "loss": 0.0434,
+      "step": 575
+    },
+    {
+      "epoch": 0.10851871947911014,
+      "grad_norm": 2.233053684234619,
+      "learning_rate": 5.989999999999999e-07,
+      "loss": 0.042,
+      "step": 600
+    },
+    {
+      "epoch": 0.11304033279073973,
+      "grad_norm": 13.10155200958252,
+      "learning_rate": 6.24e-07,
+      "loss": 0.0454,
+      "step": 625
+    },
+    {
+      "epoch": 0.11756194610236932,
+      "grad_norm": 9.450150489807129,
+      "learning_rate": 6.49e-07,
+      "loss": 0.0331,
+      "step": 650
+    },
+    {
+      "epoch": 0.12208355941399891,
+      "grad_norm": 0.5919239521026611,
+      "learning_rate": 6.74e-07,
+      "loss": 0.033,
+      "step": 675
+    },
+    {
+      "epoch": 0.12660517272562852,
+      "grad_norm": 24.587112426757812,
+      "learning_rate": 6.989999999999999e-07,
+      "loss": 0.0311,
+      "step": 700
+    },
+    {
+      "epoch": 0.13112678603725808,
+      "grad_norm": 19.61304473876953,
+      "learning_rate": 7.24e-07,
+      "loss": 0.0359,
+      "step": 725
+    },
+    {
+      "epoch": 0.13564839934888767,
+      "grad_norm": 20.30430030822754,
+      "learning_rate": 7.489999999999999e-07,
+      "loss": 0.0531,
+      "step": 750
+    },
+    {
+      "epoch": 0.14017001266051726,
+      "grad_norm": 16.841527938842773,
+      "learning_rate": 7.74e-07,
+      "loss": 0.0761,
+      "step": 775
+    },
+    {
+      "epoch": 0.14469162597214685,
+      "grad_norm": 18.845434188842773,
+      "learning_rate": 7.99e-07,
+      "loss": 0.0871,
+      "step": 800
+    },
+    {
+      "epoch": 0.14921323928377644,
+      "grad_norm": 14.67888355255127,
+      "learning_rate": 8.24e-07,
+      "loss": 0.0414,
+      "step": 825
+    },
+    {
+      "epoch": 0.15373485259540604,
+      "grad_norm": 6.383974075317383,
+      "learning_rate": 8.489999999999999e-07,
+      "loss": 0.0344,
+      "step": 850
+    },
+    {
+      "epoch": 0.15825646590703563,
+      "grad_norm": 7.403707504272461,
+      "learning_rate": 8.739999999999999e-07,
+      "loss": 0.0317,
+      "step": 875
+    },
+    {
+      "epoch": 0.16277807921866522,
+      "grad_norm": 0.7672566771507263,
+      "learning_rate": 8.99e-07,
+      "loss": 0.0319,
+      "step": 900
+    },
+    {
+      "epoch": 0.1672996925302948,
+      "grad_norm": 3.935624599456787,
+      "learning_rate": 9.24e-07,
+      "loss": 0.0364,
+      "step": 925
+    },
+    {
+      "epoch": 0.1718213058419244,
+      "grad_norm": 6.021194934844971,
+      "learning_rate": 9.489999999999999e-07,
+      "loss": 0.0378,
+      "step": 950
+    },
+    {
+      "epoch": 0.176342919153554,
+      "grad_norm": 0.7235254645347595,
+      "learning_rate": 9.74e-07,
+      "loss": 0.0386,
+      "step": 975
+    },
+    {
+      "epoch": 0.18086453246518358,
+      "grad_norm": 5.369808673858643,
+      "learning_rate": 9.989999999999999e-07,
+      "loss": 0.0307,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18538614577681317,
+      "grad_norm": 4.743622303009033,
+      "learning_rate": 9.995579296371338e-07,
+      "loss": 0.0555,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18990775908844276,
+      "grad_norm": 12.76374340057373,
+      "learning_rate": 9.99097439675815e-07,
+      "loss": 0.0478,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19442937240007235,
+      "grad_norm": 4.498536109924316,
+      "learning_rate": 9.986369497144962e-07,
+      "loss": 0.0443,
+      "step": 1075
+    },
+    {
+      "epoch": 0.19895098571170194,
+      "grad_norm": 9.623854637145996,
+      "learning_rate": 9.981764597531773e-07,
+      "loss": 0.044,
+      "step": 1100
+    },
+    {
+      "epoch": 0.20347259902333154,
+      "grad_norm": 1.4003485441207886,
+      "learning_rate": 9.977159697918585e-07,
+      "loss": 0.042,
+      "step": 1125
+    },
+    {
+      "epoch": 0.20799421233496113,
+      "grad_norm": 3.358299493789673,
+      "learning_rate": 9.972554798305397e-07,
+      "loss": 0.0422,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21251582564659072,
+      "grad_norm": 33.19310760498047,
+      "learning_rate": 9.967949898692208e-07,
+      "loss": 0.0443,
+      "step": 1175
+    },
+    {
+      "epoch": 0.21703743895822028,
+      "grad_norm": 7.122214317321777,
+      "learning_rate": 9.96334499907902e-07,
+      "loss": 0.0962,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22155905226984987,
+      "grad_norm": 8.756924629211426,
+      "learning_rate": 9.958740099465832e-07,
+      "loss": 0.0399,
+      "step": 1225
+    },
+    {
+      "epoch": 0.22608066558147946,
+      "grad_norm": 3.779580593109131,
+      "learning_rate": 9.954135199852643e-07,
+      "loss": 0.035,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23060227889310905,
+      "grad_norm": 15.683145523071289,
+      "learning_rate": 9.949530300239455e-07,
+      "loss": 0.042,
+      "step": 1275
+    },
+    {
+      "epoch": 0.23512389220473864,
+      "grad_norm": 12.767990112304688,
+      "learning_rate": 9.944925400626267e-07,
+      "loss": 0.0517,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23964550551636823,
+      "grad_norm": 0.5977092981338501,
+      "learning_rate": 9.940320501013078e-07,
+      "loss": 0.0266,
+      "step": 1325
+    },
+    {
+      "epoch": 0.24416711882799783,
+      "grad_norm": 2.613328218460083,
+      "learning_rate": 9.935715601399888e-07,
+      "loss": 0.0302,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24868873213962742,
+      "grad_norm": 3.43841290473938,
+      "learning_rate": 9.9311107017867e-07,
+      "loss": 0.0604,
+      "step": 1375
+    },
+    {
+      "epoch": 0.25321034545125704,
+      "grad_norm": 1.4975024461746216,
+      "learning_rate": 9.926505802173513e-07,
+      "loss": 0.0524,
+      "step": 1400
+    },
+    {
+      "epoch": 0.25773195876288657,
+      "grad_norm": 6.18322229385376,
+      "learning_rate": 9.921900902560325e-07,
+      "loss": 0.0318,
+      "step": 1425
+    },
+    {
+      "epoch": 0.26225357207451616,
+      "grad_norm": 3.4327890872955322,
+      "learning_rate": 9.917296002947134e-07,
+      "loss": 0.0449,
+      "step": 1450
+    },
+    {
+      "epoch": 0.26677518538614575,
+      "grad_norm": 19.299654006958008,
+      "learning_rate": 9.912691103333946e-07,
+      "loss": 0.0371,
+      "step": 1475
+    },
+    {
+      "epoch": 0.27129679869777534,
+      "grad_norm": 17.62229347229004,
+      "learning_rate": 9.908086203720758e-07,
+      "loss": 0.046,
+      "step": 1500
+    },
+    {
+      "epoch": 0.27581841200940493,
+      "grad_norm": 1.671424150466919,
+      "learning_rate": 9.90348130410757e-07,
+      "loss": 0.0391,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2803400253210345,
+      "grad_norm": 21.561243057250977,
+      "learning_rate": 9.898876404494381e-07,
+      "loss": 0.0729,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2848616386326641,
+      "grad_norm": 7.39694356918335,
+      "learning_rate": 9.894271504881193e-07,
+      "loss": 0.052,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2893832519442937,
+      "grad_norm": 1.9565777778625488,
+      "learning_rate": 9.889850801252534e-07,
+      "loss": 0.082,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2939048652559233,
+      "grad_norm": 1.8609459400177002,
+      "learning_rate": 9.885245901639343e-07,
+      "loss": 0.0446,
+      "step": 1625
+    },
+    {
+      "epoch": 0.2984264785675529,
+      "grad_norm": 4.542464733123779,
+      "learning_rate": 9.880641002026155e-07,
+      "loss": 0.0353,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3029480918791825,
+      "grad_norm": 0.291640967130661,
+      "learning_rate": 9.876036102412967e-07,
+      "loss": 0.0359,
+      "step": 1675
+    },
+    {
+      "epoch": 0.30746970519081207,
+      "grad_norm": 0.6001373529434204,
+      "learning_rate": 9.871431202799778e-07,
+      "loss": 0.031,
+      "step": 1700
+    },
+    {
+      "epoch": 0.31199131850244166,
+      "grad_norm": 12.14961051940918,
+      "learning_rate": 9.86682630318659e-07,
+      "loss": 0.0447,
+      "step": 1725
+    },
+    {
+      "epoch": 0.31651293181407125,
+      "grad_norm": 23.949186325073242,
+      "learning_rate": 9.862221403573402e-07,
+      "loss": 0.0479,
+      "step": 1750
+    },
+    {
+      "epoch": 0.32103454512570084,
+      "grad_norm": 4.505954742431641,
+      "learning_rate": 9.857616503960213e-07,
+      "loss": 0.0481,
+      "step": 1775
+    },
+    {
+      "epoch": 0.32555615843733043,
+      "grad_norm": 2.596393585205078,
+      "learning_rate": 9.853011604347025e-07,
+      "loss": 0.0482,
+      "step": 1800
+    },
+    {
+      "epoch": 0.33007777174896,
+      "grad_norm": 8.913126945495605,
+      "learning_rate": 9.848406704733837e-07,
+      "loss": 0.0612,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3345993850605896,
+      "grad_norm": 5.877014636993408,
+      "learning_rate": 9.843801805120648e-07,
+      "loss": 0.0476,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3391209983722192,
+      "grad_norm": 22.622203826904297,
+      "learning_rate": 9.83919690550746e-07,
+      "loss": 0.0447,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3436426116838488,
+      "grad_norm": 18.984943389892578,
+      "learning_rate": 9.834592005894271e-07,
+      "loss": 0.0543,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3481642249954784,
+      "grad_norm": 13.041614532470703,
+      "learning_rate": 9.829987106281083e-07,
+      "loss": 0.0393,
+      "step": 1925
+    },
+    {
+      "epoch": 0.352685838307108,
+      "grad_norm": 23.434789657592773,
+      "learning_rate": 9.825382206667893e-07,
+      "loss": 0.0488,
+      "step": 1950
+    },
+    {
+      "epoch": 0.35720745161873757,
+      "grad_norm": 7.045641899108887,
+      "learning_rate": 9.820777307054706e-07,
+      "loss": 0.0557,
+      "step": 1975
+    },
+    {
+      "epoch": 0.36172906493036716,
+      "grad_norm": 17.288244247436523,
+      "learning_rate": 9.816172407441518e-07,
+      "loss": 0.0821,
+      "step": 2000
+    },
+    {
+      "epoch": 0.36625067824199675,
+      "grad_norm": 4.284151077270508,
+      "learning_rate": 9.81156750782833e-07,
+      "loss": 0.0526,
+      "step": 2025
+    },
+    {
+      "epoch": 0.37077229155362634,
+      "grad_norm": 9.124343872070312,
+      "learning_rate": 9.80696260821514e-07,
+      "loss": 0.0309,
+      "step": 2050
+    },
+    {
+      "epoch": 0.37529390486525593,
+      "grad_norm": 4.401906490325928,
+      "learning_rate": 9.80235770860195e-07,
+      "loss": 0.046,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3798155181768855,
+      "grad_norm": 0.29334700107574463,
+      "learning_rate": 9.797752808988765e-07,
+      "loss": 0.0349,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3843371314885151,
+      "grad_norm": 4.6845197677612305,
+      "learning_rate": 9.793147909375574e-07,
+      "loss": 0.0398,
+      "step": 2125
+    },
+    {
+      "epoch": 0.3888587448001447,
+      "grad_norm": 10.254193305969238,
+      "learning_rate": 9.788543009762386e-07,
+      "loss": 0.0441,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3933803581117743,
+      "grad_norm": 9.839592933654785,
+      "learning_rate": 9.783938110149198e-07,
+      "loss": 0.0427,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3979019714234039,
+      "grad_norm": 7.145606517791748,
+      "learning_rate": 9.77933321053601e-07,
+      "loss": 0.0364,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4024235847350335,
+      "grad_norm": 13.468842506408691,
+      "learning_rate": 9.77472831092282e-07,
+      "loss": 0.0578,
+      "step": 2225
+    },
+    {
+      "epoch": 0.40694519804666307,
+      "grad_norm": 5.4638519287109375,
+      "learning_rate": 9.770123411309633e-07,
+      "loss": 0.0596,
+      "step": 2250
+    },
+    {
+      "epoch": 0.41146681135829266,
+      "grad_norm": 3.9997127056121826,
+      "learning_rate": 9.765518511696444e-07,
+      "loss": 0.0304,
+      "step": 2275
+    },
+    {
+      "epoch": 0.41598842466992225,
+      "grad_norm": 1.2957385778427124,
+      "learning_rate": 9.760913612083256e-07,
+      "loss": 0.0308,
+      "step": 2300
+    },
+    {
+      "epoch": 0.42051003798155184,
+      "grad_norm": 15.630707740783691,
+      "learning_rate": 9.756308712470068e-07,
+      "loss": 0.0538,
+      "step": 2325
+    },
+    {
+      "epoch": 0.42503165129318143,
+      "grad_norm": 0.9495472311973572,
+      "learning_rate": 9.75170381285688e-07,
+      "loss": 0.0484,
+      "step": 2350
+    },
+    {
+      "epoch": 0.42955326460481097,
+      "grad_norm": 8.299753189086914,
+      "learning_rate": 9.74709891324369e-07,
+      "loss": 0.0941,
+      "step": 2375
+    },
+    {
+      "epoch": 0.43407487791644056,
+      "grad_norm": 2.055479049682617,
+      "learning_rate": 9.742494013630503e-07,
+      "loss": 0.0871,
+      "step": 2400
+    },
+    {
+      "epoch": 0.43859649122807015,
+      "grad_norm": 2.5821428298950195,
+      "learning_rate": 9.737889114017314e-07,
+      "loss": 0.0319,
+      "step": 2425
+    },
+    {
+      "epoch": 0.44311810453969974,
+      "grad_norm": 15.306851387023926,
+      "learning_rate": 9.733284214404126e-07,
+      "loss": 0.0397,
+      "step": 2450
+    },
+    {
+      "epoch": 0.44763971785132933,
+      "grad_norm": 12.804640769958496,
+      "learning_rate": 9.728679314790938e-07,
+      "loss": 0.0397,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4521613311629589,
+      "grad_norm": 2.8858513832092285,
+      "learning_rate": 9.72407441517775e-07,
+      "loss": 0.0423,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4566829444745885,
+      "grad_norm": 5.125776767730713,
+      "learning_rate": 9.71946951556456e-07,
+      "loss": 0.0372,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4612045577862181,
+      "grad_norm": 3.445122718811035,
+      "learning_rate": 9.71486461595137e-07,
+      "loss": 0.0597,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4657261710978477,
+      "grad_norm": 1.8609939813613892,
+      "learning_rate": 9.710259716338184e-07,
+      "loss": 0.0356,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4702477844094773,
+      "grad_norm": 6.8225998878479,
+      "learning_rate": 9.705654816724996e-07,
+      "loss": 0.0341,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4747693977211069,
+      "grad_norm": 20.26767349243164,
+      "learning_rate": 9.701049917111808e-07,
+      "loss": 0.049,
+      "step": 2625
+    },
+    {
+      "epoch": 0.47929101103273647,
+      "grad_norm": 5.631345748901367,
+      "learning_rate": 9.696445017498617e-07,
+      "loss": 0.0616,
+      "step": 2650
+    },
+    {
+      "epoch": 0.48381262434436606,
+      "grad_norm": 20.55868148803711,
+      "learning_rate": 9.691840117885429e-07,
+      "loss": 0.062,
+      "step": 2675
+    },
+    {
+      "epoch": 0.48833423765599565,
+      "grad_norm": 32.73431396484375,
+      "learning_rate": 9.687235218272243e-07,
+      "loss": 0.0653,
+      "step": 2700
+    },
+    {
+      "epoch": 0.49285585096762524,
+      "grad_norm": 1.6700994968414307,
+      "learning_rate": 9.682630318659052e-07,
+      "loss": 0.0486,
+      "step": 2725
+    },
+    {
+      "epoch": 0.49737746427925483,
+      "grad_norm": 14.762917518615723,
+      "learning_rate": 9.678025419045864e-07,
+      "loss": 0.0641,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5018990775908845,
+      "grad_norm": 1.297577977180481,
+      "learning_rate": 9.673420519432675e-07,
+      "loss": 0.0625,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5064206909025141,
+      "grad_norm": 0.5399802327156067,
+      "learning_rate": 9.668815619819487e-07,
+      "loss": 0.0931,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5109423042141437,
+      "grad_norm": 7.880299091339111,
+      "learning_rate": 9.664210720206299e-07,
+      "loss": 0.0397,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 5.446402549743652,
+      "learning_rate": 9.65960582059311e-07,
+      "loss": 0.0327,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5199855308374027,
+      "grad_norm": 10.012272834777832,
+      "learning_rate": 9.655000920979922e-07,
+      "loss": 0.0405,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5245071441490323,
+      "grad_norm": 4.505229949951172,
+      "learning_rate": 9.650396021366734e-07,
+      "loss": 0.0372,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5290287574606619,
+      "grad_norm": 0.4095817804336548,
+      "learning_rate": 9.645791121753545e-07,
+      "loss": 0.0447,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5335503707722915,
+      "grad_norm": 1.1292011737823486,
+      "learning_rate": 9.641186222140357e-07,
+      "loss": 0.0306,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5380719840839211,
+      "grad_norm": 2.467514991760254,
+      "learning_rate": 9.636581322527169e-07,
+      "loss": 0.056,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5425935973955507,
+      "grad_norm": 2.929332971572876,
+      "learning_rate": 9.63197642291398e-07,
+      "loss": 0.0551,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5471152107071803,
+      "grad_norm": 4.477908134460449,
+      "learning_rate": 9.627371523300792e-07,
+      "loss": 0.0396,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5516368240188099,
+      "grad_norm": 2.9823644161224365,
+      "learning_rate": 9.622766623687604e-07,
+      "loss": 0.0394,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5561584373304395,
+      "grad_norm": 19.203933715820312,
+      "learning_rate": 9.618161724074415e-07,
+      "loss": 0.0474,
+      "step": 3075
+    },
+    {
+      "epoch": 0.560680050642069,
+      "grad_norm": 3.774156093597412,
+      "learning_rate": 9.613556824461227e-07,
+      "loss": 0.0459,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5652016639536986,
+      "grad_norm": 8.178613662719727,
+      "learning_rate": 9.608951924848039e-07,
+      "loss": 0.0704,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5697232772653282,
+      "grad_norm": 21.61182403564453,
+      "learning_rate": 9.604347025234848e-07,
+      "loss": 0.0728,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5742448905769578,
+      "grad_norm": 19.002914428710938,
+      "learning_rate": 9.599742125621662e-07,
+      "loss": 0.0839,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5787665038885874,
+      "grad_norm": 4.789153099060059,
+      "learning_rate": 9.595137226008474e-07,
+      "loss": 0.0816,
+      "step": 3200
+    },
+    {
+      "epoch": 0.583288117200217,
+      "grad_norm": 0.4933367669582367,
+      "learning_rate": 9.590532326395283e-07,
+      "loss": 0.0534,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5878097305118466,
+      "grad_norm": 1.4477636814117432,
+      "learning_rate": 9.585927426782095e-07,
+      "loss": 0.0479,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5923313438234762,
+      "grad_norm": 5.407076358795166,
+      "learning_rate": 9.581322527168907e-07,
+      "loss": 0.0328,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5968529571351058,
+      "grad_norm": 6.760631084442139,
+      "learning_rate": 9.57671762755572e-07,
+      "loss": 0.0421,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6013745704467354,
+      "grad_norm": 11.133447647094727,
+      "learning_rate": 9.57211272794253e-07,
+      "loss": 0.0315,
+      "step": 3325
+    },
+    {
+      "epoch": 0.605896183758365,
+      "grad_norm": 8.827392578125,
+      "learning_rate": 9.567507828329342e-07,
+      "loss": 0.0533,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6104177970699945,
+      "grad_norm": 4.74755334854126,
+      "learning_rate": 9.562902928716153e-07,
+      "loss": 0.0505,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6149394103816241,
+      "grad_norm": 6.094758987426758,
+      "learning_rate": 9.558298029102965e-07,
+      "loss": 0.0374,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6194610236932537,
+      "grad_norm": 7.404526710510254,
+      "learning_rate": 9.553693129489777e-07,
+      "loss": 0.0274,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6239826370048833,
+      "grad_norm": 2.237982749938965,
+      "learning_rate": 9.549088229876588e-07,
+      "loss": 0.0395,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6285042503165129,
+      "grad_norm": 20.463314056396484,
+      "learning_rate": 9.5444833302634e-07,
+      "loss": 0.0596,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6330258636281425,
+      "grad_norm": 12.629451751708984,
+      "learning_rate": 9.539878430650212e-07,
+      "loss": 0.0329,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6375474769397721,
+      "grad_norm": 19.752243041992188,
+      "learning_rate": 9.535273531037022e-07,
+      "loss": 0.0343,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6420690902514017,
+      "grad_norm": 7.476962089538574,
+      "learning_rate": 9.530668631423835e-07,
+      "loss": 0.0724,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6465907035630313,
+      "grad_norm": 31.683101654052734,
+      "learning_rate": 9.526063731810647e-07,
+      "loss": 0.07,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6511123168746609,
+      "grad_norm": 3.1800105571746826,
+      "learning_rate": 9.521458832197457e-07,
+      "loss": 0.1358,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6556339301862905,
+      "grad_norm": 7.005350589752197,
+      "learning_rate": 9.516853932584269e-07,
+      "loss": 0.0643,
+      "step": 3625
+    },
+    {
+      "epoch": 0.66015554349792,
+      "grad_norm": 5.049868583679199,
+      "learning_rate": 9.51224903297108e-07,
+      "loss": 0.0469,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6646771568095496,
+      "grad_norm": 6.004695415496826,
+      "learning_rate": 9.507644133357893e-07,
+      "loss": 0.0377,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6691987701211792,
+      "grad_norm": 14.023488998413086,
+      "learning_rate": 9.503039233744704e-07,
+      "loss": 0.0457,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6737203834328088,
+      "grad_norm": 0.5092989802360535,
+      "learning_rate": 9.498434334131515e-07,
+      "loss": 0.0467,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6782419967444384,
+      "grad_norm": 9.694477081298828,
+      "learning_rate": 9.493829434518327e-07,
+      "loss": 0.0307,
+      "step": 3750
+    },
+    {
+      "epoch": 0.682763610056068,
+      "grad_norm": 8.313312530517578,
+      "learning_rate": 9.489224534905139e-07,
+      "loss": 0.0602,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6872852233676976,
+      "grad_norm": 9.547676086425781,
+      "learning_rate": 9.48461963529195e-07,
+      "loss": 0.051,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6918068366793272,
+      "grad_norm": 5.230764865875244,
+      "learning_rate": 9.480014735678762e-07,
+      "loss": 0.0463,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6963284499909568,
+      "grad_norm": 6.57362699508667,
+      "learning_rate": 9.475409836065573e-07,
+      "loss": 0.0564,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7008500633025864,
+      "grad_norm": 8.198570251464844,
+      "learning_rate": 9.470804936452384e-07,
+      "loss": 0.0345,
+      "step": 3875
+    },
+    {
+      "epoch": 0.705371676614216,
+      "grad_norm": 12.034746170043945,
+      "learning_rate": 9.466200036839197e-07,
+      "loss": 0.0728,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7098932899258455,
+      "grad_norm": 14.528310775756836,
+      "learning_rate": 9.461595137226009e-07,
+      "loss": 0.0473,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7144149032374751,
+      "grad_norm": 1.0361833572387695,
+      "learning_rate": 9.456990237612819e-07,
+      "loss": 0.0566,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7189365165491047,
+      "grad_norm": 18.10868263244629,
+      "learning_rate": 9.452385337999631e-07,
+      "loss": 0.0647,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7234581298607343,
+      "grad_norm": 0.4723142981529236,
+      "learning_rate": 9.447780438386443e-07,
+      "loss": 0.0936,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7279797431723639,
+      "grad_norm": 2.1907668113708496,
+      "learning_rate": 9.443175538773254e-07,
+      "loss": 0.0548,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7325013564839935,
+      "grad_norm": 9.852395057678223,
+      "learning_rate": 9.438570639160066e-07,
+      "loss": 0.0568,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7370229697956231,
+      "grad_norm": 4.197534084320068,
+      "learning_rate": 9.433965739546878e-07,
+      "loss": 0.0393,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7415445831072527,
+      "grad_norm": 2.691697835922241,
+      "learning_rate": 9.429360839933689e-07,
+      "loss": 0.0486,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7460661964188823,
+      "grad_norm": 6.5368452072143555,
+      "learning_rate": 9.4247559403205e-07,
+      "loss": 0.0346,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7505878097305119,
+      "grad_norm": 5.945896148681641,
+      "learning_rate": 9.420151040707313e-07,
+      "loss": 0.0583,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7551094230421415,
+      "grad_norm": 0.8333092331886292,
+      "learning_rate": 9.415546141094124e-07,
+      "loss": 0.0532,
+      "step": 4175
+    },
+    {
+      "epoch": 0.759631036353771,
+      "grad_norm": 6.2607879638671875,
+      "learning_rate": 9.410941241480935e-07,
+      "loss": 0.0341,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7641526496654006,
+      "grad_norm": 20.586143493652344,
+      "learning_rate": 9.406336341867747e-07,
+      "loss": 0.0403,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7686742629770302,
+      "grad_norm": 7.338830947875977,
+      "learning_rate": 9.401731442254558e-07,
+      "loss": 0.0727,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7731958762886598,
+      "grad_norm": 0.2722490429878235,
+      "learning_rate": 9.39712654264137e-07,
+      "loss": 0.0489,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7777174896002894,
+      "grad_norm": 0.15249581634998322,
+      "learning_rate": 9.392521643028182e-07,
+      "loss": 0.0519,
+      "step": 4300
+    },
+    {
+      "epoch": 0.782239102911919,
+      "grad_norm": 2.320035457611084,
+      "learning_rate": 9.38810093939952e-07,
+      "loss": 0.1012,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7867607162235486,
+      "grad_norm": 21.31127166748047,
+      "learning_rate": 9.383496039786333e-07,
+      "loss": 0.0479,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7912823295351782,
+      "grad_norm": 0.8797614574432373,
+      "learning_rate": 9.378891140173144e-07,
+      "loss": 0.054,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7958039428468078,
+      "grad_norm": 12.751317977905273,
+      "learning_rate": 9.374286240559955e-07,
+      "loss": 0.1115,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8003255561584374,
+      "grad_norm": 3.1984500885009766,
+      "learning_rate": 9.369681340946767e-07,
+      "loss": 0.0431,
+      "step": 4425
+    },
+    {
+      "epoch": 0.804847169470067,
+      "grad_norm": 4.108559608459473,
+      "learning_rate": 9.365076441333579e-07,
+      "loss": 0.0326,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8093687827816965,
+      "grad_norm": 9.827606201171875,
+      "learning_rate": 9.36047154172039e-07,
+      "loss": 0.0454,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8138903960933261,
+      "grad_norm": 6.922112941741943,
+      "learning_rate": 9.355866642107202e-07,
+      "loss": 0.0472,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8184120094049557,
+      "grad_norm": 3.6166677474975586,
+      "learning_rate": 9.351261742494014e-07,
+      "loss": 0.0495,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8229336227165853,
+      "grad_norm": 6.462372303009033,
+      "learning_rate": 9.346656842880824e-07,
+      "loss": 0.0273,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8274552360282149,
+      "grad_norm": 6.296558856964111,
+      "learning_rate": 9.342051943267636e-07,
+      "loss": 0.0557,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8319768493398445,
+      "grad_norm": 4.71564245223999,
+      "learning_rate": 9.337447043654449e-07,
+      "loss": 0.0524,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8364984626514741,
+      "grad_norm": 5.095015048980713,
+      "learning_rate": 9.332842144041259e-07,
+      "loss": 0.0468,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8410200759631037,
+      "grad_norm": 1.5702929496765137,
+      "learning_rate": 9.328237244428071e-07,
+      "loss": 0.0524,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8455416892747333,
+      "grad_norm": 13.924525260925293,
+      "learning_rate": 9.323632344814883e-07,
+      "loss": 0.0681,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8500633025863629,
+      "grad_norm": 1.5610660314559937,
+      "learning_rate": 9.319027445201694e-07,
+      "loss": 0.0361,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8545849158979925,
+      "grad_norm": 15.312824249267578,
+      "learning_rate": 9.314422545588506e-07,
+      "loss": 0.0799,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8591065292096219,
+      "grad_norm": 12.912773132324219,
+      "learning_rate": 9.309817645975318e-07,
+      "loss": 0.0768,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8636281425212515,
+      "grad_norm": 0.2809099853038788,
+      "learning_rate": 9.305212746362129e-07,
+      "loss": 0.0832,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8681497558328811,
+      "grad_norm": 16.709447860717773,
+      "learning_rate": 9.30060784674894e-07,
+      "loss": 0.1101,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8726713691445107,
+      "grad_norm": 3.765148401260376,
+      "learning_rate": 9.296002947135751e-07,
+      "loss": 0.0513,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8771929824561403,
+      "grad_norm": 5.886707782745361,
+      "learning_rate": 9.291398047522564e-07,
+      "loss": 0.0414,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8817145957677699,
+      "grad_norm": 5.642549991607666,
+      "learning_rate": 9.286793147909375e-07,
+      "loss": 0.0418,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8862362090793995,
+      "grad_norm": 2.7111074924468994,
+      "learning_rate": 9.282188248296186e-07,
+      "loss": 0.0352,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8907578223910291,
+      "grad_norm": 8.905896186828613,
+      "learning_rate": 9.277583348682998e-07,
+      "loss": 0.0417,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8952794357026587,
+      "grad_norm": 5.965810775756836,
+      "learning_rate": 9.27297844906981e-07,
+      "loss": 0.056,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8998010490142883,
+      "grad_norm": 14.578080177307129,
+      "learning_rate": 9.268373549456621e-07,
+      "loss": 0.054,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9043226623259178,
+      "grad_norm": 8.001260757446289,
+      "learning_rate": 9.263768649843433e-07,
+      "loss": 0.0379,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9088442756375474,
+      "grad_norm": 10.563271522521973,
+      "learning_rate": 9.259163750230245e-07,
+      "loss": 0.0492,
+      "step": 5025
+    },
+    {
+      "epoch": 0.913365888949177,
+      "grad_norm": 6.896139621734619,
+      "learning_rate": 9.254558850617055e-07,
+      "loss": 0.0541,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9178875022608066,
+      "grad_norm": 0.6722992658615112,
+      "learning_rate": 9.249953951003868e-07,
+      "loss": 0.0523,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9224091155724362,
+      "grad_norm": 0.5435565710067749,
+      "learning_rate": 9.24534905139068e-07,
+      "loss": 0.0554,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9269307288840658,
+      "grad_norm": 17.15158462524414,
+      "learning_rate": 9.240744151777491e-07,
+      "loss": 0.1066,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9314523421956954,
+      "grad_norm": 4.256832599639893,
+      "learning_rate": 9.236139252164302e-07,
+      "loss": 0.0761,
+      "step": 5150
+    },
+    {
+      "epoch": 0.935973955507325,
+      "grad_norm": 1.2044847011566162,
+      "learning_rate": 9.231534352551114e-07,
+      "loss": 0.05,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9404955688189546,
+      "grad_norm": 36.5662956237793,
+      "learning_rate": 9.226929452937926e-07,
+      "loss": 0.1034,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9450171821305842,
+      "grad_norm": 6.578591823577881,
+      "learning_rate": 9.222324553324737e-07,
+      "loss": 0.0497,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9495387954422138,
+      "grad_norm": 0.6094868779182434,
+      "learning_rate": 9.217719653711549e-07,
+      "loss": 0.0396,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9540604087538433,
+      "grad_norm": 2.0846738815307617,
+      "learning_rate": 9.21311475409836e-07,
+      "loss": 0.0271,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9585820220654729,
+      "grad_norm": 2.4080617427825928,
+      "learning_rate": 9.208509854485171e-07,
+      "loss": 0.0568,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9631036353771025,
+      "grad_norm": 4.527153968811035,
+      "learning_rate": 9.203904954871984e-07,
+      "loss": 0.0346,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9676252486887321,
+      "grad_norm": 4.270484924316406,
+      "learning_rate": 9.199300055258795e-07,
+      "loss": 0.0649,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9721468620003617,
+      "grad_norm": 0.19271469116210938,
+      "learning_rate": 9.194695155645607e-07,
+      "loss": 0.0349,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9766684753119913,
+      "grad_norm": 3.624760389328003,
+      "learning_rate": 9.190090256032418e-07,
+      "loss": 0.0332,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9811900886236209,
+      "grad_norm": 12.83198070526123,
+      "learning_rate": 9.185485356419229e-07,
+      "loss": 0.0392,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9857117019352505,
+      "grad_norm": 2.507450580596924,
+      "learning_rate": 9.180880456806042e-07,
+      "loss": 0.0561,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9902333152468801,
+      "grad_norm": 3.1027333736419678,
+      "learning_rate": 9.176275557192853e-07,
+      "loss": 0.0411,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9947549285585097,
+      "grad_norm": 53.23537826538086,
+      "learning_rate": 9.171670657579664e-07,
+      "loss": 0.0892,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9992765418701393,
+      "grad_norm": 53.92917251586914,
+      "learning_rate": 9.167065757966476e-07,
+      "loss": 0.0849,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.19376881420612335,
+      "eval_runtime": 8685.9066,
+      "eval_samples_per_second": 1.093,
+      "eval_steps_per_second": 0.137,
+      "eval_wer": 0.10807174887892376,
+      "step": 5529
+    },
+    {
+      "epoch": 1.003798155181769,
+      "grad_norm": 1.1583398580551147,
+      "learning_rate": 9.162460858353286e-07,
+      "loss": 0.0418,
+      "step": 5550
+    },
+    {
+      "epoch": 1.0083197684933984,
+      "grad_norm": 3.819171190261841,
+      "learning_rate": 9.157855958740099e-07,
+      "loss": 0.0223,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0128413818050281,
+      "grad_norm": 12.470620155334473,
+      "learning_rate": 9.153251059126911e-07,
+      "loss": 0.0322,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0173629951166576,
+      "grad_norm": 3.7826218605041504,
+      "learning_rate": 9.148646159513723e-07,
+      "loss": 0.0159,
+      "step": 5625
+    },
+    {
+      "epoch": 1.0218846084282873,
+      "grad_norm": 3.368657112121582,
+      "learning_rate": 9.144041259900533e-07,
+      "loss": 0.0354,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0264062217399168,
+      "grad_norm": 5.757375717163086,
+      "learning_rate": 9.139436360287345e-07,
+      "loss": 0.0212,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0309278350515463,
+      "grad_norm": 0.24591568112373352,
+      "learning_rate": 9.134831460674158e-07,
+      "loss": 0.0354,
+      "step": 5700
+    },
+    {
+      "epoch": 1.035449448363176,
+      "grad_norm": 1.0728533267974854,
+      "learning_rate": 9.130226561060968e-07,
+      "loss": 0.0162,
+      "step": 5725
+    },
+    {
+      "epoch": 1.0399710616748055,
+      "grad_norm": 4.130620956420898,
+      "learning_rate": 9.12562166144778e-07,
+      "loss": 0.0286,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0444926749864352,
+      "grad_norm": 19.77248764038086,
+      "learning_rate": 9.121016761834591e-07,
+      "loss": 0.0346,
+      "step": 5775
+    },
+    {
+      "epoch": 1.0490142882980646,
+      "grad_norm": 1.552537202835083,
+      "learning_rate": 9.116411862221404e-07,
+      "loss": 0.0164,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0535359016096943,
+      "grad_norm": 0.2987182140350342,
+      "learning_rate": 9.111806962608215e-07,
+      "loss": 0.0423,
+      "step": 5825
+    },
+    {
+      "epoch": 1.0580575149213238,
+      "grad_norm": 2.336726188659668,
+      "learning_rate": 9.107202062995026e-07,
+      "loss": 0.0431,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0625791282329535,
+      "grad_norm": 1.1774622201919556,
+      "learning_rate": 9.102597163381838e-07,
+      "loss": 0.0238,
+      "step": 5875
+    },
+    {
+      "epoch": 1.067100741544583,
+      "grad_norm": 1.2562298774719238,
+      "learning_rate": 9.097992263768649e-07,
+      "loss": 0.0502,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0716223548562127,
+      "grad_norm": 15.588605880737305,
+      "learning_rate": 9.093387364155461e-07,
+      "loss": 0.0622,
+      "step": 5925
+    },
+    {
+      "epoch": 1.0761439681678422,
+      "grad_norm": 2.4600257873535156,
+      "learning_rate": 9.088782464542273e-07,
+      "loss": 0.0476,
+      "step": 5950
+    },
+    {
+      "epoch": 1.080665581479472,
+      "grad_norm": 7.753191947937012,
+      "learning_rate": 9.084177564929084e-07,
+      "loss": 0.035,
+      "step": 5975
+    },
+    {
+      "epoch": 1.0851871947911014,
+      "grad_norm": 6.821231365203857,
+      "learning_rate": 9.079572665315895e-07,
+      "loss": 0.0212,
+      "step": 6000
+    },
+    {
+      "epoch": 1.089708808102731,
+      "grad_norm": 2.2424540519714355,
+      "learning_rate": 9.074967765702707e-07,
+      "loss": 0.0163,
+      "step": 6025
+    },
+    {
+      "epoch": 1.0942304214143606,
+      "grad_norm": 23.894805908203125,
+      "learning_rate": 9.07036286608952e-07,
+      "loss": 0.0329,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0987520347259903,
+      "grad_norm": 0.27833691239356995,
+      "learning_rate": 9.06575796647633e-07,
+      "loss": 0.0378,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1032736480376197,
+      "grad_norm": 1.7873796224594116,
+      "learning_rate": 9.061153066863142e-07,
+      "loss": 0.035,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1077952613492494,
+      "grad_norm": 8.95418930053711,
+      "learning_rate": 9.056548167249954e-07,
+      "loss": 0.0346,
+      "step": 6125
+    },
+    {
+      "epoch": 1.112316874660879,
+      "grad_norm": 0.6353895664215088,
+      "learning_rate": 9.051943267636764e-07,
+      "loss": 0.0225,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1168384879725086,
+      "grad_norm": 4.148350715637207,
+      "learning_rate": 9.047338368023577e-07,
+      "loss": 0.0375,
+      "step": 6175
+    },
+    {
+      "epoch": 1.121360101284138,
+      "grad_norm": 3.327193021774292,
+      "learning_rate": 9.042733468410389e-07,
+      "loss": 0.0269,
+      "step": 6200
+    },
+    {
+      "epoch": 1.1258817145957678,
+      "grad_norm": 4.31793212890625,
+      "learning_rate": 9.038128568797199e-07,
+      "loss": 0.0242,
+      "step": 6225
+    },
+    {
+      "epoch": 1.1304033279073973,
+      "grad_norm": 20.431732177734375,
+      "learning_rate": 9.033523669184011e-07,
+      "loss": 0.0222,
+      "step": 6250
+    },
+    {
+      "epoch": 1.134924941219027,
+      "grad_norm": 6.784418106079102,
+      "learning_rate": 9.028918769570823e-07,
+      "loss": 0.0488,
+      "step": 6275
+    },
+    {
+      "epoch": 1.1394465545306565,
+      "grad_norm": 21.859533309936523,
+      "learning_rate": 9.024313869957635e-07,
+      "loss": 0.0518,
+      "step": 6300
+    },
+    {
+      "epoch": 1.1439681678422862,
+      "grad_norm": 2.1316771507263184,
+      "learning_rate": 9.019708970344446e-07,
+      "loss": 0.0403,
+      "step": 6325
+    },
+    {
+      "epoch": 1.1484897811539156,
+      "grad_norm": 2.723567247390747,
+      "learning_rate": 9.015104070731258e-07,
+      "loss": 0.0487,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1530113944655453,
+      "grad_norm": 4.712471961975098,
+      "learning_rate": 9.010499171118069e-07,
+      "loss": 0.0139,
+      "step": 6375
+    },
+    {
+      "epoch": 1.1575330077771748,
+      "grad_norm": 2.7237842082977295,
+      "learning_rate": 9.00589427150488e-07,
+      "loss": 0.053,
+      "step": 6400
+    },
+    {
+      "epoch": 1.1620546210888045,
+      "grad_norm": 3.783311605453491,
+      "learning_rate": 9.001289371891693e-07,
+      "loss": 0.0447,
+      "step": 6425
+    },
+    {
+      "epoch": 1.166576234400434,
+      "grad_norm": 4.771798610687256,
+      "learning_rate": 8.996684472278504e-07,
+      "loss": 0.0306,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1710978477120637,
+      "grad_norm": 0.3322123885154724,
+      "learning_rate": 8.992079572665316e-07,
+      "loss": 0.024,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1756194610236932,
+      "grad_norm": 17.802349090576172,
+      "learning_rate": 8.987658869036655e-07,
+      "loss": 0.0317,
+      "step": 6500
+    },
+    {
+      "epoch": 1.180141074335323,
+      "grad_norm": 0.25371697545051575,
+      "learning_rate": 8.983053969423466e-07,
+      "loss": 0.0405,
+      "step": 6525
+    },
+    {
+      "epoch": 1.1846626876469524,
+      "grad_norm": 2.539607524871826,
+      "learning_rate": 8.978449069810278e-07,
+      "loss": 0.0199,
+      "step": 6550
+    },
+    {
+      "epoch": 1.189184300958582,
+      "grad_norm": 2.3740954399108887,
+      "learning_rate": 8.973844170197089e-07,
+      "loss": 0.0252,
+      "step": 6575
+    },
+    {
+      "epoch": 1.1937059142702116,
+      "grad_norm": 2.0089480876922607,
+      "learning_rate": 8.9692392705839e-07,
+      "loss": 0.0091,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1982275275818413,
+      "grad_norm": 0.539943516254425,
+      "learning_rate": 8.964634370970713e-07,
+      "loss": 0.0479,
+      "step": 6625
+    },
+    {
+      "epoch": 1.2027491408934707,
+      "grad_norm": 15.776597023010254,
+      "learning_rate": 8.960029471357525e-07,
+      "loss": 0.0337,
+      "step": 6650
+    },
+    {
+      "epoch": 1.2072707542051004,
+      "grad_norm": 43.65888977050781,
+      "learning_rate": 8.955424571744335e-07,
+      "loss": 0.04,
+      "step": 6675
+    },
+    {
+      "epoch": 1.21179236751673,
+      "grad_norm": 0.6316529512405396,
+      "learning_rate": 8.950819672131147e-07,
+      "loss": 0.0464,
+      "step": 6700
+    },
+    {
+      "epoch": 1.2163139808283596,
+      "grad_norm": 0.06351311504840851,
+      "learning_rate": 8.946214772517959e-07,
+      "loss": 0.0596,
+      "step": 6725
+    },
+    {
+      "epoch": 1.220835594139989,
+      "grad_norm": 9.572014808654785,
+      "learning_rate": 8.94160987290477e-07,
+      "loss": 0.0398,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2253572074516188,
+      "grad_norm": 9.598252296447754,
+      "learning_rate": 8.937004973291582e-07,
+      "loss": 0.0344,
+      "step": 6775
+    },
+    {
+      "epoch": 1.2298788207632483,
+      "grad_norm": 1.5910439491271973,
+      "learning_rate": 8.932400073678394e-07,
+      "loss": 0.0265,
+      "step": 6800
+    },
+    {
+      "epoch": 1.234400434074878,
+      "grad_norm": 10.684225082397461,
+      "learning_rate": 8.927795174065205e-07,
+      "loss": 0.0264,
+      "step": 6825
+    },
+    {
+      "epoch": 1.2389220473865075,
+      "grad_norm": 11.20118236541748,
+      "learning_rate": 8.923190274452016e-07,
+      "loss": 0.0345,
+      "step": 6850
+    },
+    {
+      "epoch": 1.2434436606981372,
+      "grad_norm": 12.899706840515137,
+      "learning_rate": 8.918585374838829e-07,
+      "loss": 0.0316,
+      "step": 6875
+    },
+    {
+      "epoch": 1.2479652740097666,
+      "grad_norm": 18.632545471191406,
+      "learning_rate": 8.91398047522564e-07,
+      "loss": 0.033,
+      "step": 6900
+    },
+    {
+      "epoch": 1.2524868873213963,
+      "grad_norm": 0.7847491502761841,
+      "learning_rate": 8.909375575612451e-07,
+      "loss": 0.0292,
+      "step": 6925
+    },
+    {
+      "epoch": 1.2570085006330258,
+      "grad_norm": 1.0317540168762207,
+      "learning_rate": 8.904770675999262e-07,
+      "loss": 0.0286,
+      "step": 6950
+    },
+    {
+      "epoch": 1.2615301139446555,
+      "grad_norm": 0.7965870499610901,
+      "learning_rate": 8.900165776386075e-07,
+      "loss": 0.0222,
+      "step": 6975
+    },
+    {
+      "epoch": 1.266051727256285,
+      "grad_norm": 12.810919761657715,
+      "learning_rate": 8.895560876772886e-07,
+      "loss": 0.0369,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2705733405679147,
+      "grad_norm": 12.809525489807129,
+      "learning_rate": 8.890955977159697e-07,
+      "loss": 0.0381,
+      "step": 7025
+    },
+    {
+      "epoch": 1.2750949538795442,
+      "grad_norm": 4.802799224853516,
+      "learning_rate": 8.886351077546509e-07,
+      "loss": 0.0274,
+      "step": 7050
+    },
+    {
+      "epoch": 1.279616567191174,
+      "grad_norm": 37.88527297973633,
+      "learning_rate": 8.881746177933321e-07,
+      "loss": 0.0266,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2841381805028034,
+      "grad_norm": 23.25122833251953,
+      "learning_rate": 8.877141278320132e-07,
+      "loss": 0.0487,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2886597938144329,
+      "grad_norm": 25.70662498474121,
+      "learning_rate": 8.872536378706944e-07,
+      "loss": 0.0652,
+      "step": 7125
+    },
+    {
+      "epoch": 1.2931814071260626,
+      "grad_norm": 3.8632333278656006,
+      "learning_rate": 8.867931479093756e-07,
+      "loss": 0.0442,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2977030204376923,
+      "grad_norm": 1.3248151540756226,
+      "learning_rate": 8.863326579480566e-07,
+      "loss": 0.0229,
+      "step": 7175
+    },
+    {
+      "epoch": 1.3022246337493217,
+      "grad_norm": 1.7153706550598145,
+      "learning_rate": 8.858721679867378e-07,
+      "loss": 0.0341,
+      "step": 7200
+    },
+    {
+      "epoch": 1.3067462470609512,
+      "grad_norm": 8.373719215393066,
+      "learning_rate": 8.854116780254191e-07,
+      "loss": 0.0422,
+      "step": 7225
+    },
+    {
+      "epoch": 1.311267860372581,
+      "grad_norm": 2.403066635131836,
+      "learning_rate": 8.849511880641001e-07,
+      "loss": 0.0287,
+      "step": 7250
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 1.0376137495040894,
+      "learning_rate": 8.844906981027813e-07,
+      "loss": 0.0113,
+      "step": 7275
+    },
+    {
+      "epoch": 1.32031108699584,
+      "grad_norm": 5.192943096160889,
+      "learning_rate": 8.840302081414625e-07,
+      "loss": 0.0301,
+      "step": 7300
+    },
+    {
+      "epoch": 1.3248327003074696,
+      "grad_norm": 1.6459161043167114,
+      "learning_rate": 8.835697181801436e-07,
+      "loss": 0.044,
+      "step": 7325
+    },
+    {
+      "epoch": 1.3293543136190993,
+      "grad_norm": 9.943346977233887,
+      "learning_rate": 8.831092282188248e-07,
+      "loss": 0.0132,
+      "step": 7350
+    },
+    {
+      "epoch": 1.333875926930729,
+      "grad_norm": 5.858924865722656,
+      "learning_rate": 8.82648738257506e-07,
+      "loss": 0.0192,
+      "step": 7375
+    },
+    {
+      "epoch": 1.3383975402423585,
+      "grad_norm": 6.490448474884033,
+      "learning_rate": 8.821882482961871e-07,
+      "loss": 0.0239,
+      "step": 7400
+    },
+    {
+      "epoch": 1.342919153553988,
+      "grad_norm": 0.2476533055305481,
+      "learning_rate": 8.817277583348682e-07,
+      "loss": 0.0463,
+      "step": 7425
+    },
+    {
+      "epoch": 1.3474407668656176,
+      "grad_norm": 14.549626350402832,
+      "learning_rate": 8.812672683735494e-07,
+      "loss": 0.0507,
+      "step": 7450
+    },
+    {
+      "epoch": 1.3519623801772473,
+      "grad_norm": 2.189336061477661,
+      "learning_rate": 8.808067784122306e-07,
+      "loss": 0.0331,
+      "step": 7475
+    },
+    {
+      "epoch": 1.3564839934888768,
+      "grad_norm": 0.9585964679718018,
+      "learning_rate": 8.803462884509118e-07,
+      "loss": 0.0572,
+      "step": 7500
+    },
+    {
+      "epoch": 1.3610056068005063,
+      "grad_norm": 0.9413008093833923,
+      "learning_rate": 8.798857984895929e-07,
+      "loss": 0.0847,
+      "step": 7525
+    },
+    {
+      "epoch": 1.365527220112136,
+      "grad_norm": 1.1694642305374146,
+      "learning_rate": 8.79425308528274e-07,
+      "loss": 0.0355,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3700488334237657,
+      "grad_norm": 10.240534782409668,
+      "learning_rate": 8.789648185669552e-07,
+      "loss": 0.0264,
+      "step": 7575
+    },
+    {
+      "epoch": 1.3745704467353952,
+      "grad_norm": 0.40524232387542725,
+      "learning_rate": 8.785043286056364e-07,
+      "loss": 0.0277,
+      "step": 7600
+    },
+    {
+      "epoch": 1.3790920600470247,
+      "grad_norm": 0.6719425916671753,
+      "learning_rate": 8.780438386443175e-07,
+      "loss": 0.027,
+      "step": 7625
+    },
+    {
+      "epoch": 1.3836136733586544,
+      "grad_norm": 3.0596706867218018,
+      "learning_rate": 8.775833486829987e-07,
+      "loss": 0.0251,
+      "step": 7650
+    },
+    {
+      "epoch": 1.388135286670284,
+      "grad_norm": 3.454193592071533,
+      "learning_rate": 8.771228587216797e-07,
+      "loss": 0.0334,
+      "step": 7675
+    },
+    {
+      "epoch": 1.3926568999819136,
+      "grad_norm": 3.9943997859954834,
+      "learning_rate": 8.76662368760361e-07,
+      "loss": 0.0312,
+      "step": 7700
+    },
+    {
+      "epoch": 1.397178513293543,
+      "grad_norm": 8.968663215637207,
+      "learning_rate": 8.762018787990422e-07,
+      "loss": 0.027,
+      "step": 7725
+    },
+    {
+      "epoch": 1.4017001266051727,
+      "grad_norm": 1.615293025970459,
+      "learning_rate": 8.757413888377234e-07,
+      "loss": 0.0246,
+      "step": 7750
+    },
+    {
+      "epoch": 1.4062217399168024,
+      "grad_norm": 14.491230964660645,
+      "learning_rate": 8.752808988764044e-07,
+      "loss": 0.019,
+      "step": 7775
+    },
+    {
+      "epoch": 1.410743353228432,
+      "grad_norm": 0.586275577545166,
+      "learning_rate": 8.748204089150856e-07,
+      "loss": 0.0315,
+      "step": 7800
+    },
+    {
+      "epoch": 1.4152649665400614,
+      "grad_norm": 16.67391586303711,
+      "learning_rate": 8.743599189537669e-07,
+      "loss": 0.0468,
+      "step": 7825
+    },
+    {
+      "epoch": 1.419786579851691,
+      "grad_norm": 7.967570781707764,
+      "learning_rate": 8.738994289924479e-07,
+      "loss": 0.0443,
+      "step": 7850
+    },
+    {
+      "epoch": 1.4243081931633208,
+      "grad_norm": 3.4274497032165527,
+      "learning_rate": 8.734389390311291e-07,
+      "loss": 0.0464,
+      "step": 7875
+    },
+    {
+      "epoch": 1.4288298064749503,
+      "grad_norm": 14.454466819763184,
+      "learning_rate": 8.729784490698102e-07,
+      "loss": 0.0594,
+      "step": 7900
+    },
+    {
+      "epoch": 1.4333514197865798,
+      "grad_norm": 34.33506774902344,
+      "learning_rate": 8.725179591084913e-07,
+      "loss": 0.1098,
+      "step": 7925
+    },
+    {
+      "epoch": 1.4378730330982095,
+      "grad_norm": 5.2832770347595215,
+      "learning_rate": 8.720574691471726e-07,
+      "loss": 0.035,
+      "step": 7950
+    },
+    {
+      "epoch": 1.4423946464098392,
+      "grad_norm": 2.5055034160614014,
+      "learning_rate": 8.715969791858537e-07,
+      "loss": 0.0261,
+      "step": 7975
+    },
+    {
+      "epoch": 1.4469162597214686,
+      "grad_norm": 2.449373483657837,
+      "learning_rate": 8.711364892245349e-07,
+      "loss": 0.024,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4514378730330981,
+      "grad_norm": 3.93390154838562,
+      "learning_rate": 8.70675999263216e-07,
+      "loss": 0.0248,
+      "step": 8025
+    },
+    {
+      "epoch": 1.4559594863447278,
+      "grad_norm": 14.35305404663086,
+      "learning_rate": 8.702155093018971e-07,
+      "loss": 0.0237,
+      "step": 8050
+    },
+    {
+      "epoch": 1.4604810996563573,
+      "grad_norm": 7.074910640716553,
+      "learning_rate": 8.697550193405784e-07,
+      "loss": 0.0284,
+      "step": 8075
+    },
+    {
+      "epoch": 1.465002712967987,
+      "grad_norm": 21.563901901245117,
+      "learning_rate": 8.692945293792595e-07,
+      "loss": 0.0421,
+      "step": 8100
+    },
+    {
+      "epoch": 1.4695243262796165,
+      "grad_norm": 8.473495483398438,
+      "learning_rate": 8.688340394179406e-07,
+      "loss": 0.0194,
+      "step": 8125
+    },
+    {
+      "epoch": 1.4740459395912462,
+      "grad_norm": 0.6098468899726868,
+      "learning_rate": 8.683735494566218e-07,
+      "loss": 0.0392,
+      "step": 8150
+    },
+    {
+      "epoch": 1.4785675529028757,
+      "grad_norm": 29.32798957824707,
+      "learning_rate": 8.679130594953029e-07,
+      "loss": 0.0503,
+      "step": 8175
+    },
+    {
+      "epoch": 1.4830891662145054,
+      "grad_norm": 7.364591121673584,
+      "learning_rate": 8.674525695339841e-07,
+      "loss": 0.0227,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4876107795261349,
+      "grad_norm": 8.35572338104248,
+      "learning_rate": 8.669920795726653e-07,
+      "loss": 0.0346,
+      "step": 8225
+    },
+    {
+      "epoch": 1.4921323928377646,
+      "grad_norm": 9.818826675415039,
+      "learning_rate": 8.665315896113465e-07,
+      "loss": 0.0484,
+      "step": 8250
+    },
+    {
+      "epoch": 1.496654006149394,
+      "grad_norm": 8.03248405456543,
+      "learning_rate": 8.660710996500275e-07,
+      "loss": 0.0605,
+      "step": 8275
+    },
+    {
+      "epoch": 1.5011756194610237,
+      "grad_norm": 7.9050397872924805,
+      "learning_rate": 8.656106096887087e-07,
+      "loss": 0.0609,
+      "step": 8300
+    },
+    {
+      "epoch": 1.5056972327726532,
+      "grad_norm": 0.5489145517349243,
+      "learning_rate": 8.6515011972739e-07,
+      "loss": 0.041,
+      "step": 8325
+    },
+    {
+      "epoch": 1.510218846084283,
+      "grad_norm": 1.3398199081420898,
+      "learning_rate": 8.64689629766071e-07,
+      "loss": 0.0458,
+      "step": 8350
+    },
+    {
+      "epoch": 1.5147404593959126,
+      "grad_norm": 23.95371437072754,
+      "learning_rate": 8.642291398047522e-07,
+      "loss": 0.0322,
+      "step": 8375
+    },
+    {
+      "epoch": 1.519262072707542,
+      "grad_norm": 6.187991142272949,
+      "learning_rate": 8.637686498434334e-07,
+      "loss": 0.0186,
+      "step": 8400
+    },
+    {
+      "epoch": 1.5237836860191716,
+      "grad_norm": 2.8521807193756104,
+      "learning_rate": 8.633081598821146e-07,
+      "loss": 0.0353,
+      "step": 8425
+    },
+    {
+      "epoch": 1.5283052993308013,
+      "grad_norm": 17.128206253051758,
+      "learning_rate": 8.628476699207957e-07,
+      "loss": 0.0217,
+      "step": 8450
+    },
+    {
+      "epoch": 1.532826912642431,
+      "grad_norm": 1.29264235496521,
+      "learning_rate": 8.623871799594769e-07,
+      "loss": 0.02,
+      "step": 8475
+    },
+    {
+      "epoch": 1.5373485259540605,
+      "grad_norm": 1.0612270832061768,
+      "learning_rate": 8.61926689998158e-07,
+      "loss": 0.0273,
+      "step": 8500
+    },
+    {
+      "epoch": 1.54187013926569,
+      "grad_norm": 21.39075469970703,
+      "learning_rate": 8.614662000368391e-07,
+      "loss": 0.0326,
+      "step": 8525
+    },
+    {
+      "epoch": 1.5463917525773194,
+      "grad_norm": 22.907485961914062,
+      "learning_rate": 8.610057100755204e-07,
+      "loss": 0.0493,
+      "step": 8550
+    },
+    {
+      "epoch": 1.5509133658889491,
+      "grad_norm": 0.9095446467399597,
+      "learning_rate": 8.605452201142015e-07,
+      "loss": 0.0203,
+      "step": 8575
+    },
+    {
+      "epoch": 1.5554349792005788,
+      "grad_norm": 19.75260353088379,
+      "learning_rate": 8.600847301528826e-07,
+      "loss": 0.021,
+      "step": 8600
+    },
+    {
+      "epoch": 1.5599565925122083,
+      "grad_norm": 14.637717247009277,
+      "learning_rate": 8.596242401915637e-07,
+      "loss": 0.0392,
+      "step": 8625
+    },
+    {
+      "epoch": 1.5644782058238378,
+      "grad_norm": 16.156036376953125,
+      "learning_rate": 8.591821698286977e-07,
+      "loss": 0.0496,
+      "step": 8650
+    },
+    {
+      "epoch": 1.5689998191354675,
+      "grad_norm": 0.6795814037322998,
+      "learning_rate": 8.587216798673789e-07,
+      "loss": 0.031,
+      "step": 8675
+    },
+    {
+      "epoch": 1.5735214324470972,
+      "grad_norm": 9.835176467895508,
+      "learning_rate": 8.5826118990606e-07,
+      "loss": 0.0413,
+      "step": 8700
+    },
+    {
+      "epoch": 1.5780430457587267,
+      "grad_norm": 98.9009017944336,
+      "learning_rate": 8.578006999447411e-07,
+      "loss": 0.0764,
+      "step": 8725
+    },
+    {
+      "epoch": 1.5825646590703562,
+      "grad_norm": 11.466421127319336,
+      "learning_rate": 8.573402099834223e-07,
+      "loss": 0.0362,
+      "step": 8750
+    },
+    {
+      "epoch": 1.5870862723819859,
+      "grad_norm": 1.9184726476669312,
+      "learning_rate": 8.568797200221036e-07,
+      "loss": 0.0199,
+      "step": 8775
+    },
+    {
+      "epoch": 1.5916078856936156,
+      "grad_norm": 1.2715715169906616,
+      "learning_rate": 8.564192300607846e-07,
+      "loss": 0.0273,
+      "step": 8800
+    },
+    {
+      "epoch": 1.596129499005245,
+      "grad_norm": 0.3399398624897003,
+      "learning_rate": 8.559587400994658e-07,
+      "loss": 0.0271,
+      "step": 8825
+    },
+    {
+      "epoch": 1.6006511123168745,
+      "grad_norm": 3.984431028366089,
+      "learning_rate": 8.55498250138147e-07,
+      "loss": 0.0395,
+      "step": 8850
+    },
+    {
+      "epoch": 1.6051727256285042,
+      "grad_norm": 0.13510115444660187,
+      "learning_rate": 8.550377601768281e-07,
+      "loss": 0.0264,
+      "step": 8875
+    },
+    {
+      "epoch": 1.609694338940134,
+      "grad_norm": 1.5098395347595215,
+      "learning_rate": 8.545772702155093e-07,
+      "loss": 0.0365,
+      "step": 8900
+    },
+    {
+      "epoch": 1.6142159522517634,
+      "grad_norm": 1.5568790435791016,
+      "learning_rate": 8.541167802541905e-07,
+      "loss": 0.0348,
+      "step": 8925
+    },
+    {
+      "epoch": 1.6187375655633929,
+      "grad_norm": 15.226325988769531,
+      "learning_rate": 8.536562902928715e-07,
+      "loss": 0.032,
+      "step": 8950
+    },
+    {
+      "epoch": 1.6232591788750226,
+      "grad_norm": 4.606542587280273,
+      "learning_rate": 8.531958003315527e-07,
+      "loss": 0.0397,
+      "step": 8975
+    },
+    {
+      "epoch": 1.6277807921866523,
+      "grad_norm": 0.1431085765361786,
+      "learning_rate": 8.52735310370234e-07,
+      "loss": 0.04,
+      "step": 9000
+    },
+    {
+      "epoch": 1.6323024054982818,
+      "grad_norm": 8.834503173828125,
+      "learning_rate": 8.522748204089151e-07,
+      "loss": 0.0199,
+      "step": 9025
+    },
+    {
+      "epoch": 1.6368240188099112,
+      "grad_norm": 0.43929988145828247,
+      "learning_rate": 8.518143304475962e-07,
+      "loss": 0.0521,
+      "step": 9050
+    },
+    {
+      "epoch": 1.641345632121541,
+      "grad_norm": 0.239268958568573,
+      "learning_rate": 8.513538404862773e-07,
+      "loss": 0.028,
+      "step": 9075
+    },
+    {
+      "epoch": 1.6458672454331706,
+      "grad_norm": 7.775051593780518,
+      "learning_rate": 8.508933505249585e-07,
+      "loss": 0.0435,
+      "step": 9100
+    },
+    {
+      "epoch": 1.6503888587448001,
+      "grad_norm": 1.0783268213272095,
+      "learning_rate": 8.504328605636397e-07,
+      "loss": 0.09,
+      "step": 9125
+    },
+    {
+      "epoch": 1.6549104720564296,
+      "grad_norm": 6.303003787994385,
+      "learning_rate": 8.499723706023208e-07,
+      "loss": 0.0371,
+      "step": 9150
+    },
+    {
+      "epoch": 1.6594320853680593,
+      "grad_norm": 0.0859726294875145,
+      "learning_rate": 8.49511880641002e-07,
+      "loss": 0.0309,
+      "step": 9175
+    },
+    {
+      "epoch": 1.663953698679689,
+      "grad_norm": 7.407548427581787,
+      "learning_rate": 8.490513906796831e-07,
+      "loss": 0.0267,
+      "step": 9200
+    },
+    {
+      "epoch": 1.6684753119913185,
+      "grad_norm": 0.10481081902980804,
+      "learning_rate": 8.485909007183642e-07,
+      "loss": 0.023,
+      "step": 9225
+    },
+    {
+      "epoch": 1.672996925302948,
+      "grad_norm": 4.49674129486084,
+      "learning_rate": 8.481304107570455e-07,
+      "loss": 0.0226,
+      "step": 9250
+    },
+    {
+      "epoch": 1.6775185386145777,
+      "grad_norm": 3.7796387672424316,
+      "learning_rate": 8.476699207957267e-07,
+      "loss": 0.0452,
+      "step": 9275
+    },
+    {
+      "epoch": 1.6820401519262074,
+      "grad_norm": 4.558553218841553,
+      "learning_rate": 8.472094308344077e-07,
+      "loss": 0.0302,
+      "step": 9300
+    },
+    {
+      "epoch": 1.6865617652378369,
+      "grad_norm": 1.0844203233718872,
+      "learning_rate": 8.467489408730889e-07,
+      "loss": 0.0213,
+      "step": 9325
+    },
+    {
+      "epoch": 1.6910833785494663,
+      "grad_norm": 0.6152500510215759,
+      "learning_rate": 8.462884509117701e-07,
+      "loss": 0.028,
+      "step": 9350
+    },
+    {
+      "epoch": 1.695604991861096,
+      "grad_norm": 0.83628249168396,
+      "learning_rate": 8.458279609504512e-07,
+      "loss": 0.0249,
+      "step": 9375
+    },
+    {
+      "epoch": 1.7001266051727257,
+      "grad_norm": 4.9916157722473145,
+      "learning_rate": 8.453674709891324e-07,
+      "loss": 0.0326,
+      "step": 9400
+    },
+    {
+      "epoch": 1.7046482184843552,
+      "grad_norm": 0.2095576971769333,
+      "learning_rate": 8.449069810278136e-07,
+      "loss": 0.0494,
+      "step": 9425
+    },
+    {
+      "epoch": 1.7091698317959847,
+      "grad_norm": 16.670848846435547,
+      "learning_rate": 8.444464910664947e-07,
+      "loss": 0.0248,
+      "step": 9450
+    },
+    {
+      "epoch": 1.7136914451076144,
+      "grad_norm": 1.0165280103683472,
+      "learning_rate": 8.439860011051758e-07,
+      "loss": 0.0371,
+      "step": 9475
+    },
+    {
+      "epoch": 1.718213058419244,
+      "grad_norm": 19.8568115234375,
+      "learning_rate": 8.435255111438571e-07,
+      "loss": 0.03,
+      "step": 9500
+    },
+    {
+      "epoch": 1.7227346717308736,
+      "grad_norm": 29.754867553710938,
+      "learning_rate": 8.430650211825382e-07,
+      "loss": 0.0836,
+      "step": 9525
+    },
+    {
+      "epoch": 1.727256285042503,
+      "grad_norm": 10.585617065429688,
+      "learning_rate": 8.426045312212193e-07,
+      "loss": 0.0253,
+      "step": 9550
+    },
+    {
+      "epoch": 1.7317778983541328,
+      "grad_norm": 0.41567254066467285,
+      "learning_rate": 8.421440412599005e-07,
+      "loss": 0.0419,
+      "step": 9575
+    },
+    {
+      "epoch": 1.7362995116657625,
+      "grad_norm": 14.9446439743042,
+      "learning_rate": 8.416835512985817e-07,
+      "loss": 0.0346,
+      "step": 9600
+    },
+    {
+      "epoch": 1.740821124977392,
+      "grad_norm": 11.810590744018555,
+      "learning_rate": 8.412230613372628e-07,
+      "loss": 0.0399,
+      "step": 9625
+    },
+    {
+      "epoch": 1.7453427382890214,
+      "grad_norm": 4.900815486907959,
+      "learning_rate": 8.40762571375944e-07,
+      "loss": 0.0328,
+      "step": 9650
+    },
+    {
+      "epoch": 1.7498643516006511,
+      "grad_norm": 43.67582702636719,
+      "learning_rate": 8.403020814146251e-07,
+      "loss": 0.0329,
+      "step": 9675
+    },
+    {
+      "epoch": 1.7543859649122808,
+      "grad_norm": 9.560347557067871,
+      "learning_rate": 8.398415914533063e-07,
+      "loss": 0.0455,
+      "step": 9700
+    },
+    {
+      "epoch": 1.7589075782239103,
+      "grad_norm": 20.836200714111328,
+      "learning_rate": 8.393811014919875e-07,
+      "loss": 0.0347,
+      "step": 9725
+    },
+    {
+      "epoch": 1.7634291915355398,
+      "grad_norm": 1.0109299421310425,
+      "learning_rate": 8.389206115306686e-07,
+      "loss": 0.0171,
+      "step": 9750
+    },
+    {
+      "epoch": 1.7679508048471695,
+      "grad_norm": 15.896967887878418,
+      "learning_rate": 8.384601215693498e-07,
+      "loss": 0.0558,
+      "step": 9775
+    },
+    {
+      "epoch": 1.7724724181587992,
+      "grad_norm": 8.316397666931152,
+      "learning_rate": 8.379996316080308e-07,
+      "loss": 0.042,
+      "step": 9800
+    },
+    {
+      "epoch": 1.7769940314704287,
+      "grad_norm": 2.5963985919952393,
+      "learning_rate": 8.37539141646712e-07,
+      "loss": 0.051,
+      "step": 9825
+    },
+    {
+      "epoch": 1.7815156447820581,
+      "grad_norm": 0.8551808595657349,
+      "learning_rate": 8.370786516853933e-07,
+      "loss": 0.0379,
+      "step": 9850
+    },
+    {
+      "epoch": 1.7860372580936879,
+      "grad_norm": 7.102266311645508,
+      "learning_rate": 8.366181617240745e-07,
+      "loss": 0.0252,
+      "step": 9875
+    },
+    {
+      "epoch": 1.7905588714053176,
+      "grad_norm": 1.0050630569458008,
+      "learning_rate": 8.361576717627555e-07,
+      "loss": 0.0523,
+      "step": 9900
+    },
+    {
+      "epoch": 1.795080484716947,
+      "grad_norm": 0.23266096413135529,
+      "learning_rate": 8.356971818014367e-07,
+      "loss": 0.0771,
+      "step": 9925
+    },
+    {
+      "epoch": 1.7996020980285765,
+      "grad_norm": 2.397165298461914,
+      "learning_rate": 8.352366918401178e-07,
+      "loss": 0.0383,
+      "step": 9950
+    },
+    {
+      "epoch": 1.8041237113402062,
+      "grad_norm": 3.823277711868286,
+      "learning_rate": 8.34776201878799e-07,
+      "loss": 0.0244,
+      "step": 9975
+    },
+    {
+      "epoch": 1.808645324651836,
+      "grad_norm": 0.4726651906967163,
+      "learning_rate": 8.343157119174802e-07,
+      "loss": 0.0263,
+      "step": 10000
+    },
+    {
+      "epoch": 1.8131669379634654,
+      "grad_norm": 1.6218931674957275,
+      "learning_rate": 8.338552219561613e-07,
+      "loss": 0.0141,
+      "step": 10025
+    },
+    {
+      "epoch": 1.8176885512750949,
+      "grad_norm": 10.145578384399414,
+      "learning_rate": 8.333947319948424e-07,
+      "loss": 0.026,
+      "step": 10050
+    },
+    {
+      "epoch": 1.8222101645867246,
+      "grad_norm": 10.49772834777832,
+      "learning_rate": 8.329342420335236e-07,
+      "loss": 0.0373,
+      "step": 10075
+    },
+    {
+      "epoch": 1.8267317778983543,
+      "grad_norm": 25.12136459350586,
+      "learning_rate": 8.324737520722048e-07,
+      "loss": 0.0335,
+      "step": 10100
+    },
+    {
+      "epoch": 1.8312533912099838,
+      "grad_norm": 0.4937836229801178,
+      "learning_rate": 8.32013262110886e-07,
+      "loss": 0.038,
+      "step": 10125
+    },
+    {
+      "epoch": 1.8357750045216132,
+      "grad_norm": 0.8513910174369812,
+      "learning_rate": 8.315527721495671e-07,
+      "loss": 0.0258,
+      "step": 10150
+    },
+    {
+      "epoch": 1.840296617833243,
+      "grad_norm": 16.558271408081055,
+      "learning_rate": 8.310922821882482e-07,
+      "loss": 0.0564,
+      "step": 10175
+    },
+    {
+      "epoch": 1.8448182311448726,
+      "grad_norm": 4.88476037979126,
+      "learning_rate": 8.306317922269294e-07,
+      "loss": 0.0285,
+      "step": 10200
+    },
+    {
+      "epoch": 1.8493398444565021,
+      "grad_norm": 4.210046291351318,
+      "learning_rate": 8.301713022656106e-07,
+      "loss": 0.0293,
+      "step": 10225
+    },
+    {
+      "epoch": 1.8538614577681316,
+      "grad_norm": 11.491304397583008,
+      "learning_rate": 8.297108123042917e-07,
+      "loss": 0.0585,
+      "step": 10250
+    },
+    {
+      "epoch": 1.8583830710797613,
+      "grad_norm": 11.61664867401123,
+      "learning_rate": 8.292503223429729e-07,
+      "loss": 0.0506,
+      "step": 10275
+    },
+    {
+      "epoch": 1.862904684391391,
+      "grad_norm": 2.719242572784424,
+      "learning_rate": 8.28789832381654e-07,
+      "loss": 0.0484,
+      "step": 10300
+    },
+    {
+      "epoch": 1.8674262977030205,
+      "grad_norm": 1.5411864519119263,
+      "learning_rate": 8.283293424203352e-07,
+      "loss": 0.0472,
+      "step": 10325
+    },
+    {
+      "epoch": 1.87194791101465,
+      "grad_norm": 9.415594100952148,
+      "learning_rate": 8.278688524590164e-07,
+      "loss": 0.0792,
+      "step": 10350
+    },
+    {
+      "epoch": 1.8764695243262797,
+      "grad_norm": 7.970459938049316,
+      "learning_rate": 8.274083624976976e-07,
+      "loss": 0.0345,
+      "step": 10375
+    },
+    {
+      "epoch": 1.8809911376379094,
+      "grad_norm": 1.801161289215088,
+      "learning_rate": 8.269478725363786e-07,
+      "loss": 0.0373,
+      "step": 10400
+    },
+    {
+      "epoch": 1.8855127509495389,
+      "grad_norm": 0.596969485282898,
+      "learning_rate": 8.264873825750598e-07,
+      "loss": 0.0399,
+      "step": 10425
+    },
+    {
+      "epoch": 1.8900343642611683,
+      "grad_norm": 2.700634717941284,
+      "learning_rate": 8.260268926137411e-07,
+      "loss": 0.0269,
+      "step": 10450
+    },
+    {
+      "epoch": 1.8945559775727978,
+      "grad_norm": 5.090729713439941,
+      "learning_rate": 8.255664026524221e-07,
+      "loss": 0.026,
+      "step": 10475
+    },
+    {
+      "epoch": 1.8990775908844275,
+      "grad_norm": 2.83105206489563,
+      "learning_rate": 8.251059126911033e-07,
+      "loss": 0.0266,
+      "step": 10500
+    },
+    {
+      "epoch": 1.9035992041960572,
+      "grad_norm": 0.7878080010414124,
+      "learning_rate": 8.246454227297845e-07,
+      "loss": 0.0449,
+      "step": 10525
+    },
+    {
+      "epoch": 1.9081208175076867,
+      "grad_norm": 14.082945823669434,
+      "learning_rate": 8.241849327684655e-07,
+      "loss": 0.027,
+      "step": 10550
+    },
+    {
+      "epoch": 1.9126424308193162,
+      "grad_norm": 0.5400319695472717,
+      "learning_rate": 8.237244428071468e-07,
+      "loss": 0.0546,
+      "step": 10575
+    },
+    {
+      "epoch": 1.9171640441309459,
+      "grad_norm": 17.12287712097168,
+      "learning_rate": 8.23263952845828e-07,
+      "loss": 0.0318,
+      "step": 10600
+    },
+    {
+      "epoch": 1.9216856574425756,
+      "grad_norm": 4.993497848510742,
+      "learning_rate": 8.228034628845091e-07,
+      "loss": 0.0428,
+      "step": 10625
+    },
+    {
+      "epoch": 1.926207270754205,
+      "grad_norm": 31.576929092407227,
+      "learning_rate": 8.223429729231902e-07,
+      "loss": 0.0434,
+      "step": 10650
+    },
+    {
+      "epoch": 1.9307288840658345,
+      "grad_norm": 1.9357125759124756,
+      "learning_rate": 8.218824829618713e-07,
+      "loss": 0.033,
+      "step": 10675
+    },
+    {
+      "epoch": 1.9352504973774642,
+      "grad_norm": 0.617363691329956,
+      "learning_rate": 8.214219930005526e-07,
+      "loss": 0.0639,
+      "step": 10700
+    },
+    {
+      "epoch": 1.939772110689094,
+      "grad_norm": 0.753123939037323,
+      "learning_rate": 8.209615030392337e-07,
+      "loss": 0.0357,
+      "step": 10725
+    },
+    {
+      "epoch": 1.9442937240007234,
+      "grad_norm": 0.2875419855117798,
+      "learning_rate": 8.205010130779148e-07,
+      "loss": 0.0286,
+      "step": 10750
+    },
+    {
+      "epoch": 1.948815337312353,
+      "grad_norm": 1.6829754114151,
+      "learning_rate": 8.20040523116596e-07,
+      "loss": 0.0285,
+      "step": 10775
+    },
+    {
+      "epoch": 1.9533369506239826,
+      "grad_norm": 1.1577789783477783,
+      "learning_rate": 8.195800331552772e-07,
+      "loss": 0.018,
+      "step": 10800
+    },
+    {
+      "epoch": 1.9578585639356123,
+      "grad_norm": 9.512741088867188,
+      "learning_rate": 8.191195431939583e-07,
+      "loss": 0.031,
+      "step": 10825
+    },
+    {
+      "epoch": 1.9623801772472418,
+      "grad_norm": 2.2050418853759766,
+      "learning_rate": 8.186590532326395e-07,
+      "loss": 0.0314,
+      "step": 10850
+    },
+    {
+      "epoch": 1.9669017905588713,
+      "grad_norm": 15.566596984863281,
+      "learning_rate": 8.181985632713207e-07,
+      "loss": 0.0376,
+      "step": 10875
+    },
+    {
+      "epoch": 1.971423403870501,
+      "grad_norm": 8.69605827331543,
+      "learning_rate": 8.177380733100017e-07,
+      "loss": 0.0321,
+      "step": 10900
+    },
+    {
+      "epoch": 1.9759450171821307,
+      "grad_norm": 8.651535987854004,
+      "learning_rate": 8.172775833486829e-07,
+      "loss": 0.0212,
+      "step": 10925
+    },
+    {
+      "epoch": 1.9804666304937601,
+      "grad_norm": 1.8405441045761108,
+      "learning_rate": 8.168170933873642e-07,
+      "loss": 0.032,
+      "step": 10950
+    },
+    {
+      "epoch": 1.9849882438053896,
+      "grad_norm": 15.064764976501465,
+      "learning_rate": 8.163566034260452e-07,
+      "loss": 0.0133,
+      "step": 10975
+    },
+    {
+      "epoch": 1.9895098571170193,
+      "grad_norm": 0.0899849534034729,
+      "learning_rate": 8.158961134647264e-07,
+      "loss": 0.0469,
+      "step": 11000
+    },
+    {
+      "epoch": 1.994031470428649,
+      "grad_norm": 18.745182037353516,
+      "learning_rate": 8.154540431018604e-07,
+      "loss": 0.0688,
+      "step": 11025
+    },
+    {
+      "epoch": 1.9985530837402785,
+      "grad_norm": 0.08827279508113861,
+      "learning_rate": 8.149935531405416e-07,
+      "loss": 0.0788,
+      "step": 11050
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.22891011834144592,
+      "eval_runtime": 8887.2881,
+      "eval_samples_per_second": 1.068,
+      "eval_steps_per_second": 0.134,
+      "eval_wer": 0.10608584240871237,
+      "step": 11058
+    },
+    {
+      "epoch": 2.003074697051908,
+      "grad_norm": 1.0333837270736694,
+      "learning_rate": 8.145330631792226e-07,
+      "loss": 0.0349,
+      "step": 11075
+    },
+    {
+      "epoch": 2.007596310363538,
+      "grad_norm": 1.1579056978225708,
+      "learning_rate": 8.140725732179038e-07,
+      "loss": 0.0172,
+      "step": 11100
+    },
+    {
+      "epoch": 2.0121179236751674,
+      "grad_norm": 0.37552839517593384,
+      "learning_rate": 8.136120832565849e-07,
+      "loss": 0.0151,
+      "step": 11125
+    },
+    {
+      "epoch": 2.016639536986797,
+      "grad_norm": 23.273653030395508,
+      "learning_rate": 8.131515932952662e-07,
+      "loss": 0.0298,
+      "step": 11150
+    },
+    {
+      "epoch": 2.0211611502984264,
+      "grad_norm": 8.169917106628418,
+      "learning_rate": 8.126911033339473e-07,
+      "loss": 0.0215,
+      "step": 11175
+    },
+    {
+      "epoch": 2.0256827636100563,
+      "grad_norm": 1.1066598892211914,
+      "learning_rate": 8.122306133726284e-07,
+      "loss": 0.0222,
+      "step": 11200
+    },
+    {
+      "epoch": 2.0302043769216858,
+      "grad_norm": 0.5222472548484802,
+      "learning_rate": 8.117701234113096e-07,
+      "loss": 0.0125,
+      "step": 11225
+    },
+    {
+      "epoch": 2.0347259902333152,
+      "grad_norm": 0.707737922668457,
+      "learning_rate": 8.113096334499907e-07,
+      "loss": 0.0275,
+      "step": 11250
+    },
+    {
+      "epoch": 2.0392476035449447,
+      "grad_norm": 2.295354127883911,
+      "learning_rate": 8.108491434886719e-07,
+      "loss": 0.023,
+      "step": 11275
+    },
+    {
+      "epoch": 2.0437692168565746,
+      "grad_norm": 1.3817616701126099,
+      "learning_rate": 8.103886535273531e-07,
+      "loss": 0.0125,
+      "step": 11300
+    },
+    {
+      "epoch": 2.048290830168204,
+      "grad_norm": 2.0756027698516846,
+      "learning_rate": 8.099281635660342e-07,
+      "loss": 0.0187,
+      "step": 11325
+    },
+    {
+      "epoch": 2.0528124434798336,
+      "grad_norm": 0.05818900838494301,
+      "learning_rate": 8.094676736047153e-07,
+      "loss": 0.0278,
+      "step": 11350
+    },
+    {
+      "epoch": 2.057334056791463,
+      "grad_norm": 6.99597692489624,
+      "learning_rate": 8.090071836433965e-07,
+      "loss": 0.0241,
+      "step": 11375
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 0.15094341337680817,
+      "learning_rate": 8.085466936820778e-07,
+      "loss": 0.0223,
+      "step": 11400
+    },
+    {
+      "epoch": 2.0663772834147225,
+      "grad_norm": 0.17221419513225555,
+      "learning_rate": 8.080862037207588e-07,
+      "loss": 0.0257,
+      "step": 11425
+    },
+    {
+      "epoch": 2.070898896726352,
+      "grad_norm": 0.5105612277984619,
+      "learning_rate": 8.0762571375944e-07,
+      "loss": 0.0523,
+      "step": 11450
+    },
+    {
+      "epoch": 2.0754205100379814,
+      "grad_norm": 1.4951982498168945,
+      "learning_rate": 8.071652237981212e-07,
+      "loss": 0.0247,
+      "step": 11475
+    },
+    {
+      "epoch": 2.079942123349611,
+      "grad_norm": 5.619513988494873,
+      "learning_rate": 8.067047338368023e-07,
+      "loss": 0.0181,
+      "step": 11500
+    },
+    {
+      "epoch": 2.084463736661241,
+      "grad_norm": 21.839784622192383,
+      "learning_rate": 8.062442438754835e-07,
+      "loss": 0.0221,
+      "step": 11525
+    },
+    {
+      "epoch": 2.0889853499728703,
+      "grad_norm": 4.323668003082275,
+      "learning_rate": 8.057837539141647e-07,
+      "loss": 0.0296,
+      "step": 11550
+    },
+    {
+      "epoch": 2.0935069632845,
+      "grad_norm": 0.17072859406471252,
+      "learning_rate": 8.053232639528457e-07,
+      "loss": 0.0196,
+      "step": 11575
+    },
+    {
+      "epoch": 2.0980285765961293,
+      "grad_norm": 13.157007217407227,
+      "learning_rate": 8.048627739915269e-07,
+      "loss": 0.0238,
+      "step": 11600
+    },
+    {
+      "epoch": 2.102550189907759,
+      "grad_norm": 0.13449828326702118,
+      "learning_rate": 8.044022840302082e-07,
+      "loss": 0.0157,
+      "step": 11625
+    },
+    {
+      "epoch": 2.1070718032193887,
+      "grad_norm": 9.773667335510254,
+      "learning_rate": 8.039417940688893e-07,
+      "loss": 0.0167,
+      "step": 11650
+    },
+    {
+      "epoch": 2.111593416531018,
+      "grad_norm": 0.01844405196607113,
+      "learning_rate": 8.034813041075704e-07,
+      "loss": 0.0062,
+      "step": 11675
+    },
+    {
+      "epoch": 2.1161150298426477,
+      "grad_norm": 2.560438394546509,
+      "learning_rate": 8.030208141462516e-07,
+      "loss": 0.0193,
+      "step": 11700
+    },
+    {
+      "epoch": 2.1206366431542776,
+      "grad_norm": 0.5435400605201721,
+      "learning_rate": 8.025603241849327e-07,
+      "loss": 0.0203,
+      "step": 11725
+    },
+    {
+      "epoch": 2.125158256465907,
+      "grad_norm": 63.197235107421875,
+      "learning_rate": 8.020998342236139e-07,
+      "loss": 0.0358,
+      "step": 11750
+    },
+    {
+      "epoch": 2.1296798697775365,
+      "grad_norm": 1.3490138053894043,
+      "learning_rate": 8.016393442622951e-07,
+      "loss": 0.013,
+      "step": 11775
+    },
+    {
+      "epoch": 2.134201483089166,
+      "grad_norm": 34.209476470947266,
+      "learning_rate": 8.011788543009762e-07,
+      "loss": 0.0351,
+      "step": 11800
+    },
+    {
+      "epoch": 2.138723096400796,
+      "grad_norm": 0.7151913642883301,
+      "learning_rate": 8.007183643396574e-07,
+      "loss": 0.0258,
+      "step": 11825
+    },
+    {
+      "epoch": 2.1432447097124254,
+      "grad_norm": 0.9076653718948364,
+      "learning_rate": 8.002578743783384e-07,
+      "loss": 0.0662,
+      "step": 11850
+    },
+    {
+      "epoch": 2.147766323024055,
+      "grad_norm": 0.8418449759483337,
+      "learning_rate": 7.997973844170197e-07,
+      "loss": 0.0301,
+      "step": 11875
+    },
+    {
+      "epoch": 2.1522879363356844,
+      "grad_norm": 0.5430082082748413,
+      "learning_rate": 7.993368944557009e-07,
+      "loss": 0.0166,
+      "step": 11900
+    },
+    {
+      "epoch": 2.1568095496473143,
+      "grad_norm": 0.6853590607643127,
+      "learning_rate": 7.988764044943819e-07,
+      "loss": 0.0148,
+      "step": 11925
+    },
+    {
+      "epoch": 2.161331162958944,
+      "grad_norm": 0.5972227454185486,
+      "learning_rate": 7.984159145330631e-07,
+      "loss": 0.0177,
+      "step": 11950
+    },
+    {
+      "epoch": 2.1658527762705733,
+      "grad_norm": 2.1670873165130615,
+      "learning_rate": 7.979554245717443e-07,
+      "loss": 0.0274,
+      "step": 11975
+    },
+    {
+      "epoch": 2.1703743895822027,
+      "grad_norm": 18.318012237548828,
+      "learning_rate": 7.974949346104254e-07,
+      "loss": 0.029,
+      "step": 12000
+    },
+    {
+      "epoch": 2.1748960028938327,
+      "grad_norm": 0.31500276923179626,
+      "learning_rate": 7.970344446491066e-07,
+      "loss": 0.012,
+      "step": 12025
+    },
+    {
+      "epoch": 2.179417616205462,
+      "grad_norm": 0.21801598370075226,
+      "learning_rate": 7.965739546877878e-07,
+      "loss": 0.0121,
+      "step": 12050
+    },
+    {
+      "epoch": 2.1839392295170916,
+      "grad_norm": 1.37786865234375,
+      "learning_rate": 7.961134647264689e-07,
+      "loss": 0.0066,
+      "step": 12075
+    },
+    {
+      "epoch": 2.188460842828721,
+      "grad_norm": 32.29001998901367,
+      "learning_rate": 7.9565297476515e-07,
+      "loss": 0.0341,
+      "step": 12100
+    },
+    {
+      "epoch": 2.192982456140351,
+      "grad_norm": 2.302133083343506,
+      "learning_rate": 7.951924848038313e-07,
+      "loss": 0.0188,
+      "step": 12125
+    },
+    {
+      "epoch": 2.1975040694519805,
+      "grad_norm": 14.034008979797363,
+      "learning_rate": 7.947319948425124e-07,
+      "loss": 0.0117,
+      "step": 12150
+    },
+    {
+      "epoch": 2.20202568276361,
+      "grad_norm": 0.3767974078655243,
+      "learning_rate": 7.942715048811935e-07,
+      "loss": 0.0259,
+      "step": 12175
+    },
+    {
+      "epoch": 2.2065472960752395,
+      "grad_norm": 0.26384684443473816,
+      "learning_rate": 7.938110149198747e-07,
+      "loss": 0.0371,
+      "step": 12200
+    },
+    {
+      "epoch": 2.2110689093868694,
+      "grad_norm": 8.888740539550781,
+      "learning_rate": 7.933505249585559e-07,
+      "loss": 0.0336,
+      "step": 12225
+    },
+    {
+      "epoch": 2.215590522698499,
+      "grad_norm": 0.19948595762252808,
+      "learning_rate": 7.928900349972371e-07,
+      "loss": 0.0474,
+      "step": 12250
+    },
+    {
+      "epoch": 2.2201121360101284,
+      "grad_norm": 2.2132930755615234,
+      "learning_rate": 7.924295450359182e-07,
+      "loss": 0.0406,
+      "step": 12275
+    },
+    {
+      "epoch": 2.224633749321758,
+      "grad_norm": 11.718713760375977,
+      "learning_rate": 7.919690550745993e-07,
+      "loss": 0.0195,
+      "step": 12300
+    },
+    {
+      "epoch": 2.2291553626333878,
+      "grad_norm": 5.642462730407715,
+      "learning_rate": 7.915085651132805e-07,
+      "loss": 0.0187,
+      "step": 12325
+    },
+    {
+      "epoch": 2.2336769759450172,
+      "grad_norm": 4.084228992462158,
+      "learning_rate": 7.910480751519617e-07,
+      "loss": 0.0126,
+      "step": 12350
+    },
+    {
+      "epoch": 2.2381985892566467,
+      "grad_norm": 1.1005765199661255,
+      "learning_rate": 7.905875851906428e-07,
+      "loss": 0.0167,
+      "step": 12375
+    },
+    {
+      "epoch": 2.242720202568276,
+      "grad_norm": 9.070086479187012,
+      "learning_rate": 7.90127095229324e-07,
+      "loss": 0.0182,
+      "step": 12400
+    },
+    {
+      "epoch": 2.247241815879906,
+      "grad_norm": 15.3062162399292,
+      "learning_rate": 7.896666052680051e-07,
+      "loss": 0.0156,
+      "step": 12425
+    },
+    {
+      "epoch": 2.2517634291915356,
+      "grad_norm": 0.09264446794986725,
+      "learning_rate": 7.892061153066862e-07,
+      "loss": 0.0237,
+      "step": 12450
+    },
+    {
+      "epoch": 2.256285042503165,
+      "grad_norm": 3.3248021602630615,
+      "learning_rate": 7.887456253453675e-07,
+      "loss": 0.0139,
+      "step": 12475
+    },
+    {
+      "epoch": 2.2608066558147946,
+      "grad_norm": 3.0979135036468506,
+      "learning_rate": 7.882851353840487e-07,
+      "loss": 0.0063,
+      "step": 12500
+    },
+    {
+      "epoch": 2.2653282691264245,
+      "grad_norm": 11.612130165100098,
+      "learning_rate": 7.878246454227297e-07,
+      "loss": 0.04,
+      "step": 12525
+    },
+    {
+      "epoch": 2.269849882438054,
+      "grad_norm": 4.221678256988525,
+      "learning_rate": 7.873641554614109e-07,
+      "loss": 0.0393,
+      "step": 12550
+    },
+    {
+      "epoch": 2.2743714957496834,
+      "grad_norm": 11.065829277038574,
+      "learning_rate": 7.869036655000921e-07,
+      "loss": 0.0286,
+      "step": 12575
+    },
+    {
+      "epoch": 2.278893109061313,
+      "grad_norm": 1.366445779800415,
+      "learning_rate": 7.864431755387732e-07,
+      "loss": 0.028,
+      "step": 12600
+    },
+    {
+      "epoch": 2.283414722372943,
+      "grad_norm": 0.7951880693435669,
+      "learning_rate": 7.859826855774544e-07,
+      "loss": 0.0661,
+      "step": 12625
+    },
+    {
+      "epoch": 2.2879363356845723,
+      "grad_norm": 10.447066307067871,
+      "learning_rate": 7.855221956161356e-07,
+      "loss": 0.0484,
+      "step": 12650
+    },
+    {
+      "epoch": 2.292457948996202,
+      "grad_norm": 0.32853183150291443,
+      "learning_rate": 7.850617056548166e-07,
+      "loss": 0.0239,
+      "step": 12675
+    },
+    {
+      "epoch": 2.2969795623078313,
+      "grad_norm": 2.0133612155914307,
+      "learning_rate": 7.846012156934978e-07,
+      "loss": 0.0106,
+      "step": 12700
+    },
+    {
+      "epoch": 2.301501175619461,
+      "grad_norm": 1.5879937410354614,
+      "learning_rate": 7.841407257321791e-07,
+      "loss": 0.0151,
+      "step": 12725
+    },
+    {
+      "epoch": 2.3060227889310907,
+      "grad_norm": 3.2537899017333984,
+      "learning_rate": 7.836802357708602e-07,
+      "loss": 0.021,
+      "step": 12750
+    },
+    {
+      "epoch": 2.31054440224272,
+      "grad_norm": 0.4321633577346802,
+      "learning_rate": 7.832197458095413e-07,
+      "loss": 0.0135,
+      "step": 12775
+    },
+    {
+      "epoch": 2.3150660155543497,
+      "grad_norm": 17.50613021850586,
+      "learning_rate": 7.827592558482224e-07,
+      "loss": 0.024,
+      "step": 12800
+    },
+    {
+      "epoch": 2.319587628865979,
+      "grad_norm": 10.587005615234375,
+      "learning_rate": 7.822987658869036e-07,
+      "loss": 0.0155,
+      "step": 12825
+    },
+    {
+      "epoch": 2.324109242177609,
+      "grad_norm": 1.703659176826477,
+      "learning_rate": 7.818382759255848e-07,
+      "loss": 0.0204,
+      "step": 12850
+    },
+    {
+      "epoch": 2.3286308554892385,
+      "grad_norm": 0.26982223987579346,
+      "learning_rate": 7.813777859642659e-07,
+      "loss": 0.0087,
+      "step": 12875
+    },
+    {
+      "epoch": 2.333152468800868,
+      "grad_norm": 4.538456916809082,
+      "learning_rate": 7.809172960029471e-07,
+      "loss": 0.0177,
+      "step": 12900
+    },
+    {
+      "epoch": 2.337674082112498,
+      "grad_norm": 5.056499481201172,
+      "learning_rate": 7.804568060416282e-07,
+      "loss": 0.0163,
+      "step": 12925
+    },
+    {
+      "epoch": 2.3421956954241274,
+      "grad_norm": 0.5478576421737671,
+      "learning_rate": 7.799963160803094e-07,
+      "loss": 0.0149,
+      "step": 12950
+    },
+    {
+      "epoch": 2.346717308735757,
+      "grad_norm": 1.6986396312713623,
+      "learning_rate": 7.795358261189906e-07,
+      "loss": 0.0341,
+      "step": 12975
+    },
+    {
+      "epoch": 2.3512389220473864,
+      "grad_norm": 0.026538992300629616,
+      "learning_rate": 7.790753361576718e-07,
+      "loss": 0.0143,
+      "step": 13000
+    },
+    {
+      "epoch": 2.355760535359016,
+      "grad_norm": 0.38667038083076477,
+      "learning_rate": 7.786148461963528e-07,
+      "loss": 0.0079,
+      "step": 13025
+    },
+    {
+      "epoch": 2.360282148670646,
+      "grad_norm": 15.939850807189941,
+      "learning_rate": 7.78154356235034e-07,
+      "loss": 0.0372,
+      "step": 13050
+    },
+    {
+      "epoch": 2.3648037619822753,
+      "grad_norm": 18.573139190673828,
+      "learning_rate": 7.776938662737153e-07,
+      "loss": 0.0441,
+      "step": 13075
+    },
+    {
+      "epoch": 2.3693253752939047,
+      "grad_norm": 3.2550344467163086,
+      "learning_rate": 7.772333763123963e-07,
+      "loss": 0.027,
+      "step": 13100
+    },
+    {
+      "epoch": 2.3738469886055347,
+      "grad_norm": 1.5440771579742432,
+      "learning_rate": 7.767728863510775e-07,
+      "loss": 0.0277,
+      "step": 13125
+    },
+    {
+      "epoch": 2.378368601917164,
+      "grad_norm": 1.0547950267791748,
+      "learning_rate": 7.763123963897587e-07,
+      "loss": 0.0278,
+      "step": 13150
+    },
+    {
+      "epoch": 2.3828902152287936,
+      "grad_norm": 0.08282533288002014,
+      "learning_rate": 7.758519064284398e-07,
+      "loss": 0.0101,
+      "step": 13175
+    },
+    {
+      "epoch": 2.387411828540423,
+      "grad_norm": 0.24389539659023285,
+      "learning_rate": 7.75391416467121e-07,
+      "loss": 0.0233,
+      "step": 13200
+    },
+    {
+      "epoch": 2.3919334418520526,
+      "grad_norm": 9.251720428466797,
+      "learning_rate": 7.749309265058022e-07,
+      "loss": 0.047,
+      "step": 13225
+    },
+    {
+      "epoch": 2.3964550551636825,
+      "grad_norm": 2.2844269275665283,
+      "learning_rate": 7.744704365444833e-07,
+      "loss": 0.0162,
+      "step": 13250
+    },
+    {
+      "epoch": 2.400976668475312,
+      "grad_norm": 3.2137227058410645,
+      "learning_rate": 7.740099465831644e-07,
+      "loss": 0.0121,
+      "step": 13275
+    },
+    {
+      "epoch": 2.4054982817869415,
+      "grad_norm": 11.308737754821777,
+      "learning_rate": 7.735494566218456e-07,
+      "loss": 0.0299,
+      "step": 13300
+    },
+    {
+      "epoch": 2.4100198950985714,
+      "grad_norm": 3.3836469650268555,
+      "learning_rate": 7.730889666605268e-07,
+      "loss": 0.0133,
+      "step": 13325
+    },
+    {
+      "epoch": 2.414541508410201,
+      "grad_norm": 0.13357259333133698,
+      "learning_rate": 7.726284766992079e-07,
+      "loss": 0.0223,
+      "step": 13350
+    },
+    {
+      "epoch": 2.4190631217218304,
+      "grad_norm": 0.5216515064239502,
+      "learning_rate": 7.721864063363418e-07,
+      "loss": 0.0212,
+      "step": 13375
+    },
+    {
+      "epoch": 2.42358473503346,
+      "grad_norm": 4.0334320068359375,
+      "learning_rate": 7.71725916375023e-07,
+      "loss": 0.0367,
+      "step": 13400
+    },
+    {
+      "epoch": 2.4281063483450893,
+      "grad_norm": 8.48493766784668,
+      "learning_rate": 7.712654264137042e-07,
+      "loss": 0.0254,
+      "step": 13425
+    },
+    {
+      "epoch": 2.4326279616567192,
+      "grad_norm": 1.6405227184295654,
+      "learning_rate": 7.708049364523853e-07,
+      "loss": 0.0331,
+      "step": 13450
+    },
+    {
+      "epoch": 2.4371495749683487,
+      "grad_norm": 13.649563789367676,
+      "learning_rate": 7.703444464910664e-07,
+      "loss": 0.018,
+      "step": 13475
+    },
+    {
+      "epoch": 2.441671188279978,
+      "grad_norm": 0.7964933514595032,
+      "learning_rate": 7.698839565297476e-07,
+      "loss": 0.0235,
+      "step": 13500
+    },
+    {
+      "epoch": 2.446192801591608,
+      "grad_norm": 0.13087065517902374,
+      "learning_rate": 7.694234665684289e-07,
+      "loss": 0.0215,
+      "step": 13525
+    },
+    {
+      "epoch": 2.4507144149032376,
+      "grad_norm": 5.35853385925293,
+      "learning_rate": 7.689629766071099e-07,
+      "loss": 0.0178,
+      "step": 13550
+    },
+    {
+      "epoch": 2.455236028214867,
+      "grad_norm": 3.688849687576294,
+      "learning_rate": 7.685024866457911e-07,
+      "loss": 0.0099,
+      "step": 13575
+    },
+    {
+      "epoch": 2.4597576415264966,
+      "grad_norm": 0.2602083384990692,
+      "learning_rate": 7.680419966844723e-07,
+      "loss": 0.0205,
+      "step": 13600
+    },
+    {
+      "epoch": 2.464279254838126,
+      "grad_norm": 0.10222572088241577,
+      "learning_rate": 7.675815067231533e-07,
+      "loss": 0.012,
+      "step": 13625
+    },
+    {
+      "epoch": 2.468800868149756,
+      "grad_norm": 1.9992151260375977,
+      "learning_rate": 7.671210167618346e-07,
+      "loss": 0.0295,
+      "step": 13650
+    },
+    {
+      "epoch": 2.4733224814613854,
+      "grad_norm": 3.8986308574676514,
+      "learning_rate": 7.666605268005158e-07,
+      "loss": 0.0447,
+      "step": 13675
+    },
+    {
+      "epoch": 2.477844094773015,
+      "grad_norm": 2.5787339210510254,
+      "learning_rate": 7.662000368391968e-07,
+      "loss": 0.0233,
+      "step": 13700
+    },
+    {
+      "epoch": 2.482365708084645,
+      "grad_norm": 3.943392276763916,
+      "learning_rate": 7.65739546877878e-07,
+      "loss": 0.0214,
+      "step": 13725
+    },
+    {
+      "epoch": 2.4868873213962743,
+      "grad_norm": 4.015535831451416,
+      "learning_rate": 7.652790569165592e-07,
+      "loss": 0.0146,
+      "step": 13750
+    },
+    {
+      "epoch": 2.491408934707904,
+      "grad_norm": 1.5608233213424683,
+      "learning_rate": 7.648185669552404e-07,
+      "loss": 0.0171,
+      "step": 13775
+    },
+    {
+      "epoch": 2.4959305480195333,
+      "grad_norm": 1.696368932723999,
+      "learning_rate": 7.643580769939215e-07,
+      "loss": 0.0122,
+      "step": 13800
+    },
+    {
+      "epoch": 2.5004521613311628,
+      "grad_norm": 1.1217238903045654,
+      "learning_rate": 7.638975870326027e-07,
+      "loss": 0.0128,
+      "step": 13825
+    },
+    {
+      "epoch": 2.5049737746427927,
+      "grad_norm": 40.497745513916016,
+      "learning_rate": 7.634370970712838e-07,
+      "loss": 0.0277,
+      "step": 13850
+    },
+    {
+      "epoch": 2.509495387954422,
+      "grad_norm": 6.063665390014648,
+      "learning_rate": 7.629766071099649e-07,
+      "loss": 0.0118,
+      "step": 13875
+    },
+    {
+      "epoch": 2.5140170012660517,
+      "grad_norm": 8.675702095031738,
+      "learning_rate": 7.625161171486462e-07,
+      "loss": 0.0259,
+      "step": 13900
+    },
+    {
+      "epoch": 2.5185386145776816,
+      "grad_norm": 0.8335000872612,
+      "learning_rate": 7.620556271873273e-07,
+      "loss": 0.0147,
+      "step": 13925
+    },
+    {
+      "epoch": 2.523060227889311,
+      "grad_norm": 8.890750885009766,
+      "learning_rate": 7.615951372260084e-07,
+      "loss": 0.0226,
+      "step": 13950
+    },
+    {
+      "epoch": 2.5275818412009405,
+      "grad_norm": 0.20721301436424255,
+      "learning_rate": 7.611346472646895e-07,
+      "loss": 0.0103,
+      "step": 13975
+    },
+    {
+      "epoch": 2.53210345451257,
+      "grad_norm": 0.5705264806747437,
+      "learning_rate": 7.606741573033707e-07,
+      "loss": 0.0094,
+      "step": 14000
+    },
+    {
+      "epoch": 2.5366250678241995,
+      "grad_norm": 0.07163272053003311,
+      "learning_rate": 7.60213667342052e-07,
+      "loss": 0.0177,
+      "step": 14025
+    },
+    {
+      "epoch": 2.5411466811358294,
+      "grad_norm": 0.8082312345504761,
+      "learning_rate": 7.59753177380733e-07,
+      "loss": 0.0177,
+      "step": 14050
+    },
+    {
+      "epoch": 2.545668294447459,
+      "grad_norm": 0.3273601830005646,
+      "learning_rate": 7.592926874194142e-07,
+      "loss": 0.0429,
+      "step": 14075
+    },
+    {
+      "epoch": 2.5501899077590884,
+      "grad_norm": 1.8662065267562866,
+      "learning_rate": 7.588321974580954e-07,
+      "loss": 0.0163,
+      "step": 14100
+    },
+    {
+      "epoch": 2.5547115210707183,
+      "grad_norm": 0.7974827289581299,
+      "learning_rate": 7.583717074967764e-07,
+      "loss": 0.0159,
+      "step": 14125
+    },
+    {
+      "epoch": 2.559233134382348,
+      "grad_norm": 12.264116287231445,
+      "learning_rate": 7.579112175354577e-07,
+      "loss": 0.0329,
+      "step": 14150
+    },
+    {
+      "epoch": 2.5637547476939773,
+      "grad_norm": 0.07791896164417267,
+      "learning_rate": 7.574507275741389e-07,
+      "loss": 0.0214,
+      "step": 14175
+    },
+    {
+      "epoch": 2.5682763610056067,
+      "grad_norm": 0.7379089593887329,
+      "learning_rate": 7.5699023761282e-07,
+      "loss": 0.0366,
+      "step": 14200
+    },
+    {
+      "epoch": 2.5727979743172362,
+      "grad_norm": 91.46224975585938,
+      "learning_rate": 7.565297476515011e-07,
+      "loss": 0.027,
+      "step": 14225
+    },
+    {
+      "epoch": 2.5773195876288657,
+      "grad_norm": 14.806313514709473,
+      "learning_rate": 7.560692576901824e-07,
+      "loss": 0.0488,
+      "step": 14250
+    },
+    {
+      "epoch": 2.5818412009404956,
+      "grad_norm": 2.083322286605835,
+      "learning_rate": 7.556087677288635e-07,
+      "loss": 0.031,
+      "step": 14275
+    },
+    {
+      "epoch": 2.586362814252125,
+      "grad_norm": 4.879816055297852,
+      "learning_rate": 7.551482777675446e-07,
+      "loss": 0.0153,
+      "step": 14300
+    },
+    {
+      "epoch": 2.5908844275637546,
+      "grad_norm": 6.237574100494385,
+      "learning_rate": 7.546877878062258e-07,
+      "loss": 0.0141,
+      "step": 14325
+    },
+    {
+      "epoch": 2.5954060408753845,
+      "grad_norm": 0.12708225846290588,
+      "learning_rate": 7.542272978449069e-07,
+      "loss": 0.0101,
+      "step": 14350
+    },
+    {
+      "epoch": 2.599927654187014,
+      "grad_norm": 1.2891823053359985,
+      "learning_rate": 7.537668078835881e-07,
+      "loss": 0.0215,
+      "step": 14375
+    },
+    {
+      "epoch": 2.6044492674986435,
+      "grad_norm": 0.09198635071516037,
+      "learning_rate": 7.533063179222693e-07,
+      "loss": 0.0202,
+      "step": 14400
+    },
+    {
+      "epoch": 2.608970880810273,
+      "grad_norm": 0.42182183265686035,
+      "learning_rate": 7.528458279609504e-07,
+      "loss": 0.0115,
+      "step": 14425
+    },
+    {
+      "epoch": 2.6134924941219024,
+      "grad_norm": 0.22351473569869995,
+      "learning_rate": 7.523853379996316e-07,
+      "loss": 0.0133,
+      "step": 14450
+    },
+    {
+      "epoch": 2.6180141074335324,
+      "grad_norm": 6.165104389190674,
+      "learning_rate": 7.519248480383127e-07,
+      "loss": 0.0381,
+      "step": 14475
+    },
+    {
+      "epoch": 2.622535720745162,
+      "grad_norm": 4.592835903167725,
+      "learning_rate": 7.514643580769939e-07,
+      "loss": 0.0118,
+      "step": 14500
+    },
+    {
+      "epoch": 2.6270573340567913,
+      "grad_norm": 13.439335823059082,
+      "learning_rate": 7.510038681156751e-07,
+      "loss": 0.0324,
+      "step": 14525
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 3.1381280422210693,
+      "learning_rate": 7.505433781543562e-07,
+      "loss": 0.0383,
+      "step": 14550
+    },
+    {
+      "epoch": 2.6361005606800507,
+      "grad_norm": 3.8765087127685547,
+      "learning_rate": 7.500828881930373e-07,
+      "loss": 0.03,
+      "step": 14575
+    },
+    {
+      "epoch": 2.64062217399168,
+      "grad_norm": 12.755854606628418,
+      "learning_rate": 7.496223982317185e-07,
+      "loss": 0.0356,
+      "step": 14600
+    },
+    {
+      "epoch": 2.6451437873033097,
+      "grad_norm": 48.162261962890625,
+      "learning_rate": 7.491619082703998e-07,
+      "loss": 0.0244,
+      "step": 14625
+    },
+    {
+      "epoch": 2.649665400614939,
+      "grad_norm": 55.78163528442383,
+      "learning_rate": 7.487014183090808e-07,
+      "loss": 0.0495,
+      "step": 14650
+    },
+    {
+      "epoch": 2.654187013926569,
+      "grad_norm": 4.578949451446533,
+      "learning_rate": 7.48240928347762e-07,
+      "loss": 0.0292,
+      "step": 14675
+    },
+    {
+      "epoch": 2.6587086272381986,
+      "grad_norm": 6.509306907653809,
+      "learning_rate": 7.477804383864432e-07,
+      "loss": 0.0193,
+      "step": 14700
+    },
+    {
+      "epoch": 2.663230240549828,
+      "grad_norm": 4.97738790512085,
+      "learning_rate": 7.473199484251242e-07,
+      "loss": 0.0223,
+      "step": 14725
+    },
+    {
+      "epoch": 2.667751853861458,
+      "grad_norm": 7.346843242645264,
+      "learning_rate": 7.468594584638055e-07,
+      "loss": 0.0244,
+      "step": 14750
+    },
+    {
+      "epoch": 2.6722734671730874,
+      "grad_norm": 0.34532052278518677,
+      "learning_rate": 7.463989685024867e-07,
+      "loss": 0.0158,
+      "step": 14775
+    },
+    {
+      "epoch": 2.676795080484717,
+      "grad_norm": 1.6974883079528809,
+      "learning_rate": 7.459384785411677e-07,
+      "loss": 0.0172,
+      "step": 14800
+    },
+    {
+      "epoch": 2.6813166937963464,
+      "grad_norm": 4.6452507972717285,
+      "learning_rate": 7.454779885798489e-07,
+      "loss": 0.034,
+      "step": 14825
+    },
+    {
+      "epoch": 2.685838307107976,
+      "grad_norm": 40.83687973022461,
+      "learning_rate": 7.4501749861853e-07,
+      "loss": 0.0238,
+      "step": 14850
+    },
+    {
+      "epoch": 2.690359920419606,
+      "grad_norm": 0.27361202239990234,
+      "learning_rate": 7.445570086572113e-07,
+      "loss": 0.0217,
+      "step": 14875
+    },
+    {
+      "epoch": 2.6948815337312353,
+      "grad_norm": 0.088538758456707,
+      "learning_rate": 7.440965186958924e-07,
+      "loss": 0.0216,
+      "step": 14900
+    },
+    {
+      "epoch": 2.6994031470428648,
+      "grad_norm": 0.36908406019210815,
+      "learning_rate": 7.436360287345735e-07,
+      "loss": 0.0163,
+      "step": 14925
+    },
+    {
+      "epoch": 2.7039247603544947,
+      "grad_norm": 0.3927018344402313,
+      "learning_rate": 7.431755387732547e-07,
+      "loss": 0.0272,
+      "step": 14950
+    },
+    {
+      "epoch": 2.708446373666124,
+      "grad_norm": 0.08495494723320007,
+      "learning_rate": 7.427150488119359e-07,
+      "loss": 0.0321,
+      "step": 14975
+    },
+    {
+      "epoch": 2.7129679869777537,
+      "grad_norm": 16.534454345703125,
+      "learning_rate": 7.42254558850617e-07,
+      "loss": 0.0346,
+      "step": 15000
+    },
+    {
+      "epoch": 2.717489600289383,
+      "grad_norm": 10.199457168579102,
+      "learning_rate": 7.417940688892982e-07,
+      "loss": 0.0236,
+      "step": 15025
+    },
+    {
+      "epoch": 2.7220112136010126,
+      "grad_norm": 12.526575088500977,
+      "learning_rate": 7.413335789279793e-07,
+      "loss": 0.0344,
+      "step": 15050
+    },
+    {
+      "epoch": 2.7265328269126425,
+      "grad_norm": 0.6113793253898621,
+      "learning_rate": 7.408730889666604e-07,
+      "loss": 0.0549,
+      "step": 15075
+    },
+    {
+      "epoch": 2.731054440224272,
+      "grad_norm": 2.575866460800171,
+      "learning_rate": 7.404125990053417e-07,
+      "loss": 0.0188,
+      "step": 15100
+    },
+    {
+      "epoch": 2.7355760535359015,
+      "grad_norm": 0.5728238821029663,
+      "learning_rate": 7.399521090440229e-07,
+      "loss": 0.0248,
+      "step": 15125
+    },
+    {
+      "epoch": 2.7400976668475314,
+      "grad_norm": 5.137115478515625,
+      "learning_rate": 7.394916190827039e-07,
+      "loss": 0.0234,
+      "step": 15150
+    },
+    {
+      "epoch": 2.744619280159161,
+      "grad_norm": 2.0585458278656006,
+      "learning_rate": 7.390311291213851e-07,
+      "loss": 0.0193,
+      "step": 15175
+    },
+    {
+      "epoch": 2.7491408934707904,
+      "grad_norm": 2.6761465072631836,
+      "learning_rate": 7.385706391600663e-07,
+      "loss": 0.0242,
+      "step": 15200
+    },
+    {
+      "epoch": 2.75366250678242,
+      "grad_norm": 0.3999291956424713,
+      "learning_rate": 7.381101491987474e-07,
+      "loss": 0.0193,
+      "step": 15225
+    },
+    {
+      "epoch": 2.7581841200940493,
+      "grad_norm": 8.237870216369629,
+      "learning_rate": 7.376496592374286e-07,
+      "loss": 0.0278,
+      "step": 15250
+    },
+    {
+      "epoch": 2.7627057334056793,
+      "grad_norm": 8.460784912109375,
+      "learning_rate": 7.371891692761098e-07,
+      "loss": 0.0194,
+      "step": 15275
+    },
+    {
+      "epoch": 2.7672273467173087,
+      "grad_norm": 13.062602043151855,
+      "learning_rate": 7.367286793147908e-07,
+      "loss": 0.032,
+      "step": 15300
+    },
+    {
+      "epoch": 2.7717489600289382,
+      "grad_norm": 7.502108097076416,
+      "learning_rate": 7.36268189353472e-07,
+      "loss": 0.0076,
+      "step": 15325
+    },
+    {
+      "epoch": 2.776270573340568,
+      "grad_norm": 11.163969039916992,
+      "learning_rate": 7.358076993921533e-07,
+      "loss": 0.024,
+      "step": 15350
+    },
+    {
+      "epoch": 2.7807921866521976,
+      "grad_norm": 1.2433866262435913,
+      "learning_rate": 7.353472094308344e-07,
+      "loss": 0.0276,
+      "step": 15375
+    },
+    {
+      "epoch": 2.785313799963827,
+      "grad_norm": 0.21064484119415283,
+      "learning_rate": 7.348867194695155e-07,
+      "loss": 0.0104,
+      "step": 15400
+    },
+    {
+      "epoch": 2.7898354132754566,
+      "grad_norm": 39.60307312011719,
+      "learning_rate": 7.344446491066495e-07,
+      "loss": 0.0408,
+      "step": 15425
+    },
+    {
+      "epoch": 2.794357026587086,
+      "grad_norm": 1.9793568849563599,
+      "learning_rate": 7.339841591453306e-07,
+      "loss": 0.059,
+      "step": 15450
+    },
+    {
+      "epoch": 2.798878639898716,
+      "grad_norm": 0.8816681504249573,
+      "learning_rate": 7.335236691840118e-07,
+      "loss": 0.069,
+      "step": 15475
+    },
+    {
+      "epoch": 2.8034002532103455,
+      "grad_norm": 1.217022180557251,
+      "learning_rate": 7.330631792226929e-07,
+      "loss": 0.0182,
+      "step": 15500
+    },
+    {
+      "epoch": 2.807921866521975,
+      "grad_norm": 10.408825874328613,
+      "learning_rate": 7.32602689261374e-07,
+      "loss": 0.0192,
+      "step": 15525
+    },
+    {
+      "epoch": 2.812443479833605,
+      "grad_norm": 0.13745278120040894,
+      "learning_rate": 7.321421993000553e-07,
+      "loss": 0.02,
+      "step": 15550
+    },
+    {
+      "epoch": 2.8169650931452344,
+      "grad_norm": 4.462474346160889,
+      "learning_rate": 7.316817093387364e-07,
+      "loss": 0.0229,
+      "step": 15575
+    },
+    {
+      "epoch": 2.821486706456864,
+      "grad_norm": 0.2782382369041443,
+      "learning_rate": 7.312212193774175e-07,
+      "loss": 0.0209,
+      "step": 15600
+    },
+    {
+      "epoch": 2.8260083197684933,
+      "grad_norm": 9.691481590270996,
+      "learning_rate": 7.307607294160987e-07,
+      "loss": 0.0226,
+      "step": 15625
+    },
+    {
+      "epoch": 2.830529933080123,
+      "grad_norm": 0.038128096610307693,
+      "learning_rate": 7.303002394547798e-07,
+      "loss": 0.0076,
+      "step": 15650
+    },
+    {
+      "epoch": 2.8350515463917527,
+      "grad_norm": 3.7006304264068604,
+      "learning_rate": 7.29839749493461e-07,
+      "loss": 0.0116,
+      "step": 15675
+    },
+    {
+      "epoch": 2.839573159703382,
+      "grad_norm": 0.9358986616134644,
+      "learning_rate": 7.293792595321422e-07,
+      "loss": 0.0176,
+      "step": 15700
+    },
+    {
+      "epoch": 2.8440947730150117,
+      "grad_norm": 6.940121173858643,
+      "learning_rate": 7.289187695708234e-07,
+      "loss": 0.018,
+      "step": 15725
+    },
+    {
+      "epoch": 2.8486163863266416,
+      "grad_norm": 0.16799919307231903,
+      "learning_rate": 7.284582796095044e-07,
+      "loss": 0.0191,
+      "step": 15750
+    },
+    {
+      "epoch": 2.853137999638271,
+      "grad_norm": 8.835237503051758,
+      "learning_rate": 7.279977896481856e-07,
+      "loss": 0.023,
+      "step": 15775
+    },
+    {
+      "epoch": 2.8576596129499006,
+      "grad_norm": 41.18638229370117,
+      "learning_rate": 7.275372996868669e-07,
+      "loss": 0.0451,
+      "step": 15800
+    },
+    {
+      "epoch": 2.86218122626153,
+      "grad_norm": 7.924957752227783,
+      "learning_rate": 7.270768097255479e-07,
+      "loss": 0.017,
+      "step": 15825
+    },
+    {
+      "epoch": 2.8667028395731595,
+      "grad_norm": 0.04123552888631821,
+      "learning_rate": 7.266163197642291e-07,
+      "loss": 0.0736,
+      "step": 15850
+    },
+    {
+      "epoch": 2.8712244528847894,
+      "grad_norm": 0.38446855545043945,
+      "learning_rate": 7.261558298029103e-07,
+      "loss": 0.0227,
+      "step": 15875
+    },
+    {
+      "epoch": 2.875746066196419,
+      "grad_norm": 2.7168097496032715,
+      "learning_rate": 7.256953398415914e-07,
+      "loss": 0.0294,
+      "step": 15900
+    },
+    {
+      "epoch": 2.8802676795080484,
+      "grad_norm": 6.317012786865234,
+      "learning_rate": 7.252348498802726e-07,
+      "loss": 0.0313,
+      "step": 15925
+    },
+    {
+      "epoch": 2.8847892928196783,
+      "grad_norm": 7.5447821617126465,
+      "learning_rate": 7.247743599189538e-07,
+      "loss": 0.02,
+      "step": 15950
+    },
+    {
+      "epoch": 2.889310906131308,
+      "grad_norm": 0.06321001052856445,
+      "learning_rate": 7.243138699576349e-07,
+      "loss": 0.0197,
+      "step": 15975
+    },
+    {
+      "epoch": 2.8938325194429373,
+      "grad_norm": 0.14826831221580505,
+      "learning_rate": 7.23853379996316e-07,
+      "loss": 0.0111,
+      "step": 16000
+    },
+    {
+      "epoch": 2.8983541327545668,
+      "grad_norm": 10.19642448425293,
+      "learning_rate": 7.233928900349971e-07,
+      "loss": 0.0235,
+      "step": 16025
+    },
+    {
+      "epoch": 2.9028757460661962,
+      "grad_norm": 0.09074613451957703,
+      "learning_rate": 7.229324000736784e-07,
+      "loss": 0.0177,
+      "step": 16050
+    },
+    {
+      "epoch": 2.907397359377826,
+      "grad_norm": 50.30870819091797,
+      "learning_rate": 7.224719101123595e-07,
+      "loss": 0.0166,
+      "step": 16075
+    },
+    {
+      "epoch": 2.9119189726894557,
+      "grad_norm": 0.3094925880432129,
+      "learning_rate": 7.220114201510406e-07,
+      "loss": 0.0154,
+      "step": 16100
+    },
+    {
+      "epoch": 2.916440586001085,
+      "grad_norm": 0.837853193283081,
+      "learning_rate": 7.215509301897218e-07,
+      "loss": 0.0331,
+      "step": 16125
+    },
+    {
+      "epoch": 2.9209621993127146,
+      "grad_norm": 1.2888524532318115,
+      "learning_rate": 7.210904402284031e-07,
+      "loss": 0.0331,
+      "step": 16150
+    },
+    {
+      "epoch": 2.9254838126243445,
+      "grad_norm": 0.05001299828290939,
+      "learning_rate": 7.206299502670841e-07,
+      "loss": 0.021,
+      "step": 16175
+    },
+    {
+      "epoch": 2.930005425935974,
+      "grad_norm": 4.152213096618652,
+      "learning_rate": 7.201694603057653e-07,
+      "loss": 0.0376,
+      "step": 16200
+    },
+    {
+      "epoch": 2.9345270392476035,
+      "grad_norm": 3.798003911972046,
+      "learning_rate": 7.197089703444465e-07,
+      "loss": 0.0612,
+      "step": 16225
+    },
+    {
+      "epoch": 2.939048652559233,
+      "grad_norm": 0.4344462454319,
+      "learning_rate": 7.192484803831275e-07,
+      "loss": 0.0331,
+      "step": 16250
+    },
+    {
+      "epoch": 2.9435702658708625,
+      "grad_norm": 12.142309188842773,
+      "learning_rate": 7.187879904218088e-07,
+      "loss": 0.0369,
+      "step": 16275
+    },
+    {
+      "epoch": 2.9480918791824924,
+      "grad_norm": 1.0713512897491455,
+      "learning_rate": 7.1832750046049e-07,
+      "loss": 0.0201,
+      "step": 16300
+    },
+    {
+      "epoch": 2.952613492494122,
+      "grad_norm": 12.534268379211426,
+      "learning_rate": 7.17867010499171e-07,
+      "loss": 0.0232,
+      "step": 16325
+    },
+    {
+      "epoch": 2.9571351058057513,
+      "grad_norm": 2.9991679191589355,
+      "learning_rate": 7.174065205378522e-07,
+      "loss": 0.0212,
+      "step": 16350
+    },
+    {
+      "epoch": 2.9616567191173813,
+      "grad_norm": 0.7451911568641663,
+      "learning_rate": 7.169460305765334e-07,
+      "loss": 0.0128,
+      "step": 16375
+    },
+    {
+      "epoch": 2.9661783324290107,
+      "grad_norm": 0.15179577469825745,
+      "learning_rate": 7.164855406152146e-07,
+      "loss": 0.0148,
+      "step": 16400
+    },
+    {
+      "epoch": 2.9706999457406402,
+      "grad_norm": 0.2197951227426529,
+      "learning_rate": 7.160250506538957e-07,
+      "loss": 0.0202,
+      "step": 16425
+    },
+    {
+      "epoch": 2.9752215590522697,
+      "grad_norm": 8.621418952941895,
+      "learning_rate": 7.155645606925769e-07,
+      "loss": 0.0297,
+      "step": 16450
+    },
+    {
+      "epoch": 2.979743172363899,
+      "grad_norm": 1.0757030248641968,
+      "learning_rate": 7.15104070731258e-07,
+      "loss": 0.0164,
+      "step": 16475
+    },
+    {
+      "epoch": 2.984264785675529,
+      "grad_norm": 7.468885898590088,
+      "learning_rate": 7.146435807699391e-07,
+      "loss": 0.0222,
+      "step": 16500
+    },
+    {
+      "epoch": 2.9887863989871586,
+      "grad_norm": 22.351499557495117,
+      "learning_rate": 7.141830908086204e-07,
+      "loss": 0.032,
+      "step": 16525
+    },
+    {
+      "epoch": 2.993308012298788,
+      "grad_norm": 1.1031124591827393,
+      "learning_rate": 7.137226008473015e-07,
+      "loss": 0.0178,
+      "step": 16550
+    },
+    {
+      "epoch": 2.997829625610418,
+      "grad_norm": 0.33143824338912964,
+      "learning_rate": 7.132621108859827e-07,
+      "loss": 0.0183,
+      "step": 16575
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.28094977140426636,
+      "eval_runtime": 8563.7182,
+      "eval_samples_per_second": 1.109,
+      "eval_steps_per_second": 0.139,
+      "eval_wer": 0.10791159513132607,
+      "step": 16587
+    },
+    {
+      "epoch": 3.0023512389220475,
+      "grad_norm": 2.0394110679626465,
+      "learning_rate": 7.128016209246638e-07,
+      "loss": 0.0198,
+      "step": 16600
+    },
+    {
+      "epoch": 3.006872852233677,
+      "grad_norm": 0.34627845883369446,
+      "learning_rate": 7.123411309633449e-07,
+      "loss": 0.0163,
+      "step": 16625
+    },
+    {
+      "epoch": 3.0113944655453064,
+      "grad_norm": 0.1096586138010025,
+      "learning_rate": 7.118806410020262e-07,
+      "loss": 0.012,
+      "step": 16650
+    },
+    {
+      "epoch": 3.0159160788569364,
+      "grad_norm": 0.8496006727218628,
+      "learning_rate": 7.114201510407073e-07,
+      "loss": 0.0078,
+      "step": 16675
+    },
+    {
+      "epoch": 3.020437692168566,
+      "grad_norm": 0.45903900265693665,
+      "learning_rate": 7.109596610793884e-07,
+      "loss": 0.0113,
+      "step": 16700
+    },
+    {
+      "epoch": 3.0249593054801953,
+      "grad_norm": 0.18622107803821564,
+      "learning_rate": 7.104991711180696e-07,
+      "loss": 0.0177,
+      "step": 16725
+    },
+    {
+      "epoch": 3.029480918791825,
+      "grad_norm": 0.07214221358299255,
+      "learning_rate": 7.100386811567507e-07,
+      "loss": 0.0167,
+      "step": 16750
+    },
+    {
+      "epoch": 3.0340025321034547,
+      "grad_norm": 1.387891173362732,
+      "learning_rate": 7.095781911954319e-07,
+      "loss": 0.0085,
+      "step": 16775
+    },
+    {
+      "epoch": 3.038524145415084,
+      "grad_norm": 79.67159271240234,
+      "learning_rate": 7.091177012341131e-07,
+      "loss": 0.0121,
+      "step": 16800
+    },
+    {
+      "epoch": 3.0430457587267137,
+      "grad_norm": 11.743141174316406,
+      "learning_rate": 7.086572112727943e-07,
+      "loss": 0.0093,
+      "step": 16825
+    },
+    {
+      "epoch": 3.047567372038343,
+      "grad_norm": 18.43634605407715,
+      "learning_rate": 7.081967213114753e-07,
+      "loss": 0.0224,
+      "step": 16850
+    },
+    {
+      "epoch": 3.052088985349973,
+      "grad_norm": 0.045197684317827225,
+      "learning_rate": 7.077362313501566e-07,
+      "loss": 0.0062,
+      "step": 16875
+    },
+    {
+      "epoch": 3.0566105986616026,
+      "grad_norm": 0.12885162234306335,
+      "learning_rate": 7.072757413888378e-07,
+      "loss": 0.0214,
+      "step": 16900
+    },
+    {
+      "epoch": 3.061132211973232,
+      "grad_norm": 0.09277495741844177,
+      "learning_rate": 7.068152514275188e-07,
+      "loss": 0.0275,
+      "step": 16925
+    },
+    {
+      "epoch": 3.0656538252848615,
+      "grad_norm": 15.500121116638184,
+      "learning_rate": 7.063547614662e-07,
+      "loss": 0.0328,
+      "step": 16950
+    },
+    {
+      "epoch": 3.0701754385964914,
+      "grad_norm": 21.585752487182617,
+      "learning_rate": 7.058942715048811e-07,
+      "loss": 0.0304,
+      "step": 16975
+    },
+    {
+      "epoch": 3.074697051908121,
+      "grad_norm": 2.1149275302886963,
+      "learning_rate": 7.054337815435624e-07,
+      "loss": 0.0315,
+      "step": 17000
+    },
+    {
+      "epoch": 3.0792186652197504,
+      "grad_norm": 0.16735130548477173,
+      "learning_rate": 7.049732915822435e-07,
+      "loss": 0.0136,
+      "step": 17025
+    },
+    {
+      "epoch": 3.08374027853138,
+      "grad_norm": 0.7384225130081177,
+      "learning_rate": 7.045128016209246e-07,
+      "loss": 0.0059,
+      "step": 17050
+    },
+    {
+      "epoch": 3.0882618918430094,
+      "grad_norm": 0.19367511570453644,
+      "learning_rate": 7.040523116596058e-07,
+      "loss": 0.01,
+      "step": 17075
+    },
+    {
+      "epoch": 3.0927835051546393,
+      "grad_norm": 0.5572232604026794,
+      "learning_rate": 7.035918216982869e-07,
+      "loss": 0.0135,
+      "step": 17100
+    },
+    {
+      "epoch": 3.0973051184662688,
+      "grad_norm": 0.05385562777519226,
+      "learning_rate": 7.031313317369681e-07,
+      "loss": 0.0184,
+      "step": 17125
+    },
+    {
+      "epoch": 3.1018267317778982,
+      "grad_norm": 4.722483158111572,
+      "learning_rate": 7.026708417756493e-07,
+      "loss": 0.0039,
+      "step": 17150
+    },
+    {
+      "epoch": 3.1063483450895277,
+      "grad_norm": 0.6491204500198364,
+      "learning_rate": 7.022103518143304e-07,
+      "loss": 0.0147,
+      "step": 17175
+    },
+    {
+      "epoch": 3.1108699584011577,
+      "grad_norm": 0.014111626893281937,
+      "learning_rate": 7.017498618530115e-07,
+      "loss": 0.0119,
+      "step": 17200
+    },
+    {
+      "epoch": 3.115391571712787,
+      "grad_norm": 5.42165470123291,
+      "learning_rate": 7.012893718916927e-07,
+      "loss": 0.0116,
+      "step": 17225
+    },
+    {
+      "epoch": 3.1199131850244166,
+      "grad_norm": 7.389101982116699,
+      "learning_rate": 7.00828881930374e-07,
+      "loss": 0.0103,
+      "step": 17250
+    },
+    {
+      "epoch": 3.124434798336046,
+      "grad_norm": 0.3708292841911316,
+      "learning_rate": 7.00368391969055e-07,
+      "loss": 0.0074,
+      "step": 17275
+    },
+    {
+      "epoch": 3.128956411647676,
+      "grad_norm": 0.08553273230791092,
+      "learning_rate": 6.999079020077362e-07,
+      "loss": 0.0081,
+      "step": 17300
+    },
+    {
+      "epoch": 3.1334780249593055,
+      "grad_norm": 5.075300216674805,
+      "learning_rate": 6.994474120464174e-07,
+      "loss": 0.0204,
+      "step": 17325
+    },
+    {
+      "epoch": 3.137999638270935,
+      "grad_norm": 8.977065086364746,
+      "learning_rate": 6.989869220850984e-07,
+      "loss": 0.0095,
+      "step": 17350
+    },
+    {
+      "epoch": 3.1425212515825645,
+      "grad_norm": 3.1058037281036377,
+      "learning_rate": 6.985264321237797e-07,
+      "loss": 0.0618,
+      "step": 17375
+    },
+    {
+      "epoch": 3.1470428648941944,
+      "grad_norm": 4.081079959869385,
+      "learning_rate": 6.980659421624609e-07,
+      "loss": 0.0441,
+      "step": 17400
+    },
+    {
+      "epoch": 3.151564478205824,
+      "grad_norm": 32.8135986328125,
+      "learning_rate": 6.976054522011419e-07,
+      "loss": 0.0182,
+      "step": 17425
+    },
+    {
+      "epoch": 3.1560860915174533,
+      "grad_norm": 0.21175616979599,
+      "learning_rate": 6.971449622398231e-07,
+      "loss": 0.0125,
+      "step": 17450
+    },
+    {
+      "epoch": 3.160607704829083,
+      "grad_norm": 0.16236978769302368,
+      "learning_rate": 6.966844722785043e-07,
+      "loss": 0.0121,
+      "step": 17475
+    },
+    {
+      "epoch": 3.1651293181407127,
+      "grad_norm": 35.438865661621094,
+      "learning_rate": 6.962239823171855e-07,
+      "loss": 0.0111,
+      "step": 17500
+    },
+    {
+      "epoch": 3.1696509314523422,
+      "grad_norm": 0.0932527482509613,
+      "learning_rate": 6.957634923558666e-07,
+      "loss": 0.0118,
+      "step": 17525
+    },
+    {
+      "epoch": 3.1741725447639717,
+      "grad_norm": 9.384880065917969,
+      "learning_rate": 6.953030023945478e-07,
+      "loss": 0.0151,
+      "step": 17550
+    },
+    {
+      "epoch": 3.178694158075601,
+      "grad_norm": 0.24784910678863525,
+      "learning_rate": 6.948425124332289e-07,
+      "loss": 0.0106,
+      "step": 17575
+    },
+    {
+      "epoch": 3.183215771387231,
+      "grad_norm": 0.9011788964271545,
+      "learning_rate": 6.943820224719101e-07,
+      "loss": 0.0109,
+      "step": 17600
+    },
+    {
+      "epoch": 3.1877373846988606,
+      "grad_norm": 1.6620635986328125,
+      "learning_rate": 6.939215325105913e-07,
+      "loss": 0.0124,
+      "step": 17625
+    },
+    {
+      "epoch": 3.19225899801049,
+      "grad_norm": 0.09768769890069962,
+      "learning_rate": 6.934610425492724e-07,
+      "loss": 0.0131,
+      "step": 17650
+    },
+    {
+      "epoch": 3.1967806113221195,
+      "grad_norm": 3.7721073627471924,
+      "learning_rate": 6.930005525879535e-07,
+      "loss": 0.0315,
+      "step": 17675
+    },
+    {
+      "epoch": 3.2013022246337495,
+      "grad_norm": 4.030418872833252,
+      "learning_rate": 6.925400626266346e-07,
+      "loss": 0.0216,
+      "step": 17700
+    },
+    {
+      "epoch": 3.205823837945379,
+      "grad_norm": 16.64514923095703,
+      "learning_rate": 6.920795726653159e-07,
+      "loss": 0.0309,
+      "step": 17725
+    },
+    {
+      "epoch": 3.2103454512570084,
+      "grad_norm": 0.3508654236793518,
+      "learning_rate": 6.916190827039971e-07,
+      "loss": 0.0075,
+      "step": 17750
+    },
+    {
+      "epoch": 3.214867064568638,
+      "grad_norm": 0.19037003815174103,
+      "learning_rate": 6.911585927426781e-07,
+      "loss": 0.0547,
+      "step": 17775
+    },
+    {
+      "epoch": 3.219388677880268,
+      "grad_norm": 0.2579357624053955,
+      "learning_rate": 6.906981027813593e-07,
+      "loss": 0.0483,
+      "step": 17800
+    },
+    {
+      "epoch": 3.2239102911918973,
+      "grad_norm": 0.3658471405506134,
+      "learning_rate": 6.902376128200405e-07,
+      "loss": 0.0135,
+      "step": 17825
+    },
+    {
+      "epoch": 3.228431904503527,
+      "grad_norm": 7.055116176605225,
+      "learning_rate": 6.897771228587216e-07,
+      "loss": 0.0192,
+      "step": 17850
+    },
+    {
+      "epoch": 3.2329535178151563,
+      "grad_norm": 0.09685884416103363,
+      "learning_rate": 6.893166328974028e-07,
+      "loss": 0.0104,
+      "step": 17875
+    },
+    {
+      "epoch": 3.237475131126786,
+      "grad_norm": 0.2775970697402954,
+      "learning_rate": 6.88856142936084e-07,
+      "loss": 0.0112,
+      "step": 17900
+    },
+    {
+      "epoch": 3.2419967444384157,
+      "grad_norm": 0.04564272239804268,
+      "learning_rate": 6.883956529747651e-07,
+      "loss": 0.0167,
+      "step": 17925
+    },
+    {
+      "epoch": 3.246518357750045,
+      "grad_norm": 0.1531757116317749,
+      "learning_rate": 6.879351630134462e-07,
+      "loss": 0.0034,
+      "step": 17950
+    },
+    {
+      "epoch": 3.2510399710616746,
+      "grad_norm": 0.12003475427627563,
+      "learning_rate": 6.874746730521275e-07,
+      "loss": 0.0287,
+      "step": 17975
+    },
+    {
+      "epoch": 3.2555615843733046,
+      "grad_norm": 0.1408698409795761,
+      "learning_rate": 6.870141830908086e-07,
+      "loss": 0.0058,
+      "step": 18000
+    },
+    {
+      "epoch": 3.260083197684934,
+      "grad_norm": 0.21101155877113342,
+      "learning_rate": 6.865536931294897e-07,
+      "loss": 0.0224,
+      "step": 18025
+    },
+    {
+      "epoch": 3.2646048109965635,
+      "grad_norm": 3.5580930709838867,
+      "learning_rate": 6.861116227666237e-07,
+      "loss": 0.0336,
+      "step": 18050
+    },
+    {
+      "epoch": 3.269126424308193,
+      "grad_norm": 3.6865243911743164,
+      "learning_rate": 6.856511328053049e-07,
+      "loss": 0.0253,
+      "step": 18075
+    },
+    {
+      "epoch": 3.273648037619823,
+      "grad_norm": 0.7238378524780273,
+      "learning_rate": 6.85190642843986e-07,
+      "loss": 0.0028,
+      "step": 18100
+    },
+    {
+      "epoch": 3.2781696509314524,
+      "grad_norm": 0.032546211034059525,
+      "learning_rate": 6.847301528826671e-07,
+      "loss": 0.011,
+      "step": 18125
+    },
+    {
+      "epoch": 3.282691264243082,
+      "grad_norm": 0.2295057624578476,
+      "learning_rate": 6.842696629213482e-07,
+      "loss": 0.0241,
+      "step": 18150
+    },
+    {
+      "epoch": 3.2872128775547114,
+      "grad_norm": 0.37478190660476685,
+      "learning_rate": 6.838091729600295e-07,
+      "loss": 0.0487,
+      "step": 18175
+    },
+    {
+      "epoch": 3.2917344908663413,
+      "grad_norm": 2.0155792236328125,
+      "learning_rate": 6.833486829987106e-07,
+      "loss": 0.0273,
+      "step": 18200
+    },
+    {
+      "epoch": 3.2962561041779708,
+      "grad_norm": 1.1389836072921753,
+      "learning_rate": 6.828881930373917e-07,
+      "loss": 0.0132,
+      "step": 18225
+    },
+    {
+      "epoch": 3.3007777174896002,
+      "grad_norm": 7.253364086151123,
+      "learning_rate": 6.824277030760729e-07,
+      "loss": 0.0085,
+      "step": 18250
+    },
+    {
+      "epoch": 3.3052993308012297,
+      "grad_norm": 17.829906463623047,
+      "learning_rate": 6.819672131147541e-07,
+      "loss": 0.0189,
+      "step": 18275
+    },
+    {
+      "epoch": 3.3098209441128597,
+      "grad_norm": 3.49690842628479,
+      "learning_rate": 6.815067231534352e-07,
+      "loss": 0.01,
+      "step": 18300
+    },
+    {
+      "epoch": 3.314342557424489,
+      "grad_norm": 0.6538553237915039,
+      "learning_rate": 6.810462331921164e-07,
+      "loss": 0.0097,
+      "step": 18325
+    },
+    {
+      "epoch": 3.3188641707361186,
+      "grad_norm": 29.16781234741211,
+      "learning_rate": 6.805857432307976e-07,
+      "loss": 0.0162,
+      "step": 18350
+    },
+    {
+      "epoch": 3.323385784047748,
+      "grad_norm": 0.38442912697792053,
+      "learning_rate": 6.801252532694786e-07,
+      "loss": 0.013,
+      "step": 18375
+    },
+    {
+      "epoch": 3.327907397359378,
+      "grad_norm": 16.521053314208984,
+      "learning_rate": 6.796647633081598e-07,
+      "loss": 0.0153,
+      "step": 18400
+    },
+    {
+      "epoch": 3.3324290106710075,
+      "grad_norm": 1.947344422340393,
+      "learning_rate": 6.792042733468411e-07,
+      "loss": 0.0088,
+      "step": 18425
+    },
+    {
+      "epoch": 3.336950623982637,
+      "grad_norm": 0.08833196014165878,
+      "learning_rate": 6.787437833855221e-07,
+      "loss": 0.0169,
+      "step": 18450
+    },
+    {
+      "epoch": 3.3414722372942665,
+      "grad_norm": 0.04213396832346916,
+      "learning_rate": 6.782832934242033e-07,
+      "loss": 0.0048,
+      "step": 18475
+    },
+    {
+      "epoch": 3.345993850605896,
+      "grad_norm": 14.339350700378418,
+      "learning_rate": 6.778228034628845e-07,
+      "loss": 0.0322,
+      "step": 18500
+    },
+    {
+      "epoch": 3.350515463917526,
+      "grad_norm": 2.3550896644592285,
+      "learning_rate": 6.773623135015656e-07,
+      "loss": 0.0079,
+      "step": 18525
+    },
+    {
+      "epoch": 3.3550370772291553,
+      "grad_norm": 0.04377694055438042,
+      "learning_rate": 6.769018235402468e-07,
+      "loss": 0.0408,
+      "step": 18550
+    },
+    {
+      "epoch": 3.359558690540785,
+      "grad_norm": 66.76377868652344,
+      "learning_rate": 6.76441333578928e-07,
+      "loss": 0.0646,
+      "step": 18575
+    },
+    {
+      "epoch": 3.3640803038524147,
+      "grad_norm": 24.474994659423828,
+      "learning_rate": 6.759808436176091e-07,
+      "loss": 0.0251,
+      "step": 18600
+    },
+    {
+      "epoch": 3.3686019171640442,
+      "grad_norm": 4.777419090270996,
+      "learning_rate": 6.755203536562902e-07,
+      "loss": 0.0127,
+      "step": 18625
+    },
+    {
+      "epoch": 3.3731235304756737,
+      "grad_norm": 1.3446674346923828,
+      "learning_rate": 6.750598636949714e-07,
+      "loss": 0.0246,
+      "step": 18650
+    },
+    {
+      "epoch": 3.377645143787303,
+      "grad_norm": 5.482788562774658,
+      "learning_rate": 6.745993737336526e-07,
+      "loss": 0.0141,
+      "step": 18675
+    },
+    {
+      "epoch": 3.3821667570989327,
+      "grad_norm": 1.5254578590393066,
+      "learning_rate": 6.741388837723337e-07,
+      "loss": 0.019,
+      "step": 18700
+    },
+    {
+      "epoch": 3.3866883704105626,
+      "grad_norm": 2.620288372039795,
+      "learning_rate": 6.736783938110149e-07,
+      "loss": 0.0257,
+      "step": 18725
+    },
+    {
+      "epoch": 3.391209983722192,
+      "grad_norm": 5.762913703918457,
+      "learning_rate": 6.73217903849696e-07,
+      "loss": 0.0033,
+      "step": 18750
+    },
+    {
+      "epoch": 3.3957315970338215,
+      "grad_norm": 5.787083625793457,
+      "learning_rate": 6.727574138883773e-07,
+      "loss": 0.0093,
+      "step": 18775
+    },
+    {
+      "epoch": 3.4002532103454515,
+      "grad_norm": 2.400695562362671,
+      "learning_rate": 6.722969239270584e-07,
+      "loss": 0.0103,
+      "step": 18800
+    },
+    {
+      "epoch": 3.404774823657081,
+      "grad_norm": 0.05644823983311653,
+      "learning_rate": 6.718364339657395e-07,
+      "loss": 0.0058,
+      "step": 18825
+    },
+    {
+      "epoch": 3.4092964369687104,
+      "grad_norm": 0.2891055941581726,
+      "learning_rate": 6.713759440044207e-07,
+      "loss": 0.0188,
+      "step": 18850
+    },
+    {
+      "epoch": 3.41381805028034,
+      "grad_norm": 0.8349182605743408,
+      "learning_rate": 6.709154540431017e-07,
+      "loss": 0.0295,
+      "step": 18875
+    },
+    {
+      "epoch": 3.4183396635919694,
+      "grad_norm": 0.06508205085992813,
+      "learning_rate": 6.70454964081783e-07,
+      "loss": 0.0079,
+      "step": 18900
+    },
+    {
+      "epoch": 3.4228612769035993,
+      "grad_norm": 8.653030395507812,
+      "learning_rate": 6.699944741204642e-07,
+      "loss": 0.0152,
+      "step": 18925
+    },
+    {
+      "epoch": 3.427382890215229,
+      "grad_norm": 0.2311927080154419,
+      "learning_rate": 6.695339841591454e-07,
+      "loss": 0.0414,
+      "step": 18950
+    },
+    {
+      "epoch": 3.4319045035268583,
+      "grad_norm": 55.472957611083984,
+      "learning_rate": 6.690734941978264e-07,
+      "loss": 0.0386,
+      "step": 18975
+    },
+    {
+      "epoch": 3.436426116838488,
+      "grad_norm": 0.4529309570789337,
+      "learning_rate": 6.686130042365076e-07,
+      "loss": 0.0494,
+      "step": 19000
+    },
+    {
+      "epoch": 3.4409477301501177,
+      "grad_norm": 1.5187815427780151,
+      "learning_rate": 6.681525142751889e-07,
+      "loss": 0.0166,
+      "step": 19025
+    },
+    {
+      "epoch": 3.445469343461747,
+      "grad_norm": 0.964755654335022,
+      "learning_rate": 6.676920243138699e-07,
+      "loss": 0.0076,
+      "step": 19050
+    },
+    {
+      "epoch": 3.4499909567733766,
+      "grad_norm": 22.352828979492188,
+      "learning_rate": 6.672315343525511e-07,
+      "loss": 0.019,
+      "step": 19075
+    },
+    {
+      "epoch": 3.454512570085006,
+      "grad_norm": 0.08743809163570404,
+      "learning_rate": 6.667710443912322e-07,
+      "loss": 0.0132,
+      "step": 19100
+    },
+    {
+      "epoch": 3.459034183396636,
+      "grad_norm": 0.2849852740764618,
+      "learning_rate": 6.663105544299133e-07,
+      "loss": 0.0206,
+      "step": 19125
+    },
+    {
+      "epoch": 3.4635557967082655,
+      "grad_norm": 1.0750819444656372,
+      "learning_rate": 6.658500644685946e-07,
+      "loss": 0.0206,
+      "step": 19150
+    },
+    {
+      "epoch": 3.468077410019895,
+      "grad_norm": 2.5783636569976807,
+      "learning_rate": 6.653895745072757e-07,
+      "loss": 0.0086,
+      "step": 19175
+    },
+    {
+      "epoch": 3.472599023331525,
+      "grad_norm": 3.71582293510437,
+      "learning_rate": 6.649290845459569e-07,
+      "loss": 0.0055,
+      "step": 19200
+    },
+    {
+      "epoch": 3.4771206366431544,
+      "grad_norm": 0.3580998182296753,
+      "learning_rate": 6.64468594584638e-07,
+      "loss": 0.0128,
+      "step": 19225
+    },
+    {
+      "epoch": 3.481642249954784,
+      "grad_norm": 20.23494529724121,
+      "learning_rate": 6.640081046233191e-07,
+      "loss": 0.032,
+      "step": 19250
+    },
+    {
+      "epoch": 3.4861638632664134,
+      "grad_norm": 0.634772539138794,
+      "learning_rate": 6.635476146620004e-07,
+      "loss": 0.0041,
+      "step": 19275
+    },
+    {
+      "epoch": 3.490685476578043,
+      "grad_norm": 9.67199993133545,
+      "learning_rate": 6.630871247006815e-07,
+      "loss": 0.0264,
+      "step": 19300
+    },
+    {
+      "epoch": 3.4952070898896728,
+      "grad_norm": 0.13094273209571838,
+      "learning_rate": 6.626266347393626e-07,
+      "loss": 0.0177,
+      "step": 19325
+    },
+    {
+      "epoch": 3.4997287032013022,
+      "grad_norm": 57.88798141479492,
+      "learning_rate": 6.621661447780438e-07,
+      "loss": 0.0171,
+      "step": 19350
+    },
+    {
+      "epoch": 3.5042503165129317,
+      "grad_norm": 0.32330283522605896,
+      "learning_rate": 6.617056548167249e-07,
+      "loss": 0.0333,
+      "step": 19375
+    },
+    {
+      "epoch": 3.5087719298245617,
+      "grad_norm": 2.477999210357666,
+      "learning_rate": 6.612451648554061e-07,
+      "loss": 0.0257,
+      "step": 19400
+    },
+    {
+      "epoch": 3.513293543136191,
+      "grad_norm": 4.913309097290039,
+      "learning_rate": 6.607846748940873e-07,
+      "loss": 0.014,
+      "step": 19425
+    },
+    {
+      "epoch": 3.5178151564478206,
+      "grad_norm": 1.5322345495224,
+      "learning_rate": 6.603241849327685e-07,
+      "loss": 0.0214,
+      "step": 19450
+    },
+    {
+      "epoch": 3.52233676975945,
+      "grad_norm": 0.07133228331804276,
+      "learning_rate": 6.598636949714495e-07,
+      "loss": 0.0148,
+      "step": 19475
+    },
+    {
+      "epoch": 3.5268583830710796,
+      "grad_norm": 13.74171257019043,
+      "learning_rate": 6.594032050101308e-07,
+      "loss": 0.0121,
+      "step": 19500
+    },
+    {
+      "epoch": 3.5313799963827095,
+      "grad_norm": 27.33115577697754,
+      "learning_rate": 6.58942715048812e-07,
+      "loss": 0.013,
+      "step": 19525
+    },
+    {
+      "epoch": 3.535901609694339,
+      "grad_norm": 0.22666044533252716,
+      "learning_rate": 6.58482225087493e-07,
+      "loss": 0.0145,
+      "step": 19550
+    },
+    {
+      "epoch": 3.5404232230059685,
+      "grad_norm": 13.335356712341309,
+      "learning_rate": 6.580217351261742e-07,
+      "loss": 0.0061,
+      "step": 19575
+    },
+    {
+      "epoch": 3.5449448363175984,
+      "grad_norm": 4.350337505340576,
+      "learning_rate": 6.575612451648554e-07,
+      "loss": 0.0038,
+      "step": 19600
+    },
+    {
+      "epoch": 3.549466449629228,
+      "grad_norm": 0.13537144660949707,
+      "learning_rate": 6.571007552035366e-07,
+      "loss": 0.0125,
+      "step": 19625
+    },
+    {
+      "epoch": 3.5539880629408573,
+      "grad_norm": 56.16394805908203,
+      "learning_rate": 6.566402652422177e-07,
+      "loss": 0.0213,
+      "step": 19650
+    },
+    {
+      "epoch": 3.558509676252487,
+      "grad_norm": 0.10288344323635101,
+      "learning_rate": 6.561797752808989e-07,
+      "loss": 0.0156,
+      "step": 19675
+    },
+    {
+      "epoch": 3.5630312895641163,
+      "grad_norm": 3.5337467193603516,
+      "learning_rate": 6.5571928531958e-07,
+      "loss": 0.019,
+      "step": 19700
+    },
+    {
+      "epoch": 3.5675529028757462,
+      "grad_norm": 0.20034463703632355,
+      "learning_rate": 6.552587953582611e-07,
+      "loss": 0.05,
+      "step": 19725
+    },
+    {
+      "epoch": 3.5720745161873757,
+      "grad_norm": 0.22690874338150024,
+      "learning_rate": 6.547983053969424e-07,
+      "loss": 0.0217,
+      "step": 19750
+    },
+    {
+      "epoch": 3.576596129499005,
+      "grad_norm": 26.994596481323242,
+      "learning_rate": 6.543378154356235e-07,
+      "loss": 0.0343,
+      "step": 19775
+    },
+    {
+      "epoch": 3.581117742810635,
+      "grad_norm": 5.310791492462158,
+      "learning_rate": 6.538773254743046e-07,
+      "loss": 0.0325,
+      "step": 19800
+    },
+    {
+      "epoch": 3.5856393561222646,
+      "grad_norm": 1.536670207977295,
+      "learning_rate": 6.534168355129857e-07,
+      "loss": 0.0057,
+      "step": 19825
+    },
+    {
+      "epoch": 3.590160969433894,
+      "grad_norm": 4.136430263519287,
+      "learning_rate": 6.529563455516669e-07,
+      "loss": 0.0292,
+      "step": 19850
+    },
+    {
+      "epoch": 3.5946825827455235,
+      "grad_norm": 0.5038244724273682,
+      "learning_rate": 6.524958555903482e-07,
+      "loss": 0.0159,
+      "step": 19875
+    },
+    {
+      "epoch": 3.599204196057153,
+      "grad_norm": 0.03355490043759346,
+      "learning_rate": 6.520353656290292e-07,
+      "loss": 0.0137,
+      "step": 19900
+    },
+    {
+      "epoch": 3.6037258093687825,
+      "grad_norm": 1.1418379545211792,
+      "learning_rate": 6.515748756677104e-07,
+      "loss": 0.0043,
+      "step": 19925
+    },
+    {
+      "epoch": 3.6082474226804124,
+      "grad_norm": 56.00255584716797,
+      "learning_rate": 6.511143857063916e-07,
+      "loss": 0.034,
+      "step": 19950
+    },
+    {
+      "epoch": 3.612769035992042,
+      "grad_norm": 0.5989301204681396,
+      "learning_rate": 6.506538957450726e-07,
+      "loss": 0.0094,
+      "step": 19975
+    },
+    {
+      "epoch": 3.6172906493036714,
+      "grad_norm": 5.834647178649902,
+      "learning_rate": 6.501934057837539e-07,
+      "loss": 0.0104,
+      "step": 20000
+    },
+    {
+      "epoch": 3.6218122626153013,
+      "grad_norm": 0.1490633189678192,
+      "learning_rate": 6.497329158224351e-07,
+      "loss": 0.0167,
+      "step": 20025
+    },
+    {
+      "epoch": 3.626333875926931,
+      "grad_norm": 18.232946395874023,
+      "learning_rate": 6.492724258611161e-07,
+      "loss": 0.0139,
+      "step": 20050
+    },
+    {
+      "epoch": 3.6308554892385603,
+      "grad_norm": 0.08834528177976608,
+      "learning_rate": 6.488119358997973e-07,
+      "loss": 0.0253,
+      "step": 20075
+    },
+    {
+      "epoch": 3.6353771025501898,
+      "grad_norm": 0.3102453351020813,
+      "learning_rate": 6.483698655369313e-07,
+      "loss": 0.0146,
+      "step": 20100
+    },
+    {
+      "epoch": 3.6398987158618192,
+      "grad_norm": 0.26993054151535034,
+      "learning_rate": 6.479093755756125e-07,
+      "loss": 0.0187,
+      "step": 20125
+    },
+    {
+      "epoch": 3.644420329173449,
+      "grad_norm": 0.2688346803188324,
+      "learning_rate": 6.474488856142935e-07,
+      "loss": 0.0304,
+      "step": 20150
+    },
+    {
+      "epoch": 3.6489419424850786,
+      "grad_norm": 0.11477080732584,
+      "learning_rate": 6.469883956529747e-07,
+      "loss": 0.0515,
+      "step": 20175
+    },
+    {
+      "epoch": 3.653463555796708,
+      "grad_norm": 2.2574214935302734,
+      "learning_rate": 6.46527905691656e-07,
+      "loss": 0.0248,
+      "step": 20200
+    },
+    {
+      "epoch": 3.657985169108338,
+      "grad_norm": 0.9582886099815369,
+      "learning_rate": 6.460674157303371e-07,
+      "loss": 0.0128,
+      "step": 20225
+    },
+    {
+      "epoch": 3.6625067824199675,
+      "grad_norm": 17.577070236206055,
+      "learning_rate": 6.456069257690182e-07,
+      "loss": 0.0189,
+      "step": 20250
+    },
+    {
+      "epoch": 3.667028395731597,
+      "grad_norm": 8.872756004333496,
+      "learning_rate": 6.451464358076993e-07,
+      "loss": 0.0113,
+      "step": 20275
+    },
+    {
+      "epoch": 3.6715500090432265,
+      "grad_norm": 0.408764123916626,
+      "learning_rate": 6.446859458463805e-07,
+      "loss": 0.0155,
+      "step": 20300
+    },
+    {
+      "epoch": 3.676071622354856,
+      "grad_norm": 0.4208977520465851,
+      "learning_rate": 6.442254558850617e-07,
+      "loss": 0.0044,
+      "step": 20325
+    },
+    {
+      "epoch": 3.680593235666486,
+      "grad_norm": 0.39012089371681213,
+      "learning_rate": 6.437649659237428e-07,
+      "loss": 0.0161,
+      "step": 20350
+    },
+    {
+      "epoch": 3.6851148489781154,
+      "grad_norm": 0.0061714984476566315,
+      "learning_rate": 6.43304475962424e-07,
+      "loss": 0.0113,
+      "step": 20375
+    },
+    {
+      "epoch": 3.689636462289745,
+      "grad_norm": 0.03614750877022743,
+      "learning_rate": 6.428439860011051e-07,
+      "loss": 0.0049,
+      "step": 20400
+    },
+    {
+      "epoch": 3.6941580756013748,
+      "grad_norm": 0.015124999918043613,
+      "learning_rate": 6.423834960397862e-07,
+      "loss": 0.0111,
+      "step": 20425
+    },
+    {
+      "epoch": 3.6986796889130042,
+      "grad_norm": 1.7233740091323853,
+      "learning_rate": 6.419230060784675e-07,
+      "loss": 0.0185,
+      "step": 20450
+    },
+    {
+      "epoch": 3.7032013022246337,
+      "grad_norm": 1.9191359281539917,
+      "learning_rate": 6.414625161171487e-07,
+      "loss": 0.0043,
+      "step": 20475
+    },
+    {
+      "epoch": 3.707722915536263,
+      "grad_norm": 0.6458703875541687,
+      "learning_rate": 6.410020261558297e-07,
+      "loss": 0.0183,
+      "step": 20500
+    },
+    {
+      "epoch": 3.7122445288478927,
+      "grad_norm": 0.0197773240506649,
+      "learning_rate": 6.405415361945109e-07,
+      "loss": 0.0244,
+      "step": 20525
+    },
+    {
+      "epoch": 3.7167661421595226,
+      "grad_norm": 0.8223174214363098,
+      "learning_rate": 6.400810462331921e-07,
+      "loss": 0.0311,
+      "step": 20550
+    },
+    {
+      "epoch": 3.721287755471152,
+      "grad_norm": 42.07274627685547,
+      "learning_rate": 6.396205562718732e-07,
+      "loss": 0.0578,
+      "step": 20575
+    },
+    {
+      "epoch": 3.7258093687827816,
+      "grad_norm": 0.2349138706922531,
+      "learning_rate": 6.391600663105544e-07,
+      "loss": 0.0253,
+      "step": 20600
+    },
+    {
+      "epoch": 3.7303309820944115,
+      "grad_norm": 0.18693946301937103,
+      "learning_rate": 6.386995763492356e-07,
+      "loss": 0.0104,
+      "step": 20625
+    },
+    {
+      "epoch": 3.734852595406041,
+      "grad_norm": 3.6481404304504395,
+      "learning_rate": 6.382390863879167e-07,
+      "loss": 0.0067,
+      "step": 20650
+    },
+    {
+      "epoch": 3.7393742087176705,
+      "grad_norm": 0.06991686671972275,
+      "learning_rate": 6.377785964265979e-07,
+      "loss": 0.0086,
+      "step": 20675
+    },
+    {
+      "epoch": 3.7438958220293,
+      "grad_norm": 0.06903531402349472,
+      "learning_rate": 6.373181064652791e-07,
+      "loss": 0.012,
+      "step": 20700
+    },
+    {
+      "epoch": 3.7484174353409294,
+      "grad_norm": 0.3118179142475128,
+      "learning_rate": 6.368576165039602e-07,
+      "loss": 0.0145,
+      "step": 20725
+    },
+    {
+      "epoch": 3.7529390486525593,
+      "grad_norm": 0.13410675525665283,
+      "learning_rate": 6.363971265426413e-07,
+      "loss": 0.0116,
+      "step": 20750
+    },
+    {
+      "epoch": 3.757460661964189,
+      "grad_norm": 6.128607273101807,
+      "learning_rate": 6.359366365813225e-07,
+      "loss": 0.0147,
+      "step": 20775
+    },
+    {
+      "epoch": 3.7619822752758183,
+      "grad_norm": 16.329174041748047,
+      "learning_rate": 6.354761466200037e-07,
+      "loss": 0.0289,
+      "step": 20800
+    },
+    {
+      "epoch": 3.7665038885874482,
+      "grad_norm": 0.022927312180399895,
+      "learning_rate": 6.350156566586848e-07,
+      "loss": 0.0194,
+      "step": 20825
+    },
+    {
+      "epoch": 3.7710255018990777,
+      "grad_norm": 0.08002068102359772,
+      "learning_rate": 6.34555166697366e-07,
+      "loss": 0.0276,
+      "step": 20850
+    },
+    {
+      "epoch": 3.775547115210707,
+      "grad_norm": 6.620906352996826,
+      "learning_rate": 6.340946767360471e-07,
+      "loss": 0.0303,
+      "step": 20875
+    },
+    {
+      "epoch": 3.7800687285223367,
+      "grad_norm": 20.082733154296875,
+      "learning_rate": 6.336341867747283e-07,
+      "loss": 0.0285,
+      "step": 20900
+    },
+    {
+      "epoch": 3.784590341833966,
+      "grad_norm": 0.4795994758605957,
+      "learning_rate": 6.331736968134095e-07,
+      "loss": 0.0164,
+      "step": 20925
+    },
+    {
+      "epoch": 3.789111955145596,
+      "grad_norm": 0.2975751757621765,
+      "learning_rate": 6.327132068520906e-07,
+      "loss": 0.0246,
+      "step": 20950
+    },
+    {
+      "epoch": 3.7936335684572255,
+      "grad_norm": 1.9404077529907227,
+      "learning_rate": 6.322527168907718e-07,
+      "loss": 0.0274,
+      "step": 20975
+    },
+    {
+      "epoch": 3.798155181768855,
+      "grad_norm": 1.5092577934265137,
+      "learning_rate": 6.317922269294528e-07,
+      "loss": 0.0286,
+      "step": 21000
+    },
+    {
+      "epoch": 3.802676795080485,
+      "grad_norm": 3.025921106338501,
+      "learning_rate": 6.31331736968134e-07,
+      "loss": 0.0139,
+      "step": 21025
+    },
+    {
+      "epoch": 3.8071984083921144,
+      "grad_norm": 0.11709149926900864,
+      "learning_rate": 6.308712470068153e-07,
+      "loss": 0.0067,
+      "step": 21050
+    },
+    {
+      "epoch": 3.811720021703744,
+      "grad_norm": 3.3579108715057373,
+      "learning_rate": 6.304107570454963e-07,
+      "loss": 0.0077,
+      "step": 21075
+    },
+    {
+      "epoch": 3.8162416350153734,
+      "grad_norm": 4.9033050537109375,
+      "learning_rate": 6.299502670841775e-07,
+      "loss": 0.0124,
+      "step": 21100
+    },
+    {
+      "epoch": 3.820763248327003,
+      "grad_norm": 0.054667115211486816,
+      "learning_rate": 6.294897771228587e-07,
+      "loss": 0.0147,
+      "step": 21125
+    },
+    {
+      "epoch": 3.825284861638633,
+      "grad_norm": 20.414718627929688,
+      "learning_rate": 6.290292871615398e-07,
+      "loss": 0.018,
+      "step": 21150
+    },
+    {
+      "epoch": 3.8298064749502623,
+      "grad_norm": 0.12169066816568375,
+      "learning_rate": 6.28568797200221e-07,
+      "loss": 0.0266,
+      "step": 21175
+    },
+    {
+      "epoch": 3.8343280882618918,
+      "grad_norm": 0.21765799820423126,
+      "learning_rate": 6.281083072389022e-07,
+      "loss": 0.0061,
+      "step": 21200
+    },
+    {
+      "epoch": 3.8388497015735217,
+      "grad_norm": 0.43716397881507874,
+      "learning_rate": 6.276478172775833e-07,
+      "loss": 0.0114,
+      "step": 21225
+    },
+    {
+      "epoch": 3.843371314885151,
+      "grad_norm": 1.5071337223052979,
+      "learning_rate": 6.271873273162644e-07,
+      "loss": 0.0111,
+      "step": 21250
+    },
+    {
+      "epoch": 3.8478929281967806,
+      "grad_norm": 1.3820738792419434,
+      "learning_rate": 6.267268373549456e-07,
+      "loss": 0.0088,
+      "step": 21275
+    },
+    {
+      "epoch": 3.85241454150841,
+      "grad_norm": 18.006547927856445,
+      "learning_rate": 6.262663473936268e-07,
+      "loss": 0.0628,
+      "step": 21300
+    },
+    {
+      "epoch": 3.8569361548200396,
+      "grad_norm": 0.09640432149171829,
+      "learning_rate": 6.25805857432308e-07,
+      "loss": 0.0324,
+      "step": 21325
+    },
+    {
+      "epoch": 3.8614577681316695,
+      "grad_norm": 86.30760192871094,
+      "learning_rate": 6.253453674709891e-07,
+      "loss": 0.0348,
+      "step": 21350
+    },
+    {
+      "epoch": 3.865979381443299,
+      "grad_norm": 20.232952117919922,
+      "learning_rate": 6.248848775096702e-07,
+      "loss": 0.0564,
+      "step": 21375
+    },
+    {
+      "epoch": 3.8705009947549285,
+      "grad_norm": 0.07119341939687729,
+      "learning_rate": 6.244243875483515e-07,
+      "loss": 0.0311,
+      "step": 21400
+    },
+    {
+      "epoch": 3.8750226080665584,
+      "grad_norm": 1.7042196989059448,
+      "learning_rate": 6.239638975870326e-07,
+      "loss": 0.0114,
+      "step": 21425
+    },
+    {
+      "epoch": 3.879544221378188,
+      "grad_norm": 0.36282435059547424,
+      "learning_rate": 6.235034076257137e-07,
+      "loss": 0.0227,
+      "step": 21450
+    },
+    {
+      "epoch": 3.8840658346898174,
+      "grad_norm": 0.15856370329856873,
+      "learning_rate": 6.230429176643949e-07,
+      "loss": 0.0134,
+      "step": 21475
+    },
+    {
+      "epoch": 3.888587448001447,
+      "grad_norm": 22.399805068969727,
+      "learning_rate": 6.22582427703076e-07,
+      "loss": 0.0165,
+      "step": 21500
+    },
+    {
+      "epoch": 3.8931090613130763,
+      "grad_norm": 1.0881842374801636,
+      "learning_rate": 6.221219377417572e-07,
+      "loss": 0.0081,
+      "step": 21525
+    },
+    {
+      "epoch": 3.8976306746247062,
+      "grad_norm": 0.025745024904608727,
+      "learning_rate": 6.216614477804384e-07,
+      "loss": 0.012,
+      "step": 21550
+    },
+    {
+      "epoch": 3.9021522879363357,
+      "grad_norm": 0.05021649971604347,
+      "learning_rate": 6.212009578191196e-07,
+      "loss": 0.0035,
+      "step": 21575
+    },
+    {
+      "epoch": 3.906673901247965,
+      "grad_norm": 0.08221199363470078,
+      "learning_rate": 6.207404678578006e-07,
+      "loss": 0.0131,
+      "step": 21600
+    },
+    {
+      "epoch": 3.911195514559595,
+      "grad_norm": 0.9112662672996521,
+      "learning_rate": 6.202799778964818e-07,
+      "loss": 0.0087,
+      "step": 21625
+    },
+    {
+      "epoch": 3.9157171278712246,
+      "grad_norm": 3.2704386711120605,
+      "learning_rate": 6.198194879351631e-07,
+      "loss": 0.0057,
+      "step": 21650
+    },
+    {
+      "epoch": 3.920238741182854,
+      "grad_norm": 0.2193220853805542,
+      "learning_rate": 6.193589979738441e-07,
+      "loss": 0.0154,
+      "step": 21675
+    },
+    {
+      "epoch": 3.9247603544944836,
+      "grad_norm": 1.7677456140518188,
+      "learning_rate": 6.188985080125253e-07,
+      "loss": 0.0338,
+      "step": 21700
+    },
+    {
+      "epoch": 3.929281967806113,
+      "grad_norm": 4.569864273071289,
+      "learning_rate": 6.184380180512065e-07,
+      "loss": 0.045,
+      "step": 21725
+    },
+    {
+      "epoch": 3.933803581117743,
+      "grad_norm": 1.5949956178665161,
+      "learning_rate": 6.179775280898875e-07,
+      "loss": 0.0126,
+      "step": 21750
+    },
+    {
+      "epoch": 3.9383251944293725,
+      "grad_norm": 0.5718483924865723,
+      "learning_rate": 6.175170381285688e-07,
+      "loss": 0.0407,
+      "step": 21775
+    },
+    {
+      "epoch": 3.942846807741002,
+      "grad_norm": 0.3438205122947693,
+      "learning_rate": 6.1705654816725e-07,
+      "loss": 0.0312,
+      "step": 21800
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.23237770795822144,
+      "learning_rate": 6.165960582059311e-07,
+      "loss": 0.0075,
+      "step": 21825
+    },
+    {
+      "epoch": 3.9518900343642613,
+      "grad_norm": 0.4722042381763458,
+      "learning_rate": 6.161355682446122e-07,
+      "loss": 0.0142,
+      "step": 21850
+    },
+    {
+      "epoch": 3.956411647675891,
+      "grad_norm": 3.872162342071533,
+      "learning_rate": 6.156750782832933e-07,
+      "loss": 0.0061,
+      "step": 21875
+    },
+    {
+      "epoch": 3.9609332609875203,
+      "grad_norm": 12.684093475341797,
+      "learning_rate": 6.152145883219746e-07,
+      "loss": 0.0123,
+      "step": 21900
+    },
+    {
+      "epoch": 3.96545487429915,
+      "grad_norm": 1.105630874633789,
+      "learning_rate": 6.147540983606557e-07,
+      "loss": 0.0066,
+      "step": 21925
+    },
+    {
+      "epoch": 3.9699764876107793,
+      "grad_norm": 0.050917476415634155,
+      "learning_rate": 6.142936083993368e-07,
+      "loss": 0.0095,
+      "step": 21950
+    },
+    {
+      "epoch": 3.974498100922409,
+      "grad_norm": 6.288554668426514,
+      "learning_rate": 6.13833118438018e-07,
+      "loss": 0.0039,
+      "step": 21975
+    },
+    {
+      "epoch": 3.9790197142340387,
+      "grad_norm": 0.09361649304628372,
+      "learning_rate": 6.133726284766991e-07,
+      "loss": 0.0174,
+      "step": 22000
+    },
+    {
+      "epoch": 3.983541327545668,
+      "grad_norm": 0.36472177505493164,
+      "learning_rate": 6.129121385153803e-07,
+      "loss": 0.012,
+      "step": 22025
+    },
+    {
+      "epoch": 3.988062940857298,
+      "grad_norm": 0.4556725025177002,
+      "learning_rate": 6.124516485540615e-07,
+      "loss": 0.012,
+      "step": 22050
+    },
+    {
+      "epoch": 3.9925845541689275,
+      "grad_norm": 4.3723225593566895,
+      "learning_rate": 6.119911585927427e-07,
+      "loss": 0.0076,
+      "step": 22075
+    },
+    {
+      "epoch": 3.997106167480557,
+      "grad_norm": 0.04663983732461929,
+      "learning_rate": 6.115490882298766e-07,
+      "loss": 0.0322,
+      "step": 22100
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.3087974786758423,
+      "eval_runtime": 8497.8744,
+      "eval_samples_per_second": 1.117,
+      "eval_steps_per_second": 0.14,
+      "eval_wer": 0.10576553491351698,
+      "step": 22116
+    },
+    {
+      "epoch": 4.0016277807921865,
+      "grad_norm": 0.3466642498970032,
+      "learning_rate": 6.110885982685577e-07,
+      "loss": 0.0404,
+      "step": 22125
+    },
+    {
+      "epoch": 4.006149394103816,
+      "grad_norm": 0.15252432227134705,
+      "learning_rate": 6.106281083072389e-07,
+      "loss": 0.0096,
+      "step": 22150
+    },
+    {
+      "epoch": 4.0106710074154455,
+      "grad_norm": 19.444740295410156,
+      "learning_rate": 6.101676183459201e-07,
+      "loss": 0.0159,
+      "step": 22175
+    },
+    {
+      "epoch": 4.015192620727076,
+      "grad_norm": 10.760107040405273,
+      "learning_rate": 6.097071283846011e-07,
+      "loss": 0.0068,
+      "step": 22200
+    },
+    {
+      "epoch": 4.019714234038705,
+      "grad_norm": 0.07839302718639374,
+      "learning_rate": 6.092466384232824e-07,
+      "loss": 0.0026,
+      "step": 22225
+    },
+    {
+      "epoch": 4.024235847350335,
+      "grad_norm": 0.3171479105949402,
+      "learning_rate": 6.087861484619636e-07,
+      "loss": 0.0058,
+      "step": 22250
+    },
+    {
+      "epoch": 4.028757460661964,
+      "grad_norm": 0.2635548412799835,
+      "learning_rate": 6.083256585006446e-07,
+      "loss": 0.0104,
+      "step": 22275
+    },
+    {
+      "epoch": 4.033279073973594,
+      "grad_norm": 0.7347187995910645,
+      "learning_rate": 6.078651685393258e-07,
+      "loss": 0.0041,
+      "step": 22300
+    },
+    {
+      "epoch": 4.037800687285223,
+      "grad_norm": 0.03082045540213585,
+      "learning_rate": 6.074046785780069e-07,
+      "loss": 0.0023,
+      "step": 22325
+    },
+    {
+      "epoch": 4.042322300596853,
+      "grad_norm": 0.030160456895828247,
+      "learning_rate": 6.069441886166882e-07,
+      "loss": 0.0023,
+      "step": 22350
+    },
+    {
+      "epoch": 4.046843913908482,
+      "grad_norm": 22.204673767089844,
+      "learning_rate": 6.064836986553693e-07,
+      "loss": 0.0166,
+      "step": 22375
+    },
+    {
+      "epoch": 4.051365527220113,
+      "grad_norm": 5.911830425262451,
+      "learning_rate": 6.060232086940504e-07,
+      "loss": 0.0026,
+      "step": 22400
+    },
+    {
+      "epoch": 4.055887140531742,
+      "grad_norm": 0.5570561289787292,
+      "learning_rate": 6.055627187327316e-07,
+      "loss": 0.0142,
+      "step": 22425
+    },
+    {
+      "epoch": 4.0604087538433715,
+      "grad_norm": 0.12296438962221146,
+      "learning_rate": 6.051022287714127e-07,
+      "loss": 0.0138,
+      "step": 22450
+    },
+    {
+      "epoch": 4.064930367155001,
+      "grad_norm": 1.0237263441085815,
+      "learning_rate": 6.046417388100939e-07,
+      "loss": 0.0244,
+      "step": 22475
+    },
+    {
+      "epoch": 4.0694519804666305,
+      "grad_norm": 15.541071891784668,
+      "learning_rate": 6.041812488487751e-07,
+      "loss": 0.0174,
+      "step": 22500
+    },
+    {
+      "epoch": 4.07397359377826,
+      "grad_norm": 2.008063793182373,
+      "learning_rate": 6.037207588874562e-07,
+      "loss": 0.0325,
+      "step": 22525
+    },
+    {
+      "epoch": 4.078495207089889,
+      "grad_norm": 1.934888243675232,
+      "learning_rate": 6.032602689261373e-07,
+      "loss": 0.0051,
+      "step": 22550
+    },
+    {
+      "epoch": 4.083016820401519,
+      "grad_norm": 14.471735000610352,
+      "learning_rate": 6.027997789648186e-07,
+      "loss": 0.0077,
+      "step": 22575
+    },
+    {
+      "epoch": 4.087538433713149,
+      "grad_norm": 0.05337606742978096,
+      "learning_rate": 6.023392890034998e-07,
+      "loss": 0.012,
+      "step": 22600
+    },
+    {
+      "epoch": 4.092060047024779,
+      "grad_norm": 0.21558235585689545,
+      "learning_rate": 6.018787990421808e-07,
+      "loss": 0.0116,
+      "step": 22625
+    },
+    {
+      "epoch": 4.096581660336408,
+      "grad_norm": 11.160529136657715,
+      "learning_rate": 6.01418309080862e-07,
+      "loss": 0.0083,
+      "step": 22650
+    },
+    {
+      "epoch": 4.101103273648038,
+      "grad_norm": 0.33503735065460205,
+      "learning_rate": 6.009578191195432e-07,
+      "loss": 0.0071,
+      "step": 22675
+    },
+    {
+      "epoch": 4.105624886959667,
+      "grad_norm": 0.04343891143798828,
+      "learning_rate": 6.004973291582243e-07,
+      "loss": 0.0057,
+      "step": 22700
+    },
+    {
+      "epoch": 4.110146500271297,
+      "grad_norm": 0.017316769808530807,
+      "learning_rate": 6.000368391969055e-07,
+      "loss": 0.0085,
+      "step": 22725
+    },
+    {
+      "epoch": 4.114668113582926,
+      "grad_norm": 2.94707989692688,
+      "learning_rate": 5.995763492355867e-07,
+      "loss": 0.0085,
+      "step": 22750
+    },
+    {
+      "epoch": 4.119189726894556,
+      "grad_norm": 0.5166690349578857,
+      "learning_rate": 5.991158592742677e-07,
+      "loss": 0.0085,
+      "step": 22775
+    },
+    {
+      "epoch": 4.123711340206185,
+      "grad_norm": 0.27040883898735046,
+      "learning_rate": 5.986553693129489e-07,
+      "loss": 0.0115,
+      "step": 22800
+    },
+    {
+      "epoch": 4.1282329535178155,
+      "grad_norm": 0.2504747807979584,
+      "learning_rate": 5.981948793516302e-07,
+      "loss": 0.0193,
+      "step": 22825
+    },
+    {
+      "epoch": 4.132754566829445,
+      "grad_norm": 0.23941832780838013,
+      "learning_rate": 5.977343893903113e-07,
+      "loss": 0.0184,
+      "step": 22850
+    },
+    {
+      "epoch": 4.1372761801410745,
+      "grad_norm": 0.060054145753383636,
+      "learning_rate": 5.972738994289924e-07,
+      "loss": 0.0327,
+      "step": 22875
+    },
+    {
+      "epoch": 4.141797793452704,
+      "grad_norm": 2.8317267894744873,
+      "learning_rate": 5.968134094676736e-07,
+      "loss": 0.0261,
+      "step": 22900
+    },
+    {
+      "epoch": 4.146319406764333,
+      "grad_norm": 0.7358568906784058,
+      "learning_rate": 5.963529195063547e-07,
+      "loss": 0.0361,
+      "step": 22925
+    },
+    {
+      "epoch": 4.150841020075963,
+      "grad_norm": 0.17320603132247925,
+      "learning_rate": 5.958924295450359e-07,
+      "loss": 0.0153,
+      "step": 22950
+    },
+    {
+      "epoch": 4.155362633387592,
+      "grad_norm": 7.578225612640381,
+      "learning_rate": 5.954319395837171e-07,
+      "loss": 0.0065,
+      "step": 22975
+    },
+    {
+      "epoch": 4.159884246699222,
+      "grad_norm": 0.20573076605796814,
+      "learning_rate": 5.949714496223982e-07,
+      "loss": 0.008,
+      "step": 23000
+    },
+    {
+      "epoch": 4.164405860010852,
+      "grad_norm": 0.031806666404008865,
+      "learning_rate": 5.945109596610793e-07,
+      "loss": 0.0039,
+      "step": 23025
+    },
+    {
+      "epoch": 4.168927473322482,
+      "grad_norm": 1.0935386419296265,
+      "learning_rate": 5.940504696997604e-07,
+      "loss": 0.0039,
+      "step": 23050
+    },
+    {
+      "epoch": 4.173449086634111,
+      "grad_norm": 3.248675584793091,
+      "learning_rate": 5.935899797384417e-07,
+      "loss": 0.0084,
+      "step": 23075
+    },
+    {
+      "epoch": 4.177970699945741,
+      "grad_norm": 0.0767347663640976,
+      "learning_rate": 5.931294897771229e-07,
+      "loss": 0.0062,
+      "step": 23100
+    },
+    {
+      "epoch": 4.18249231325737,
+      "grad_norm": 0.03708457574248314,
+      "learning_rate": 5.92668999815804e-07,
+      "loss": 0.0085,
+      "step": 23125
+    },
+    {
+      "epoch": 4.187013926569,
+      "grad_norm": 0.16320359706878662,
+      "learning_rate": 5.922085098544851e-07,
+      "loss": 0.0053,
+      "step": 23150
+    },
+    {
+      "epoch": 4.191535539880629,
+      "grad_norm": 0.7727710008621216,
+      "learning_rate": 5.917480198931663e-07,
+      "loss": 0.0291,
+      "step": 23175
+    },
+    {
+      "epoch": 4.196057153192259,
+      "grad_norm": 0.0653238445520401,
+      "learning_rate": 5.912875299318474e-07,
+      "loss": 0.0165,
+      "step": 23200
+    },
+    {
+      "epoch": 4.200578766503889,
+      "grad_norm": 18.652076721191406,
+      "learning_rate": 5.908270399705286e-07,
+      "loss": 0.0119,
+      "step": 23225
+    },
+    {
+      "epoch": 4.205100379815518,
+      "grad_norm": 1.3249595165252686,
+      "learning_rate": 5.903665500092098e-07,
+      "loss": 0.0127,
+      "step": 23250
+    },
+    {
+      "epoch": 4.209621993127148,
+      "grad_norm": 0.1082233116030693,
+      "learning_rate": 5.899060600478909e-07,
+      "loss": 0.0312,
+      "step": 23275
+    },
+    {
+      "epoch": 4.214143606438777,
+      "grad_norm": 0.7809641361236572,
+      "learning_rate": 5.894455700865721e-07,
+      "loss": 0.0206,
+      "step": 23300
+    },
+    {
+      "epoch": 4.218665219750407,
+      "grad_norm": 0.12851662933826447,
+      "learning_rate": 5.889850801252533e-07,
+      "loss": 0.0284,
+      "step": 23325
+    },
+    {
+      "epoch": 4.223186833062036,
+      "grad_norm": 1.1004611253738403,
+      "learning_rate": 5.885245901639344e-07,
+      "loss": 0.0192,
+      "step": 23350
+    },
+    {
+      "epoch": 4.227708446373666,
+      "grad_norm": 22.368629455566406,
+      "learning_rate": 5.880641002026155e-07,
+      "loss": 0.0076,
+      "step": 23375
+    },
+    {
+      "epoch": 4.232230059685295,
+      "grad_norm": 0.16481293737888336,
+      "learning_rate": 5.876036102412967e-07,
+      "loss": 0.0121,
+      "step": 23400
+    },
+    {
+      "epoch": 4.236751672996926,
+      "grad_norm": 0.09105231612920761,
+      "learning_rate": 5.871431202799779e-07,
+      "loss": 0.0055,
+      "step": 23425
+    },
+    {
+      "epoch": 4.241273286308555,
+      "grad_norm": 0.06610502302646637,
+      "learning_rate": 5.86682630318659e-07,
+      "loss": 0.0051,
+      "step": 23450
+    },
+    {
+      "epoch": 4.245794899620185,
+      "grad_norm": 1.563685417175293,
+      "learning_rate": 5.862221403573402e-07,
+      "loss": 0.0034,
+      "step": 23475
+    },
+    {
+      "epoch": 4.250316512931814,
+      "grad_norm": 0.07692436873912811,
+      "learning_rate": 5.857616503960213e-07,
+      "loss": 0.0049,
+      "step": 23500
+    },
+    {
+      "epoch": 4.254838126243444,
+      "grad_norm": 0.42418307065963745,
+      "learning_rate": 5.853011604347025e-07,
+      "loss": 0.0021,
+      "step": 23525
+    },
+    {
+      "epoch": 4.259359739555073,
+      "grad_norm": 0.6450216174125671,
+      "learning_rate": 5.848406704733837e-07,
+      "loss": 0.0117,
+      "step": 23550
+    },
+    {
+      "epoch": 4.263881352866703,
+      "grad_norm": 3.0769999027252197,
+      "learning_rate": 5.843801805120648e-07,
+      "loss": 0.0021,
+      "step": 23575
+    },
+    {
+      "epoch": 4.268402966178332,
+      "grad_norm": 0.060876231640577316,
+      "learning_rate": 5.83919690550746e-07,
+      "loss": 0.0241,
+      "step": 23600
+    },
+    {
+      "epoch": 4.272924579489962,
+      "grad_norm": 10.53986930847168,
+      "learning_rate": 5.834592005894271e-07,
+      "loss": 0.0232,
+      "step": 23625
+    },
+    {
+      "epoch": 4.277446192801592,
+      "grad_norm": 17.494129180908203,
+      "learning_rate": 5.829987106281082e-07,
+      "loss": 0.0222,
+      "step": 23650
+    },
+    {
+      "epoch": 4.281967806113221,
+      "grad_norm": 0.1705417037010193,
+      "learning_rate": 5.825382206667895e-07,
+      "loss": 0.0187,
+      "step": 23675
+    },
+    {
+      "epoch": 4.286489419424851,
+      "grad_norm": 2.707817316055298,
+      "learning_rate": 5.820777307054707e-07,
+      "loss": 0.0327,
+      "step": 23700
+    },
+    {
+      "epoch": 4.29101103273648,
+      "grad_norm": 1.4121204614639282,
+      "learning_rate": 5.816172407441517e-07,
+      "loss": 0.0243,
+      "step": 23725
+    },
+    {
+      "epoch": 4.29553264604811,
+      "grad_norm": 7.615264415740967,
+      "learning_rate": 5.811567507828329e-07,
+      "loss": 0.0081,
+      "step": 23750
+    },
+    {
+      "epoch": 4.300054259359739,
+      "grad_norm": 0.06863299757242203,
+      "learning_rate": 5.806962608215141e-07,
+      "loss": 0.0112,
+      "step": 23775
+    },
+    {
+      "epoch": 4.304575872671369,
+      "grad_norm": 0.018180107697844505,
+      "learning_rate": 5.802357708601952e-07,
+      "loss": 0.0046,
+      "step": 23800
+    },
+    {
+      "epoch": 4.309097485982999,
+      "grad_norm": 0.43104514479637146,
+      "learning_rate": 5.797752808988764e-07,
+      "loss": 0.0055,
+      "step": 23825
+    },
+    {
+      "epoch": 4.313619099294629,
+      "grad_norm": 0.0967748612165451,
+      "learning_rate": 5.793147909375576e-07,
+      "loss": 0.0046,
+      "step": 23850
+    },
+    {
+      "epoch": 4.318140712606258,
+      "grad_norm": 0.15935564041137695,
+      "learning_rate": 5.788543009762386e-07,
+      "loss": 0.0086,
+      "step": 23875
+    },
+    {
+      "epoch": 4.322662325917888,
+      "grad_norm": 0.141365185379982,
+      "learning_rate": 5.783938110149198e-07,
+      "loss": 0.0167,
+      "step": 23900
+    },
+    {
+      "epoch": 4.327183939229517,
+      "grad_norm": 0.17591165006160736,
+      "learning_rate": 5.779333210536011e-07,
+      "loss": 0.0064,
+      "step": 23925
+    },
+    {
+      "epoch": 4.3317055525411465,
+      "grad_norm": 6.465728759765625,
+      "learning_rate": 5.774728310922822e-07,
+      "loss": 0.004,
+      "step": 23950
+    },
+    {
+      "epoch": 4.336227165852776,
+      "grad_norm": 0.12863588333129883,
+      "learning_rate": 5.770123411309633e-07,
+      "loss": 0.0163,
+      "step": 23975
+    },
+    {
+      "epoch": 4.3407487791644055,
+      "grad_norm": 0.3128308355808258,
+      "learning_rate": 5.765518511696444e-07,
+      "loss": 0.0216,
+      "step": 24000
+    },
+    {
+      "epoch": 4.345270392476036,
+      "grad_norm": 6.302773475646973,
+      "learning_rate": 5.760913612083256e-07,
+      "loss": 0.0115,
+      "step": 24025
+    },
+    {
+      "epoch": 4.349792005787665,
+      "grad_norm": 0.05059755593538284,
+      "learning_rate": 5.756308712470068e-07,
+      "loss": 0.0064,
+      "step": 24050
+    },
+    {
+      "epoch": 4.354313619099295,
+      "grad_norm": 0.18986685574054718,
+      "learning_rate": 5.75170381285688e-07,
+      "loss": 0.0277,
+      "step": 24075
+    },
+    {
+      "epoch": 4.358835232410924,
+      "grad_norm": 9.643892288208008,
+      "learning_rate": 5.747098913243691e-07,
+      "loss": 0.0373,
+      "step": 24100
+    },
+    {
+      "epoch": 4.363356845722554,
+      "grad_norm": 1.0030667781829834,
+      "learning_rate": 5.742494013630502e-07,
+      "loss": 0.0188,
+      "step": 24125
+    },
+    {
+      "epoch": 4.367878459034183,
+      "grad_norm": 0.44967132806777954,
+      "learning_rate": 5.737889114017314e-07,
+      "loss": 0.0095,
+      "step": 24150
+    },
+    {
+      "epoch": 4.372400072345813,
+      "grad_norm": 2.169846534729004,
+      "learning_rate": 5.733284214404126e-07,
+      "loss": 0.0168,
+      "step": 24175
+    },
+    {
+      "epoch": 4.376921685657442,
+      "grad_norm": 0.01923939771950245,
+      "learning_rate": 5.728679314790938e-07,
+      "loss": 0.0095,
+      "step": 24200
+    },
+    {
+      "epoch": 4.381443298969073,
+      "grad_norm": 9.444828033447266,
+      "learning_rate": 5.724074415177748e-07,
+      "loss": 0.0053,
+      "step": 24225
+    },
+    {
+      "epoch": 4.385964912280702,
+      "grad_norm": 60.225868225097656,
+      "learning_rate": 5.71946951556456e-07,
+      "loss": 0.0052,
+      "step": 24250
+    },
+    {
+      "epoch": 4.3904865255923315,
+      "grad_norm": 5.413934230804443,
+      "learning_rate": 5.714864615951373e-07,
+      "loss": 0.0125,
+      "step": 24275
+    },
+    {
+      "epoch": 4.395008138903961,
+      "grad_norm": 0.14292512834072113,
+      "learning_rate": 5.710259716338183e-07,
+      "loss": 0.0211,
+      "step": 24300
+    },
+    {
+      "epoch": 4.3995297522155905,
+      "grad_norm": 0.2194162905216217,
+      "learning_rate": 5.705654816724995e-07,
+      "loss": 0.0125,
+      "step": 24325
+    },
+    {
+      "epoch": 4.40405136552722,
+      "grad_norm": 0.029263151809573174,
+      "learning_rate": 5.701049917111807e-07,
+      "loss": 0.0045,
+      "step": 24350
+    },
+    {
+      "epoch": 4.4085729788388495,
+      "grad_norm": 0.2611495554447174,
+      "learning_rate": 5.696445017498617e-07,
+      "loss": 0.0031,
+      "step": 24375
+    },
+    {
+      "epoch": 4.413094592150479,
+      "grad_norm": 0.025238435715436935,
+      "learning_rate": 5.69184011788543e-07,
+      "loss": 0.0155,
+      "step": 24400
+    },
+    {
+      "epoch": 4.417616205462108,
+      "grad_norm": 0.043699052184820175,
+      "learning_rate": 5.687235218272242e-07,
+      "loss": 0.0156,
+      "step": 24425
+    },
+    {
+      "epoch": 4.422137818773739,
+      "grad_norm": 0.04374052584171295,
+      "learning_rate": 5.682630318659053e-07,
+      "loss": 0.0133,
+      "step": 24450
+    },
+    {
+      "epoch": 4.426659432085368,
+      "grad_norm": 3.1334991455078125,
+      "learning_rate": 5.678025419045864e-07,
+      "loss": 0.0383,
+      "step": 24475
+    },
+    {
+      "epoch": 4.431181045396998,
+      "grad_norm": 0.17290575802326202,
+      "learning_rate": 5.673420519432676e-07,
+      "loss": 0.0098,
+      "step": 24500
+    },
+    {
+      "epoch": 4.435702658708627,
+      "grad_norm": 0.5227589011192322,
+      "learning_rate": 5.668815619819488e-07,
+      "loss": 0.0863,
+      "step": 24525
+    },
+    {
+      "epoch": 4.440224272020257,
+      "grad_norm": 1.993817687034607,
+      "learning_rate": 5.664210720206299e-07,
+      "loss": 0.0161,
+      "step": 24550
+    },
+    {
+      "epoch": 4.444745885331886,
+      "grad_norm": 0.021811481565237045,
+      "learning_rate": 5.659605820593111e-07,
+      "loss": 0.0101,
+      "step": 24575
+    },
+    {
+      "epoch": 4.449267498643516,
+      "grad_norm": 0.055265914648771286,
+      "learning_rate": 5.655000920979922e-07,
+      "loss": 0.0117,
+      "step": 24600
+    },
+    {
+      "epoch": 4.453789111955146,
+      "grad_norm": 0.027848972007632256,
+      "learning_rate": 5.650396021366734e-07,
+      "loss": 0.0046,
+      "step": 24625
+    },
+    {
+      "epoch": 4.4583107252667755,
+      "grad_norm": 0.613524317741394,
+      "learning_rate": 5.645791121753546e-07,
+      "loss": 0.0079,
+      "step": 24650
+    },
+    {
+      "epoch": 4.462832338578405,
+      "grad_norm": 0.329345166683197,
+      "learning_rate": 5.641186222140357e-07,
+      "loss": 0.0096,
+      "step": 24675
+    },
+    {
+      "epoch": 4.4673539518900345,
+      "grad_norm": 0.07431361824274063,
+      "learning_rate": 5.636581322527169e-07,
+      "loss": 0.0033,
+      "step": 24700
+    },
+    {
+      "epoch": 4.471875565201664,
+      "grad_norm": 0.01413232646882534,
+      "learning_rate": 5.63197642291398e-07,
+      "loss": 0.0145,
+      "step": 24725
+    },
+    {
+      "epoch": 4.476397178513293,
+      "grad_norm": 0.03763017803430557,
+      "learning_rate": 5.627371523300791e-07,
+      "loss": 0.0093,
+      "step": 24750
+    },
+    {
+      "epoch": 4.480918791824923,
+      "grad_norm": 0.8524808287620544,
+      "learning_rate": 5.622766623687604e-07,
+      "loss": 0.0118,
+      "step": 24775
+    },
+    {
+      "epoch": 4.485440405136552,
+      "grad_norm": 0.10143906623125076,
+      "learning_rate": 5.618161724074415e-07,
+      "loss": 0.0203,
+      "step": 24800
+    },
+    {
+      "epoch": 4.489962018448182,
+      "grad_norm": 3.6819851398468018,
+      "learning_rate": 5.613556824461226e-07,
+      "loss": 0.0046,
+      "step": 24825
+    },
+    {
+      "epoch": 4.494483631759812,
+      "grad_norm": 17.471813201904297,
+      "learning_rate": 5.608951924848038e-07,
+      "loss": 0.0092,
+      "step": 24850
+    },
+    {
+      "epoch": 4.499005245071442,
+      "grad_norm": 0.03576793521642685,
+      "learning_rate": 5.604347025234851e-07,
+      "loss": 0.0171,
+      "step": 24875
+    },
+    {
+      "epoch": 4.503526858383071,
+      "grad_norm": 1.3747385740280151,
+      "learning_rate": 5.599742125621661e-07,
+      "loss": 0.0433,
+      "step": 24900
+    },
+    {
+      "epoch": 4.508048471694701,
+      "grad_norm": 0.658429741859436,
+      "learning_rate": 5.595137226008473e-07,
+      "loss": 0.0272,
+      "step": 24925
+    },
+    {
+      "epoch": 4.51257008500633,
+      "grad_norm": 2.7316272258758545,
+      "learning_rate": 5.590532326395284e-07,
+      "loss": 0.0061,
+      "step": 24950
+    },
+    {
+      "epoch": 4.51709169831796,
+      "grad_norm": 0.3981630206108093,
+      "learning_rate": 5.585927426782095e-07,
+      "loss": 0.0078,
+      "step": 24975
+    },
+    {
+      "epoch": 4.521613311629589,
+      "grad_norm": 0.21992315351963043,
+      "learning_rate": 5.581322527168908e-07,
+      "loss": 0.0076,
+      "step": 25000
+    },
+    {
+      "epoch": 4.5261349249412195,
+      "grad_norm": 7.3034796714782715,
+      "learning_rate": 5.576717627555719e-07,
+      "loss": 0.0073,
+      "step": 25025
+    },
+    {
+      "epoch": 4.530656538252849,
+      "grad_norm": 0.10995008796453476,
+      "learning_rate": 5.572112727942531e-07,
+      "loss": 0.0064,
+      "step": 25050
+    },
+    {
+      "epoch": 4.5351781515644785,
+      "grad_norm": 0.31085583567619324,
+      "learning_rate": 5.567507828329342e-07,
+      "loss": 0.0081,
+      "step": 25075
+    },
+    {
+      "epoch": 4.539699764876108,
+      "grad_norm": 0.062460754066705704,
+      "learning_rate": 5.562902928716153e-07,
+      "loss": 0.0158,
+      "step": 25100
+    },
+    {
+      "epoch": 4.544221378187737,
+      "grad_norm": 0.5231966376304626,
+      "learning_rate": 5.558298029102966e-07,
+      "loss": 0.0048,
+      "step": 25125
+    },
+    {
+      "epoch": 4.548742991499367,
+      "grad_norm": 0.23824147880077362,
+      "learning_rate": 5.553693129489777e-07,
+      "loss": 0.0172,
+      "step": 25150
+    },
+    {
+      "epoch": 4.553264604810996,
+      "grad_norm": 0.060738705098629,
+      "learning_rate": 5.549088229876588e-07,
+      "loss": 0.0127,
+      "step": 25175
+    },
+    {
+      "epoch": 4.557786218122626,
+      "grad_norm": 0.017272014170885086,
+      "learning_rate": 5.5444833302634e-07,
+      "loss": 0.0124,
+      "step": 25200
+    },
+    {
+      "epoch": 4.562307831434255,
+      "grad_norm": 0.1754794865846634,
+      "learning_rate": 5.539878430650211e-07,
+      "loss": 0.0148,
+      "step": 25225
+    },
+    {
+      "epoch": 4.566829444745886,
+      "grad_norm": 0.2678035795688629,
+      "learning_rate": 5.535273531037023e-07,
+      "loss": 0.0127,
+      "step": 25250
+    },
+    {
+      "epoch": 4.571351058057515,
+      "grad_norm": 10.760842323303223,
+      "learning_rate": 5.530668631423835e-07,
+      "loss": 0.0245,
+      "step": 25275
+    },
+    {
+      "epoch": 4.575872671369145,
+      "grad_norm": 0.16525591909885406,
+      "learning_rate": 5.526247927795174e-07,
+      "loss": 0.0371,
+      "step": 25300
+    },
+    {
+      "epoch": 4.580394284680774,
+      "grad_norm": 1.5673601627349854,
+      "learning_rate": 5.521643028181985e-07,
+      "loss": 0.0264,
+      "step": 25325
+    },
+    {
+      "epoch": 4.584915897992404,
+      "grad_norm": 0.20957525074481964,
+      "learning_rate": 5.517038128568797e-07,
+      "loss": 0.0167,
+      "step": 25350
+    },
+    {
+      "epoch": 4.589437511304033,
+      "grad_norm": 0.5806902050971985,
+      "learning_rate": 5.512433228955609e-07,
+      "loss": 0.0187,
+      "step": 25375
+    },
+    {
+      "epoch": 4.593959124615663,
+      "grad_norm": 0.487657368183136,
+      "learning_rate": 5.507828329342419e-07,
+      "loss": 0.0049,
+      "step": 25400
+    },
+    {
+      "epoch": 4.598480737927293,
+      "grad_norm": 0.2314232587814331,
+      "learning_rate": 5.503223429729231e-07,
+      "loss": 0.0065,
+      "step": 25425
+    },
+    {
+      "epoch": 4.603002351238922,
+      "grad_norm": 0.11486486345529556,
+      "learning_rate": 5.498618530116044e-07,
+      "loss": 0.0032,
+      "step": 25450
+    },
+    {
+      "epoch": 4.607523964550552,
+      "grad_norm": 0.23516380786895752,
+      "learning_rate": 5.494013630502855e-07,
+      "loss": 0.0088,
+      "step": 25475
+    },
+    {
+      "epoch": 4.612045577862181,
+      "grad_norm": 0.26957616209983826,
+      "learning_rate": 5.489408730889666e-07,
+      "loss": 0.011,
+      "step": 25500
+    },
+    {
+      "epoch": 4.616567191173811,
+      "grad_norm": 0.9121783375740051,
+      "learning_rate": 5.484803831276478e-07,
+      "loss": 0.0125,
+      "step": 25525
+    },
+    {
+      "epoch": 4.62108880448544,
+      "grad_norm": 0.0874478742480278,
+      "learning_rate": 5.480198931663289e-07,
+      "loss": 0.0015,
+      "step": 25550
+    },
+    {
+      "epoch": 4.62561041779707,
+      "grad_norm": 0.13613806664943695,
+      "learning_rate": 5.475594032050101e-07,
+      "loss": 0.005,
+      "step": 25575
+    },
+    {
+      "epoch": 4.630132031108699,
+      "grad_norm": 12.003376960754395,
+      "learning_rate": 5.470989132436913e-07,
+      "loss": 0.0104,
+      "step": 25600
+    },
+    {
+      "epoch": 4.634653644420329,
+      "grad_norm": 0.1077638640999794,
+      "learning_rate": 5.466384232823724e-07,
+      "loss": 0.0035,
+      "step": 25625
+    },
+    {
+      "epoch": 4.639175257731958,
+      "grad_norm": 0.20498277246952057,
+      "learning_rate": 5.461779333210536e-07,
+      "loss": 0.0327,
+      "step": 25650
+    },
+    {
+      "epoch": 4.643696871043589,
+      "grad_norm": 0.12293770909309387,
+      "learning_rate": 5.457174433597347e-07,
+      "loss": 0.0208,
+      "step": 25675
+    },
+    {
+      "epoch": 4.648218484355218,
+      "grad_norm": 0.13977064192295074,
+      "learning_rate": 5.452569533984159e-07,
+      "loss": 0.0279,
+      "step": 25700
+    },
+    {
+      "epoch": 4.652740097666848,
+      "grad_norm": 0.14199872314929962,
+      "learning_rate": 5.447964634370971e-07,
+      "loss": 0.0407,
+      "step": 25725
+    },
+    {
+      "epoch": 4.657261710978477,
+      "grad_norm": 0.16517263650894165,
+      "learning_rate": 5.443359734757782e-07,
+      "loss": 0.0084,
+      "step": 25750
+    },
+    {
+      "epoch": 4.661783324290107,
+      "grad_norm": 4.908987045288086,
+      "learning_rate": 5.438754835144593e-07,
+      "loss": 0.0096,
+      "step": 25775
+    },
+    {
+      "epoch": 4.666304937601736,
+      "grad_norm": 2.0926215648651123,
+      "learning_rate": 5.434149935531405e-07,
+      "loss": 0.0069,
+      "step": 25800
+    },
+    {
+      "epoch": 4.6708265509133655,
+      "grad_norm": 37.50233840942383,
+      "learning_rate": 5.429545035918217e-07,
+      "loss": 0.0127,
+      "step": 25825
+    },
+    {
+      "epoch": 4.675348164224996,
+      "grad_norm": 3.6432156562805176,
+      "learning_rate": 5.424940136305028e-07,
+      "loss": 0.0037,
+      "step": 25850
+    },
+    {
+      "epoch": 4.679869777536625,
+      "grad_norm": 0.005684335716068745,
+      "learning_rate": 5.42033523669184e-07,
+      "loss": 0.016,
+      "step": 25875
+    },
+    {
+      "epoch": 4.684391390848255,
+      "grad_norm": 0.06579804420471191,
+      "learning_rate": 5.415730337078652e-07,
+      "loss": 0.0046,
+      "step": 25900
+    },
+    {
+      "epoch": 4.688913004159884,
+      "grad_norm": 18.67939567565918,
+      "learning_rate": 5.411125437465462e-07,
+      "loss": 0.0067,
+      "step": 25925
+    },
+    {
+      "epoch": 4.693434617471514,
+      "grad_norm": 0.035297270864248276,
+      "learning_rate": 5.406520537852275e-07,
+      "loss": 0.0092,
+      "step": 25950
+    },
+    {
+      "epoch": 4.697956230783143,
+      "grad_norm": 0.08447632938623428,
+      "learning_rate": 5.401915638239087e-07,
+      "loss": 0.0152,
+      "step": 25975
+    },
+    {
+      "epoch": 4.702477844094773,
+      "grad_norm": 0.2718825042247772,
+      "learning_rate": 5.397310738625897e-07,
+      "loss": 0.0267,
+      "step": 26000
+    },
+    {
+      "epoch": 4.706999457406402,
+      "grad_norm": 0.02587636187672615,
+      "learning_rate": 5.392705839012709e-07,
+      "loss": 0.0032,
+      "step": 26025
+    },
+    {
+      "epoch": 4.711521070718032,
+      "grad_norm": 0.295260488986969,
+      "learning_rate": 5.388100939399522e-07,
+      "loss": 0.0249,
+      "step": 26050
+    },
+    {
+      "epoch": 4.716042684029662,
+      "grad_norm": 0.10036912560462952,
+      "learning_rate": 5.383496039786333e-07,
+      "loss": 0.0248,
+      "step": 26075
+    },
+    {
+      "epoch": 4.720564297341292,
+      "grad_norm": 19.557443618774414,
+      "learning_rate": 5.378891140173144e-07,
+      "loss": 0.0269,
+      "step": 26100
+    },
+    {
+      "epoch": 4.725085910652921,
+      "grad_norm": 13.175033569335938,
+      "learning_rate": 5.374286240559955e-07,
+      "loss": 0.0286,
+      "step": 26125
+    },
+    {
+      "epoch": 4.7296075239645505,
+      "grad_norm": 28.173160552978516,
+      "learning_rate": 5.369681340946767e-07,
+      "loss": 0.0069,
+      "step": 26150
+    },
+    {
+      "epoch": 4.73412913727618,
+      "grad_norm": 0.07096391171216965,
+      "learning_rate": 5.365076441333579e-07,
+      "loss": 0.0089,
+      "step": 26175
+    },
+    {
+      "epoch": 4.7386507505878095,
+      "grad_norm": 0.07818924635648727,
+      "learning_rate": 5.36047154172039e-07,
+      "loss": 0.004,
+      "step": 26200
+    },
+    {
+      "epoch": 4.743172363899439,
+      "grad_norm": 2.088768243789673,
+      "learning_rate": 5.355866642107202e-07,
+      "loss": 0.0118,
+      "step": 26225
+    },
+    {
+      "epoch": 4.747693977211069,
+      "grad_norm": 0.3656996488571167,
+      "learning_rate": 5.351261742494013e-07,
+      "loss": 0.0029,
+      "step": 26250
+    },
+    {
+      "epoch": 4.752215590522699,
+      "grad_norm": 0.016400739550590515,
+      "learning_rate": 5.346656842880824e-07,
+      "loss": 0.0068,
+      "step": 26275
+    },
+    {
+      "epoch": 4.756737203834328,
+      "grad_norm": 0.023505523800849915,
+      "learning_rate": 5.342051943267637e-07,
+      "loss": 0.0053,
+      "step": 26300
+    },
+    {
+      "epoch": 4.761258817145958,
+      "grad_norm": 0.033994659781455994,
+      "learning_rate": 5.337447043654449e-07,
+      "loss": 0.0084,
+      "step": 26325
+    },
+    {
+      "epoch": 4.765780430457587,
+      "grad_norm": 0.4282836318016052,
+      "learning_rate": 5.332842144041259e-07,
+      "loss": 0.0178,
+      "step": 26350
+    },
+    {
+      "epoch": 4.770302043769217,
+      "grad_norm": 0.9596586227416992,
+      "learning_rate": 5.328237244428071e-07,
+      "loss": 0.0055,
+      "step": 26375
+    },
+    {
+      "epoch": 4.774823657080846,
+      "grad_norm": 0.07100539654493332,
+      "learning_rate": 5.323632344814883e-07,
+      "loss": 0.0035,
+      "step": 26400
+    },
+    {
+      "epoch": 4.779345270392476,
+      "grad_norm": 20.814708709716797,
+      "learning_rate": 5.319027445201694e-07,
+      "loss": 0.0018,
+      "step": 26425
+    },
+    {
+      "epoch": 4.783866883704105,
+      "grad_norm": 0.1001884788274765,
+      "learning_rate": 5.314422545588506e-07,
+      "loss": 0.0169,
+      "step": 26450
+    },
+    {
+      "epoch": 4.7883884970157355,
+      "grad_norm": 10.18614387512207,
+      "learning_rate": 5.309817645975318e-07,
+      "loss": 0.0833,
+      "step": 26475
+    },
+    {
+      "epoch": 4.792910110327365,
+      "grad_norm": 5.15158224105835,
+      "learning_rate": 5.305212746362128e-07,
+      "loss": 0.0395,
+      "step": 26500
+    },
+    {
+      "epoch": 4.7974317236389945,
+      "grad_norm": 0.10631446540355682,
+      "learning_rate": 5.30060784674894e-07,
+      "loss": 0.019,
+      "step": 26525
+    },
+    {
+      "epoch": 4.801953336950624,
+      "grad_norm": 0.15538524091243744,
+      "learning_rate": 5.296002947135753e-07,
+      "loss": 0.0075,
+      "step": 26550
+    },
+    {
+      "epoch": 4.8064749502622535,
+      "grad_norm": 1.2634289264678955,
+      "learning_rate": 5.291398047522564e-07,
+      "loss": 0.0111,
+      "step": 26575
+    },
+    {
+      "epoch": 4.810996563573883,
+      "grad_norm": 0.11970185488462448,
+      "learning_rate": 5.286793147909375e-07,
+      "loss": 0.009,
+      "step": 26600
+    },
+    {
+      "epoch": 4.815518176885512,
+      "grad_norm": 3.1821625232696533,
+      "learning_rate": 5.282188248296187e-07,
+      "loss": 0.008,
+      "step": 26625
+    },
+    {
+      "epoch": 4.820039790197143,
+      "grad_norm": 1.2722951173782349,
+      "learning_rate": 5.277583348682998e-07,
+      "loss": 0.0083,
+      "step": 26650
+    },
+    {
+      "epoch": 4.824561403508772,
+      "grad_norm": 0.16242074966430664,
+      "learning_rate": 5.27297844906981e-07,
+      "loss": 0.0093,
+      "step": 26675
+    },
+    {
+      "epoch": 4.829083016820402,
+      "grad_norm": 0.08178609609603882,
+      "learning_rate": 5.268373549456622e-07,
+      "loss": 0.0163,
+      "step": 26700
+    },
+    {
+      "epoch": 4.833604630132031,
+      "grad_norm": 0.05239911004900932,
+      "learning_rate": 5.263768649843433e-07,
+      "loss": 0.0047,
+      "step": 26725
+    },
+    {
+      "epoch": 4.838126243443661,
+      "grad_norm": 0.11282465606927872,
+      "learning_rate": 5.259163750230244e-07,
+      "loss": 0.0057,
+      "step": 26750
+    },
+    {
+      "epoch": 4.84264785675529,
+      "grad_norm": 0.11248558014631271,
+      "learning_rate": 5.254558850617057e-07,
+      "loss": 0.011,
+      "step": 26775
+    },
+    {
+      "epoch": 4.84716947006692,
+      "grad_norm": 0.020478971302509308,
+      "learning_rate": 5.249953951003868e-07,
+      "loss": 0.013,
+      "step": 26800
+    },
+    {
+      "epoch": 4.851691083378549,
+      "grad_norm": 0.0279947929084301,
+      "learning_rate": 5.24534905139068e-07,
+      "loss": 0.016,
+      "step": 26825
+    },
+    {
+      "epoch": 4.856212696690179,
+      "grad_norm": 2.424943208694458,
+      "learning_rate": 5.24074415177749e-07,
+      "loss": 0.0214,
+      "step": 26850
+    },
+    {
+      "epoch": 4.860734310001809,
+      "grad_norm": 0.05842322111129761,
+      "learning_rate": 5.236139252164302e-07,
+      "loss": 0.0067,
+      "step": 26875
+    },
+    {
+      "epoch": 4.8652559233134385,
+      "grad_norm": 11.81139087677002,
+      "learning_rate": 5.231534352551115e-07,
+      "loss": 0.023,
+      "step": 26900
+    },
+    {
+      "epoch": 4.869777536625068,
+      "grad_norm": 4.321495056152344,
+      "learning_rate": 5.226929452937926e-07,
+      "loss": 0.0276,
+      "step": 26925
+    },
+    {
+      "epoch": 4.874299149936697,
+      "grad_norm": 26.13640022277832,
+      "learning_rate": 5.222324553324737e-07,
+      "loss": 0.0133,
+      "step": 26950
+    },
+    {
+      "epoch": 4.878820763248327,
+      "grad_norm": 0.09919006377458572,
+      "learning_rate": 5.217719653711549e-07,
+      "loss": 0.01,
+      "step": 26975
+    },
+    {
+      "epoch": 4.883342376559956,
+      "grad_norm": 2.89017653465271,
+      "learning_rate": 5.21311475409836e-07,
+      "loss": 0.0199,
+      "step": 27000
+    },
+    {
+      "epoch": 4.887863989871586,
+      "grad_norm": 0.1923578828573227,
+      "learning_rate": 5.208509854485172e-07,
+      "loss": 0.0055,
+      "step": 27025
+    },
+    {
+      "epoch": 4.892385603183216,
+      "grad_norm": 0.05905037745833397,
+      "learning_rate": 5.203904954871984e-07,
+      "loss": 0.007,
+      "step": 27050
+    },
+    {
+      "epoch": 4.896907216494846,
+      "grad_norm": 4.7976837158203125,
+      "learning_rate": 5.199300055258795e-07,
+      "loss": 0.0095,
+      "step": 27075
+    },
+    {
+      "epoch": 4.901428829806475,
+      "grad_norm": 0.5176482200622559,
+      "learning_rate": 5.194695155645606e-07,
+      "loss": 0.0068,
+      "step": 27100
+    },
+    {
+      "epoch": 4.905950443118105,
+      "grad_norm": 0.15327677130699158,
+      "learning_rate": 5.190090256032418e-07,
+      "loss": 0.0016,
+      "step": 27125
+    },
+    {
+      "epoch": 4.910472056429734,
+      "grad_norm": 23.123516082763672,
+      "learning_rate": 5.18548535641923e-07,
+      "loss": 0.0126,
+      "step": 27150
+    },
+    {
+      "epoch": 4.914993669741364,
+      "grad_norm": 5.405014514923096,
+      "learning_rate": 5.180880456806041e-07,
+      "loss": 0.0063,
+      "step": 27175
+    },
+    {
+      "epoch": 4.919515283052993,
+      "grad_norm": 0.03432038426399231,
+      "learning_rate": 5.176275557192853e-07,
+      "loss": 0.0085,
+      "step": 27200
+    },
+    {
+      "epoch": 4.924036896364623,
+      "grad_norm": 0.012576217763125896,
+      "learning_rate": 5.171670657579664e-07,
+      "loss": 0.0089,
+      "step": 27225
+    },
+    {
+      "epoch": 4.928558509676252,
+      "grad_norm": 0.03322821483016014,
+      "learning_rate": 5.167065757966476e-07,
+      "loss": 0.0117,
+      "step": 27250
+    },
+    {
+      "epoch": 4.9330801229878825,
+      "grad_norm": 0.14308449625968933,
+      "learning_rate": 5.162460858353288e-07,
+      "loss": 0.0083,
+      "step": 27275
+    },
+    {
+      "epoch": 4.937601736299512,
+      "grad_norm": 39.83473205566406,
+      "learning_rate": 5.157855958740099e-07,
+      "loss": 0.0412,
+      "step": 27300
+    },
+    {
+      "epoch": 4.942123349611141,
+      "grad_norm": 0.15658709406852722,
+      "learning_rate": 5.153251059126911e-07,
+      "loss": 0.048,
+      "step": 27325
+    },
+    {
+      "epoch": 4.946644962922771,
+      "grad_norm": 0.5169302821159363,
+      "learning_rate": 5.148646159513722e-07,
+      "loss": 0.0125,
+      "step": 27350
+    },
+    {
+      "epoch": 4.9511665762344,
+      "grad_norm": 0.021764138713479042,
+      "learning_rate": 5.144041259900533e-07,
+      "loss": 0.0044,
+      "step": 27375
+    },
+    {
+      "epoch": 4.95568818954603,
+      "grad_norm": 0.2565416395664215,
+      "learning_rate": 5.139436360287346e-07,
+      "loss": 0.0134,
+      "step": 27400
+    },
+    {
+      "epoch": 4.960209802857659,
+      "grad_norm": 3.5746593475341797,
+      "learning_rate": 5.134831460674158e-07,
+      "loss": 0.0144,
+      "step": 27425
+    },
+    {
+      "epoch": 4.96473141616929,
+      "grad_norm": 1.1976155042648315,
+      "learning_rate": 5.130226561060968e-07,
+      "loss": 0.0015,
+      "step": 27450
+    },
+    {
+      "epoch": 4.969253029480919,
+      "grad_norm": 1.560697078704834,
+      "learning_rate": 5.12562166144778e-07,
+      "loss": 0.0105,
+      "step": 27475
+    },
+    {
+      "epoch": 4.973774642792549,
+      "grad_norm": 0.21413296461105347,
+      "learning_rate": 5.121016761834593e-07,
+      "loss": 0.0086,
+      "step": 27500
+    },
+    {
+      "epoch": 4.978296256104178,
+      "grad_norm": 0.1948905736207962,
+      "learning_rate": 5.116411862221403e-07,
+      "loss": 0.0142,
+      "step": 27525
+    },
+    {
+      "epoch": 4.982817869415808,
+      "grad_norm": 0.0875588059425354,
+      "learning_rate": 5.111806962608215e-07,
+      "loss": 0.0086,
+      "step": 27550
+    },
+    {
+      "epoch": 4.987339482727437,
+      "grad_norm": 0.18821197748184204,
+      "learning_rate": 5.107202062995027e-07,
+      "loss": 0.0343,
+      "step": 27575
+    },
+    {
+      "epoch": 4.991861096039067,
+      "grad_norm": 3.243891954421997,
+      "learning_rate": 5.102597163381837e-07,
+      "loss": 0.0066,
+      "step": 27600
+    },
+    {
+      "epoch": 4.996382709350696,
+      "grad_norm": 0.04505661129951477,
+      "learning_rate": 5.09799226376865e-07,
+      "loss": 0.0273,
+      "step": 27625
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.32219141721725464,
+      "eval_runtime": 8558.7304,
+      "eval_samples_per_second": 1.109,
+      "eval_steps_per_second": 0.139,
+      "eval_wer": 0.1037636130685458,
+      "step": 27645
+    },
+    {
+      "epoch": 5.0009043226623255,
+      "grad_norm": 3.0571129322052,
+      "learning_rate": 5.093387364155462e-07,
+      "loss": 0.0432,
+      "step": 27650
+    },
+    {
+      "epoch": 5.005425935973956,
+      "grad_norm": 0.28156787157058716,
+      "learning_rate": 5.088782464542273e-07,
+      "loss": 0.015,
+      "step": 27675
+    },
+    {
+      "epoch": 5.009947549285585,
+      "grad_norm": 1.9075210094451904,
+      "learning_rate": 5.084177564929084e-07,
+      "loss": 0.0056,
+      "step": 27700
+    },
+    {
+      "epoch": 5.014469162597215,
+      "grad_norm": 0.8747662305831909,
+      "learning_rate": 5.079572665315896e-07,
+      "loss": 0.0115,
+      "step": 27725
+    },
+    {
+      "epoch": 5.018990775908844,
+      "grad_norm": 0.4982765316963196,
+      "learning_rate": 5.074967765702708e-07,
+      "loss": 0.0029,
+      "step": 27750
+    },
+    {
+      "epoch": 5.023512389220474,
+      "grad_norm": 0.06921929121017456,
+      "learning_rate": 5.070362866089519e-07,
+      "loss": 0.0133,
+      "step": 27775
+    },
+    {
+      "epoch": 5.028034002532103,
+      "grad_norm": 0.3169139623641968,
+      "learning_rate": 5.06575796647633e-07,
+      "loss": 0.0059,
+      "step": 27800
+    },
+    {
+      "epoch": 5.032555615843733,
+      "grad_norm": 0.11565407365560532,
+      "learning_rate": 5.061153066863142e-07,
+      "loss": 0.0092,
+      "step": 27825
+    },
+    {
+      "epoch": 5.037077229155362,
+      "grad_norm": 0.03337372466921806,
+      "learning_rate": 5.056548167249953e-07,
+      "loss": 0.0045,
+      "step": 27850
+    },
+    {
+      "epoch": 5.041598842466993,
+      "grad_norm": 0.03493885695934296,
+      "learning_rate": 5.051943267636765e-07,
+      "loss": 0.0024,
+      "step": 27875
+    },
+    {
+      "epoch": 5.046120455778622,
+      "grad_norm": 0.018299061805009842,
+      "learning_rate": 5.047338368023577e-07,
+      "loss": 0.0072,
+      "step": 27900
+    },
+    {
+      "epoch": 5.050642069090252,
+      "grad_norm": 0.10773641616106033,
+      "learning_rate": 5.042733468410389e-07,
+      "loss": 0.0122,
+      "step": 27925
+    },
+    {
+      "epoch": 5.055163682401881,
+      "grad_norm": 0.8263195157051086,
+      "learning_rate": 5.038128568797199e-07,
+      "loss": 0.0025,
+      "step": 27950
+    },
+    {
+      "epoch": 5.0596852957135106,
+      "grad_norm": 0.03416213020682335,
+      "learning_rate": 5.033523669184011e-07,
+      "loss": 0.0017,
+      "step": 27975
+    },
+    {
+      "epoch": 5.06420690902514,
+      "grad_norm": 0.05288705974817276,
+      "learning_rate": 5.028918769570824e-07,
+      "loss": 0.0196,
+      "step": 28000
+    },
+    {
+      "epoch": 5.0687285223367695,
+      "grad_norm": 1.1040371656417847,
+      "learning_rate": 5.024313869957634e-07,
+      "loss": 0.0162,
+      "step": 28025
+    },
+    {
+      "epoch": 5.073250135648399,
+      "grad_norm": 1.0380799770355225,
+      "learning_rate": 5.019708970344446e-07,
+      "loss": 0.0872,
+      "step": 28050
+    },
+    {
+      "epoch": 5.077771748960029,
+      "grad_norm": 1.2944865226745605,
+      "learning_rate": 5.015104070731258e-07,
+      "loss": 0.0196,
+      "step": 28075
+    },
+    {
+      "epoch": 5.082293362271659,
+      "grad_norm": 0.10234280675649643,
+      "learning_rate": 5.010499171118068e-07,
+      "loss": 0.0039,
+      "step": 28100
+    },
+    {
+      "epoch": 5.086814975583288,
+      "grad_norm": 0.2737126350402832,
+      "learning_rate": 5.005894271504881e-07,
+      "loss": 0.0138,
+      "step": 28125
+    },
+    {
+      "epoch": 5.091336588894918,
+      "grad_norm": 0.04862267151474953,
+      "learning_rate": 5.001289371891693e-07,
+      "loss": 0.0084,
+      "step": 28150
+    },
+    {
+      "epoch": 5.095858202206547,
+      "grad_norm": 0.12347660213708878,
+      "learning_rate": 4.996684472278504e-07,
+      "loss": 0.0039,
+      "step": 28175
+    },
+    {
+      "epoch": 5.100379815518177,
+      "grad_norm": 1.5366371870040894,
+      "learning_rate": 4.992079572665316e-07,
+      "loss": 0.0024,
+      "step": 28200
+    },
+    {
+      "epoch": 5.104901428829806,
+      "grad_norm": 0.11356142163276672,
+      "learning_rate": 4.987474673052128e-07,
+      "loss": 0.0069,
+      "step": 28225
+    },
+    {
+      "epoch": 5.109423042141436,
+      "grad_norm": 0.019394779577851295,
+      "learning_rate": 4.982869773438938e-07,
+      "loss": 0.0026,
+      "step": 28250
+    },
+    {
+      "epoch": 5.113944655453066,
+      "grad_norm": 12.068894386291504,
+      "learning_rate": 4.97826487382575e-07,
+      "loss": 0.0064,
+      "step": 28275
+    },
+    {
+      "epoch": 5.118466268764696,
+      "grad_norm": 3.972506523132324,
+      "learning_rate": 4.973659974212562e-07,
+      "loss": 0.004,
+      "step": 28300
+    },
+    {
+      "epoch": 5.122987882076325,
+      "grad_norm": 0.024081800132989883,
+      "learning_rate": 4.969055074599373e-07,
+      "loss": 0.0079,
+      "step": 28325
+    },
+    {
+      "epoch": 5.1275094953879545,
+      "grad_norm": 0.2852739095687866,
+      "learning_rate": 4.964450174986185e-07,
+      "loss": 0.0111,
+      "step": 28350
+    },
+    {
+      "epoch": 5.132031108699584,
+      "grad_norm": 0.1766853779554367,
+      "learning_rate": 4.959845275372997e-07,
+      "loss": 0.0042,
+      "step": 28375
+    },
+    {
+      "epoch": 5.1365527220112135,
+      "grad_norm": 0.1434965282678604,
+      "learning_rate": 4.955240375759808e-07,
+      "loss": 0.0018,
+      "step": 28400
+    },
+    {
+      "epoch": 5.141074335322843,
+      "grad_norm": 15.574809074401855,
+      "learning_rate": 4.95063547614662e-07,
+      "loss": 0.0219,
+      "step": 28425
+    },
+    {
+      "epoch": 5.1455959486344724,
+      "grad_norm": 0.07658454775810242,
+      "learning_rate": 4.946030576533432e-07,
+      "loss": 0.0361,
+      "step": 28450
+    },
+    {
+      "epoch": 5.150117561946102,
+      "grad_norm": 2.8722949028015137,
+      "learning_rate": 4.941425676920243e-07,
+      "loss": 0.0081,
+      "step": 28475
+    },
+    {
+      "epoch": 5.154639175257732,
+      "grad_norm": 0.16390861570835114,
+      "learning_rate": 4.936820777307055e-07,
+      "loss": 0.0024,
+      "step": 28500
+    },
+    {
+      "epoch": 5.159160788569362,
+      "grad_norm": 0.5398291945457458,
+      "learning_rate": 4.932215877693866e-07,
+      "loss": 0.0106,
+      "step": 28525
+    },
+    {
+      "epoch": 5.163682401880991,
+      "grad_norm": 10.156366348266602,
+      "learning_rate": 4.927610978080677e-07,
+      "loss": 0.0137,
+      "step": 28550
+    },
+    {
+      "epoch": 5.168204015192621,
+      "grad_norm": 0.013218329288065434,
+      "learning_rate": 4.923006078467489e-07,
+      "loss": 0.0104,
+      "step": 28575
+    },
+    {
+      "epoch": 5.17272562850425,
+      "grad_norm": 8.923846244812012,
+      "learning_rate": 4.9184011788543e-07,
+      "loss": 0.0106,
+      "step": 28600
+    },
+    {
+      "epoch": 5.17724724181588,
+      "grad_norm": 0.07841784507036209,
+      "learning_rate": 4.913796279241112e-07,
+      "loss": 0.0065,
+      "step": 28625
+    },
+    {
+      "epoch": 5.181768855127509,
+      "grad_norm": 0.3685607314109802,
+      "learning_rate": 4.909191379627924e-07,
+      "loss": 0.0118,
+      "step": 28650
+    },
+    {
+      "epoch": 5.186290468439139,
+      "grad_norm": 1.6632095575332642,
+      "learning_rate": 4.904586480014736e-07,
+      "loss": 0.0019,
+      "step": 28675
+    },
+    {
+      "epoch": 5.190812081750769,
+      "grad_norm": 0.10080606490373611,
+      "learning_rate": 4.899981580401547e-07,
+      "loss": 0.0027,
+      "step": 28700
+    },
+    {
+      "epoch": 5.1953336950623985,
+      "grad_norm": 4.444336891174316,
+      "learning_rate": 4.895376680788359e-07,
+      "loss": 0.0029,
+      "step": 28725
+    },
+    {
+      "epoch": 5.199855308374028,
+      "grad_norm": 5.826324939727783,
+      "learning_rate": 4.89077178117517e-07,
+      "loss": 0.0058,
+      "step": 28750
+    },
+    {
+      "epoch": 5.2043769216856575,
+      "grad_norm": 0.09290173649787903,
+      "learning_rate": 4.886166881561982e-07,
+      "loss": 0.0015,
+      "step": 28775
+    },
+    {
+      "epoch": 5.208898534997287,
+      "grad_norm": 0.0847366601228714,
+      "learning_rate": 4.881561981948794e-07,
+      "loss": 0.0143,
+      "step": 28800
+    },
+    {
+      "epoch": 5.213420148308916,
+      "grad_norm": 0.4414973855018616,
+      "learning_rate": 4.876957082335604e-07,
+      "loss": 0.012,
+      "step": 28825
+    },
+    {
+      "epoch": 5.217941761620546,
+      "grad_norm": 1.8146724700927734,
+      "learning_rate": 4.872352182722416e-07,
+      "loss": 0.0213,
+      "step": 28850
+    },
+    {
+      "epoch": 5.222463374932175,
+      "grad_norm": 0.3021218478679657,
+      "learning_rate": 4.867747283109228e-07,
+      "loss": 0.0052,
+      "step": 28875
+    },
+    {
+      "epoch": 5.226984988243806,
+      "grad_norm": 0.32385706901550293,
+      "learning_rate": 4.863142383496039e-07,
+      "loss": 0.006,
+      "step": 28900
+    },
+    {
+      "epoch": 5.231506601555435,
+      "grad_norm": 0.3742549419403076,
+      "learning_rate": 4.858537483882851e-07,
+      "loss": 0.0088,
+      "step": 28925
+    },
+    {
+      "epoch": 5.236028214867065,
+      "grad_norm": 0.1462940126657486,
+      "learning_rate": 4.853932584269663e-07,
+      "loss": 0.003,
+      "step": 28950
+    },
+    {
+      "epoch": 5.240549828178694,
+      "grad_norm": 0.248609721660614,
+      "learning_rate": 4.849327684656474e-07,
+      "loss": 0.0124,
+      "step": 28975
+    },
+    {
+      "epoch": 5.245071441490324,
+      "grad_norm": 0.13235850632190704,
+      "learning_rate": 4.844722785043286e-07,
+      "loss": 0.0111,
+      "step": 29000
+    },
+    {
+      "epoch": 5.249593054801953,
+      "grad_norm": 0.1916670799255371,
+      "learning_rate": 4.840117885430098e-07,
+      "loss": 0.0034,
+      "step": 29025
+    },
+    {
+      "epoch": 5.254114668113583,
+      "grad_norm": 0.0536821186542511,
+      "learning_rate": 4.835512985816909e-07,
+      "loss": 0.0025,
+      "step": 29050
+    },
+    {
+      "epoch": 5.258636281425212,
+      "grad_norm": 0.03594660758972168,
+      "learning_rate": 4.83090808620372e-07,
+      "loss": 0.0005,
+      "step": 29075
+    },
+    {
+      "epoch": 5.2631578947368425,
+      "grad_norm": 0.012173148803412914,
+      "learning_rate": 4.826303186590533e-07,
+      "loss": 0.0037,
+      "step": 29100
+    },
+    {
+      "epoch": 5.267679508048472,
+      "grad_norm": 0.033210985362529755,
+      "learning_rate": 4.821698286977343e-07,
+      "loss": 0.0014,
+      "step": 29125
+    },
+    {
+      "epoch": 5.272201121360101,
+      "grad_norm": 0.131244495511055,
+      "learning_rate": 4.817093387364155e-07,
+      "loss": 0.0023,
+      "step": 29150
+    },
+    {
+      "epoch": 5.276722734671731,
+      "grad_norm": 6.246963977813721,
+      "learning_rate": 4.812488487750967e-07,
+      "loss": 0.002,
+      "step": 29175
+    },
+    {
+      "epoch": 5.28124434798336,
+      "grad_norm": 0.4974225163459778,
+      "learning_rate": 4.807883588137778e-07,
+      "loss": 0.0056,
+      "step": 29200
+    },
+    {
+      "epoch": 5.28576596129499,
+      "grad_norm": 16.300535202026367,
+      "learning_rate": 4.80327868852459e-07,
+      "loss": 0.0446,
+      "step": 29225
+    },
+    {
+      "epoch": 5.290287574606619,
+      "grad_norm": 2.680767059326172,
+      "learning_rate": 4.798673788911402e-07,
+      "loss": 0.0483,
+      "step": 29250
+    },
+    {
+      "epoch": 5.294809187918249,
+      "grad_norm": 8.133724212646484,
+      "learning_rate": 4.794068889298213e-07,
+      "loss": 0.0121,
+      "step": 29275
+    },
+    {
+      "epoch": 5.299330801229879,
+      "grad_norm": 0.09077363461256027,
+      "learning_rate": 4.789463989685025e-07,
+      "loss": 0.0048,
+      "step": 29300
+    },
+    {
+      "epoch": 5.303852414541509,
+      "grad_norm": 36.61612319946289,
+      "learning_rate": 4.784859090071836e-07,
+      "loss": 0.0182,
+      "step": 29325
+    },
+    {
+      "epoch": 5.308374027853138,
+      "grad_norm": 6.475874900817871,
+      "learning_rate": 4.780254190458648e-07,
+      "loss": 0.0075,
+      "step": 29350
+    },
+    {
+      "epoch": 5.312895641164768,
+      "grad_norm": 0.09107718616724014,
+      "learning_rate": 4.775649290845459e-07,
+      "loss": 0.003,
+      "step": 29375
+    },
+    {
+      "epoch": 5.317417254476397,
+      "grad_norm": 0.0676039382815361,
+      "learning_rate": 4.771044391232271e-07,
+      "loss": 0.0139,
+      "step": 29400
+    },
+    {
+      "epoch": 5.321938867788027,
+      "grad_norm": 0.01544503029435873,
+      "learning_rate": 4.7664394916190827e-07,
+      "loss": 0.0118,
+      "step": 29425
+    },
+    {
+      "epoch": 5.326460481099656,
+      "grad_norm": 65.15055847167969,
+      "learning_rate": 4.761834592005894e-07,
+      "loss": 0.0057,
+      "step": 29450
+    },
+    {
+      "epoch": 5.330982094411286,
+      "grad_norm": 0.08361693471670151,
+      "learning_rate": 4.7572296923927055e-07,
+      "loss": 0.0023,
+      "step": 29475
+    },
+    {
+      "epoch": 5.335503707722916,
+      "grad_norm": 0.10150730609893799,
+      "learning_rate": 4.752624792779517e-07,
+      "loss": 0.0152,
+      "step": 29500
+    },
+    {
+      "epoch": 5.340025321034545,
+      "grad_norm": 0.7418703436851501,
+      "learning_rate": 4.748019893166329e-07,
+      "loss": 0.0181,
+      "step": 29525
+    },
+    {
+      "epoch": 5.344546934346175,
+      "grad_norm": 0.09599591046571732,
+      "learning_rate": 4.7434149935531405e-07,
+      "loss": 0.0076,
+      "step": 29550
+    },
+    {
+      "epoch": 5.349068547657804,
+      "grad_norm": 0.28607824444770813,
+      "learning_rate": 4.7388100939399516e-07,
+      "loss": 0.0053,
+      "step": 29575
+    },
+    {
+      "epoch": 5.353590160969434,
+      "grad_norm": 0.02649850957095623,
+      "learning_rate": 4.7342051943267633e-07,
+      "loss": 0.0114,
+      "step": 29600
+    },
+    {
+      "epoch": 5.358111774281063,
+      "grad_norm": 19.19407081604004,
+      "learning_rate": 4.729600294713575e-07,
+      "loss": 0.0629,
+      "step": 29625
+    },
+    {
+      "epoch": 5.362633387592693,
+      "grad_norm": 0.16849057376384735,
+      "learning_rate": 4.7251795910849143e-07,
+      "loss": 0.0308,
+      "step": 29650
+    },
+    {
+      "epoch": 5.367155000904322,
+      "grad_norm": 4.074181079864502,
+      "learning_rate": 4.720574691471726e-07,
+      "loss": 0.0024,
+      "step": 29675
+    },
+    {
+      "epoch": 5.371676614215953,
+      "grad_norm": 0.3396848142147064,
+      "learning_rate": 4.715969791858537e-07,
+      "loss": 0.019,
+      "step": 29700
+    },
+    {
+      "epoch": 5.376198227527582,
+      "grad_norm": 1.8326916694641113,
+      "learning_rate": 4.7113648922453487e-07,
+      "loss": 0.0116,
+      "step": 29725
+    },
+    {
+      "epoch": 5.380719840839212,
+      "grad_norm": 1.045644760131836,
+      "learning_rate": 4.7067599926321604e-07,
+      "loss": 0.0014,
+      "step": 29750
+    },
+    {
+      "epoch": 5.385241454150841,
+      "grad_norm": 0.034128542989492416,
+      "learning_rate": 4.702155093018972e-07,
+      "loss": 0.0032,
+      "step": 29775
+    },
+    {
+      "epoch": 5.389763067462471,
+      "grad_norm": 0.24559779465198517,
+      "learning_rate": 4.6975501934057837e-07,
+      "loss": 0.0092,
+      "step": 29800
+    },
+    {
+      "epoch": 5.3942846807741,
+      "grad_norm": 0.08045164495706558,
+      "learning_rate": 4.692945293792595e-07,
+      "loss": 0.0039,
+      "step": 29825
+    },
+    {
+      "epoch": 5.3988062940857295,
+      "grad_norm": 0.2354724407196045,
+      "learning_rate": 4.6883403941794065e-07,
+      "loss": 0.0062,
+      "step": 29850
+    },
+    {
+      "epoch": 5.403327907397359,
+      "grad_norm": 0.02797023393213749,
+      "learning_rate": 4.683735494566218e-07,
+      "loss": 0.0071,
+      "step": 29875
+    },
+    {
+      "epoch": 5.407849520708989,
+      "grad_norm": 2.4346461296081543,
+      "learning_rate": 4.67913059495303e-07,
+      "loss": 0.005,
+      "step": 29900
+    },
+    {
+      "epoch": 5.412371134020619,
+      "grad_norm": 0.08410675823688507,
+      "learning_rate": 4.6745256953398415e-07,
+      "loss": 0.0008,
+      "step": 29925
+    },
+    {
+      "epoch": 5.416892747332248,
+      "grad_norm": 1.741323709487915,
+      "learning_rate": 4.6699207957266526e-07,
+      "loss": 0.0129,
+      "step": 29950
+    },
+    {
+      "epoch": 5.421414360643878,
+      "grad_norm": 0.08235138654708862,
+      "learning_rate": 4.665315896113465e-07,
+      "loss": 0.0283,
+      "step": 29975
+    },
+    {
+      "epoch": 5.425935973955507,
+      "grad_norm": 20.520862579345703,
+      "learning_rate": 4.660710996500276e-07,
+      "loss": 0.0526,
+      "step": 30000
+    },
+    {
+      "epoch": 5.430457587267137,
+      "grad_norm": 0.11165034025907516,
+      "learning_rate": 4.6561060968870876e-07,
+      "loss": 0.0368,
+      "step": 30025
+    },
+    {
+      "epoch": 5.434979200578766,
+      "grad_norm": 2.317666530609131,
+      "learning_rate": 4.6515011972738993e-07,
+      "loss": 0.0265,
+      "step": 30050
+    },
+    {
+      "epoch": 5.439500813890396,
+      "grad_norm": 9.61573314666748,
+      "learning_rate": 4.6468962976607104e-07,
+      "loss": 0.0167,
+      "step": 30075
+    },
+    {
+      "epoch": 5.444022427202025,
+      "grad_norm": 0.6778357028961182,
+      "learning_rate": 4.6422913980475226e-07,
+      "loss": 0.0099,
+      "step": 30100
+    },
+    {
+      "epoch": 5.448544040513656,
+      "grad_norm": 9.778840065002441,
+      "learning_rate": 4.637686498434334e-07,
+      "loss": 0.0052,
+      "step": 30125
+    },
+    {
+      "epoch": 5.453065653825285,
+      "grad_norm": 1.3489108085632324,
+      "learning_rate": 4.6330815988211454e-07,
+      "loss": 0.0032,
+      "step": 30150
+    },
+    {
+      "epoch": 5.4575872671369146,
+      "grad_norm": 0.05276188254356384,
+      "learning_rate": 4.628476699207957e-07,
+      "loss": 0.0079,
+      "step": 30175
+    },
+    {
+      "epoch": 5.462108880448544,
+      "grad_norm": 2.3165993690490723,
+      "learning_rate": 4.6238717995947687e-07,
+      "loss": 0.0239,
+      "step": 30200
+    },
+    {
+      "epoch": 5.4666304937601735,
+      "grad_norm": 0.06526318937540054,
+      "learning_rate": 4.6192668999815804e-07,
+      "loss": 0.0006,
+      "step": 30225
+    },
+    {
+      "epoch": 5.471152107071803,
+      "grad_norm": 12.61938190460205,
+      "learning_rate": 4.6146620003683915e-07,
+      "loss": 0.0061,
+      "step": 30250
+    },
+    {
+      "epoch": 5.4756737203834325,
+      "grad_norm": 1.8737666606903076,
+      "learning_rate": 4.6100571007552037e-07,
+      "loss": 0.0126,
+      "step": 30275
+    },
+    {
+      "epoch": 5.480195333695063,
+      "grad_norm": 0.7588403820991516,
+      "learning_rate": 4.605452201142015e-07,
+      "loss": 0.0029,
+      "step": 30300
+    },
+    {
+      "epoch": 5.484716947006692,
+      "grad_norm": 0.04945962131023407,
+      "learning_rate": 4.6008473015288265e-07,
+      "loss": 0.0232,
+      "step": 30325
+    },
+    {
+      "epoch": 5.489238560318322,
+      "grad_norm": 0.03633030131459236,
+      "learning_rate": 4.596242401915638e-07,
+      "loss": 0.0049,
+      "step": 30350
+    },
+    {
+      "epoch": 5.493760173629951,
+      "grad_norm": 0.049966856837272644,
+      "learning_rate": 4.5916375023024493e-07,
+      "loss": 0.0095,
+      "step": 30375
+    },
+    {
+      "epoch": 5.498281786941581,
+      "grad_norm": 0.017245082184672356,
+      "learning_rate": 4.5870326026892615e-07,
+      "loss": 0.0172,
+      "step": 30400
+    },
+    {
+      "epoch": 5.50280340025321,
+      "grad_norm": 4.209415435791016,
+      "learning_rate": 4.5824277030760726e-07,
+      "loss": 0.0076,
+      "step": 30425
+    },
+    {
+      "epoch": 5.50732501356484,
+      "grad_norm": 0.1449451893568039,
+      "learning_rate": 4.5778228034628843e-07,
+      "loss": 0.0268,
+      "step": 30450
+    },
+    {
+      "epoch": 5.511846626876469,
+      "grad_norm": 6.67424201965332,
+      "learning_rate": 4.573217903849696e-07,
+      "loss": 0.0155,
+      "step": 30475
+    },
+    {
+      "epoch": 5.516368240188099,
+      "grad_norm": 0.1047978401184082,
+      "learning_rate": 4.568613004236507e-07,
+      "loss": 0.003,
+      "step": 30500
+    },
+    {
+      "epoch": 5.520889853499729,
+      "grad_norm": 11.987727165222168,
+      "learning_rate": 4.5640081046233193e-07,
+      "loss": 0.006,
+      "step": 30525
+    },
+    {
+      "epoch": 5.5254114668113585,
+      "grad_norm": 0.01995599828660488,
+      "learning_rate": 4.5594032050101304e-07,
+      "loss": 0.0024,
+      "step": 30550
+    },
+    {
+      "epoch": 5.529933080122988,
+      "grad_norm": 0.24483761191368103,
+      "learning_rate": 4.554798305396942e-07,
+      "loss": 0.0054,
+      "step": 30575
+    },
+    {
+      "epoch": 5.5344546934346175,
+      "grad_norm": 0.03671187534928322,
+      "learning_rate": 4.550193405783754e-07,
+      "loss": 0.014,
+      "step": 30600
+    },
+    {
+      "epoch": 5.538976306746247,
+      "grad_norm": 13.608981132507324,
+      "learning_rate": 4.545588506170565e-07,
+      "loss": 0.0025,
+      "step": 30625
+    },
+    {
+      "epoch": 5.5434979200578764,
+      "grad_norm": 68.79994201660156,
+      "learning_rate": 4.540983606557377e-07,
+      "loss": 0.0027,
+      "step": 30650
+    },
+    {
+      "epoch": 5.548019533369506,
+      "grad_norm": 0.06913666427135468,
+      "learning_rate": 4.536378706944188e-07,
+      "loss": 0.0039,
+      "step": 30675
+    },
+    {
+      "epoch": 5.552541146681136,
+      "grad_norm": 0.04184752330183983,
+      "learning_rate": 4.5317738073310004e-07,
+      "loss": 0.0105,
+      "step": 30700
+    },
+    {
+      "epoch": 5.557062759992766,
+      "grad_norm": 22.15312957763672,
+      "learning_rate": 4.5271689077178115e-07,
+      "loss": 0.0089,
+      "step": 30725
+    },
+    {
+      "epoch": 5.561584373304395,
+      "grad_norm": 0.17298342287540436,
+      "learning_rate": 4.5225640081046227e-07,
+      "loss": 0.0253,
+      "step": 30750
+    },
+    {
+      "epoch": 5.566105986616025,
+      "grad_norm": 17.002382278442383,
+      "learning_rate": 4.517959108491435e-07,
+      "loss": 0.024,
+      "step": 30775
+    },
+    {
+      "epoch": 5.570627599927654,
+      "grad_norm": 0.07881984114646912,
+      "learning_rate": 4.513354208878246e-07,
+      "loss": 0.0185,
+      "step": 30800
+    },
+    {
+      "epoch": 5.575149213239284,
+      "grad_norm": 1.0953749418258667,
+      "learning_rate": 4.508749309265058e-07,
+      "loss": 0.0238,
+      "step": 30825
+    },
+    {
+      "epoch": 5.579670826550913,
+      "grad_norm": 0.5996679067611694,
+      "learning_rate": 4.5041444096518693e-07,
+      "loss": 0.0408,
+      "step": 30850
+    },
+    {
+      "epoch": 5.584192439862543,
+      "grad_norm": 0.037622638046741486,
+      "learning_rate": 4.499539510038681e-07,
+      "loss": 0.0082,
+      "step": 30875
+    },
+    {
+      "epoch": 5.588714053174172,
+      "grad_norm": 0.8418008685112,
+      "learning_rate": 4.4949346104254926e-07,
+      "loss": 0.0049,
+      "step": 30900
+    },
+    {
+      "epoch": 5.5932356664858025,
+      "grad_norm": 0.01251909602433443,
+      "learning_rate": 4.490329710812304e-07,
+      "loss": 0.0055,
+      "step": 30925
+    },
+    {
+      "epoch": 5.597757279797432,
+      "grad_norm": 2.921856641769409,
+      "learning_rate": 4.485724811199116e-07,
+      "loss": 0.0089,
+      "step": 30950
+    },
+    {
+      "epoch": 5.6022788931090615,
+      "grad_norm": 45.41884231567383,
+      "learning_rate": 4.481119911585927e-07,
+      "loss": 0.0098,
+      "step": 30975
+    },
+    {
+      "epoch": 5.606800506420691,
+      "grad_norm": 11.943808555603027,
+      "learning_rate": 4.4765150119727393e-07,
+      "loss": 0.0201,
+      "step": 31000
+    },
+    {
+      "epoch": 5.61132211973232,
+      "grad_norm": 0.03331173211336136,
+      "learning_rate": 4.4719101123595504e-07,
+      "loss": 0.0075,
+      "step": 31025
+    },
+    {
+      "epoch": 5.61584373304395,
+      "grad_norm": 19.69361686706543,
+      "learning_rate": 4.4673052127463615e-07,
+      "loss": 0.011,
+      "step": 31050
+    },
+    {
+      "epoch": 5.620365346355579,
+      "grad_norm": 1.9068259000778198,
+      "learning_rate": 4.4627003131331737e-07,
+      "loss": 0.0067,
+      "step": 31075
+    },
+    {
+      "epoch": 5.62488695966721,
+      "grad_norm": 0.019552985206246376,
+      "learning_rate": 4.458095413519985e-07,
+      "loss": 0.0192,
+      "step": 31100
+    },
+    {
+      "epoch": 5.629408572978839,
+      "grad_norm": 0.02558848075568676,
+      "learning_rate": 4.453490513906797e-07,
+      "loss": 0.0064,
+      "step": 31125
+    },
+    {
+      "epoch": 5.633930186290469,
+      "grad_norm": 0.023367932066321373,
+      "learning_rate": 4.448885614293608e-07,
+      "loss": 0.0093,
+      "step": 31150
+    },
+    {
+      "epoch": 5.638451799602098,
+      "grad_norm": 0.04951472207903862,
+      "learning_rate": 4.4442807146804193e-07,
+      "loss": 0.0056,
+      "step": 31175
+    },
+    {
+      "epoch": 5.642973412913728,
+      "grad_norm": 0.042895007878541946,
+      "learning_rate": 4.4396758150672315e-07,
+      "loss": 0.0086,
+      "step": 31200
+    },
+    {
+      "epoch": 5.647495026225357,
+      "grad_norm": 3.2334864139556885,
+      "learning_rate": 4.4350709154540426e-07,
+      "loss": 0.0169,
+      "step": 31225
+    },
+    {
+      "epoch": 5.652016639536987,
+      "grad_norm": 4.943774700164795,
+      "learning_rate": 4.430466015840855e-07,
+      "loss": 0.0401,
+      "step": 31250
+    },
+    {
+      "epoch": 5.656538252848616,
+      "grad_norm": 0.135379821062088,
+      "learning_rate": 4.425861116227666e-07,
+      "loss": 0.0066,
+      "step": 31275
+    },
+    {
+      "epoch": 5.661059866160246,
+      "grad_norm": 0.023317676037549973,
+      "learning_rate": 4.421256216614477e-07,
+      "loss": 0.0035,
+      "step": 31300
+    },
+    {
+      "epoch": 5.665581479471875,
+      "grad_norm": 0.1724722981452942,
+      "learning_rate": 4.4166513170012893e-07,
+      "loss": 0.0106,
+      "step": 31325
+    },
+    {
+      "epoch": 5.670103092783505,
+      "grad_norm": 0.0674796774983406,
+      "learning_rate": 4.4120464173881004e-07,
+      "loss": 0.0078,
+      "step": 31350
+    },
+    {
+      "epoch": 5.674624706095135,
+      "grad_norm": 0.15505366027355194,
+      "learning_rate": 4.4074415177749126e-07,
+      "loss": 0.0069,
+      "step": 31375
+    },
+    {
+      "epoch": 5.679146319406764,
+      "grad_norm": 7.8905134201049805,
+      "learning_rate": 4.402836618161724e-07,
+      "loss": 0.0128,
+      "step": 31400
+    },
+    {
+      "epoch": 5.683667932718394,
+      "grad_norm": 24.007476806640625,
+      "learning_rate": 4.398231718548536e-07,
+      "loss": 0.0102,
+      "step": 31425
+    },
+    {
+      "epoch": 5.688189546030023,
+      "grad_norm": 0.03173492103815079,
+      "learning_rate": 4.393626818935347e-07,
+      "loss": 0.0056,
+      "step": 31450
+    },
+    {
+      "epoch": 5.692711159341653,
+      "grad_norm": 0.4066329300403595,
+      "learning_rate": 4.389021919322158e-07,
+      "loss": 0.0019,
+      "step": 31475
+    },
+    {
+      "epoch": 5.697232772653282,
+      "grad_norm": 10.2057466506958,
+      "learning_rate": 4.3844170197089704e-07,
+      "loss": 0.0063,
+      "step": 31500
+    },
+    {
+      "epoch": 5.701754385964913,
+      "grad_norm": 0.02208542451262474,
+      "learning_rate": 4.3798121200957815e-07,
+      "loss": 0.0202,
+      "step": 31525
+    },
+    {
+      "epoch": 5.706275999276542,
+      "grad_norm": 0.38067665696144104,
+      "learning_rate": 4.3752072204825937e-07,
+      "loss": 0.0076,
+      "step": 31550
+    },
+    {
+      "epoch": 5.710797612588172,
+      "grad_norm": 0.04764077439904213,
+      "learning_rate": 4.370602320869405e-07,
+      "loss": 0.0016,
+      "step": 31575
+    },
+    {
+      "epoch": 5.715319225899801,
+      "grad_norm": 0.04114853963255882,
+      "learning_rate": 4.365997421256216e-07,
+      "loss": 0.0109,
+      "step": 31600
+    },
+    {
+      "epoch": 5.719840839211431,
+      "grad_norm": 0.229792058467865,
+      "learning_rate": 4.361392521643028e-07,
+      "loss": 0.0155,
+      "step": 31625
+    },
+    {
+      "epoch": 5.72436245252306,
+      "grad_norm": 2.7913074493408203,
+      "learning_rate": 4.3567876220298393e-07,
+      "loss": 0.0199,
+      "step": 31650
+    },
+    {
+      "epoch": 5.72888406583469,
+      "grad_norm": 0.0458095520734787,
+      "learning_rate": 4.3521827224166515e-07,
+      "loss": 0.0096,
+      "step": 31675
+    },
+    {
+      "epoch": 5.733405679146319,
+      "grad_norm": 2.6488194465637207,
+      "learning_rate": 4.3475778228034626e-07,
+      "loss": 0.0095,
+      "step": 31700
+    },
+    {
+      "epoch": 5.7379272924579485,
+      "grad_norm": 0.013299129903316498,
+      "learning_rate": 4.3429729231902743e-07,
+      "loss": 0.0081,
+      "step": 31725
+    },
+    {
+      "epoch": 5.742448905769579,
+      "grad_norm": 0.6266424059867859,
+      "learning_rate": 4.338368023577086e-07,
+      "loss": 0.0009,
+      "step": 31750
+    },
+    {
+      "epoch": 5.746970519081208,
+      "grad_norm": 13.529029846191406,
+      "learning_rate": 4.333763123963897e-07,
+      "loss": 0.0102,
+      "step": 31775
+    },
+    {
+      "epoch": 5.751492132392838,
+      "grad_norm": 0.0697498694062233,
+      "learning_rate": 4.3291582243507093e-07,
+      "loss": 0.0053,
+      "step": 31800
+    },
+    {
+      "epoch": 5.756013745704467,
+      "grad_norm": 0.053803663700819016,
+      "learning_rate": 4.3245533247375204e-07,
+      "loss": 0.0016,
+      "step": 31825
+    },
+    {
+      "epoch": 5.760535359016097,
+      "grad_norm": 0.23834507167339325,
+      "learning_rate": 4.319948425124332e-07,
+      "loss": 0.0018,
+      "step": 31850
+    },
+    {
+      "epoch": 5.765056972327726,
+      "grad_norm": 0.01928309164941311,
+      "learning_rate": 4.315343525511144e-07,
+      "loss": 0.019,
+      "step": 31875
+    },
+    {
+      "epoch": 5.769578585639356,
+      "grad_norm": 0.20884621143341064,
+      "learning_rate": 4.310738625897955e-07,
+      "loss": 0.0331,
+      "step": 31900
+    },
+    {
+      "epoch": 5.774100198950986,
+      "grad_norm": 28.804271697998047,
+      "learning_rate": 4.306133726284767e-07,
+      "loss": 0.0103,
+      "step": 31925
+    },
+    {
+      "epoch": 5.778621812262616,
+      "grad_norm": 16.59792137145996,
+      "learning_rate": 4.301528826671578e-07,
+      "loss": 0.005,
+      "step": 31950
+    },
+    {
+      "epoch": 5.783143425574245,
+      "grad_norm": 13.525711059570312,
+      "learning_rate": 4.29692392705839e-07,
+      "loss": 0.0197,
+      "step": 31975
+    },
+    {
+      "epoch": 5.787665038885875,
+      "grad_norm": 0.2153875082731247,
+      "learning_rate": 4.2923190274452015e-07,
+      "loss": 0.0028,
+      "step": 32000
+    },
+    {
+      "epoch": 5.792186652197504,
+      "grad_norm": 21.80754852294922,
+      "learning_rate": 4.2877141278320127e-07,
+      "loss": 0.0509,
+      "step": 32025
+    },
+    {
+      "epoch": 5.7967082655091335,
+      "grad_norm": 0.027552086859941483,
+      "learning_rate": 4.283109228218825e-07,
+      "loss": 0.0455,
+      "step": 32050
+    },
+    {
+      "epoch": 5.801229878820763,
+      "grad_norm": 0.5974065065383911,
+      "learning_rate": 4.278504328605636e-07,
+      "loss": 0.003,
+      "step": 32075
+    },
+    {
+      "epoch": 5.8057514921323925,
+      "grad_norm": 0.0764407068490982,
+      "learning_rate": 4.273899428992448e-07,
+      "loss": 0.0073,
+      "step": 32100
+    },
+    {
+      "epoch": 5.810273105444022,
+      "grad_norm": 0.00713876448571682,
+      "learning_rate": 4.2692945293792593e-07,
+      "loss": 0.0016,
+      "step": 32125
+    },
+    {
+      "epoch": 5.814794718755652,
+      "grad_norm": 2.5511715412139893,
+      "learning_rate": 4.264689629766071e-07,
+      "loss": 0.0018,
+      "step": 32150
+    },
+    {
+      "epoch": 5.819316332067282,
+      "grad_norm": 0.024245627224445343,
+      "learning_rate": 4.2600847301528826e-07,
+      "loss": 0.0072,
+      "step": 32175
+    },
+    {
+      "epoch": 5.823837945378911,
+      "grad_norm": 0.7642532587051392,
+      "learning_rate": 4.255479830539694e-07,
+      "loss": 0.0063,
+      "step": 32200
+    },
+    {
+      "epoch": 5.828359558690541,
+      "grad_norm": 0.008916143327951431,
+      "learning_rate": 4.250874930926506e-07,
+      "loss": 0.0016,
+      "step": 32225
+    },
+    {
+      "epoch": 5.83288117200217,
+      "grad_norm": 0.04523088410496712,
+      "learning_rate": 4.246270031313317e-07,
+      "loss": 0.0068,
+      "step": 32250
+    },
+    {
+      "epoch": 5.8374027853138,
+      "grad_norm": 0.01864382065832615,
+      "learning_rate": 4.241665131700129e-07,
+      "loss": 0.0062,
+      "step": 32275
+    },
+    {
+      "epoch": 5.841924398625429,
+      "grad_norm": 0.09099319577217102,
+      "learning_rate": 4.2370602320869404e-07,
+      "loss": 0.0081,
+      "step": 32300
+    },
+    {
+      "epoch": 5.84644601193706,
+      "grad_norm": 0.14763249456882477,
+      "learning_rate": 4.2324553324737516e-07,
+      "loss": 0.008,
+      "step": 32325
+    },
+    {
+      "epoch": 5.850967625248689,
+      "grad_norm": 0.08158791810274124,
+      "learning_rate": 4.227850432860564e-07,
+      "loss": 0.0096,
+      "step": 32350
+    },
+    {
+      "epoch": 5.8554892385603186,
+      "grad_norm": 0.14870816469192505,
+      "learning_rate": 4.223245533247375e-07,
+      "loss": 0.0019,
+      "step": 32375
+    },
+    {
+      "epoch": 5.860010851871948,
+      "grad_norm": 0.02300347574055195,
+      "learning_rate": 4.2186406336341865e-07,
+      "loss": 0.0054,
+      "step": 32400
+    },
+    {
+      "epoch": 5.8645324651835775,
+      "grad_norm": 1.2022995948791504,
+      "learning_rate": 4.214035734020998e-07,
+      "loss": 0.0355,
+      "step": 32425
+    },
+    {
+      "epoch": 5.869054078495207,
+      "grad_norm": 1.5367909669876099,
+      "learning_rate": 4.20943083440781e-07,
+      "loss": 0.0319,
+      "step": 32450
+    },
+    {
+      "epoch": 5.8735756918068365,
+      "grad_norm": 0.14563749730587006,
+      "learning_rate": 4.2048259347946215e-07,
+      "loss": 0.0059,
+      "step": 32475
+    },
+    {
+      "epoch": 5.878097305118466,
+      "grad_norm": 0.21176332235336304,
+      "learning_rate": 4.2002210351814327e-07,
+      "loss": 0.0023,
+      "step": 32500
+    },
+    {
+      "epoch": 5.882618918430095,
+      "grad_norm": 0.14737099409103394,
+      "learning_rate": 4.1956161355682443e-07,
+      "loss": 0.0029,
+      "step": 32525
+    },
+    {
+      "epoch": 5.887140531741726,
+      "grad_norm": 0.05561397597193718,
+      "learning_rate": 4.191011235955056e-07,
+      "loss": 0.0041,
+      "step": 32550
+    },
+    {
+      "epoch": 5.891662145053355,
+      "grad_norm": 0.1818033903837204,
+      "learning_rate": 4.1864063363418676e-07,
+      "loss": 0.0079,
+      "step": 32575
+    },
+    {
+      "epoch": 5.896183758364985,
+      "grad_norm": 6.998167514801025,
+      "learning_rate": 4.1818014367286793e-07,
+      "loss": 0.0076,
+      "step": 32600
+    },
+    {
+      "epoch": 5.900705371676614,
+      "grad_norm": 0.168256476521492,
+      "learning_rate": 4.1771965371154904e-07,
+      "loss": 0.0046,
+      "step": 32625
+    },
+    {
+      "epoch": 5.905226984988244,
+      "grad_norm": 0.041657544672489166,
+      "learning_rate": 4.172591637502302e-07,
+      "loss": 0.0054,
+      "step": 32650
+    },
+    {
+      "epoch": 5.909748598299873,
+      "grad_norm": 40.278446197509766,
+      "learning_rate": 4.167986737889114e-07,
+      "loss": 0.0125,
+      "step": 32675
+    },
+    {
+      "epoch": 5.914270211611503,
+      "grad_norm": 0.1068694144487381,
+      "learning_rate": 4.1633818382759254e-07,
+      "loss": 0.0023,
+      "step": 32700
+    },
+    {
+      "epoch": 5.918791824923133,
+      "grad_norm": 0.24333126842975616,
+      "learning_rate": 4.158776938662737e-07,
+      "loss": 0.0289,
+      "step": 32725
+    },
+    {
+      "epoch": 5.9233134382347625,
+      "grad_norm": 0.16926701366901398,
+      "learning_rate": 4.154172039049548e-07,
+      "loss": 0.0123,
+      "step": 32750
+    },
+    {
+      "epoch": 5.927835051546392,
+      "grad_norm": 3.9394893646240234,
+      "learning_rate": 4.1495671394363604e-07,
+      "loss": 0.0244,
+      "step": 32775
+    },
+    {
+      "epoch": 5.9323566648580215,
+      "grad_norm": 0.06976808607578278,
+      "learning_rate": 4.1449622398231715e-07,
+      "loss": 0.0098,
+      "step": 32800
+    },
+    {
+      "epoch": 5.936878278169651,
+      "grad_norm": 24.640043258666992,
+      "learning_rate": 4.140357340209983e-07,
+      "loss": 0.0225,
+      "step": 32825
+    },
+    {
+      "epoch": 5.9413998914812804,
+      "grad_norm": 4.303247451782227,
+      "learning_rate": 4.135752440596795e-07,
+      "loss": 0.0136,
+      "step": 32850
+    },
+    {
+      "epoch": 5.94592150479291,
+      "grad_norm": 0.46884116530418396,
+      "learning_rate": 4.1311475409836065e-07,
+      "loss": 0.0156,
+      "step": 32875
+    },
+    {
+      "epoch": 5.950443118104539,
+      "grad_norm": 0.9783020615577698,
+      "learning_rate": 4.126542641370418e-07,
+      "loss": 0.0114,
+      "step": 32900
+    },
+    {
+      "epoch": 5.954964731416169,
+      "grad_norm": 0.5886393785476685,
+      "learning_rate": 4.1219377417572293e-07,
+      "loss": 0.0077,
+      "step": 32925
+    },
+    {
+      "epoch": 5.959486344727799,
+      "grad_norm": 0.019434532150626183,
+      "learning_rate": 4.117332842144041e-07,
+      "loss": 0.0025,
+      "step": 32950
+    },
+    {
+      "epoch": 5.964007958039429,
+      "grad_norm": 2.6205480098724365,
+      "learning_rate": 4.1127279425308527e-07,
+      "loss": 0.0026,
+      "step": 32975
+    },
+    {
+      "epoch": 5.968529571351058,
+      "grad_norm": 0.25357627868652344,
+      "learning_rate": 4.1081230429176643e-07,
+      "loss": 0.0056,
+      "step": 33000
+    },
+    {
+      "epoch": 5.973051184662688,
+      "grad_norm": 9.764650344848633,
+      "learning_rate": 4.103518143304476e-07,
+      "loss": 0.005,
+      "step": 33025
+    },
+    {
+      "epoch": 5.977572797974317,
+      "grad_norm": 0.15983889997005463,
+      "learning_rate": 4.098913243691287e-07,
+      "loss": 0.0093,
+      "step": 33050
+    },
+    {
+      "epoch": 5.982094411285947,
+      "grad_norm": 0.44210419058799744,
+      "learning_rate": 4.094308344078099e-07,
+      "loss": 0.006,
+      "step": 33075
+    },
+    {
+      "epoch": 5.986616024597576,
+      "grad_norm": 7.316641807556152,
+      "learning_rate": 4.0897034444649104e-07,
+      "loss": 0.0096,
+      "step": 33100
+    },
+    {
+      "epoch": 5.9911376379092065,
+      "grad_norm": 0.040955204516649246,
+      "learning_rate": 4.085098544851722e-07,
+      "loss": 0.002,
+      "step": 33125
+    },
+    {
+      "epoch": 5.995659251220836,
+      "grad_norm": 28.565380096435547,
+      "learning_rate": 4.080493645238534e-07,
+      "loss": 0.0204,
+      "step": 33150
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.35324448347091675,
+      "eval_runtime": 8693.2226,
+      "eval_samples_per_second": 1.092,
+      "eval_steps_per_second": 0.137,
+      "eval_wer": 0.10656630365150545,
+      "step": 33174
+    },
+    {
+      "epoch": 6.0001808645324655,
+      "grad_norm": 1.004451870918274,
+      "learning_rate": 4.0760729416098725e-07,
+      "loss": 0.0388,
+      "step": 33175
+    },
+    {
+      "epoch": 6.004702477844095,
+      "grad_norm": 0.3171859383583069,
+      "learning_rate": 4.071468041996684e-07,
+      "loss": 0.0116,
+      "step": 33200
+    },
+    {
+      "epoch": 6.009224091155724,
+      "grad_norm": 0.3744235634803772,
+      "learning_rate": 4.066863142383496e-07,
+      "loss": 0.0065,
+      "step": 33225
+    },
+    {
+      "epoch": 6.013745704467354,
+      "grad_norm": 0.04872240498661995,
+      "learning_rate": 4.0622582427703075e-07,
+      "loss": 0.0118,
+      "step": 33250
+    },
+    {
+      "epoch": 6.018267317778983,
+      "grad_norm": 0.20710858702659607,
+      "learning_rate": 4.057653343157119e-07,
+      "loss": 0.0018,
+      "step": 33275
+    },
+    {
+      "epoch": 6.022788931090613,
+      "grad_norm": 0.6177895665168762,
+      "learning_rate": 4.0530484435439303e-07,
+      "loss": 0.0079,
+      "step": 33300
+    },
+    {
+      "epoch": 6.027310544402242,
+      "grad_norm": 0.042505986988544464,
+      "learning_rate": 4.0484435439307425e-07,
+      "loss": 0.0122,
+      "step": 33325
+    },
+    {
+      "epoch": 6.031832157713873,
+      "grad_norm": 0.018311861902475357,
+      "learning_rate": 4.0438386443175536e-07,
+      "loss": 0.0019,
+      "step": 33350
+    },
+    {
+      "epoch": 6.036353771025502,
+      "grad_norm": 0.05716663971543312,
+      "learning_rate": 4.0392337447043653e-07,
+      "loss": 0.0006,
+      "step": 33375
+    },
+    {
+      "epoch": 6.040875384337132,
+      "grad_norm": 0.03125373646616936,
+      "learning_rate": 4.034628845091177e-07,
+      "loss": 0.0008,
+      "step": 33400
+    },
+    {
+      "epoch": 6.045396997648761,
+      "grad_norm": 0.04944000765681267,
+      "learning_rate": 4.030023945477988e-07,
+      "loss": 0.0112,
+      "step": 33425
+    },
+    {
+      "epoch": 6.049918610960391,
+      "grad_norm": 3.482018232345581,
+      "learning_rate": 4.0254190458648003e-07,
+      "loss": 0.0068,
+      "step": 33450
+    },
+    {
+      "epoch": 6.05444022427202,
+      "grad_norm": 0.051414769142866135,
+      "learning_rate": 4.0208141462516114e-07,
+      "loss": 0.0032,
+      "step": 33475
+    },
+    {
+      "epoch": 6.05896183758365,
+      "grad_norm": 1.4644255638122559,
+      "learning_rate": 4.016209246638423e-07,
+      "loss": 0.0123,
+      "step": 33500
+    },
+    {
+      "epoch": 6.063483450895279,
+      "grad_norm": 0.038726359605789185,
+      "learning_rate": 4.011604347025235e-07,
+      "loss": 0.0028,
+      "step": 33525
+    },
+    {
+      "epoch": 6.068005064206909,
+      "grad_norm": 0.45498254895210266,
+      "learning_rate": 4.006999447412046e-07,
+      "loss": 0.019,
+      "step": 33550
+    },
+    {
+      "epoch": 6.072526677518539,
+      "grad_norm": 2.0771706104278564,
+      "learning_rate": 4.002394547798858e-07,
+      "loss": 0.0319,
+      "step": 33575
+    },
+    {
+      "epoch": 6.077048290830168,
+      "grad_norm": 0.0325147807598114,
+      "learning_rate": 3.997789648185669e-07,
+      "loss": 0.0073,
+      "step": 33600
+    },
+    {
+      "epoch": 6.081569904141798,
+      "grad_norm": 0.1332240253686905,
+      "learning_rate": 3.993184748572481e-07,
+      "loss": 0.0051,
+      "step": 33625
+    },
+    {
+      "epoch": 6.086091517453427,
+      "grad_norm": 0.13362543284893036,
+      "learning_rate": 3.9885798489592925e-07,
+      "loss": 0.0042,
+      "step": 33650
+    },
+    {
+      "epoch": 6.090613130765057,
+      "grad_norm": 0.09919251501560211,
+      "learning_rate": 3.9839749493461037e-07,
+      "loss": 0.0028,
+      "step": 33675
+    },
+    {
+      "epoch": 6.095134744076686,
+      "grad_norm": 0.2754238545894623,
+      "learning_rate": 3.979370049732916e-07,
+      "loss": 0.0031,
+      "step": 33700
+    },
+    {
+      "epoch": 6.099656357388316,
+      "grad_norm": 0.028968511149287224,
+      "learning_rate": 3.974765150119727e-07,
+      "loss": 0.0141,
+      "step": 33725
+    },
+    {
+      "epoch": 6.104177970699946,
+      "grad_norm": 0.2524532675743103,
+      "learning_rate": 3.970160250506539e-07,
+      "loss": 0.0071,
+      "step": 33750
+    },
+    {
+      "epoch": 6.108699584011576,
+      "grad_norm": 0.031802672892808914,
+      "learning_rate": 3.9655553508933503e-07,
+      "loss": 0.0012,
+      "step": 33775
+    },
+    {
+      "epoch": 6.113221197323205,
+      "grad_norm": 0.05196872353553772,
+      "learning_rate": 3.9609504512801614e-07,
+      "loss": 0.0072,
+      "step": 33800
+    },
+    {
+      "epoch": 6.117742810634835,
+      "grad_norm": 0.03085019811987877,
+      "learning_rate": 3.9563455516669736e-07,
+      "loss": 0.002,
+      "step": 33825
+    },
+    {
+      "epoch": 6.122264423946464,
+      "grad_norm": 0.0648372620344162,
+      "learning_rate": 3.951740652053785e-07,
+      "loss": 0.0254,
+      "step": 33850
+    },
+    {
+      "epoch": 6.126786037258094,
+      "grad_norm": 0.00866938941180706,
+      "learning_rate": 3.947135752440597e-07,
+      "loss": 0.0108,
+      "step": 33875
+    },
+    {
+      "epoch": 6.131307650569723,
+      "grad_norm": 0.19999012351036072,
+      "learning_rate": 3.942530852827408e-07,
+      "loss": 0.0069,
+      "step": 33900
+    },
+    {
+      "epoch": 6.1358292638813525,
+      "grad_norm": 0.04982515051960945,
+      "learning_rate": 3.93792595321422e-07,
+      "loss": 0.0016,
+      "step": 33925
+    },
+    {
+      "epoch": 6.140350877192983,
+      "grad_norm": 2.1079819202423096,
+      "learning_rate": 3.9333210536010314e-07,
+      "loss": 0.0117,
+      "step": 33950
+    },
+    {
+      "epoch": 6.144872490504612,
+      "grad_norm": 0.12225229293107986,
+      "learning_rate": 3.9287161539878426e-07,
+      "loss": 0.0389,
+      "step": 33975
+    },
+    {
+      "epoch": 6.149394103816242,
+      "grad_norm": 0.014936073683202267,
+      "learning_rate": 3.924111254374655e-07,
+      "loss": 0.0173,
+      "step": 34000
+    },
+    {
+      "epoch": 6.153915717127871,
+      "grad_norm": 0.05773633345961571,
+      "learning_rate": 3.919506354761466e-07,
+      "loss": 0.0046,
+      "step": 34025
+    },
+    {
+      "epoch": 6.158437330439501,
+      "grad_norm": 0.7598387598991394,
+      "learning_rate": 3.914901455148278e-07,
+      "loss": 0.013,
+      "step": 34050
+    },
+    {
+      "epoch": 6.16295894375113,
+      "grad_norm": 48.75155258178711,
+      "learning_rate": 3.910296555535089e-07,
+      "loss": 0.0044,
+      "step": 34075
+    },
+    {
+      "epoch": 6.16748055706276,
+      "grad_norm": 0.024905025959014893,
+      "learning_rate": 3.9056916559219003e-07,
+      "loss": 0.0021,
+      "step": 34100
+    },
+    {
+      "epoch": 6.172002170374389,
+      "grad_norm": 0.01038370281457901,
+      "learning_rate": 3.9010867563087125e-07,
+      "loss": 0.0016,
+      "step": 34125
+    },
+    {
+      "epoch": 6.176523783686019,
+      "grad_norm": 2.034224510192871,
+      "learning_rate": 3.8964818566955237e-07,
+      "loss": 0.0096,
+      "step": 34150
+    },
+    {
+      "epoch": 6.181045396997649,
+      "grad_norm": 1.9965308904647827,
+      "learning_rate": 3.891876957082336e-07,
+      "loss": 0.0176,
+      "step": 34175
+    },
+    {
+      "epoch": 6.185567010309279,
+      "grad_norm": 108.9871826171875,
+      "learning_rate": 3.887272057469147e-07,
+      "loss": 0.0172,
+      "step": 34200
+    },
+    {
+      "epoch": 6.190088623620908,
+      "grad_norm": 0.04911862686276436,
+      "learning_rate": 3.882667157855958e-07,
+      "loss": 0.0012,
+      "step": 34225
+    },
+    {
+      "epoch": 6.1946102369325375,
+      "grad_norm": 2.3518688678741455,
+      "learning_rate": 3.8780622582427703e-07,
+      "loss": 0.003,
+      "step": 34250
+    },
+    {
+      "epoch": 6.199131850244167,
+      "grad_norm": 0.08073610067367554,
+      "learning_rate": 3.8734573586295814e-07,
+      "loss": 0.0228,
+      "step": 34275
+    },
+    {
+      "epoch": 6.2036534635557965,
+      "grad_norm": 0.04245547577738762,
+      "learning_rate": 3.8688524590163936e-07,
+      "loss": 0.0008,
+      "step": 34300
+    },
+    {
+      "epoch": 6.208175076867426,
+      "grad_norm": 0.04015149176120758,
+      "learning_rate": 3.864247559403205e-07,
+      "loss": 0.003,
+      "step": 34325
+    },
+    {
+      "epoch": 6.2126966901790555,
+      "grad_norm": 0.040025342255830765,
+      "learning_rate": 3.859642659790016e-07,
+      "loss": 0.0283,
+      "step": 34350
+    },
+    {
+      "epoch": 6.217218303490686,
+      "grad_norm": 4.162205696105957,
+      "learning_rate": 3.855037760176828e-07,
+      "loss": 0.0259,
+      "step": 34375
+    },
+    {
+      "epoch": 6.221739916802315,
+      "grad_norm": 0.10629791021347046,
+      "learning_rate": 3.850432860563639e-07,
+      "loss": 0.0046,
+      "step": 34400
+    },
+    {
+      "epoch": 6.226261530113945,
+      "grad_norm": 0.05209145322442055,
+      "learning_rate": 3.8458279609504514e-07,
+      "loss": 0.0063,
+      "step": 34425
+    },
+    {
+      "epoch": 6.230783143425574,
+      "grad_norm": 0.03029513917863369,
+      "learning_rate": 3.8412230613372625e-07,
+      "loss": 0.0035,
+      "step": 34450
+    },
+    {
+      "epoch": 6.235304756737204,
+      "grad_norm": 0.0416143536567688,
+      "learning_rate": 3.836618161724075e-07,
+      "loss": 0.0064,
+      "step": 34475
+    },
+    {
+      "epoch": 6.239826370048833,
+      "grad_norm": 17.89125633239746,
+      "learning_rate": 3.832013262110886e-07,
+      "loss": 0.0103,
+      "step": 34500
+    },
+    {
+      "epoch": 6.244347983360463,
+      "grad_norm": 0.08937986195087433,
+      "learning_rate": 3.827408362497697e-07,
+      "loss": 0.0102,
+      "step": 34525
+    },
+    {
+      "epoch": 6.248869596672092,
+      "grad_norm": 0.012076592072844505,
+      "learning_rate": 3.822803462884509e-07,
+      "loss": 0.0012,
+      "step": 34550
+    },
+    {
+      "epoch": 6.2533912099837226,
+      "grad_norm": 13.431784629821777,
+      "learning_rate": 3.8181985632713203e-07,
+      "loss": 0.0029,
+      "step": 34575
+    },
+    {
+      "epoch": 6.257912823295352,
+      "grad_norm": 0.15255410969257355,
+      "learning_rate": 3.8135936636581325e-07,
+      "loss": 0.0056,
+      "step": 34600
+    },
+    {
+      "epoch": 6.2624344366069815,
+      "grad_norm": 0.7227760553359985,
+      "learning_rate": 3.8089887640449437e-07,
+      "loss": 0.0048,
+      "step": 34625
+    },
+    {
+      "epoch": 6.266956049918611,
+      "grad_norm": 0.022182561457157135,
+      "learning_rate": 3.804383864431755e-07,
+      "loss": 0.0055,
+      "step": 34650
+    },
+    {
+      "epoch": 6.2714776632302405,
+      "grad_norm": 0.10832487791776657,
+      "learning_rate": 3.799778964818567e-07,
+      "loss": 0.0141,
+      "step": 34675
+    },
+    {
+      "epoch": 6.27599927654187,
+      "grad_norm": 0.08272892981767654,
+      "learning_rate": 3.795174065205378e-07,
+      "loss": 0.022,
+      "step": 34700
+    },
+    {
+      "epoch": 6.280520889853499,
+      "grad_norm": 0.10777498036623001,
+      "learning_rate": 3.7905691655921903e-07,
+      "loss": 0.0062,
+      "step": 34725
+    },
+    {
+      "epoch": 6.285042503165129,
+      "grad_norm": 0.1464318186044693,
+      "learning_rate": 3.7859642659790014e-07,
+      "loss": 0.0078,
+      "step": 34750
+    },
+    {
+      "epoch": 6.289564116476759,
+      "grad_norm": 2.59875750541687,
+      "learning_rate": 3.781359366365813e-07,
+      "loss": 0.0364,
+      "step": 34775
+    },
+    {
+      "epoch": 6.294085729788389,
+      "grad_norm": 0.6128166913986206,
+      "learning_rate": 3.776754466752625e-07,
+      "loss": 0.0119,
+      "step": 34800
+    },
+    {
+      "epoch": 6.298607343100018,
+      "grad_norm": 0.3331466019153595,
+      "learning_rate": 3.772149567139436e-07,
+      "loss": 0.0147,
+      "step": 34825
+    },
+    {
+      "epoch": 6.303128956411648,
+      "grad_norm": 1.0157824754714966,
+      "learning_rate": 3.767544667526248e-07,
+      "loss": 0.005,
+      "step": 34850
+    },
+    {
+      "epoch": 6.307650569723277,
+      "grad_norm": 0.20274540781974792,
+      "learning_rate": 3.762939767913059e-07,
+      "loss": 0.0027,
+      "step": 34875
+    },
+    {
+      "epoch": 6.312172183034907,
+      "grad_norm": 0.29477596282958984,
+      "learning_rate": 3.758334868299871e-07,
+      "loss": 0.006,
+      "step": 34900
+    },
+    {
+      "epoch": 6.316693796346536,
+      "grad_norm": 0.10228332132101059,
+      "learning_rate": 3.7537299686866825e-07,
+      "loss": 0.002,
+      "step": 34925
+    },
+    {
+      "epoch": 6.321215409658166,
+      "grad_norm": 0.08996782451868057,
+      "learning_rate": 3.7491250690734937e-07,
+      "loss": 0.0048,
+      "step": 34950
+    },
+    {
+      "epoch": 6.325737022969796,
+      "grad_norm": 0.012035650201141834,
+      "learning_rate": 3.744520169460306e-07,
+      "loss": 0.0157,
+      "step": 34975
+    },
+    {
+      "epoch": 6.3302586362814255,
+      "grad_norm": 0.035582542419433594,
+      "learning_rate": 3.739915269847117e-07,
+      "loss": 0.0112,
+      "step": 35000
+    },
+    {
+      "epoch": 6.334780249593055,
+      "grad_norm": 0.01056807953864336,
+      "learning_rate": 3.7353103702339287e-07,
+      "loss": 0.0008,
+      "step": 35025
+    },
+    {
+      "epoch": 6.3393018629046844,
+      "grad_norm": 0.2861970067024231,
+      "learning_rate": 3.7307054706207403e-07,
+      "loss": 0.0057,
+      "step": 35050
+    },
+    {
+      "epoch": 6.343823476216314,
+      "grad_norm": 0.013209059834480286,
+      "learning_rate": 3.7261005710075515e-07,
+      "loss": 0.0134,
+      "step": 35075
+    },
+    {
+      "epoch": 6.348345089527943,
+      "grad_norm": 0.1460546851158142,
+      "learning_rate": 3.7214956713943636e-07,
+      "loss": 0.0073,
+      "step": 35100
+    },
+    {
+      "epoch": 6.352866702839573,
+      "grad_norm": 0.09581249952316284,
+      "learning_rate": 3.716890771781175e-07,
+      "loss": 0.0104,
+      "step": 35125
+    },
+    {
+      "epoch": 6.357388316151202,
+      "grad_norm": 0.08900044858455658,
+      "learning_rate": 3.712285872167987e-07,
+      "loss": 0.0052,
+      "step": 35150
+    },
+    {
+      "epoch": 6.361909929462833,
+      "grad_norm": 0.3841894865036011,
+      "learning_rate": 3.707680972554798e-07,
+      "loss": 0.0195,
+      "step": 35175
+    },
+    {
+      "epoch": 6.366431542774462,
+      "grad_norm": 0.09828022122383118,
+      "learning_rate": 3.70307607294161e-07,
+      "loss": 0.012,
+      "step": 35200
+    },
+    {
+      "epoch": 6.370953156086092,
+      "grad_norm": 6.788888931274414,
+      "learning_rate": 3.6984711733284214e-07,
+      "loss": 0.023,
+      "step": 35225
+    },
+    {
+      "epoch": 6.375474769397721,
+      "grad_norm": 2.6568901538848877,
+      "learning_rate": 3.6938662737152326e-07,
+      "loss": 0.0014,
+      "step": 35250
+    },
+    {
+      "epoch": 6.379996382709351,
+      "grad_norm": 0.021253295242786407,
+      "learning_rate": 3.689261374102045e-07,
+      "loss": 0.0089,
+      "step": 35275
+    },
+    {
+      "epoch": 6.38451799602098,
+      "grad_norm": 0.7334450483322144,
+      "learning_rate": 3.684656474488856e-07,
+      "loss": 0.0017,
+      "step": 35300
+    },
+    {
+      "epoch": 6.38903960933261,
+      "grad_norm": 0.41396549344062805,
+      "learning_rate": 3.6800515748756676e-07,
+      "loss": 0.0033,
+      "step": 35325
+    },
+    {
+      "epoch": 6.393561222644239,
+      "grad_norm": 1.214400291442871,
+      "learning_rate": 3.675446675262479e-07,
+      "loss": 0.0016,
+      "step": 35350
+    },
+    {
+      "epoch": 6.3980828359558695,
+      "grad_norm": 0.03277917578816414,
+      "learning_rate": 3.6708417756492903e-07,
+      "loss": 0.0009,
+      "step": 35375
+    },
+    {
+      "epoch": 6.402604449267499,
+      "grad_norm": 0.1065434068441391,
+      "learning_rate": 3.6662368760361025e-07,
+      "loss": 0.0034,
+      "step": 35400
+    },
+    {
+      "epoch": 6.407126062579128,
+      "grad_norm": 1.2677608728408813,
+      "learning_rate": 3.6616319764229137e-07,
+      "loss": 0.0024,
+      "step": 35425
+    },
+    {
+      "epoch": 6.411647675890758,
+      "grad_norm": 0.034879542887210846,
+      "learning_rate": 3.6570270768097253e-07,
+      "loss": 0.0147,
+      "step": 35450
+    },
+    {
+      "epoch": 6.416169289202387,
+      "grad_norm": 0.009821565821766853,
+      "learning_rate": 3.652422177196537e-07,
+      "loss": 0.0084,
+      "step": 35475
+    },
+    {
+      "epoch": 6.420690902514017,
+      "grad_norm": 0.15989068150520325,
+      "learning_rate": 3.6478172775833487e-07,
+      "loss": 0.0253,
+      "step": 35500
+    },
+    {
+      "epoch": 6.425212515825646,
+      "grad_norm": 0.021616969257593155,
+      "learning_rate": 3.6432123779701603e-07,
+      "loss": 0.0385,
+      "step": 35525
+    },
+    {
+      "epoch": 6.429734129137276,
+      "grad_norm": 64.16004180908203,
+      "learning_rate": 3.6386074783569715e-07,
+      "loss": 0.0218,
+      "step": 35550
+    },
+    {
+      "epoch": 6.434255742448906,
+      "grad_norm": 2.1929166316986084,
+      "learning_rate": 3.634002578743783e-07,
+      "loss": 0.0276,
+      "step": 35575
+    },
+    {
+      "epoch": 6.438777355760536,
+      "grad_norm": 0.3426229655742645,
+      "learning_rate": 3.629397679130595e-07,
+      "loss": 0.0047,
+      "step": 35600
+    },
+    {
+      "epoch": 6.443298969072165,
+      "grad_norm": 0.9598920345306396,
+      "learning_rate": 3.6247927795174064e-07,
+      "loss": 0.0171,
+      "step": 35625
+    },
+    {
+      "epoch": 6.447820582383795,
+      "grad_norm": 0.39060238003730774,
+      "learning_rate": 3.620187879904218e-07,
+      "loss": 0.011,
+      "step": 35650
+    },
+    {
+      "epoch": 6.452342195695424,
+      "grad_norm": 0.9227154850959778,
+      "learning_rate": 3.615582980291029e-07,
+      "loss": 0.023,
+      "step": 35675
+    },
+    {
+      "epoch": 6.456863809007054,
+      "grad_norm": 5.520915985107422,
+      "learning_rate": 3.610978080677841e-07,
+      "loss": 0.0012,
+      "step": 35700
+    },
+    {
+      "epoch": 6.461385422318683,
+      "grad_norm": 0.09858091920614243,
+      "learning_rate": 3.6063731810646526e-07,
+      "loss": 0.0038,
+      "step": 35725
+    },
+    {
+      "epoch": 6.4659070356303125,
+      "grad_norm": 0.02428305707871914,
+      "learning_rate": 3.601768281451464e-07,
+      "loss": 0.004,
+      "step": 35750
+    },
+    {
+      "epoch": 6.470428648941942,
+      "grad_norm": 0.07016027718782425,
+      "learning_rate": 3.597163381838276e-07,
+      "loss": 0.0006,
+      "step": 35775
+    },
+    {
+      "epoch": 6.474950262253572,
+      "grad_norm": 0.1356564462184906,
+      "learning_rate": 3.592558482225087e-07,
+      "loss": 0.0025,
+      "step": 35800
+    },
+    {
+      "epoch": 6.479471875565202,
+      "grad_norm": 0.056458380073308945,
+      "learning_rate": 3.587953582611899e-07,
+      "loss": 0.0128,
+      "step": 35825
+    },
+    {
+      "epoch": 6.483993488876831,
+      "grad_norm": 2.0865302085876465,
+      "learning_rate": 3.5833486829987103e-07,
+      "loss": 0.0052,
+      "step": 35850
+    },
+    {
+      "epoch": 6.488515102188461,
+      "grad_norm": 0.2598702013492584,
+      "learning_rate": 3.578743783385522e-07,
+      "loss": 0.0079,
+      "step": 35875
+    },
+    {
+      "epoch": 6.49303671550009,
+      "grad_norm": 2.376058578491211,
+      "learning_rate": 3.5741388837723337e-07,
+      "loss": 0.0096,
+      "step": 35900
+    },
+    {
+      "epoch": 6.49755832881172,
+      "grad_norm": 0.022796526551246643,
+      "learning_rate": 3.5695339841591453e-07,
+      "loss": 0.0178,
+      "step": 35925
+    },
+    {
+      "epoch": 6.502079942123349,
+      "grad_norm": 17.341182708740234,
+      "learning_rate": 3.564929084545957e-07,
+      "loss": 0.0093,
+      "step": 35950
+    },
+    {
+      "epoch": 6.50660155543498,
+      "grad_norm": 0.10820627212524414,
+      "learning_rate": 3.560324184932768e-07,
+      "loss": 0.0426,
+      "step": 35975
+    },
+    {
+      "epoch": 6.511123168746609,
+      "grad_norm": 0.10570292919874191,
+      "learning_rate": 3.55571928531958e-07,
+      "loss": 0.0118,
+      "step": 36000
+    },
+    {
+      "epoch": 6.515644782058239,
+      "grad_norm": 5.8104658126831055,
+      "learning_rate": 3.5511143857063914e-07,
+      "loss": 0.0083,
+      "step": 36025
+    },
+    {
+      "epoch": 6.520166395369868,
+      "grad_norm": 0.9492124915122986,
+      "learning_rate": 3.546509486093203e-07,
+      "loss": 0.0157,
+      "step": 36050
+    },
+    {
+      "epoch": 6.524688008681498,
+      "grad_norm": 0.022533750161528587,
+      "learning_rate": 3.541904586480015e-07,
+      "loss": 0.0021,
+      "step": 36075
+    },
+    {
+      "epoch": 6.529209621993127,
+      "grad_norm": 0.35308748483657837,
+      "learning_rate": 3.537299686866826e-07,
+      "loss": 0.0015,
+      "step": 36100
+    },
+    {
+      "epoch": 6.5337312353047565,
+      "grad_norm": 1.7888010740280151,
+      "learning_rate": 3.5326947872536376e-07,
+      "loss": 0.004,
+      "step": 36125
+    },
+    {
+      "epoch": 6.538252848616386,
+      "grad_norm": 0.1663166582584381,
+      "learning_rate": 3.528089887640449e-07,
+      "loss": 0.0048,
+      "step": 36150
+    },
+    {
+      "epoch": 6.5427744619280155,
+      "grad_norm": 0.4084545969963074,
+      "learning_rate": 3.523484988027261e-07,
+      "loss": 0.0004,
+      "step": 36175
+    },
+    {
+      "epoch": 6.547296075239646,
+      "grad_norm": 0.039476677775382996,
+      "learning_rate": 3.5188800884140726e-07,
+      "loss": 0.0006,
+      "step": 36200
+    },
+    {
+      "epoch": 6.551817688551275,
+      "grad_norm": 0.0072963847778737545,
+      "learning_rate": 3.514275188800884e-07,
+      "loss": 0.008,
+      "step": 36225
+    },
+    {
+      "epoch": 6.556339301862905,
+      "grad_norm": 0.01364920660853386,
+      "learning_rate": 3.5096702891876954e-07,
+      "loss": 0.0077,
+      "step": 36250
+    },
+    {
+      "epoch": 6.560860915174534,
+      "grad_norm": 0.796404242515564,
+      "learning_rate": 3.505065389574507e-07,
+      "loss": 0.0229,
+      "step": 36275
+    },
+    {
+      "epoch": 6.565382528486164,
+      "grad_norm": 0.11998113989830017,
+      "learning_rate": 3.5004604899613187e-07,
+      "loss": 0.0266,
+      "step": 36300
+    },
+    {
+      "epoch": 6.569904141797793,
+      "grad_norm": 0.41521695256233215,
+      "learning_rate": 3.4958555903481303e-07,
+      "loss": 0.0025,
+      "step": 36325
+    },
+    {
+      "epoch": 6.574425755109423,
+      "grad_norm": 0.07342156767845154,
+      "learning_rate": 3.491250690734942e-07,
+      "loss": 0.0189,
+      "step": 36350
+    },
+    {
+      "epoch": 6.578947368421053,
+      "grad_norm": 0.369840145111084,
+      "learning_rate": 3.486645791121753e-07,
+      "loss": 0.0209,
+      "step": 36375
+    },
+    {
+      "epoch": 6.583468981732683,
+      "grad_norm": 0.04731602966785431,
+      "learning_rate": 3.482040891508565e-07,
+      "loss": 0.0051,
+      "step": 36400
+    },
+    {
+      "epoch": 6.587990595044312,
+      "grad_norm": 0.21776604652404785,
+      "learning_rate": 3.4774359918953765e-07,
+      "loss": 0.0017,
+      "step": 36425
+    },
+    {
+      "epoch": 6.5925122083559415,
+      "grad_norm": 0.05085720121860504,
+      "learning_rate": 3.472831092282188e-07,
+      "loss": 0.0014,
+      "step": 36450
+    },
+    {
+      "epoch": 6.597033821667571,
+      "grad_norm": 1.6889132261276245,
+      "learning_rate": 3.468226192669e-07,
+      "loss": 0.0029,
+      "step": 36475
+    },
+    {
+      "epoch": 6.6015554349792005,
+      "grad_norm": 7.945947170257568,
+      "learning_rate": 3.4636212930558114e-07,
+      "loss": 0.0041,
+      "step": 36500
+    },
+    {
+      "epoch": 6.60607704829083,
+      "grad_norm": 0.7570422291755676,
+      "learning_rate": 3.4590163934426226e-07,
+      "loss": 0.0015,
+      "step": 36525
+    },
+    {
+      "epoch": 6.6105986616024595,
+      "grad_norm": 0.05319717898964882,
+      "learning_rate": 3.454411493829434e-07,
+      "loss": 0.0042,
+      "step": 36550
+    },
+    {
+      "epoch": 6.615120274914089,
+      "grad_norm": 0.012651624158024788,
+      "learning_rate": 3.449806594216246e-07,
+      "loss": 0.0028,
+      "step": 36575
+    },
+    {
+      "epoch": 6.619641888225719,
+      "grad_norm": 0.06157555803656578,
+      "learning_rate": 3.4452016946030576e-07,
+      "loss": 0.0004,
+      "step": 36600
+    },
+    {
+      "epoch": 6.624163501537349,
+      "grad_norm": 0.009101797826588154,
+      "learning_rate": 3.440596794989869e-07,
+      "loss": 0.0071,
+      "step": 36625
+    },
+    {
+      "epoch": 6.628685114848978,
+      "grad_norm": 0.08508434146642685,
+      "learning_rate": 3.435991895376681e-07,
+      "loss": 0.0046,
+      "step": 36650
+    },
+    {
+      "epoch": 6.633206728160608,
+      "grad_norm": 0.026564456522464752,
+      "learning_rate": 3.431386995763492e-07,
+      "loss": 0.0059,
+      "step": 36675
+    },
+    {
+      "epoch": 6.637728341472237,
+      "grad_norm": 0.935742199420929,
+      "learning_rate": 3.4267820961503037e-07,
+      "loss": 0.0014,
+      "step": 36700
+    },
+    {
+      "epoch": 6.642249954783867,
+      "grad_norm": 7.623640060424805,
+      "learning_rate": 3.4221771965371153e-07,
+      "loss": 0.009,
+      "step": 36725
+    },
+    {
+      "epoch": 6.646771568095496,
+      "grad_norm": 0.023885022848844528,
+      "learning_rate": 3.417572296923927e-07,
+      "loss": 0.0017,
+      "step": 36750
+    },
+    {
+      "epoch": 6.6512931814071266,
+      "grad_norm": 0.20136716961860657,
+      "learning_rate": 3.4129673973107387e-07,
+      "loss": 0.0513,
+      "step": 36775
+    },
+    {
+      "epoch": 6.655814794718756,
+      "grad_norm": 0.06223003938794136,
+      "learning_rate": 3.40836249769755e-07,
+      "loss": 0.0029,
+      "step": 36800
+    },
+    {
+      "epoch": 6.6603364080303855,
+      "grad_norm": 0.20519289374351501,
+      "learning_rate": 3.4037575980843615e-07,
+      "loss": 0.0041,
+      "step": 36825
+    },
+    {
+      "epoch": 6.664858021342015,
+      "grad_norm": 101.21697998046875,
+      "learning_rate": 3.399152698471173e-07,
+      "loss": 0.0043,
+      "step": 36850
+    },
+    {
+      "epoch": 6.6693796346536445,
+      "grad_norm": 0.015112169086933136,
+      "learning_rate": 3.394547798857985e-07,
+      "loss": 0.0037,
+      "step": 36875
+    },
+    {
+      "epoch": 6.673901247965274,
+      "grad_norm": 0.8060915470123291,
+      "learning_rate": 3.3899428992447965e-07,
+      "loss": 0.0029,
+      "step": 36900
+    },
+    {
+      "epoch": 6.678422861276903,
+      "grad_norm": 0.3591912090778351,
+      "learning_rate": 3.3853379996316076e-07,
+      "loss": 0.0042,
+      "step": 36925
+    },
+    {
+      "epoch": 6.682944474588533,
+      "grad_norm": 0.3402554392814636,
+      "learning_rate": 3.38073310001842e-07,
+      "loss": 0.0005,
+      "step": 36950
+    },
+    {
+      "epoch": 6.687466087900162,
+      "grad_norm": 0.0494844950735569,
+      "learning_rate": 3.376128200405231e-07,
+      "loss": 0.0041,
+      "step": 36975
+    },
+    {
+      "epoch": 6.691987701211792,
+      "grad_norm": 0.08879272639751434,
+      "learning_rate": 3.3715233007920426e-07,
+      "loss": 0.0045,
+      "step": 37000
+    },
+    {
+      "epoch": 6.696509314523422,
+      "grad_norm": 8.207866668701172,
+      "learning_rate": 3.366918401178854e-07,
+      "loss": 0.0093,
+      "step": 37025
+    },
+    {
+      "epoch": 6.701030927835052,
+      "grad_norm": 0.1084265485405922,
+      "learning_rate": 3.3624976975501935e-07,
+      "loss": 0.0175,
+      "step": 37050
+    },
+    {
+      "epoch": 6.705552541146681,
+      "grad_norm": 0.03057611919939518,
+      "learning_rate": 3.3578927979370047e-07,
+      "loss": 0.0027,
+      "step": 37075
+    },
+    {
+      "epoch": 6.710074154458311,
+      "grad_norm": 0.14145521819591522,
+      "learning_rate": 3.353287898323817e-07,
+      "loss": 0.0155,
+      "step": 37100
+    },
+    {
+      "epoch": 6.71459576776994,
+      "grad_norm": 0.03683609515428543,
+      "learning_rate": 3.348682998710628e-07,
+      "loss": 0.0121,
+      "step": 37125
+    },
+    {
+      "epoch": 6.71911738108157,
+      "grad_norm": 0.2000865340232849,
+      "learning_rate": 3.344078099097439e-07,
+      "loss": 0.0104,
+      "step": 37150
+    },
+    {
+      "epoch": 6.723638994393199,
+      "grad_norm": 1.1748223304748535,
+      "learning_rate": 3.3394731994842513e-07,
+      "loss": 0.0266,
+      "step": 37175
+    },
+    {
+      "epoch": 6.7281606077048295,
+      "grad_norm": 0.04432156682014465,
+      "learning_rate": 3.3348682998710625e-07,
+      "loss": 0.0047,
+      "step": 37200
+    },
+    {
+      "epoch": 6.732682221016459,
+      "grad_norm": 0.1042926087975502,
+      "learning_rate": 3.3302634002578746e-07,
+      "loss": 0.0017,
+      "step": 37225
+    },
+    {
+      "epoch": 6.7372038343280884,
+      "grad_norm": 0.025923365727066994,
+      "learning_rate": 3.325658500644686e-07,
+      "loss": 0.0023,
+      "step": 37250
+    },
+    {
+      "epoch": 6.741725447639718,
+      "grad_norm": 5.998639106750488,
+      "learning_rate": 3.321053601031497e-07,
+      "loss": 0.0066,
+      "step": 37275
+    },
+    {
+      "epoch": 6.746247060951347,
+      "grad_norm": 0.038715049624443054,
+      "learning_rate": 3.316448701418309e-07,
+      "loss": 0.0114,
+      "step": 37300
+    },
+    {
+      "epoch": 6.750768674262977,
+      "grad_norm": 0.2970021963119507,
+      "learning_rate": 3.31184380180512e-07,
+      "loss": 0.0028,
+      "step": 37325
+    },
+    {
+      "epoch": 6.755290287574606,
+      "grad_norm": 0.021920220926404,
+      "learning_rate": 3.3072389021919324e-07,
+      "loss": 0.0028,
+      "step": 37350
+    },
+    {
+      "epoch": 6.759811900886236,
+      "grad_norm": 0.009460126049816608,
+      "learning_rate": 3.3026340025787436e-07,
+      "loss": 0.0038,
+      "step": 37375
+    },
+    {
+      "epoch": 6.764333514197865,
+      "grad_norm": 10.395684242248535,
+      "learning_rate": 3.298029102965555e-07,
+      "loss": 0.004,
+      "step": 37400
+    },
+    {
+      "epoch": 6.768855127509496,
+      "grad_norm": 0.3913702070713043,
+      "learning_rate": 3.293424203352367e-07,
+      "loss": 0.007,
+      "step": 37425
+    },
+    {
+      "epoch": 6.773376740821125,
+      "grad_norm": 0.009950965642929077,
+      "learning_rate": 3.288819303739178e-07,
+      "loss": 0.0049,
+      "step": 37450
+    },
+    {
+      "epoch": 6.777898354132755,
+      "grad_norm": 0.06827165186405182,
+      "learning_rate": 3.28421440412599e-07,
+      "loss": 0.0006,
+      "step": 37475
+    },
+    {
+      "epoch": 6.782419967444384,
+      "grad_norm": 0.044129110872745514,
+      "learning_rate": 3.2796095045128013e-07,
+      "loss": 0.0108,
+      "step": 37500
+    },
+    {
+      "epoch": 6.786941580756014,
+      "grad_norm": 0.08213396370410919,
+      "learning_rate": 3.2750046048996135e-07,
+      "loss": 0.001,
+      "step": 37525
+    },
+    {
+      "epoch": 6.791463194067643,
+      "grad_norm": 0.8949390053749084,
+      "learning_rate": 3.2703997052864247e-07,
+      "loss": 0.0508,
+      "step": 37550
+    },
+    {
+      "epoch": 6.795984807379273,
+      "grad_norm": 3.0352182388305664,
+      "learning_rate": 3.265794805673236e-07,
+      "loss": 0.0366,
+      "step": 37575
+    },
+    {
+      "epoch": 6.800506420690903,
+      "grad_norm": 0.03377075120806694,
+      "learning_rate": 3.261189906060048e-07,
+      "loss": 0.0061,
+      "step": 37600
+    },
+    {
+      "epoch": 6.805028034002532,
+      "grad_norm": 0.012610839679837227,
+      "learning_rate": 3.256585006446859e-07,
+      "loss": 0.0029,
+      "step": 37625
+    },
+    {
+      "epoch": 6.809549647314162,
+      "grad_norm": 0.3035992681980133,
+      "learning_rate": 3.2519801068336713e-07,
+      "loss": 0.0075,
+      "step": 37650
+    },
+    {
+      "epoch": 6.814071260625791,
+      "grad_norm": 0.06034184619784355,
+      "learning_rate": 3.2473752072204824e-07,
+      "loss": 0.0042,
+      "step": 37675
+    },
+    {
+      "epoch": 6.818592873937421,
+      "grad_norm": 0.10909536480903625,
+      "learning_rate": 3.2427703076072936e-07,
+      "loss": 0.0116,
+      "step": 37700
+    },
+    {
+      "epoch": 6.82311448724905,
+      "grad_norm": 13.51667594909668,
+      "learning_rate": 3.238165407994106e-07,
+      "loss": 0.0039,
+      "step": 37725
+    },
+    {
+      "epoch": 6.82763610056068,
+      "grad_norm": 0.026187343522906303,
+      "learning_rate": 3.233560508380917e-07,
+      "loss": 0.0117,
+      "step": 37750
+    },
+    {
+      "epoch": 6.832157713872309,
+      "grad_norm": 0.04201328754425049,
+      "learning_rate": 3.228955608767729e-07,
+      "loss": 0.0197,
+      "step": 37775
+    },
+    {
+      "epoch": 6.836679327183939,
+      "grad_norm": 0.05836571007966995,
+      "learning_rate": 3.22435070915454e-07,
+      "loss": 0.0122,
+      "step": 37800
+    },
+    {
+      "epoch": 6.841200940495569,
+      "grad_norm": 0.005761469714343548,
+      "learning_rate": 3.219745809541352e-07,
+      "loss": 0.0194,
+      "step": 37825
+    },
+    {
+      "epoch": 6.845722553807199,
+      "grad_norm": 0.019468627870082855,
+      "learning_rate": 3.2151409099281636e-07,
+      "loss": 0.0053,
+      "step": 37850
+    },
+    {
+      "epoch": 6.850244167118828,
+      "grad_norm": 0.005167699884623289,
+      "learning_rate": 3.2105360103149747e-07,
+      "loss": 0.0013,
+      "step": 37875
+    },
+    {
+      "epoch": 6.854765780430458,
+      "grad_norm": 0.06815607845783234,
+      "learning_rate": 3.205931110701787e-07,
+      "loss": 0.0078,
+      "step": 37900
+    },
+    {
+      "epoch": 6.859287393742087,
+      "grad_norm": 0.05461608245968819,
+      "learning_rate": 3.201326211088598e-07,
+      "loss": 0.0046,
+      "step": 37925
+    },
+    {
+      "epoch": 6.8638090070537165,
+      "grad_norm": 0.3847118020057678,
+      "learning_rate": 3.1967213114754097e-07,
+      "loss": 0.0215,
+      "step": 37950
+    },
+    {
+      "epoch": 6.868330620365346,
+      "grad_norm": 0.15587353706359863,
+      "learning_rate": 3.1921164118622213e-07,
+      "loss": 0.03,
+      "step": 37975
+    },
+    {
+      "epoch": 6.872852233676976,
+      "grad_norm": 0.06245379522442818,
+      "learning_rate": 3.1875115122490325e-07,
+      "loss": 0.0102,
+      "step": 38000
+    },
+    {
+      "epoch": 6.877373846988606,
+      "grad_norm": 0.03094295971095562,
+      "learning_rate": 3.1829066126358447e-07,
+      "loss": 0.0065,
+      "step": 38025
+    },
+    {
+      "epoch": 6.881895460300235,
+      "grad_norm": 0.1180916577577591,
+      "learning_rate": 3.178301713022656e-07,
+      "loss": 0.0013,
+      "step": 38050
+    },
+    {
+      "epoch": 6.886417073611865,
+      "grad_norm": 0.5813055634498596,
+      "learning_rate": 3.1736968134094675e-07,
+      "loss": 0.0071,
+      "step": 38075
+    },
+    {
+      "epoch": 6.890938686923494,
+      "grad_norm": 0.14167048037052155,
+      "learning_rate": 3.169091913796279e-07,
+      "loss": 0.0033,
+      "step": 38100
+    },
+    {
+      "epoch": 6.895460300235124,
+      "grad_norm": 0.1619337499141693,
+      "learning_rate": 3.164487014183091e-07,
+      "loss": 0.0045,
+      "step": 38125
+    },
+    {
+      "epoch": 6.899981913546753,
+      "grad_norm": 0.03847223520278931,
+      "learning_rate": 3.1598821145699024e-07,
+      "loss": 0.0043,
+      "step": 38150
+    },
+    {
+      "epoch": 6.904503526858383,
+      "grad_norm": 0.11694706231355667,
+      "learning_rate": 3.1552772149567136e-07,
+      "loss": 0.0097,
+      "step": 38175
+    },
+    {
+      "epoch": 6.909025140170012,
+      "grad_norm": 0.07207904756069183,
+      "learning_rate": 3.150672315343526e-07,
+      "loss": 0.0004,
+      "step": 38200
+    },
+    {
+      "epoch": 6.913546753481643,
+      "grad_norm": 1.5510050058364868,
+      "learning_rate": 3.146067415730337e-07,
+      "loss": 0.007,
+      "step": 38225
+    },
+    {
+      "epoch": 6.918068366793272,
+      "grad_norm": 0.13146886229515076,
+      "learning_rate": 3.1414625161171486e-07,
+      "loss": 0.0121,
+      "step": 38250
+    },
+    {
+      "epoch": 6.922589980104902,
+      "grad_norm": 0.09875814616680145,
+      "learning_rate": 3.13685761650396e-07,
+      "loss": 0.0128,
+      "step": 38275
+    },
+    {
+      "epoch": 6.927111593416531,
+      "grad_norm": 0.02351992577314377,
+      "learning_rate": 3.1322527168907714e-07,
+      "loss": 0.017,
+      "step": 38300
+    },
+    {
+      "epoch": 6.9316332067281605,
+      "grad_norm": 1.6462205648422241,
+      "learning_rate": 3.1276478172775835e-07,
+      "loss": 0.0443,
+      "step": 38325
+    },
+    {
+      "epoch": 6.93615482003979,
+      "grad_norm": 1.7699528932571411,
+      "learning_rate": 3.1230429176643947e-07,
+      "loss": 0.0535,
+      "step": 38350
+    },
+    {
+      "epoch": 6.9406764333514195,
+      "grad_norm": 0.2231673002243042,
+      "learning_rate": 3.1184380180512063e-07,
+      "loss": 0.0369,
+      "step": 38375
+    },
+    {
+      "epoch": 6.94519804666305,
+      "grad_norm": 4.6962738037109375,
+      "learning_rate": 3.113833118438018e-07,
+      "loss": 0.006,
+      "step": 38400
+    },
+    {
+      "epoch": 6.949719659974679,
+      "grad_norm": 0.08258053660392761,
+      "learning_rate": 3.109228218824829e-07,
+      "loss": 0.0052,
+      "step": 38425
+    },
+    {
+      "epoch": 6.954241273286309,
+      "grad_norm": 1.8889635801315308,
+      "learning_rate": 3.1046233192116413e-07,
+      "loss": 0.0061,
+      "step": 38450
+    },
+    {
+      "epoch": 6.958762886597938,
+      "grad_norm": 0.09585348516702652,
+      "learning_rate": 3.1000184195984525e-07,
+      "loss": 0.0047,
+      "step": 38475
+    },
+    {
+      "epoch": 6.963284499909568,
+      "grad_norm": 0.10662294924259186,
+      "learning_rate": 3.095413519985264e-07,
+      "loss": 0.0076,
+      "step": 38500
+    },
+    {
+      "epoch": 6.967806113221197,
+      "grad_norm": 0.006054690573364496,
+      "learning_rate": 3.090808620372076e-07,
+      "loss": 0.0095,
+      "step": 38525
+    },
+    {
+      "epoch": 6.972327726532827,
+      "grad_norm": 27.81556510925293,
+      "learning_rate": 3.0862037207588875e-07,
+      "loss": 0.0016,
+      "step": 38550
+    },
+    {
+      "epoch": 6.976849339844456,
+      "grad_norm": 0.030143573880195618,
+      "learning_rate": 3.081598821145699e-07,
+      "loss": 0.0136,
+      "step": 38575
+    },
+    {
+      "epoch": 6.981370953156086,
+      "grad_norm": 2.358839750289917,
+      "learning_rate": 3.07699392153251e-07,
+      "loss": 0.0021,
+      "step": 38600
+    },
+    {
+      "epoch": 6.985892566467716,
+      "grad_norm": 12.556412696838379,
+      "learning_rate": 3.072389021919322e-07,
+      "loss": 0.0012,
+      "step": 38625
+    },
+    {
+      "epoch": 6.9904141797793455,
+      "grad_norm": 0.13631652295589447,
+      "learning_rate": 3.0677841223061336e-07,
+      "loss": 0.0066,
+      "step": 38650
+    },
+    {
+      "epoch": 6.994935793090975,
+      "grad_norm": 41.28492736816406,
+      "learning_rate": 3.063179222692945e-07,
+      "loss": 0.0265,
+      "step": 38675
+    },
+    {
+      "epoch": 6.9994574064026045,
+      "grad_norm": 0.12053684145212173,
+      "learning_rate": 3.058574323079757e-07,
+      "loss": 0.0605,
+      "step": 38700
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.3541729748249054,
+      "eval_runtime": 8626.509,
+      "eval_samples_per_second": 1.101,
+      "eval_steps_per_second": 0.138,
+      "eval_wer": 0.10525304292120435,
+      "step": 38703
+    },
+    {
+      "epoch": 7.003979019714234,
+      "grad_norm": 1.1688233613967896,
+      "learning_rate": 3.053969423466568e-07,
+      "loss": 0.0075,
+      "step": 38725
+    },
+    {
+      "epoch": 7.0085006330258635,
+      "grad_norm": 0.06878869980573654,
+      "learning_rate": 3.0493645238533797e-07,
+      "loss": 0.0013,
+      "step": 38750
+    },
+    {
+      "epoch": 7.013022246337493,
+      "grad_norm": 0.12803910672664642,
+      "learning_rate": 3.0447596242401914e-07,
+      "loss": 0.006,
+      "step": 38775
+    },
+    {
+      "epoch": 7.017543859649122,
+      "grad_norm": 11.873225212097168,
+      "learning_rate": 3.040154724627003e-07,
+      "loss": 0.0017,
+      "step": 38800
+    },
+    {
+      "epoch": 7.022065472960753,
+      "grad_norm": 0.09927644580602646,
+      "learning_rate": 3.0355498250138147e-07,
+      "loss": 0.0011,
+      "step": 38825
+    },
+    {
+      "epoch": 7.026587086272382,
+      "grad_norm": 0.014069181866943836,
+      "learning_rate": 3.030944925400626e-07,
+      "loss": 0.0007,
+      "step": 38850
+    },
+    {
+      "epoch": 7.031108699584012,
+      "grad_norm": 3.6158812046051025,
+      "learning_rate": 3.026340025787438e-07,
+      "loss": 0.0076,
+      "step": 38875
+    },
+    {
+      "epoch": 7.035630312895641,
+      "grad_norm": 0.049011897295713425,
+      "learning_rate": 3.021735126174249e-07,
+      "loss": 0.0027,
+      "step": 38900
+    },
+    {
+      "epoch": 7.040151926207271,
+      "grad_norm": 0.02783900685608387,
+      "learning_rate": 3.017130226561061e-07,
+      "loss": 0.0026,
+      "step": 38925
+    },
+    {
+      "epoch": 7.0446735395189,
+      "grad_norm": 0.024276690557599068,
+      "learning_rate": 3.0125253269478725e-07,
+      "loss": 0.0018,
+      "step": 38950
+    },
+    {
+      "epoch": 7.04919515283053,
+      "grad_norm": 0.052325211465358734,
+      "learning_rate": 3.007920427334684e-07,
+      "loss": 0.0089,
+      "step": 38975
+    },
+    {
+      "epoch": 7.053716766142159,
+      "grad_norm": 0.021905574947595596,
+      "learning_rate": 3.003315527721496e-07,
+      "loss": 0.0027,
+      "step": 39000
+    },
+    {
+      "epoch": 7.0582383794537895,
+      "grad_norm": 0.013508542440831661,
+      "learning_rate": 2.998710628108307e-07,
+      "loss": 0.002,
+      "step": 39025
+    },
+    {
+      "epoch": 7.062759992765419,
+      "grad_norm": 0.052605342119932175,
+      "learning_rate": 2.9941057284951186e-07,
+      "loss": 0.0069,
+      "step": 39050
+    },
+    {
+      "epoch": 7.0672816060770485,
+      "grad_norm": 0.578596830368042,
+      "learning_rate": 2.98950082888193e-07,
+      "loss": 0.0269,
+      "step": 39075
+    },
+    {
+      "epoch": 7.071803219388678,
+      "grad_norm": 21.0313663482666,
+      "learning_rate": 2.984895929268742e-07,
+      "loss": 0.0125,
+      "step": 39100
+    },
+    {
+      "epoch": 7.076324832700307,
+      "grad_norm": 0.19108355045318604,
+      "learning_rate": 2.9802910296555536e-07,
+      "loss": 0.0074,
+      "step": 39125
+    },
+    {
+      "epoch": 7.080846446011937,
+      "grad_norm": 2.0000555515289307,
+      "learning_rate": 2.9756861300423647e-07,
+      "loss": 0.0028,
+      "step": 39150
+    },
+    {
+      "epoch": 7.085368059323566,
+      "grad_norm": 0.03665238618850708,
+      "learning_rate": 2.9710812304291764e-07,
+      "loss": 0.0046,
+      "step": 39175
+    },
+    {
+      "epoch": 7.089889672635196,
+      "grad_norm": 0.03791901841759682,
+      "learning_rate": 2.966476330815988e-07,
+      "loss": 0.0016,
+      "step": 39200
+    },
+    {
+      "epoch": 7.094411285946826,
+      "grad_norm": 0.016226934269070625,
+      "learning_rate": 2.9618714312027997e-07,
+      "loss": 0.0028,
+      "step": 39225
+    },
+    {
+      "epoch": 7.098932899258456,
+      "grad_norm": 12.398660659790039,
+      "learning_rate": 2.9572665315896113e-07,
+      "loss": 0.0017,
+      "step": 39250
+    },
+    {
+      "epoch": 7.103454512570085,
+      "grad_norm": 0.012261465191841125,
+      "learning_rate": 2.952661631976423e-07,
+      "loss": 0.001,
+      "step": 39275
+    },
+    {
+      "epoch": 7.107976125881715,
+      "grad_norm": 14.556396484375,
+      "learning_rate": 2.948056732363234e-07,
+      "loss": 0.0109,
+      "step": 39300
+    },
+    {
+      "epoch": 7.112497739193344,
+      "grad_norm": 0.11572438478469849,
+      "learning_rate": 2.943451832750046e-07,
+      "loss": 0.0049,
+      "step": 39325
+    },
+    {
+      "epoch": 7.117019352504974,
+      "grad_norm": 0.28125593066215515,
+      "learning_rate": 2.9388469331368575e-07,
+      "loss": 0.0114,
+      "step": 39350
+    },
+    {
+      "epoch": 7.121540965816603,
+      "grad_norm": 0.05056861415505409,
+      "learning_rate": 2.934242033523669e-07,
+      "loss": 0.0012,
+      "step": 39375
+    },
+    {
+      "epoch": 7.126062579128233,
+      "grad_norm": 0.22000250220298767,
+      "learning_rate": 2.929637133910481e-07,
+      "loss": 0.0305,
+      "step": 39400
+    },
+    {
+      "epoch": 7.130584192439863,
+      "grad_norm": 0.020098086446523666,
+      "learning_rate": 2.925032234297292e-07,
+      "loss": 0.0013,
+      "step": 39425
+    },
+    {
+      "epoch": 7.1351058057514924,
+      "grad_norm": 0.026488734409213066,
+      "learning_rate": 2.9204273346841036e-07,
+      "loss": 0.0356,
+      "step": 39450
+    },
+    {
+      "epoch": 7.139627419063122,
+      "grad_norm": 0.024454880505800247,
+      "learning_rate": 2.915822435070915e-07,
+      "loss": 0.0077,
+      "step": 39475
+    },
+    {
+      "epoch": 7.144149032374751,
+      "grad_norm": 0.017868295311927795,
+      "learning_rate": 2.911217535457727e-07,
+      "loss": 0.0215,
+      "step": 39500
+    },
+    {
+      "epoch": 7.148670645686381,
+      "grad_norm": 1.8522053956985474,
+      "learning_rate": 2.9066126358445386e-07,
+      "loss": 0.0043,
+      "step": 39525
+    },
+    {
+      "epoch": 7.15319225899801,
+      "grad_norm": 0.20884989202022552,
+      "learning_rate": 2.90200773623135e-07,
+      "loss": 0.0055,
+      "step": 39550
+    },
+    {
+      "epoch": 7.15771387230964,
+      "grad_norm": 0.1342606246471405,
+      "learning_rate": 2.8974028366181614e-07,
+      "loss": 0.0012,
+      "step": 39575
+    },
+    {
+      "epoch": 7.162235485621269,
+      "grad_norm": 0.14675575494766235,
+      "learning_rate": 2.892797937004973e-07,
+      "loss": 0.0013,
+      "step": 39600
+    },
+    {
+      "epoch": 7.1667570989329,
+      "grad_norm": 0.034802380949258804,
+      "learning_rate": 2.8881930373917847e-07,
+      "loss": 0.0014,
+      "step": 39625
+    },
+    {
+      "epoch": 7.171278712244529,
+      "grad_norm": 0.03734385594725609,
+      "learning_rate": 2.8835881377785964e-07,
+      "loss": 0.0026,
+      "step": 39650
+    },
+    {
+      "epoch": 7.175800325556159,
+      "grad_norm": 0.7827721834182739,
+      "learning_rate": 2.878983238165408e-07,
+      "loss": 0.0009,
+      "step": 39675
+    },
+    {
+      "epoch": 7.180321938867788,
+      "grad_norm": 0.06746553629636765,
+      "learning_rate": 2.8743783385522197e-07,
+      "loss": 0.0024,
+      "step": 39700
+    },
+    {
+      "epoch": 7.184843552179418,
+      "grad_norm": 0.2701489329338074,
+      "learning_rate": 2.869773438939031e-07,
+      "loss": 0.0058,
+      "step": 39725
+    },
+    {
+      "epoch": 7.189365165491047,
+      "grad_norm": 0.021761702373623848,
+      "learning_rate": 2.8651685393258425e-07,
+      "loss": 0.0078,
+      "step": 39750
+    },
+    {
+      "epoch": 7.193886778802677,
+      "grad_norm": 0.004122884478420019,
+      "learning_rate": 2.860563639712654e-07,
+      "loss": 0.0009,
+      "step": 39775
+    },
+    {
+      "epoch": 7.198408392114306,
+      "grad_norm": 0.05977817252278328,
+      "learning_rate": 2.855958740099466e-07,
+      "loss": 0.0061,
+      "step": 39800
+    },
+    {
+      "epoch": 7.2029300054259355,
+      "grad_norm": 0.7055386900901794,
+      "learning_rate": 2.8513538404862775e-07,
+      "loss": 0.0058,
+      "step": 39825
+    },
+    {
+      "epoch": 7.207451618737566,
+      "grad_norm": 13.389866828918457,
+      "learning_rate": 2.8467489408730886e-07,
+      "loss": 0.0142,
+      "step": 39850
+    },
+    {
+      "epoch": 7.211973232049195,
+      "grad_norm": 0.13937248289585114,
+      "learning_rate": 2.8421440412599e-07,
+      "loss": 0.0028,
+      "step": 39875
+    },
+    {
+      "epoch": 7.216494845360825,
+      "grad_norm": 0.6235303282737732,
+      "learning_rate": 2.837539141646712e-07,
+      "loss": 0.0083,
+      "step": 39900
+    },
+    {
+      "epoch": 7.221016458672454,
+      "grad_norm": 2.9348771572113037,
+      "learning_rate": 2.8329342420335236e-07,
+      "loss": 0.0374,
+      "step": 39925
+    },
+    {
+      "epoch": 7.225538071984084,
+      "grad_norm": 3.077686071395874,
+      "learning_rate": 2.828329342420335e-07,
+      "loss": 0.0082,
+      "step": 39950
+    },
+    {
+      "epoch": 7.230059685295713,
+      "grad_norm": 0.22787833213806152,
+      "learning_rate": 2.8237244428071464e-07,
+      "loss": 0.0012,
+      "step": 39975
+    },
+    {
+      "epoch": 7.234581298607343,
+      "grad_norm": 0.06242289021611214,
+      "learning_rate": 2.8191195431939586e-07,
+      "loss": 0.0061,
+      "step": 40000
+    },
+    {
+      "epoch": 7.239102911918972,
+      "grad_norm": 0.9802669882774353,
+      "learning_rate": 2.8145146435807697e-07,
+      "loss": 0.0092,
+      "step": 40025
+    },
+    {
+      "epoch": 7.243624525230603,
+      "grad_norm": 0.012614204548299313,
+      "learning_rate": 2.8099097439675814e-07,
+      "loss": 0.0005,
+      "step": 40050
+    },
+    {
+      "epoch": 7.248146138542232,
+      "grad_norm": 0.013021476566791534,
+      "learning_rate": 2.805304844354393e-07,
+      "loss": 0.0101,
+      "step": 40075
+    },
+    {
+      "epoch": 7.252667751853862,
+      "grad_norm": 0.008002633228898048,
+      "learning_rate": 2.800699944741204e-07,
+      "loss": 0.0083,
+      "step": 40100
+    },
+    {
+      "epoch": 7.257189365165491,
+      "grad_norm": 0.04996323958039284,
+      "learning_rate": 2.7960950451280164e-07,
+      "loss": 0.0011,
+      "step": 40125
+    },
+    {
+      "epoch": 7.2617109784771205,
+      "grad_norm": 11.185647964477539,
+      "learning_rate": 2.7914901455148275e-07,
+      "loss": 0.0034,
+      "step": 40150
+    },
+    {
+      "epoch": 7.26623259178875,
+      "grad_norm": 0.5441507697105408,
+      "learning_rate": 2.786885245901639e-07,
+      "loss": 0.012,
+      "step": 40175
+    },
+    {
+      "epoch": 7.2707542051003795,
+      "grad_norm": 0.020439432933926582,
+      "learning_rate": 2.782280346288451e-07,
+      "loss": 0.0094,
+      "step": 40200
+    },
+    {
+      "epoch": 7.275275818412009,
+      "grad_norm": 1.3434791564941406,
+      "learning_rate": 2.77785964265979e-07,
+      "loss": 0.0097,
+      "step": 40225
+    },
+    {
+      "epoch": 7.279797431723639,
+      "grad_norm": 0.302569180727005,
+      "learning_rate": 2.773254743046601e-07,
+      "loss": 0.0277,
+      "step": 40250
+    },
+    {
+      "epoch": 7.284319045035269,
+      "grad_norm": 0.06300198286771774,
+      "learning_rate": 2.7686498434334134e-07,
+      "loss": 0.0133,
+      "step": 40275
+    },
+    {
+      "epoch": 7.288840658346898,
+      "grad_norm": 22.47533416748047,
+      "learning_rate": 2.7640449438202246e-07,
+      "loss": 0.0173,
+      "step": 40300
+    },
+    {
+      "epoch": 7.293362271658528,
+      "grad_norm": 0.6076884865760803,
+      "learning_rate": 2.7594400442070357e-07,
+      "loss": 0.003,
+      "step": 40325
+    },
+    {
+      "epoch": 7.297883884970157,
+      "grad_norm": 0.18815143406391144,
+      "learning_rate": 2.754835144593848e-07,
+      "loss": 0.0089,
+      "step": 40350
+    },
+    {
+      "epoch": 7.302405498281787,
+      "grad_norm": 0.004078809637576342,
+      "learning_rate": 2.750230244980659e-07,
+      "loss": 0.0007,
+      "step": 40375
+    },
+    {
+      "epoch": 7.306927111593416,
+      "grad_norm": 0.06133987382054329,
+      "learning_rate": 2.745625345367471e-07,
+      "loss": 0.0008,
+      "step": 40400
+    },
+    {
+      "epoch": 7.311448724905046,
+      "grad_norm": 0.43629419803619385,
+      "learning_rate": 2.7410204457542824e-07,
+      "loss": 0.0055,
+      "step": 40425
+    },
+    {
+      "epoch": 7.315970338216676,
+      "grad_norm": 16.26517105102539,
+      "learning_rate": 2.736415546141094e-07,
+      "loss": 0.0016,
+      "step": 40450
+    },
+    {
+      "epoch": 7.320491951528306,
+      "grad_norm": 0.05941370874643326,
+      "learning_rate": 2.7318106465279057e-07,
+      "loss": 0.0071,
+      "step": 40475
+    },
+    {
+      "epoch": 7.325013564839935,
+      "grad_norm": 0.15563565492630005,
+      "learning_rate": 2.727205746914717e-07,
+      "loss": 0.0013,
+      "step": 40500
+    },
+    {
+      "epoch": 7.3295351781515645,
+      "grad_norm": 0.5691679120063782,
+      "learning_rate": 2.722600847301529e-07,
+      "loss": 0.0005,
+      "step": 40525
+    },
+    {
+      "epoch": 7.334056791463194,
+      "grad_norm": 0.004803875926882029,
+      "learning_rate": 2.71799594768834e-07,
+      "loss": 0.008,
+      "step": 40550
+    },
+    {
+      "epoch": 7.3385784047748235,
+      "grad_norm": 0.08438849449157715,
+      "learning_rate": 2.7133910480751523e-07,
+      "loss": 0.0012,
+      "step": 40575
+    },
+    {
+      "epoch": 7.343100018086453,
+      "grad_norm": 10.578065872192383,
+      "learning_rate": 2.7087861484619635e-07,
+      "loss": 0.0042,
+      "step": 40600
+    },
+    {
+      "epoch": 7.347621631398082,
+      "grad_norm": 0.00701162638142705,
+      "learning_rate": 2.7041812488487746e-07,
+      "loss": 0.0098,
+      "step": 40625
+    },
+    {
+      "epoch": 7.352143244709713,
+      "grad_norm": 0.029753483831882477,
+      "learning_rate": 2.699576349235587e-07,
+      "loss": 0.0032,
+      "step": 40650
+    },
+    {
+      "epoch": 7.356664858021342,
+      "grad_norm": 0.18117420375347137,
+      "learning_rate": 2.694971449622398e-07,
+      "loss": 0.0027,
+      "step": 40675
+    },
+    {
+      "epoch": 7.361186471332972,
+      "grad_norm": 0.9770309329032898,
+      "learning_rate": 2.69036655000921e-07,
+      "loss": 0.0229,
+      "step": 40700
+    },
+    {
+      "epoch": 7.365708084644601,
+      "grad_norm": 0.0630781352519989,
+      "learning_rate": 2.685761650396021e-07,
+      "loss": 0.0112,
+      "step": 40725
+    },
+    {
+      "epoch": 7.370229697956231,
+      "grad_norm": 0.0655864030122757,
+      "learning_rate": 2.6811567507828324e-07,
+      "loss": 0.0023,
+      "step": 40750
+    },
+    {
+      "epoch": 7.37475131126786,
+      "grad_norm": 0.28010591864585876,
+      "learning_rate": 2.6765518511696446e-07,
+      "loss": 0.0016,
+      "step": 40775
+    },
+    {
+      "epoch": 7.37927292457949,
+      "grad_norm": 0.04630829766392708,
+      "learning_rate": 2.6719469515564557e-07,
+      "loss": 0.0011,
+      "step": 40800
+    },
+    {
+      "epoch": 7.383794537891119,
+      "grad_norm": 0.03749445080757141,
+      "learning_rate": 2.667342051943268e-07,
+      "loss": 0.0038,
+      "step": 40825
+    },
+    {
+      "epoch": 7.3883161512027495,
+      "grad_norm": 0.22030295431613922,
+      "learning_rate": 2.662737152330079e-07,
+      "loss": 0.0049,
+      "step": 40850
+    },
+    {
+      "epoch": 7.392837764514379,
+      "grad_norm": 0.07691410928964615,
+      "learning_rate": 2.6581322527168907e-07,
+      "loss": 0.0004,
+      "step": 40875
+    },
+    {
+      "epoch": 7.3973593778260085,
+      "grad_norm": 0.027430010959506035,
+      "learning_rate": 2.6535273531037023e-07,
+      "loss": 0.0022,
+      "step": 40900
+    },
+    {
+      "epoch": 7.401880991137638,
+      "grad_norm": 0.07501472532749176,
+      "learning_rate": 2.6489224534905135e-07,
+      "loss": 0.0013,
+      "step": 40925
+    },
+    {
+      "epoch": 7.4064026044492675,
+      "grad_norm": 0.004892929922789335,
+      "learning_rate": 2.6443175538773257e-07,
+      "loss": 0.0044,
+      "step": 40950
+    },
+    {
+      "epoch": 7.410924217760897,
+      "grad_norm": 0.11225175857543945,
+      "learning_rate": 2.639712654264137e-07,
+      "loss": 0.0145,
+      "step": 40975
+    },
+    {
+      "epoch": 7.415445831072526,
+      "grad_norm": 0.01234606932848692,
+      "learning_rate": 2.6351077546509485e-07,
+      "loss": 0.0007,
+      "step": 41000
+    },
+    {
+      "epoch": 7.419967444384156,
+      "grad_norm": 20.348722457885742,
+      "learning_rate": 2.63050285503776e-07,
+      "loss": 0.0173,
+      "step": 41025
+    },
+    {
+      "epoch": 7.424489057695786,
+      "grad_norm": 0.01828809268772602,
+      "learning_rate": 2.625897955424571e-07,
+      "loss": 0.0154,
+      "step": 41050
+    },
+    {
+      "epoch": 7.429010671007416,
+      "grad_norm": 0.1676705926656723,
+      "learning_rate": 2.6212930558113835e-07,
+      "loss": 0.0254,
+      "step": 41075
+    },
+    {
+      "epoch": 7.433532284319045,
+      "grad_norm": 0.04343516007065773,
+      "learning_rate": 2.6166881561981946e-07,
+      "loss": 0.0271,
+      "step": 41100
+    },
+    {
+      "epoch": 7.438053897630675,
+      "grad_norm": 0.06619753688573837,
+      "learning_rate": 2.612083256585006e-07,
+      "loss": 0.0081,
+      "step": 41125
+    },
+    {
+      "epoch": 7.442575510942304,
+      "grad_norm": 0.2102379947900772,
+      "learning_rate": 2.607478356971818e-07,
+      "loss": 0.0045,
+      "step": 41150
+    },
+    {
+      "epoch": 7.447097124253934,
+      "grad_norm": 0.020573345944285393,
+      "learning_rate": 2.6028734573586296e-07,
+      "loss": 0.0027,
+      "step": 41175
+    },
+    {
+      "epoch": 7.451618737565563,
+      "grad_norm": 0.0597822479903698,
+      "learning_rate": 2.598268557745441e-07,
+      "loss": 0.0018,
+      "step": 41200
+    },
+    {
+      "epoch": 7.456140350877193,
+      "grad_norm": 2.0168440341949463,
+      "learning_rate": 2.5936636581322524e-07,
+      "loss": 0.0125,
+      "step": 41225
+    },
+    {
+      "epoch": 7.460661964188823,
+      "grad_norm": 0.42403435707092285,
+      "learning_rate": 2.5890587585190646e-07,
+      "loss": 0.0032,
+      "step": 41250
+    },
+    {
+      "epoch": 7.4651835775004525,
+      "grad_norm": 0.03630862757563591,
+      "learning_rate": 2.5844538589058757e-07,
+      "loss": 0.0011,
+      "step": 41275
+    },
+    {
+      "epoch": 7.469705190812082,
+      "grad_norm": 0.008765432052314281,
+      "learning_rate": 2.5798489592926874e-07,
+      "loss": 0.0095,
+      "step": 41300
+    },
+    {
+      "epoch": 7.474226804123711,
+      "grad_norm": 0.019622275605797768,
+      "learning_rate": 2.575244059679499e-07,
+      "loss": 0.0237,
+      "step": 41325
+    },
+    {
+      "epoch": 7.478748417435341,
+      "grad_norm": 0.14788568019866943,
+      "learning_rate": 2.57063916006631e-07,
+      "loss": 0.0071,
+      "step": 41350
+    },
+    {
+      "epoch": 7.48327003074697,
+      "grad_norm": 9.201263427734375,
+      "learning_rate": 2.5660342604531223e-07,
+      "loss": 0.0059,
+      "step": 41375
+    },
+    {
+      "epoch": 7.4877916440586,
+      "grad_norm": 9.586228370666504,
+      "learning_rate": 2.5614293608399335e-07,
+      "loss": 0.001,
+      "step": 41400
+    },
+    {
+      "epoch": 7.492313257370229,
+      "grad_norm": 0.035080842673778534,
+      "learning_rate": 2.556824461226745e-07,
+      "loss": 0.008,
+      "step": 41425
+    },
+    {
+      "epoch": 7.496834870681859,
+      "grad_norm": 0.09500127285718918,
+      "learning_rate": 2.552219561613557e-07,
+      "loss": 0.0203,
+      "step": 41450
+    },
+    {
+      "epoch": 7.501356483993489,
+      "grad_norm": 0.030213013291358948,
+      "learning_rate": 2.547614662000368e-07,
+      "loss": 0.0226,
+      "step": 41475
+    },
+    {
+      "epoch": 7.505878097305119,
+      "grad_norm": 0.16554242372512817,
+      "learning_rate": 2.54300976238718e-07,
+      "loss": 0.0396,
+      "step": 41500
+    },
+    {
+      "epoch": 7.510399710616748,
+      "grad_norm": 0.049443017691373825,
+      "learning_rate": 2.538404862773991e-07,
+      "loss": 0.0085,
+      "step": 41525
+    },
+    {
+      "epoch": 7.514921323928378,
+      "grad_norm": 0.045605577528476715,
+      "learning_rate": 2.533799963160803e-07,
+      "loss": 0.0045,
+      "step": 41550
+    },
+    {
+      "epoch": 7.519442937240007,
+      "grad_norm": 0.023337364196777344,
+      "learning_rate": 2.5291950635476146e-07,
+      "loss": 0.0102,
+      "step": 41575
+    },
+    {
+      "epoch": 7.523964550551637,
+      "grad_norm": 0.78485107421875,
+      "learning_rate": 2.524590163934426e-07,
+      "loss": 0.0024,
+      "step": 41600
+    },
+    {
+      "epoch": 7.528486163863266,
+      "grad_norm": 0.014340350404381752,
+      "learning_rate": 2.519985264321238e-07,
+      "loss": 0.0019,
+      "step": 41625
+    },
+    {
+      "epoch": 7.5330077771748964,
+      "grad_norm": 0.004216828849166632,
+      "learning_rate": 2.515380364708049e-07,
+      "loss": 0.003,
+      "step": 41650
+    },
+    {
+      "epoch": 7.537529390486526,
+      "grad_norm": 14.32552719116211,
+      "learning_rate": 2.5107754650948607e-07,
+      "loss": 0.0066,
+      "step": 41675
+    },
+    {
+      "epoch": 7.542051003798155,
+      "grad_norm": 0.160021111369133,
+      "learning_rate": 2.5061705654816724e-07,
+      "loss": 0.0053,
+      "step": 41700
+    },
+    {
+      "epoch": 7.546572617109785,
+      "grad_norm": 0.020444253459572792,
+      "learning_rate": 2.501565665868484e-07,
+      "loss": 0.0011,
+      "step": 41725
+    },
+    {
+      "epoch": 7.551094230421414,
+      "grad_norm": 1.1539157629013062,
+      "learning_rate": 2.4969607662552957e-07,
+      "loss": 0.0011,
+      "step": 41750
+    },
+    {
+      "epoch": 7.555615843733044,
+      "grad_norm": 0.10138887166976929,
+      "learning_rate": 2.4923558666421074e-07,
+      "loss": 0.001,
+      "step": 41775
+    },
+    {
+      "epoch": 7.560137457044673,
+      "grad_norm": 0.034172624349594116,
+      "learning_rate": 2.4877509670289185e-07,
+      "loss": 0.0013,
+      "step": 41800
+    },
+    {
+      "epoch": 7.564659070356303,
+      "grad_norm": 0.02480352483689785,
+      "learning_rate": 2.48314606741573e-07,
+      "loss": 0.01,
+      "step": 41825
+    },
+    {
+      "epoch": 7.569180683667932,
+      "grad_norm": 0.03536754474043846,
+      "learning_rate": 2.478541167802542e-07,
+      "loss": 0.0208,
+      "step": 41850
+    },
+    {
+      "epoch": 7.573702296979563,
+      "grad_norm": 0.20276236534118652,
+      "learning_rate": 2.4739362681893535e-07,
+      "loss": 0.0131,
+      "step": 41875
+    },
+    {
+      "epoch": 7.578223910291192,
+      "grad_norm": 0.03076460212469101,
+      "learning_rate": 2.469331368576165e-07,
+      "loss": 0.0082,
+      "step": 41900
+    },
+    {
+      "epoch": 7.582745523602822,
+      "grad_norm": 0.057225510478019714,
+      "learning_rate": 2.464726468962977e-07,
+      "loss": 0.0082,
+      "step": 41925
+    },
+    {
+      "epoch": 7.587267136914451,
+      "grad_norm": 0.07822202891111374,
+      "learning_rate": 2.460121569349788e-07,
+      "loss": 0.0038,
+      "step": 41950
+    },
+    {
+      "epoch": 7.591788750226081,
+      "grad_norm": 0.01712547056376934,
+      "learning_rate": 2.4555166697365996e-07,
+      "loss": 0.0063,
+      "step": 41975
+    },
+    {
+      "epoch": 7.59631036353771,
+      "grad_norm": 0.06029786914587021,
+      "learning_rate": 2.450911770123411e-07,
+      "loss": 0.0069,
+      "step": 42000
+    },
+    {
+      "epoch": 7.6008319768493395,
+      "grad_norm": 0.025808099657297134,
+      "learning_rate": 2.446306870510223e-07,
+      "loss": 0.0042,
+      "step": 42025
+    },
+    {
+      "epoch": 7.60535359016097,
+      "grad_norm": 0.01600913517177105,
+      "learning_rate": 2.4417019708970346e-07,
+      "loss": 0.0007,
+      "step": 42050
+    },
+    {
+      "epoch": 7.609875203472599,
+      "grad_norm": 7.5506815910339355,
+      "learning_rate": 2.4370970712838457e-07,
+      "loss": 0.0054,
+      "step": 42075
+    },
+    {
+      "epoch": 7.614396816784229,
+      "grad_norm": 0.04834285005927086,
+      "learning_rate": 2.4324921716706574e-07,
+      "loss": 0.0015,
+      "step": 42100
+    },
+    {
+      "epoch": 7.618918430095858,
+      "grad_norm": 0.04366715997457504,
+      "learning_rate": 2.427887272057469e-07,
+      "loss": 0.0037,
+      "step": 42125
+    },
+    {
+      "epoch": 7.623440043407488,
+      "grad_norm": 0.01586577482521534,
+      "learning_rate": 2.4232823724442807e-07,
+      "loss": 0.017,
+      "step": 42150
+    },
+    {
+      "epoch": 7.627961656719117,
+      "grad_norm": 21.55689239501953,
+      "learning_rate": 2.4186774728310924e-07,
+      "loss": 0.0098,
+      "step": 42175
+    },
+    {
+      "epoch": 7.632483270030747,
+      "grad_norm": 0.019474711269140244,
+      "learning_rate": 2.4140725732179035e-07,
+      "loss": 0.008,
+      "step": 42200
+    },
+    {
+      "epoch": 7.637004883342376,
+      "grad_norm": 0.20361949503421783,
+      "learning_rate": 2.409467673604715e-07,
+      "loss": 0.0021,
+      "step": 42225
+    },
+    {
+      "epoch": 7.641526496654006,
+      "grad_norm": 0.061369773000478745,
+      "learning_rate": 2.404862773991527e-07,
+      "loss": 0.0167,
+      "step": 42250
+    },
+    {
+      "epoch": 7.646048109965636,
+      "grad_norm": 0.03836612030863762,
+      "learning_rate": 2.400442070362866e-07,
+      "loss": 0.0152,
+      "step": 42275
+    },
+    {
+      "epoch": 7.650569723277266,
+      "grad_norm": 0.6904728412628174,
+      "learning_rate": 2.395837170749677e-07,
+      "loss": 0.0254,
+      "step": 42300
+    },
+    {
+      "epoch": 7.655091336588895,
+      "grad_norm": 9.530180931091309,
+      "learning_rate": 2.391232271136489e-07,
+      "loss": 0.0247,
+      "step": 42325
+    },
+    {
+      "epoch": 7.6596129499005245,
+      "grad_norm": 0.010562293231487274,
+      "learning_rate": 2.3866273715233006e-07,
+      "loss": 0.0021,
+      "step": 42350
+    },
+    {
+      "epoch": 7.664134563212154,
+      "grad_norm": 0.032406773418188095,
+      "learning_rate": 2.3820224719101122e-07,
+      "loss": 0.0026,
+      "step": 42375
+    },
+    {
+      "epoch": 7.6686561765237835,
+      "grad_norm": 7.6475958824157715,
+      "learning_rate": 2.377417572296924e-07,
+      "loss": 0.0095,
+      "step": 42400
+    },
+    {
+      "epoch": 7.673177789835413,
+      "grad_norm": 0.03494368493556976,
+      "learning_rate": 2.3728126726837353e-07,
+      "loss": 0.0029,
+      "step": 42425
+    },
+    {
+      "epoch": 7.677699403147043,
+      "grad_norm": 14.583782196044922,
+      "learning_rate": 2.368207773070547e-07,
+      "loss": 0.0128,
+      "step": 42450
+    },
+    {
+      "epoch": 7.682221016458673,
+      "grad_norm": 0.12016825377941132,
+      "learning_rate": 2.3636028734573584e-07,
+      "loss": 0.0124,
+      "step": 42475
+    },
+    {
+      "epoch": 7.686742629770302,
+      "grad_norm": 0.11401164531707764,
+      "learning_rate": 2.35899797384417e-07,
+      "loss": 0.0006,
+      "step": 42500
+    },
+    {
+      "epoch": 7.691264243081932,
+      "grad_norm": 0.030007168650627136,
+      "learning_rate": 2.3543930742309817e-07,
+      "loss": 0.0013,
+      "step": 42525
+    },
+    {
+      "epoch": 7.695785856393561,
+      "grad_norm": 0.01560743898153305,
+      "learning_rate": 2.3497881746177933e-07,
+      "loss": 0.0132,
+      "step": 42550
+    },
+    {
+      "epoch": 7.700307469705191,
+      "grad_norm": 0.03253242000937462,
+      "learning_rate": 2.3451832750046047e-07,
+      "loss": 0.0104,
+      "step": 42575
+    },
+    {
+      "epoch": 7.70482908301682,
+      "grad_norm": 0.004338828381150961,
+      "learning_rate": 2.3405783753914164e-07,
+      "loss": 0.0108,
+      "step": 42600
+    },
+    {
+      "epoch": 7.70935069632845,
+      "grad_norm": 0.06748280674219131,
+      "learning_rate": 2.3359734757782278e-07,
+      "loss": 0.0169,
+      "step": 42625
+    },
+    {
+      "epoch": 7.713872309640079,
+      "grad_norm": 0.0437793992459774,
+      "learning_rate": 2.3313685761650395e-07,
+      "loss": 0.0142,
+      "step": 42650
+    },
+    {
+      "epoch": 7.718393922951709,
+      "grad_norm": 0.06706016510725021,
+      "learning_rate": 2.326763676551851e-07,
+      "loss": 0.0117,
+      "step": 42675
+    },
+    {
+      "epoch": 7.722915536263339,
+      "grad_norm": 25.65509796142578,
+      "learning_rate": 2.3221587769386628e-07,
+      "loss": 0.0218,
+      "step": 42700
+    },
+    {
+      "epoch": 7.7274371495749685,
+      "grad_norm": 0.0024367687292397022,
+      "learning_rate": 2.3175538773254742e-07,
+      "loss": 0.0028,
+      "step": 42725
+    },
+    {
+      "epoch": 7.731958762886598,
+      "grad_norm": 0.3435537815093994,
+      "learning_rate": 2.3129489777122856e-07,
+      "loss": 0.0041,
+      "step": 42750
+    },
+    {
+      "epoch": 7.7364803761982275,
+      "grad_norm": 0.23171645402908325,
+      "learning_rate": 2.3083440780990972e-07,
+      "loss": 0.0054,
+      "step": 42775
+    },
+    {
+      "epoch": 7.741001989509857,
+      "grad_norm": 0.06181083992123604,
+      "learning_rate": 2.303739178485909e-07,
+      "loss": 0.0006,
+      "step": 42800
+    },
+    {
+      "epoch": 7.745523602821486,
+      "grad_norm": 0.10873863101005554,
+      "learning_rate": 2.2991342788727206e-07,
+      "loss": 0.0035,
+      "step": 42825
+    },
+    {
+      "epoch": 7.750045216133116,
+      "grad_norm": 0.0161910280585289,
+      "learning_rate": 2.2945293792595322e-07,
+      "loss": 0.0048,
+      "step": 42850
+    },
+    {
+      "epoch": 7.754566829444746,
+      "grad_norm": 0.07214026153087616,
+      "learning_rate": 2.2899244796463434e-07,
+      "loss": 0.002,
+      "step": 42875
+    },
+    {
+      "epoch": 7.759088442756376,
+      "grad_norm": 0.01364390179514885,
+      "learning_rate": 2.285319580033155e-07,
+      "loss": 0.0007,
+      "step": 42900
+    },
+    {
+      "epoch": 7.763610056068005,
+      "grad_norm": 0.023643679916858673,
+      "learning_rate": 2.2807146804199667e-07,
+      "loss": 0.0022,
+      "step": 42925
+    },
+    {
+      "epoch": 7.768131669379635,
+      "grad_norm": 0.021116966381669044,
+      "learning_rate": 2.2761097808067784e-07,
+      "loss": 0.0101,
+      "step": 42950
+    },
+    {
+      "epoch": 7.772653282691264,
+      "grad_norm": 0.10896778851747513,
+      "learning_rate": 2.27150488119359e-07,
+      "loss": 0.0031,
+      "step": 42975
+    },
+    {
+      "epoch": 7.777174896002894,
+      "grad_norm": 0.010054018348455429,
+      "learning_rate": 2.2668999815804017e-07,
+      "loss": 0.0057,
+      "step": 43000
+    },
+    {
+      "epoch": 7.781696509314523,
+      "grad_norm": 0.06566222012042999,
+      "learning_rate": 2.2622950819672128e-07,
+      "loss": 0.007,
+      "step": 43025
+    },
+    {
+      "epoch": 7.786218122626153,
+      "grad_norm": 0.05465374514460564,
+      "learning_rate": 2.2576901823540245e-07,
+      "loss": 0.0005,
+      "step": 43050
+    },
+    {
+      "epoch": 7.790739735937782,
+      "grad_norm": 0.6215785145759583,
+      "learning_rate": 2.2530852827408361e-07,
+      "loss": 0.0265,
+      "step": 43075
+    },
+    {
+      "epoch": 7.7952613492494125,
+      "grad_norm": 3.957580089569092,
+      "learning_rate": 2.2484803831276478e-07,
+      "loss": 0.0708,
+      "step": 43100
+    },
+    {
+      "epoch": 7.799782962561042,
+      "grad_norm": 0.034929584711790085,
+      "learning_rate": 2.2438754835144595e-07,
+      "loss": 0.0043,
+      "step": 43125
+    },
+    {
+      "epoch": 7.8043045758726715,
+      "grad_norm": 0.06470832228660583,
+      "learning_rate": 2.2392705839012706e-07,
+      "loss": 0.0059,
+      "step": 43150
+    },
+    {
+      "epoch": 7.808826189184301,
+      "grad_norm": 3.028578758239746,
+      "learning_rate": 2.2346656842880823e-07,
+      "loss": 0.0054,
+      "step": 43175
+    },
+    {
+      "epoch": 7.81334780249593,
+      "grad_norm": 0.060492563992738724,
+      "learning_rate": 2.230060784674894e-07,
+      "loss": 0.0099,
+      "step": 43200
+    },
+    {
+      "epoch": 7.81786941580756,
+      "grad_norm": 1.7288694381713867,
+      "learning_rate": 2.2254558850617056e-07,
+      "loss": 0.0016,
+      "step": 43225
+    },
+    {
+      "epoch": 7.822391029119189,
+      "grad_norm": 0.06679031252861023,
+      "learning_rate": 2.2208509854485172e-07,
+      "loss": 0.0042,
+      "step": 43250
+    },
+    {
+      "epoch": 7.82691264243082,
+      "grad_norm": 17.381139755249023,
+      "learning_rate": 2.216246085835329e-07,
+      "loss": 0.0047,
+      "step": 43275
+    },
+    {
+      "epoch": 7.831434255742449,
+      "grad_norm": 0.03778848424553871,
+      "learning_rate": 2.21164118622214e-07,
+      "loss": 0.0077,
+      "step": 43300
+    },
+    {
+      "epoch": 7.835955869054079,
+      "grad_norm": 0.009451249614357948,
+      "learning_rate": 2.2070362866089517e-07,
+      "loss": 0.0021,
+      "step": 43325
+    },
+    {
+      "epoch": 7.840477482365708,
+      "grad_norm": 25.109268188476562,
+      "learning_rate": 2.2024313869957634e-07,
+      "loss": 0.0042,
+      "step": 43350
+    },
+    {
+      "epoch": 7.844999095677338,
+      "grad_norm": 0.015822693705558777,
+      "learning_rate": 2.197826487382575e-07,
+      "loss": 0.0063,
+      "step": 43375
+    },
+    {
+      "epoch": 7.849520708988967,
+      "grad_norm": 0.016655854880809784,
+      "learning_rate": 2.1932215877693867e-07,
+      "loss": 0.0081,
+      "step": 43400
+    },
+    {
+      "epoch": 7.854042322300597,
+      "grad_norm": 0.25390344858169556,
+      "learning_rate": 2.188616688156198e-07,
+      "loss": 0.0037,
+      "step": 43425
+    },
+    {
+      "epoch": 7.858563935612226,
+      "grad_norm": 0.09659765660762787,
+      "learning_rate": 2.1840117885430095e-07,
+      "loss": 0.0146,
+      "step": 43450
+    },
+    {
+      "epoch": 7.863085548923856,
+      "grad_norm": 0.028894655406475067,
+      "learning_rate": 2.1794068889298211e-07,
+      "loss": 0.0147,
+      "step": 43475
+    },
+    {
+      "epoch": 7.867607162235486,
+      "grad_norm": 0.015346791595220566,
+      "learning_rate": 2.1748019893166328e-07,
+      "loss": 0.0241,
+      "step": 43500
+    },
+    {
+      "epoch": 7.872128775547115,
+      "grad_norm": 0.43620771169662476,
+      "learning_rate": 2.1701970897034445e-07,
+      "loss": 0.0101,
+      "step": 43525
+    },
+    {
+      "epoch": 7.876650388858745,
+      "grad_norm": 0.023062733933329582,
+      "learning_rate": 2.1655921900902561e-07,
+      "loss": 0.0048,
+      "step": 43550
+    },
+    {
+      "epoch": 7.881172002170374,
+      "grad_norm": 0.011817359365522861,
+      "learning_rate": 2.1609872904770675e-07,
+      "loss": 0.0033,
+      "step": 43575
+    },
+    {
+      "epoch": 7.885693615482004,
+      "grad_norm": 0.3834367096424103,
+      "learning_rate": 2.156382390863879e-07,
+      "loss": 0.0012,
+      "step": 43600
+    },
+    {
+      "epoch": 7.890215228793633,
+      "grad_norm": 4.274073600769043,
+      "learning_rate": 2.1517774912506906e-07,
+      "loss": 0.0077,
+      "step": 43625
+    },
+    {
+      "epoch": 7.894736842105263,
+      "grad_norm": 0.024657847359776497,
+      "learning_rate": 2.1471725916375023e-07,
+      "loss": 0.0069,
+      "step": 43650
+    },
+    {
+      "epoch": 7.899258455416893,
+      "grad_norm": 7.262208461761475,
+      "learning_rate": 2.142567692024314e-07,
+      "loss": 0.0073,
+      "step": 43675
+    },
+    {
+      "epoch": 7.903780068728523,
+      "grad_norm": 0.035356614738702774,
+      "learning_rate": 2.1379627924111253e-07,
+      "loss": 0.0162,
+      "step": 43700
+    },
+    {
+      "epoch": 7.908301682040152,
+      "grad_norm": 0.07969994097948074,
+      "learning_rate": 2.133357892797937e-07,
+      "loss": 0.0014,
+      "step": 43725
+    },
+    {
+      "epoch": 7.912823295351782,
+      "grad_norm": 0.013595969416201115,
+      "learning_rate": 2.1287529931847484e-07,
+      "loss": 0.0012,
+      "step": 43750
+    },
+    {
+      "epoch": 7.917344908663411,
+      "grad_norm": 0.02247740514576435,
+      "learning_rate": 2.12414809357156e-07,
+      "loss": 0.0012,
+      "step": 43775
+    },
+    {
+      "epoch": 7.921866521975041,
+      "grad_norm": 0.07739664614200592,
+      "learning_rate": 2.1195431939583717e-07,
+      "loss": 0.0047,
+      "step": 43800
+    },
+    {
+      "epoch": 7.92638813528667,
+      "grad_norm": 0.32251256704330444,
+      "learning_rate": 2.114938294345183e-07,
+      "loss": 0.0056,
+      "step": 43825
+    },
+    {
+      "epoch": 7.9309097485983,
+      "grad_norm": 0.007755752187222242,
+      "learning_rate": 2.1103333947319948e-07,
+      "loss": 0.0188,
+      "step": 43850
+    },
+    {
+      "epoch": 7.935431361909929,
+      "grad_norm": 8.146416664123535,
+      "learning_rate": 2.1057284951188062e-07,
+      "loss": 0.014,
+      "step": 43875
+    },
+    {
+      "epoch": 7.939952975221559,
+      "grad_norm": 22.875139236450195,
+      "learning_rate": 2.1011235955056178e-07,
+      "loss": 0.0218,
+      "step": 43900
+    },
+    {
+      "epoch": 7.944474588533189,
+      "grad_norm": 0.0895143672823906,
+      "learning_rate": 2.0965186958924295e-07,
+      "loss": 0.0331,
+      "step": 43925
+    },
+    {
+      "epoch": 7.948996201844818,
+      "grad_norm": 0.019100898876786232,
+      "learning_rate": 2.0919137962792411e-07,
+      "loss": 0.0013,
+      "step": 43950
+    },
+    {
+      "epoch": 7.953517815156448,
+      "grad_norm": 0.0297053474932909,
+      "learning_rate": 2.0873088966660525e-07,
+      "loss": 0.0007,
+      "step": 43975
+    },
+    {
+      "epoch": 7.958039428468077,
+      "grad_norm": 0.4429755210876465,
+      "learning_rate": 2.0827039970528642e-07,
+      "loss": 0.006,
+      "step": 44000
+    },
+    {
+      "epoch": 7.962561041779707,
+      "grad_norm": 0.021714074537158012,
+      "learning_rate": 2.0780990974396756e-07,
+      "loss": 0.0126,
+      "step": 44025
+    },
+    {
+      "epoch": 7.967082655091336,
+      "grad_norm": 0.048985399305820465,
+      "learning_rate": 2.0734941978264873e-07,
+      "loss": 0.0055,
+      "step": 44050
+    },
+    {
+      "epoch": 7.971604268402967,
+      "grad_norm": 0.22394807636737823,
+      "learning_rate": 2.068889298213299e-07,
+      "loss": 0.0006,
+      "step": 44075
+    },
+    {
+      "epoch": 7.976125881714596,
+      "grad_norm": 0.030540427193045616,
+      "learning_rate": 2.0642843986001103e-07,
+      "loss": 0.0146,
+      "step": 44100
+    },
+    {
+      "epoch": 7.980647495026226,
+      "grad_norm": 1.285352110862732,
+      "learning_rate": 2.059679498986922e-07,
+      "loss": 0.0027,
+      "step": 44125
+    },
+    {
+      "epoch": 7.985169108337855,
+      "grad_norm": 0.025696493685245514,
+      "learning_rate": 2.0550745993737336e-07,
+      "loss": 0.0072,
+      "step": 44150
+    },
+    {
+      "epoch": 7.989690721649485,
+      "grad_norm": 0.1506178379058838,
+      "learning_rate": 2.050469699760545e-07,
+      "loss": 0.0024,
+      "step": 44175
+    },
+    {
+      "epoch": 7.994212334961114,
+      "grad_norm": 0.03548096492886543,
+      "learning_rate": 2.0458648001473567e-07,
+      "loss": 0.0208,
+      "step": 44200
+    },
+    {
+      "epoch": 7.9987339482727435,
+      "grad_norm": 1.0646295547485352,
+      "learning_rate": 2.0412599005341684e-07,
+      "loss": 0.043,
+      "step": 44225
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.36690396070480347,
+      "eval_runtime": 8374.4331,
+      "eval_samples_per_second": 1.134,
+      "eval_steps_per_second": 0.142,
+      "eval_wer": 0.10485265855221013,
+      "step": 44232
+    },
+    {
+      "epoch": 8.003255561584373,
+      "grad_norm": 0.08020028471946716,
+      "learning_rate": 2.0366550009209798e-07,
+      "loss": 0.02,
+      "step": 44250
+    },
+    {
+      "epoch": 8.007777174896002,
+      "grad_norm": 0.013830600306391716,
+      "learning_rate": 2.0320501013077914e-07,
+      "loss": 0.0032,
+      "step": 44275
+    },
+    {
+      "epoch": 8.012298788207632,
+      "grad_norm": 0.05772462114691734,
+      "learning_rate": 2.027445201694603e-07,
+      "loss": 0.0008,
+      "step": 44300
+    },
+    {
+      "epoch": 8.016820401519261,
+      "grad_norm": 0.021993961185216904,
+      "learning_rate": 2.0228403020814145e-07,
+      "loss": 0.0041,
+      "step": 44325
+    },
+    {
+      "epoch": 8.021342014830891,
+      "grad_norm": 5.4001383781433105,
+      "learning_rate": 2.0182354024682262e-07,
+      "loss": 0.0048,
+      "step": 44350
+    },
+    {
+      "epoch": 8.02586362814252,
+      "grad_norm": 0.08318978548049927,
+      "learning_rate": 2.0136305028550375e-07,
+      "loss": 0.0045,
+      "step": 44375
+    },
+    {
+      "epoch": 8.030385241454152,
+      "grad_norm": 0.09250658750534058,
+      "learning_rate": 2.0090256032418492e-07,
+      "loss": 0.0032,
+      "step": 44400
+    },
+    {
+      "epoch": 8.034906854765781,
+      "grad_norm": 0.23829813301563263,
+      "learning_rate": 2.004420703628661e-07,
+      "loss": 0.0019,
+      "step": 44425
+    },
+    {
+      "epoch": 8.03942846807741,
+      "grad_norm": 0.016779888421297073,
+      "learning_rate": 1.9998158040154725e-07,
+      "loss": 0.0033,
+      "step": 44450
+    },
+    {
+      "epoch": 8.04395008138904,
+      "grad_norm": 0.08098112046718597,
+      "learning_rate": 1.995210904402284e-07,
+      "loss": 0.0017,
+      "step": 44475
+    },
+    {
+      "epoch": 8.04847169470067,
+      "grad_norm": 0.042912207543849945,
+      "learning_rate": 1.9906060047890953e-07,
+      "loss": 0.0038,
+      "step": 44500
+    },
+    {
+      "epoch": 8.052993308012299,
+      "grad_norm": 0.052388470619916916,
+      "learning_rate": 1.986001105175907e-07,
+      "loss": 0.0004,
+      "step": 44525
+    },
+    {
+      "epoch": 8.057514921323929,
+      "grad_norm": 3.6011130809783936,
+      "learning_rate": 1.9813962055627187e-07,
+      "loss": 0.0064,
+      "step": 44550
+    },
+    {
+      "epoch": 8.062036534635558,
+      "grad_norm": 0.05901797115802765,
+      "learning_rate": 1.9767913059495303e-07,
+      "loss": 0.0098,
+      "step": 44575
+    },
+    {
+      "epoch": 8.066558147947188,
+      "grad_norm": 0.08754336833953857,
+      "learning_rate": 1.9721864063363417e-07,
+      "loss": 0.0342,
+      "step": 44600
+    },
+    {
+      "epoch": 8.071079761258817,
+      "grad_norm": 33.210994720458984,
+      "learning_rate": 1.9675815067231534e-07,
+      "loss": 0.0427,
+      "step": 44625
+    },
+    {
+      "epoch": 8.075601374570446,
+      "grad_norm": 0.3902020752429962,
+      "learning_rate": 1.9629766071099648e-07,
+      "loss": 0.0331,
+      "step": 44650
+    },
+    {
+      "epoch": 8.080122987882076,
+      "grad_norm": 0.5214375853538513,
+      "learning_rate": 1.9583717074967764e-07,
+      "loss": 0.0022,
+      "step": 44675
+    },
+    {
+      "epoch": 8.084644601193705,
+      "grad_norm": 0.03465314209461212,
+      "learning_rate": 1.953766807883588e-07,
+      "loss": 0.0085,
+      "step": 44700
+    },
+    {
+      "epoch": 8.089166214505335,
+      "grad_norm": 5.756930828094482,
+      "learning_rate": 1.9491619082703998e-07,
+      "loss": 0.0024,
+      "step": 44725
+    },
+    {
+      "epoch": 8.093687827816964,
+      "grad_norm": 0.10550093650817871,
+      "learning_rate": 1.9445570086572112e-07,
+      "loss": 0.0109,
+      "step": 44750
+    },
+    {
+      "epoch": 8.098209441128594,
+      "grad_norm": 2.2892873287200928,
+      "learning_rate": 1.9399521090440226e-07,
+      "loss": 0.0022,
+      "step": 44775
+    },
+    {
+      "epoch": 8.102731054440225,
+      "grad_norm": 0.008341608569025993,
+      "learning_rate": 1.9353472094308342e-07,
+      "loss": 0.0004,
+      "step": 44800
+    },
+    {
+      "epoch": 8.107252667751855,
+      "grad_norm": 4.5169782638549805,
+      "learning_rate": 1.930742309817646e-07,
+      "loss": 0.0021,
+      "step": 44825
+    },
+    {
+      "epoch": 8.111774281063484,
+      "grad_norm": 0.0315685048699379,
+      "learning_rate": 1.9261374102044575e-07,
+      "loss": 0.007,
+      "step": 44850
+    },
+    {
+      "epoch": 8.116295894375114,
+      "grad_norm": 0.12863166630268097,
+      "learning_rate": 1.9215325105912692e-07,
+      "loss": 0.0056,
+      "step": 44875
+    },
+    {
+      "epoch": 8.120817507686743,
+      "grad_norm": 0.6251371502876282,
+      "learning_rate": 1.9169276109780806e-07,
+      "loss": 0.0006,
+      "step": 44900
+    },
+    {
+      "epoch": 8.125339120998373,
+      "grad_norm": 0.8310458064079285,
+      "learning_rate": 1.912322711364892e-07,
+      "loss": 0.0084,
+      "step": 44925
+    },
+    {
+      "epoch": 8.129860734310002,
+      "grad_norm": 0.02388242445886135,
+      "learning_rate": 1.9077178117517037e-07,
+      "loss": 0.0113,
+      "step": 44950
+    },
+    {
+      "epoch": 8.134382347621631,
+      "grad_norm": 0.00754801370203495,
+      "learning_rate": 1.9031129121385153e-07,
+      "loss": 0.0031,
+      "step": 44975
+    },
+    {
+      "epoch": 8.138903960933261,
+      "grad_norm": 0.41686248779296875,
+      "learning_rate": 1.898508012525327e-07,
+      "loss": 0.0042,
+      "step": 45000
+    },
+    {
+      "epoch": 8.14342557424489,
+      "grad_norm": 3.3315136432647705,
+      "learning_rate": 1.8939031129121387e-07,
+      "loss": 0.0198,
+      "step": 45025
+    },
+    {
+      "epoch": 8.14794718755652,
+      "grad_norm": 0.44836312532424927,
+      "learning_rate": 1.8892982132989498e-07,
+      "loss": 0.0191,
+      "step": 45050
+    },
+    {
+      "epoch": 8.15246880086815,
+      "grad_norm": 0.06881590187549591,
+      "learning_rate": 1.8846933136857614e-07,
+      "loss": 0.0016,
+      "step": 45075
+    },
+    {
+      "epoch": 8.156990414179779,
+      "grad_norm": 0.020737141370773315,
+      "learning_rate": 1.880088414072573e-07,
+      "loss": 0.0054,
+      "step": 45100
+    },
+    {
+      "epoch": 8.161512027491408,
+      "grad_norm": 0.033259179443120956,
+      "learning_rate": 1.8754835144593848e-07,
+      "loss": 0.0007,
+      "step": 45125
+    },
+    {
+      "epoch": 8.166033640803038,
+      "grad_norm": 14.501016616821289,
+      "learning_rate": 1.8708786148461964e-07,
+      "loss": 0.0066,
+      "step": 45150
+    },
+    {
+      "epoch": 8.170555254114667,
+      "grad_norm": 0.03625442460179329,
+      "learning_rate": 1.866273715233008e-07,
+      "loss": 0.0036,
+      "step": 45175
+    },
+    {
+      "epoch": 8.175076867426299,
+      "grad_norm": 0.027331039309501648,
+      "learning_rate": 1.8616688156198192e-07,
+      "loss": 0.0058,
+      "step": 45200
+    },
+    {
+      "epoch": 8.179598480737928,
+      "grad_norm": 1.0191278457641602,
+      "learning_rate": 1.857063916006631e-07,
+      "loss": 0.0185,
+      "step": 45225
+    },
+    {
+      "epoch": 8.184120094049558,
+      "grad_norm": 0.65291428565979,
+      "learning_rate": 1.8524590163934426e-07,
+      "loss": 0.0031,
+      "step": 45250
+    },
+    {
+      "epoch": 8.188641707361187,
+      "grad_norm": 0.14517787098884583,
+      "learning_rate": 1.8478541167802542e-07,
+      "loss": 0.0004,
+      "step": 45275
+    },
+    {
+      "epoch": 8.193163320672816,
+      "grad_norm": 0.016481177881360054,
+      "learning_rate": 1.8434334131515933e-07,
+      "loss": 0.0271,
+      "step": 45300
+    },
+    {
+      "epoch": 8.197684933984446,
+      "grad_norm": 2.0038397312164307,
+      "learning_rate": 1.838828513538405e-07,
+      "loss": 0.0157,
+      "step": 45325
+    },
+    {
+      "epoch": 8.202206547296075,
+      "grad_norm": 0.07896800339221954,
+      "learning_rate": 1.8342236139252163e-07,
+      "loss": 0.0092,
+      "step": 45350
+    },
+    {
+      "epoch": 8.206728160607705,
+      "grad_norm": 0.04571648687124252,
+      "learning_rate": 1.829618714312028e-07,
+      "loss": 0.0112,
+      "step": 45375
+    },
+    {
+      "epoch": 8.211249773919334,
+      "grad_norm": 0.03423347696661949,
+      "learning_rate": 1.8250138146988394e-07,
+      "loss": 0.013,
+      "step": 45400
+    },
+    {
+      "epoch": 8.215771387230964,
+      "grad_norm": 34.946712493896484,
+      "learning_rate": 1.820408915085651e-07,
+      "loss": 0.0322,
+      "step": 45425
+    },
+    {
+      "epoch": 8.220293000542593,
+      "grad_norm": 0.0242567490786314,
+      "learning_rate": 1.8158040154724627e-07,
+      "loss": 0.0108,
+      "step": 45450
+    },
+    {
+      "epoch": 8.224814613854223,
+      "grad_norm": 0.04625415802001953,
+      "learning_rate": 1.8111991158592744e-07,
+      "loss": 0.0005,
+      "step": 45475
+    },
+    {
+      "epoch": 8.229336227165852,
+      "grad_norm": 0.16521072387695312,
+      "learning_rate": 1.8065942162460858e-07,
+      "loss": 0.0038,
+      "step": 45500
+    },
+    {
+      "epoch": 8.233857840477482,
+      "grad_norm": 0.01613481342792511,
+      "learning_rate": 1.8019893166328972e-07,
+      "loss": 0.0036,
+      "step": 45525
+    },
+    {
+      "epoch": 8.238379453789111,
+      "grad_norm": 8.090846061706543,
+      "learning_rate": 1.7973844170197088e-07,
+      "loss": 0.011,
+      "step": 45550
+    },
+    {
+      "epoch": 8.24290106710074,
+      "grad_norm": 0.062387678772211075,
+      "learning_rate": 1.7927795174065205e-07,
+      "loss": 0.0022,
+      "step": 45575
+    },
+    {
+      "epoch": 8.24742268041237,
+      "grad_norm": 0.11970767378807068,
+      "learning_rate": 1.7881746177933321e-07,
+      "loss": 0.0078,
+      "step": 45600
+    },
+    {
+      "epoch": 8.251944293724002,
+      "grad_norm": 0.30487823486328125,
+      "learning_rate": 1.7835697181801435e-07,
+      "loss": 0.0005,
+      "step": 45625
+    },
+    {
+      "epoch": 8.256465907035631,
+      "grad_norm": 0.12851175665855408,
+      "learning_rate": 1.7789648185669552e-07,
+      "loss": 0.0041,
+      "step": 45650
+    },
+    {
+      "epoch": 8.26098752034726,
+      "grad_norm": 0.7285523414611816,
+      "learning_rate": 1.7743599189537666e-07,
+      "loss": 0.0012,
+      "step": 45675
+    },
+    {
+      "epoch": 8.26550913365889,
+      "grad_norm": 0.006563634146004915,
+      "learning_rate": 1.7697550193405783e-07,
+      "loss": 0.0006,
+      "step": 45700
+    },
+    {
+      "epoch": 8.27003074697052,
+      "grad_norm": 0.08779245615005493,
+      "learning_rate": 1.76515011972739e-07,
+      "loss": 0.0176,
+      "step": 45725
+    },
+    {
+      "epoch": 8.274552360282149,
+      "grad_norm": 1.7844347953796387,
+      "learning_rate": 1.7605452201142016e-07,
+      "loss": 0.0035,
+      "step": 45750
+    },
+    {
+      "epoch": 8.279073973593778,
+      "grad_norm": 0.026561090722680092,
+      "learning_rate": 1.755940320501013e-07,
+      "loss": 0.0154,
+      "step": 45775
+    },
+    {
+      "epoch": 8.283595586905408,
+      "grad_norm": 0.08869388699531555,
+      "learning_rate": 1.7513354208878244e-07,
+      "loss": 0.0189,
+      "step": 45800
+    },
+    {
+      "epoch": 8.288117200217037,
+      "grad_norm": 0.09968849271535873,
+      "learning_rate": 1.746730521274636e-07,
+      "loss": 0.0057,
+      "step": 45825
+    },
+    {
+      "epoch": 8.292638813528667,
+      "grad_norm": 0.7243178486824036,
+      "learning_rate": 1.7421256216614477e-07,
+      "loss": 0.0214,
+      "step": 45850
+    },
+    {
+      "epoch": 8.297160426840296,
+      "grad_norm": 0.019970480352640152,
+      "learning_rate": 1.7375207220482594e-07,
+      "loss": 0.0058,
+      "step": 45875
+    },
+    {
+      "epoch": 8.301682040151926,
+      "grad_norm": 0.08685352653265,
+      "learning_rate": 1.732915822435071e-07,
+      "loss": 0.0053,
+      "step": 45900
+    },
+    {
+      "epoch": 8.306203653463555,
+      "grad_norm": 0.02851727232336998,
+      "learning_rate": 1.7283109228218822e-07,
+      "loss": 0.0085,
+      "step": 45925
+    },
+    {
+      "epoch": 8.310725266775185,
+      "grad_norm": 0.10517676919698715,
+      "learning_rate": 1.7237060232086938e-07,
+      "loss": 0.0043,
+      "step": 45950
+    },
+    {
+      "epoch": 8.315246880086814,
+      "grad_norm": 0.03421909734606743,
+      "learning_rate": 1.7191011235955055e-07,
+      "loss": 0.0009,
+      "step": 45975
+    },
+    {
+      "epoch": 8.319768493398444,
+      "grad_norm": 8.051795959472656,
+      "learning_rate": 1.7144962239823171e-07,
+      "loss": 0.0012,
+      "step": 46000
+    },
+    {
+      "epoch": 8.324290106710075,
+      "grad_norm": 0.04528782516717911,
+      "learning_rate": 1.7098913243691288e-07,
+      "loss": 0.0049,
+      "step": 46025
+    },
+    {
+      "epoch": 8.328811720021704,
+      "grad_norm": 0.024573039263486862,
+      "learning_rate": 1.7052864247559405e-07,
+      "loss": 0.0055,
+      "step": 46050
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 13.220183372497559,
+      "learning_rate": 1.7006815251427516e-07,
+      "loss": 0.0102,
+      "step": 46075
+    },
+    {
+      "epoch": 8.337854946644963,
+      "grad_norm": 3.289721727371216,
+      "learning_rate": 1.6960766255295633e-07,
+      "loss": 0.0292,
+      "step": 46100
+    },
+    {
+      "epoch": 8.342376559956593,
+      "grad_norm": 10.872947692871094,
+      "learning_rate": 1.691471725916375e-07,
+      "loss": 0.0028,
+      "step": 46125
+    },
+    {
+      "epoch": 8.346898173268222,
+      "grad_norm": 0.3125823438167572,
+      "learning_rate": 1.6868668263031866e-07,
+      "loss": 0.0007,
+      "step": 46150
+    },
+    {
+      "epoch": 8.351419786579852,
+      "grad_norm": 42.284271240234375,
+      "learning_rate": 1.6822619266899983e-07,
+      "loss": 0.0214,
+      "step": 46175
+    },
+    {
+      "epoch": 8.355941399891481,
+      "grad_norm": 40.723052978515625,
+      "learning_rate": 1.6776570270768097e-07,
+      "loss": 0.0119,
+      "step": 46200
+    },
+    {
+      "epoch": 8.36046301320311,
+      "grad_norm": 15.188941955566406,
+      "learning_rate": 1.673052127463621e-07,
+      "loss": 0.0174,
+      "step": 46225
+    },
+    {
+      "epoch": 8.36498462651474,
+      "grad_norm": 0.19278444349765778,
+      "learning_rate": 1.6684472278504327e-07,
+      "loss": 0.0107,
+      "step": 46250
+    },
+    {
+      "epoch": 8.36950623982637,
+      "grad_norm": 0.961004912853241,
+      "learning_rate": 1.6638423282372444e-07,
+      "loss": 0.0023,
+      "step": 46275
+    },
+    {
+      "epoch": 8.374027853138,
+      "grad_norm": 0.09470321238040924,
+      "learning_rate": 1.659237428624056e-07,
+      "loss": 0.0023,
+      "step": 46300
+    },
+    {
+      "epoch": 8.378549466449629,
+      "grad_norm": 0.10626068711280823,
+      "learning_rate": 1.6546325290108677e-07,
+      "loss": 0.0022,
+      "step": 46325
+    },
+    {
+      "epoch": 8.383071079761258,
+      "grad_norm": 0.07527956366539001,
+      "learning_rate": 1.6500276293976788e-07,
+      "loss": 0.0021,
+      "step": 46350
+    },
+    {
+      "epoch": 8.387592693072888,
+      "grad_norm": 0.07098814100027084,
+      "learning_rate": 1.6454227297844905e-07,
+      "loss": 0.0032,
+      "step": 46375
+    },
+    {
+      "epoch": 8.392114306384517,
+      "grad_norm": 0.06297345459461212,
+      "learning_rate": 1.6408178301713022e-07,
+      "loss": 0.0009,
+      "step": 46400
+    },
+    {
+      "epoch": 8.396635919696148,
+      "grad_norm": 0.6129382252693176,
+      "learning_rate": 1.6362129305581138e-07,
+      "loss": 0.0023,
+      "step": 46425
+    },
+    {
+      "epoch": 8.401157533007778,
+      "grad_norm": 0.07391338050365448,
+      "learning_rate": 1.6316080309449255e-07,
+      "loss": 0.0192,
+      "step": 46450
+    },
+    {
+      "epoch": 8.405679146319407,
+      "grad_norm": 0.009829241782426834,
+      "learning_rate": 1.627003131331737e-07,
+      "loss": 0.0009,
+      "step": 46475
+    },
+    {
+      "epoch": 8.410200759631037,
+      "grad_norm": 0.02008306048810482,
+      "learning_rate": 1.6223982317185483e-07,
+      "loss": 0.0012,
+      "step": 46500
+    },
+    {
+      "epoch": 8.414722372942666,
+      "grad_norm": 0.032892853021621704,
+      "learning_rate": 1.61779333210536e-07,
+      "loss": 0.0117,
+      "step": 46525
+    },
+    {
+      "epoch": 8.419243986254296,
+      "grad_norm": 0.3454951047897339,
+      "learning_rate": 1.6131884324921716e-07,
+      "loss": 0.0041,
+      "step": 46550
+    },
+    {
+      "epoch": 8.423765599565925,
+      "grad_norm": 41.06159973144531,
+      "learning_rate": 1.6085835328789833e-07,
+      "loss": 0.0189,
+      "step": 46575
+    },
+    {
+      "epoch": 8.428287212877555,
+      "grad_norm": 11.987919807434082,
+      "learning_rate": 1.603978633265795e-07,
+      "loss": 0.01,
+      "step": 46600
+    },
+    {
+      "epoch": 8.432808826189184,
+      "grad_norm": 0.24590350687503815,
+      "learning_rate": 1.5993737336526063e-07,
+      "loss": 0.0019,
+      "step": 46625
+    },
+    {
+      "epoch": 8.437330439500814,
+      "grad_norm": 0.11725660413503647,
+      "learning_rate": 1.5947688340394177e-07,
+      "loss": 0.0196,
+      "step": 46650
+    },
+    {
+      "epoch": 8.441852052812443,
+      "grad_norm": 0.04116886481642723,
+      "learning_rate": 1.5901639344262294e-07,
+      "loss": 0.0011,
+      "step": 46675
+    },
+    {
+      "epoch": 8.446373666124073,
+      "grad_norm": 0.03313690423965454,
+      "learning_rate": 1.585559034813041e-07,
+      "loss": 0.0051,
+      "step": 46700
+    },
+    {
+      "epoch": 8.450895279435702,
+      "grad_norm": 0.1109694391489029,
+      "learning_rate": 1.5809541351998527e-07,
+      "loss": 0.0042,
+      "step": 46725
+    },
+    {
+      "epoch": 8.455416892747332,
+      "grad_norm": 0.2245902568101883,
+      "learning_rate": 1.576349235586664e-07,
+      "loss": 0.0059,
+      "step": 46750
+    },
+    {
+      "epoch": 8.459938506058961,
+      "grad_norm": 0.014154641889035702,
+      "learning_rate": 1.5717443359734758e-07,
+      "loss": 0.0023,
+      "step": 46775
+    },
+    {
+      "epoch": 8.46446011937059,
+      "grad_norm": 0.16116198897361755,
+      "learning_rate": 1.5671394363602872e-07,
+      "loss": 0.0158,
+      "step": 46800
+    },
+    {
+      "epoch": 8.468981732682222,
+      "grad_norm": 0.00956847332417965,
+      "learning_rate": 1.5625345367470988e-07,
+      "loss": 0.0024,
+      "step": 46825
+    },
+    {
+      "epoch": 8.473503345993851,
+      "grad_norm": 0.007467388175427914,
+      "learning_rate": 1.5579296371339105e-07,
+      "loss": 0.0074,
+      "step": 46850
+    },
+    {
+      "epoch": 8.47802495930548,
+      "grad_norm": 0.028611036017537117,
+      "learning_rate": 1.553324737520722e-07,
+      "loss": 0.0052,
+      "step": 46875
+    },
+    {
+      "epoch": 8.48254657261711,
+      "grad_norm": 0.23077279329299927,
+      "learning_rate": 1.5487198379075336e-07,
+      "loss": 0.0137,
+      "step": 46900
+    },
+    {
+      "epoch": 8.48706818592874,
+      "grad_norm": 0.1357181966304779,
+      "learning_rate": 1.5441149382943452e-07,
+      "loss": 0.0082,
+      "step": 46925
+    },
+    {
+      "epoch": 8.49158979924037,
+      "grad_norm": 12.11319351196289,
+      "learning_rate": 1.5395100386811566e-07,
+      "loss": 0.0021,
+      "step": 46950
+    },
+    {
+      "epoch": 8.496111412551999,
+      "grad_norm": 1.0207489728927612,
+      "learning_rate": 1.5349051390679683e-07,
+      "loss": 0.0012,
+      "step": 46975
+    },
+    {
+      "epoch": 8.500633025863628,
+      "grad_norm": 0.05066627636551857,
+      "learning_rate": 1.53030023945478e-07,
+      "loss": 0.0033,
+      "step": 47000
+    },
+    {
+      "epoch": 8.505154639175258,
+      "grad_norm": 0.18461652100086212,
+      "learning_rate": 1.5256953398415913e-07,
+      "loss": 0.0051,
+      "step": 47025
+    },
+    {
+      "epoch": 8.509676252486887,
+      "grad_norm": 0.058383241295814514,
+      "learning_rate": 1.521090440228403e-07,
+      "loss": 0.0079,
+      "step": 47050
+    },
+    {
+      "epoch": 8.514197865798517,
+      "grad_norm": 0.08401685953140259,
+      "learning_rate": 1.5164855406152144e-07,
+      "loss": 0.005,
+      "step": 47075
+    },
+    {
+      "epoch": 8.518719479110146,
+      "grad_norm": 0.12280333787202835,
+      "learning_rate": 1.511880641002026e-07,
+      "loss": 0.0027,
+      "step": 47100
+    },
+    {
+      "epoch": 8.523241092421776,
+      "grad_norm": 0.0439714677631855,
+      "learning_rate": 1.5072757413888377e-07,
+      "loss": 0.002,
+      "step": 47125
+    },
+    {
+      "epoch": 8.527762705733405,
+      "grad_norm": 0.02209680713713169,
+      "learning_rate": 1.502670841775649e-07,
+      "loss": 0.0009,
+      "step": 47150
+    },
+    {
+      "epoch": 8.532284319045035,
+      "grad_norm": 6.216536521911621,
+      "learning_rate": 1.4980659421624608e-07,
+      "loss": 0.0197,
+      "step": 47175
+    },
+    {
+      "epoch": 8.536805932356664,
+      "grad_norm": 0.030912378802895546,
+      "learning_rate": 1.4934610425492724e-07,
+      "loss": 0.0003,
+      "step": 47200
+    },
+    {
+      "epoch": 8.541327545668295,
+      "grad_norm": 0.02260858565568924,
+      "learning_rate": 1.4888561429360838e-07,
+      "loss": 0.0013,
+      "step": 47225
+    },
+    {
+      "epoch": 8.545849158979925,
+      "grad_norm": 0.04053138568997383,
+      "learning_rate": 1.4842512433228955e-07,
+      "loss": 0.0036,
+      "step": 47250
+    },
+    {
+      "epoch": 8.550370772291554,
+      "grad_norm": 0.10423174500465393,
+      "learning_rate": 1.4796463437097072e-07,
+      "loss": 0.0005,
+      "step": 47275
+    },
+    {
+      "epoch": 8.554892385603184,
+      "grad_norm": 0.060733165591955185,
+      "learning_rate": 1.4750414440965186e-07,
+      "loss": 0.0017,
+      "step": 47300
+    },
+    {
+      "epoch": 8.559413998914813,
+      "grad_norm": 0.07485374808311462,
+      "learning_rate": 1.4704365444833302e-07,
+      "loss": 0.0007,
+      "step": 47325
+    },
+    {
+      "epoch": 8.563935612226443,
+      "grad_norm": 0.03115830197930336,
+      "learning_rate": 1.465831644870142e-07,
+      "loss": 0.0112,
+      "step": 47350
+    },
+    {
+      "epoch": 8.568457225538072,
+      "grad_norm": 0.02164495922625065,
+      "learning_rate": 1.4612267452569533e-07,
+      "loss": 0.0006,
+      "step": 47375
+    },
+    {
+      "epoch": 8.572978838849702,
+      "grad_norm": 0.024691320955753326,
+      "learning_rate": 1.456621845643765e-07,
+      "loss": 0.0259,
+      "step": 47400
+    },
+    {
+      "epoch": 8.577500452161331,
+      "grad_norm": 17.621110916137695,
+      "learning_rate": 1.4520169460305763e-07,
+      "loss": 0.0218,
+      "step": 47425
+    },
+    {
+      "epoch": 8.58202206547296,
+      "grad_norm": 4.349635601043701,
+      "learning_rate": 1.447412046417388e-07,
+      "loss": 0.0055,
+      "step": 47450
+    },
+    {
+      "epoch": 8.58654367878459,
+      "grad_norm": 1.4642325639724731,
+      "learning_rate": 1.4428071468041997e-07,
+      "loss": 0.0065,
+      "step": 47475
+    },
+    {
+      "epoch": 8.59106529209622,
+      "grad_norm": 0.0913548618555069,
+      "learning_rate": 1.4382022471910113e-07,
+      "loss": 0.0034,
+      "step": 47500
+    },
+    {
+      "epoch": 8.595586905407849,
+      "grad_norm": 0.025698378682136536,
+      "learning_rate": 1.4335973475778227e-07,
+      "loss": 0.0023,
+      "step": 47525
+    },
+    {
+      "epoch": 8.600108518719479,
+      "grad_norm": 0.07771757990121841,
+      "learning_rate": 1.428992447964634e-07,
+      "loss": 0.0022,
+      "step": 47550
+    },
+    {
+      "epoch": 8.604630132031108,
+      "grad_norm": 5.4854254722595215,
+      "learning_rate": 1.4243875483514458e-07,
+      "loss": 0.005,
+      "step": 47575
+    },
+    {
+      "epoch": 8.609151745342738,
+      "grad_norm": 0.031100405380129814,
+      "learning_rate": 1.4197826487382574e-07,
+      "loss": 0.0067,
+      "step": 47600
+    },
+    {
+      "epoch": 8.613673358654367,
+      "grad_norm": 0.015138168819248676,
+      "learning_rate": 1.415177749125069e-07,
+      "loss": 0.0094,
+      "step": 47625
+    },
+    {
+      "epoch": 8.618194971965998,
+      "grad_norm": 1.461853265762329,
+      "learning_rate": 1.4105728495118808e-07,
+      "loss": 0.002,
+      "step": 47650
+    },
+    {
+      "epoch": 8.622716585277628,
+      "grad_norm": 0.05030835047364235,
+      "learning_rate": 1.4059679498986922e-07,
+      "loss": 0.0014,
+      "step": 47675
+    },
+    {
+      "epoch": 8.627238198589257,
+      "grad_norm": 0.15744911134243011,
+      "learning_rate": 1.4013630502855036e-07,
+      "loss": 0.0006,
+      "step": 47700
+    },
+    {
+      "epoch": 8.631759811900887,
+      "grad_norm": 0.6732610464096069,
+      "learning_rate": 1.3967581506723152e-07,
+      "loss": 0.0024,
+      "step": 47725
+    },
+    {
+      "epoch": 8.636281425212516,
+      "grad_norm": 3.470303773880005,
+      "learning_rate": 1.392153251059127e-07,
+      "loss": 0.0079,
+      "step": 47750
+    },
+    {
+      "epoch": 8.640803038524146,
+      "grad_norm": 27.621957778930664,
+      "learning_rate": 1.3875483514459386e-07,
+      "loss": 0.0164,
+      "step": 47775
+    },
+    {
+      "epoch": 8.645324651835775,
+      "grad_norm": 0.01883932389318943,
+      "learning_rate": 1.38294345183275e-07,
+      "loss": 0.0145,
+      "step": 47800
+    },
+    {
+      "epoch": 8.649846265147405,
+      "grad_norm": 0.08440113812685013,
+      "learning_rate": 1.3783385522195614e-07,
+      "loss": 0.0188,
+      "step": 47825
+    },
+    {
+      "epoch": 8.654367878459034,
+      "grad_norm": 0.0456276573240757,
+      "learning_rate": 1.373733652606373e-07,
+      "loss": 0.0116,
+      "step": 47850
+    },
+    {
+      "epoch": 8.658889491770664,
+      "grad_norm": 8.457013130187988,
+      "learning_rate": 1.3691287529931847e-07,
+      "loss": 0.005,
+      "step": 47875
+    },
+    {
+      "epoch": 8.663411105082293,
+      "grad_norm": 2.870605707168579,
+      "learning_rate": 1.3645238533799963e-07,
+      "loss": 0.0051,
+      "step": 47900
+    },
+    {
+      "epoch": 8.667932718393923,
+      "grad_norm": 0.0406540222465992,
+      "learning_rate": 1.359918953766808e-07,
+      "loss": 0.0103,
+      "step": 47925
+    },
+    {
+      "epoch": 8.672454331705552,
+      "grad_norm": 0.12393586337566376,
+      "learning_rate": 1.3553140541536194e-07,
+      "loss": 0.0089,
+      "step": 47950
+    },
+    {
+      "epoch": 8.676975945017182,
+      "grad_norm": 0.04684567451477051,
+      "learning_rate": 1.3507091545404308e-07,
+      "loss": 0.005,
+      "step": 47975
+    },
+    {
+      "epoch": 8.681497558328811,
+      "grad_norm": 0.005455177277326584,
+      "learning_rate": 1.3461042549272425e-07,
+      "loss": 0.0011,
+      "step": 48000
+    },
+    {
+      "epoch": 8.686019171640442,
+      "grad_norm": 0.0036073036026209593,
+      "learning_rate": 1.341499355314054e-07,
+      "loss": 0.0109,
+      "step": 48025
+    },
+    {
+      "epoch": 8.690540784952072,
+      "grad_norm": 0.015157288871705532,
+      "learning_rate": 1.3368944557008658e-07,
+      "loss": 0.0028,
+      "step": 48050
+    },
+    {
+      "epoch": 8.695062398263701,
+      "grad_norm": 0.045515723526477814,
+      "learning_rate": 1.3322895560876774e-07,
+      "loss": 0.0061,
+      "step": 48075
+    },
+    {
+      "epoch": 8.69958401157533,
+      "grad_norm": 0.3657865822315216,
+      "learning_rate": 1.3276846564744886e-07,
+      "loss": 0.0024,
+      "step": 48100
+    },
+    {
+      "epoch": 8.70410562488696,
+      "grad_norm": 0.009580901823937893,
+      "learning_rate": 1.3230797568613002e-07,
+      "loss": 0.0132,
+      "step": 48125
+    },
+    {
+      "epoch": 8.70862723819859,
+      "grad_norm": 0.033869609236717224,
+      "learning_rate": 1.318474857248112e-07,
+      "loss": 0.0049,
+      "step": 48150
+    },
+    {
+      "epoch": 8.71314885151022,
+      "grad_norm": 11.958950996398926,
+      "learning_rate": 1.3138699576349236e-07,
+      "loss": 0.0098,
+      "step": 48175
+    },
+    {
+      "epoch": 8.717670464821849,
+      "grad_norm": 0.02979426644742489,
+      "learning_rate": 1.3092650580217352e-07,
+      "loss": 0.0011,
+      "step": 48200
+    },
+    {
+      "epoch": 8.722192078133478,
+      "grad_norm": 1.0056304931640625,
+      "learning_rate": 1.304660158408547e-07,
+      "loss": 0.0352,
+      "step": 48225
+    },
+    {
+      "epoch": 8.726713691445108,
+      "grad_norm": 1.8282511234283447,
+      "learning_rate": 1.300055258795358e-07,
+      "loss": 0.0204,
+      "step": 48250
+    },
+    {
+      "epoch": 8.731235304756737,
+      "grad_norm": 0.0187783632427454,
+      "learning_rate": 1.2954503591821697e-07,
+      "loss": 0.0129,
+      "step": 48275
+    },
+    {
+      "epoch": 8.735756918068367,
+      "grad_norm": 0.3944437801837921,
+      "learning_rate": 1.2908454595689813e-07,
+      "loss": 0.0009,
+      "step": 48300
+    },
+    {
+      "epoch": 8.740278531379996,
+      "grad_norm": 11.324723243713379,
+      "learning_rate": 1.286240559955793e-07,
+      "loss": 0.0015,
+      "step": 48325
+    },
+    {
+      "epoch": 8.744800144691625,
+      "grad_norm": 0.03956815227866173,
+      "learning_rate": 1.2816356603426047e-07,
+      "loss": 0.003,
+      "step": 48350
+    },
+    {
+      "epoch": 8.749321758003255,
+      "grad_norm": 0.06833972781896591,
+      "learning_rate": 1.277030760729416e-07,
+      "loss": 0.0005,
+      "step": 48375
+    },
+    {
+      "epoch": 8.753843371314884,
+      "grad_norm": 0.029176251962780952,
+      "learning_rate": 1.2724258611162275e-07,
+      "loss": 0.001,
+      "step": 48400
+    },
+    {
+      "epoch": 8.758364984626514,
+      "grad_norm": 0.02110159769654274,
+      "learning_rate": 1.267820961503039e-07,
+      "loss": 0.0009,
+      "step": 48425
+    },
+    {
+      "epoch": 8.762886597938145,
+      "grad_norm": 0.023354971781373024,
+      "learning_rate": 1.2632160618898508e-07,
+      "loss": 0.0012,
+      "step": 48450
+    },
+    {
+      "epoch": 8.767408211249775,
+      "grad_norm": 0.0024972474202513695,
+      "learning_rate": 1.2586111622766625e-07,
+      "loss": 0.0014,
+      "step": 48475
+    },
+    {
+      "epoch": 8.771929824561404,
+      "grad_norm": 10.001263618469238,
+      "learning_rate": 1.2540062626634739e-07,
+      "loss": 0.0187,
+      "step": 48500
+    },
+    {
+      "epoch": 8.776451437873034,
+      "grad_norm": 0.003959618508815765,
+      "learning_rate": 1.2494013630502855e-07,
+      "loss": 0.0181,
+      "step": 48525
+    },
+    {
+      "epoch": 8.780973051184663,
+      "grad_norm": 0.05257127806544304,
+      "learning_rate": 1.2447964634370972e-07,
+      "loss": 0.0041,
+      "step": 48550
+    },
+    {
+      "epoch": 8.785494664496293,
+      "grad_norm": 1.0013961791992188,
+      "learning_rate": 1.2401915638239086e-07,
+      "loss": 0.0028,
+      "step": 48575
+    },
+    {
+      "epoch": 8.790016277807922,
+      "grad_norm": 0.04151635989546776,
+      "learning_rate": 1.2355866642107202e-07,
+      "loss": 0.0439,
+      "step": 48600
+    },
+    {
+      "epoch": 8.794537891119552,
+      "grad_norm": 0.07857254147529602,
+      "learning_rate": 1.230981764597532e-07,
+      "loss": 0.0176,
+      "step": 48625
+    },
+    {
+      "epoch": 8.799059504431181,
+      "grad_norm": 0.0398498959839344,
+      "learning_rate": 1.2263768649843433e-07,
+      "loss": 0.0035,
+      "step": 48650
+    },
+    {
+      "epoch": 8.80358111774281,
+      "grad_norm": 0.18911878764629364,
+      "learning_rate": 1.221771965371155e-07,
+      "loss": 0.0041,
+      "step": 48675
+    },
+    {
+      "epoch": 8.80810273105444,
+      "grad_norm": 0.5053550601005554,
+      "learning_rate": 1.2171670657579666e-07,
+      "loss": 0.0029,
+      "step": 48700
+    },
+    {
+      "epoch": 8.81262434436607,
+      "grad_norm": 0.11117665469646454,
+      "learning_rate": 1.212562166144778e-07,
+      "loss": 0.0022,
+      "step": 48725
+    },
+    {
+      "epoch": 8.817145957677699,
+      "grad_norm": 0.03761090338230133,
+      "learning_rate": 1.2079572665315897e-07,
+      "loss": 0.0014,
+      "step": 48750
+    },
+    {
+      "epoch": 8.821667570989328,
+      "grad_norm": 0.0750727653503418,
+      "learning_rate": 1.203352366918401e-07,
+      "loss": 0.0051,
+      "step": 48775
+    },
+    {
+      "epoch": 8.826189184300958,
+      "grad_norm": 0.1517859250307083,
+      "learning_rate": 1.1987474673052127e-07,
+      "loss": 0.0072,
+      "step": 48800
+    },
+    {
+      "epoch": 8.830710797612587,
+      "grad_norm": 0.02724577859044075,
+      "learning_rate": 1.1941425676920244e-07,
+      "loss": 0.0025,
+      "step": 48825
+    },
+    {
+      "epoch": 8.835232410924217,
+      "grad_norm": 0.02841174229979515,
+      "learning_rate": 1.1895376680788358e-07,
+      "loss": 0.0017,
+      "step": 48850
+    },
+    {
+      "epoch": 8.839754024235848,
+      "grad_norm": 0.03608907014131546,
+      "learning_rate": 1.1849327684656473e-07,
+      "loss": 0.0041,
+      "step": 48875
+    },
+    {
+      "epoch": 8.844275637547478,
+      "grad_norm": 0.022049210965633392,
+      "learning_rate": 1.180327868852459e-07,
+      "loss": 0.0011,
+      "step": 48900
+    },
+    {
+      "epoch": 8.848797250859107,
+      "grad_norm": 13.626410484313965,
+      "learning_rate": 1.1757229692392705e-07,
+      "loss": 0.0021,
+      "step": 48925
+    },
+    {
+      "epoch": 8.853318864170737,
+      "grad_norm": 0.021010667085647583,
+      "learning_rate": 1.171118069626082e-07,
+      "loss": 0.0022,
+      "step": 48950
+    },
+    {
+      "epoch": 8.857840477482366,
+      "grad_norm": 0.009972944855690002,
+      "learning_rate": 1.1665131700128937e-07,
+      "loss": 0.0047,
+      "step": 48975
+    },
+    {
+      "epoch": 8.862362090793996,
+      "grad_norm": 0.04962944611907005,
+      "learning_rate": 1.1619082703997052e-07,
+      "loss": 0.0036,
+      "step": 49000
+    },
+    {
+      "epoch": 8.866883704105625,
+      "grad_norm": 1.5837059020996094,
+      "learning_rate": 1.1573033707865168e-07,
+      "loss": 0.0017,
+      "step": 49025
+    },
+    {
+      "epoch": 8.871405317417254,
+      "grad_norm": 4.073685646057129,
+      "learning_rate": 1.1526984711733284e-07,
+      "loss": 0.0092,
+      "step": 49050
+    },
+    {
+      "epoch": 8.875926930728884,
+      "grad_norm": 0.0222158282995224,
+      "learning_rate": 1.1480935715601398e-07,
+      "loss": 0.0051,
+      "step": 49075
+    },
+    {
+      "epoch": 8.880448544040513,
+      "grad_norm": 0.055091459304094315,
+      "learning_rate": 1.1434886719469515e-07,
+      "loss": 0.0077,
+      "step": 49100
+    },
+    {
+      "epoch": 8.884970157352143,
+      "grad_norm": 0.4387997090816498,
+      "learning_rate": 1.1388837723337632e-07,
+      "loss": 0.0013,
+      "step": 49125
+    },
+    {
+      "epoch": 8.889491770663772,
+      "grad_norm": 0.020865125581622124,
+      "learning_rate": 1.1342788727205746e-07,
+      "loss": 0.0005,
+      "step": 49150
+    },
+    {
+      "epoch": 8.894013383975402,
+      "grad_norm": 0.05737067386507988,
+      "learning_rate": 1.1296739731073862e-07,
+      "loss": 0.0016,
+      "step": 49175
+    },
+    {
+      "epoch": 8.898534997287031,
+      "grad_norm": 0.04351663962006569,
+      "learning_rate": 1.1250690734941979e-07,
+      "loss": 0.0049,
+      "step": 49200
+    },
+    {
+      "epoch": 8.90305661059866,
+      "grad_norm": 0.09150709211826324,
+      "learning_rate": 1.1204641738810093e-07,
+      "loss": 0.0047,
+      "step": 49225
+    },
+    {
+      "epoch": 8.907578223910292,
+      "grad_norm": 0.019495923072099686,
+      "learning_rate": 1.115859274267821e-07,
+      "loss": 0.0004,
+      "step": 49250
+    },
+    {
+      "epoch": 8.912099837221922,
+      "grad_norm": 0.002537541324272752,
+      "learning_rate": 1.1112543746546326e-07,
+      "loss": 0.0021,
+      "step": 49275
+    },
+    {
+      "epoch": 8.916621450533551,
+      "grad_norm": 0.00815527979284525,
+      "learning_rate": 1.106649475041444e-07,
+      "loss": 0.0002,
+      "step": 49300
+    },
+    {
+      "epoch": 8.92114306384518,
+      "grad_norm": 4.410140037536621,
+      "learning_rate": 1.1020445754282557e-07,
+      "loss": 0.0086,
+      "step": 49325
+    },
+    {
+      "epoch": 8.92566467715681,
+      "grad_norm": 0.3327310383319855,
+      "learning_rate": 1.0974396758150672e-07,
+      "loss": 0.0179,
+      "step": 49350
+    },
+    {
+      "epoch": 8.93018629046844,
+      "grad_norm": 0.052819494158029556,
+      "learning_rate": 1.0930189721864062e-07,
+      "loss": 0.0199,
+      "step": 49375
+    },
+    {
+      "epoch": 8.934707903780069,
+      "grad_norm": 29.91364097595215,
+      "learning_rate": 1.0884140725732179e-07,
+      "loss": 0.0274,
+      "step": 49400
+    },
+    {
+      "epoch": 8.939229517091698,
+      "grad_norm": Infinity,
+      "learning_rate": 1.0839933689445569e-07,
+      "loss": 0.031,
+      "step": 49425
+    },
+    {
+      "epoch": 8.943751130403328,
+      "grad_norm": 0.09751866012811661,
+      "learning_rate": 1.0793884693313686e-07,
+      "loss": 0.025,
+      "step": 49450
+    },
+    {
+      "epoch": 8.948272743714957,
+      "grad_norm": 0.38296425342559814,
+      "learning_rate": 1.0747835697181801e-07,
+      "loss": 0.004,
+      "step": 49475
+    },
+    {
+      "epoch": 8.952794357026587,
+      "grad_norm": 0.05933375656604767,
+      "learning_rate": 1.0701786701049917e-07,
+      "loss": 0.0018,
+      "step": 49500
+    },
+    {
+      "epoch": 8.957315970338216,
+      "grad_norm": 0.14764128625392914,
+      "learning_rate": 1.0655737704918032e-07,
+      "loss": 0.0005,
+      "step": 49525
+    },
+    {
+      "epoch": 8.961837583649846,
+      "grad_norm": 0.034122079610824585,
+      "learning_rate": 1.0609688708786148e-07,
+      "loss": 0.0029,
+      "step": 49550
+    },
+    {
+      "epoch": 8.966359196961475,
+      "grad_norm": 0.04565449431538582,
+      "learning_rate": 1.0563639712654264e-07,
+      "loss": 0.0004,
+      "step": 49575
+    },
+    {
+      "epoch": 8.970880810273105,
+      "grad_norm": 0.01443118043243885,
+      "learning_rate": 1.0517590716522379e-07,
+      "loss": 0.0041,
+      "step": 49600
+    },
+    {
+      "epoch": 8.975402423584734,
+      "grad_norm": 0.005753234960138798,
+      "learning_rate": 1.0471541720390494e-07,
+      "loss": 0.0026,
+      "step": 49625
+    },
+    {
+      "epoch": 8.979924036896364,
+      "grad_norm": 0.02034948766231537,
+      "learning_rate": 1.0425492724258611e-07,
+      "loss": 0.0007,
+      "step": 49650
+    },
+    {
+      "epoch": 8.984445650207995,
+      "grad_norm": 0.012954095378518105,
+      "learning_rate": 1.0379443728126726e-07,
+      "loss": 0.0028,
+      "step": 49675
+    },
+    {
+      "epoch": 8.988967263519624,
+      "grad_norm": 11.640181541442871,
+      "learning_rate": 1.0333394731994842e-07,
+      "loss": 0.0033,
+      "step": 49700
+    },
+    {
+      "epoch": 8.993488876831254,
+      "grad_norm": 0.07725071907043457,
+      "learning_rate": 1.0287345735862958e-07,
+      "loss": 0.0125,
+      "step": 49725
+    },
+    {
+      "epoch": 8.998010490142883,
+      "grad_norm": 0.12377389520406723,
+      "learning_rate": 1.0241296739731073e-07,
+      "loss": 0.0204,
+      "step": 49750
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.370661199092865,
+      "eval_runtime": 8351.69,
+      "eval_samples_per_second": 1.137,
+      "eval_steps_per_second": 0.142,
+      "eval_wer": 0.10357142857142858,
+      "step": 49761
+    },
+    {
+      "epoch": 9.002532103454513,
+      "grad_norm": 0.027261001989245415,
+      "learning_rate": 1.0195247743599189e-07,
+      "loss": 0.0049,
+      "step": 49775
+    },
+    {
+      "epoch": 9.007053716766142,
+      "grad_norm": 0.1393139511346817,
+      "learning_rate": 1.0149198747467305e-07,
+      "loss": 0.0031,
+      "step": 49800
+    },
+    {
+      "epoch": 9.011575330077772,
+      "grad_norm": 0.31711727380752563,
+      "learning_rate": 1.010314975133542e-07,
+      "loss": 0.0015,
+      "step": 49825
+    },
+    {
+      "epoch": 9.016096943389401,
+      "grad_norm": 0.021990245208144188,
+      "learning_rate": 1.0057100755203536e-07,
+      "loss": 0.0064,
+      "step": 49850
+    },
+    {
+      "epoch": 9.02061855670103,
+      "grad_norm": 0.028167344629764557,
+      "learning_rate": 1.0011051759071653e-07,
+      "loss": 0.0019,
+      "step": 49875
+    },
+    {
+      "epoch": 9.02514017001266,
+      "grad_norm": 12.446762084960938,
+      "learning_rate": 9.965002762939767e-08,
+      "loss": 0.0111,
+      "step": 49900
+    },
+    {
+      "epoch": 9.02966178332429,
+      "grad_norm": 0.024779673665761948,
+      "learning_rate": 9.918953766807883e-08,
+      "loss": 0.0027,
+      "step": 49925
+    },
+    {
+      "epoch": 9.03418339663592,
+      "grad_norm": 2.8062970638275146,
+      "learning_rate": 9.872904770676e-08,
+      "loss": 0.0046,
+      "step": 49950
+    },
+    {
+      "epoch": 9.038705009947549,
+      "grad_norm": 0.011668199673295021,
+      "learning_rate": 9.826855774544114e-08,
+      "loss": 0.0315,
+      "step": 49975
+    },
+    {
+      "epoch": 9.043226623259178,
+      "grad_norm": 0.005240909289568663,
+      "learning_rate": 9.78080677841223e-08,
+      "loss": 0.0067,
+      "step": 50000
+    },
+    {
+      "epoch": 9.047748236570808,
+      "grad_norm": 0.015816286206245422,
+      "learning_rate": 9.734757782280347e-08,
+      "loss": 0.0018,
+      "step": 50025
+    },
+    {
+      "epoch": 9.052269849882437,
+      "grad_norm": 0.0809655636548996,
+      "learning_rate": 9.688708786148461e-08,
+      "loss": 0.0169,
+      "step": 50050
+    },
+    {
+      "epoch": 9.056791463194068,
+      "grad_norm": 0.1855606585741043,
+      "learning_rate": 9.642659790016578e-08,
+      "loss": 0.0003,
+      "step": 50075
+    },
+    {
+      "epoch": 9.061313076505698,
+      "grad_norm": 1.0818848609924316,
+      "learning_rate": 9.596610793884693e-08,
+      "loss": 0.0114,
+      "step": 50100
+    },
+    {
+      "epoch": 9.065834689817327,
+      "grad_norm": 1.6838890314102173,
+      "learning_rate": 9.550561797752808e-08,
+      "loss": 0.0142,
+      "step": 50125
+    },
+    {
+      "epoch": 9.070356303128957,
+      "grad_norm": 0.051931653171777725,
+      "learning_rate": 9.504512801620925e-08,
+      "loss": 0.0267,
+      "step": 50150
+    },
+    {
+      "epoch": 9.074877916440586,
+      "grad_norm": 0.03803849592804909,
+      "learning_rate": 9.45846380548904e-08,
+      "loss": 0.0079,
+      "step": 50175
+    },
+    {
+      "epoch": 9.079399529752216,
+      "grad_norm": 0.17854449152946472,
+      "learning_rate": 9.412414809357155e-08,
+      "loss": 0.0032,
+      "step": 50200
+    },
+    {
+      "epoch": 9.083921143063845,
+      "grad_norm": 0.07141181081533432,
+      "learning_rate": 9.366365813225272e-08,
+      "loss": 0.0009,
+      "step": 50225
+    },
+    {
+      "epoch": 9.088442756375475,
+      "grad_norm": 0.11261973530054092,
+      "learning_rate": 9.320316817093386e-08,
+      "loss": 0.0006,
+      "step": 50250
+    },
+    {
+      "epoch": 9.092964369687104,
+      "grad_norm": 0.009887372143566608,
+      "learning_rate": 9.274267820961503e-08,
+      "loss": 0.0028,
+      "step": 50275
+    },
+    {
+      "epoch": 9.097485982998734,
+      "grad_norm": 0.03133253753185272,
+      "learning_rate": 9.228218824829618e-08,
+      "loss": 0.0106,
+      "step": 50300
+    },
+    {
+      "epoch": 9.102007596310363,
+      "grad_norm": 0.11163907498121262,
+      "learning_rate": 9.182169828697733e-08,
+      "loss": 0.0012,
+      "step": 50325
+    },
+    {
+      "epoch": 9.106529209621993,
+      "grad_norm": 0.004720740485936403,
+      "learning_rate": 9.13612083256585e-08,
+      "loss": 0.0051,
+      "step": 50350
+    },
+    {
+      "epoch": 9.111050822933622,
+      "grad_norm": 0.031010426580905914,
+      "learning_rate": 9.090071836433965e-08,
+      "loss": 0.0008,
+      "step": 50375
+    },
+    {
+      "epoch": 9.115572436245252,
+      "grad_norm": 0.02335488423705101,
+      "learning_rate": 9.04402284030208e-08,
+      "loss": 0.0129,
+      "step": 50400
+    },
+    {
+      "epoch": 9.120094049556881,
+      "grad_norm": 0.0240317415446043,
+      "learning_rate": 8.997973844170197e-08,
+      "loss": 0.0024,
+      "step": 50425
+    },
+    {
+      "epoch": 9.12461566286851,
+      "grad_norm": 0.03213008865714073,
+      "learning_rate": 8.951924848038312e-08,
+      "loss": 0.0052,
+      "step": 50450
+    },
+    {
+      "epoch": 9.129137276180142,
+      "grad_norm": 0.04678371921181679,
+      "learning_rate": 8.905875851906428e-08,
+      "loss": 0.0014,
+      "step": 50475
+    },
+    {
+      "epoch": 9.133658889491771,
+      "grad_norm": 1.1161173582077026,
+      "learning_rate": 8.859826855774543e-08,
+      "loss": 0.0005,
+      "step": 50500
+    },
+    {
+      "epoch": 9.1381805028034,
+      "grad_norm": 0.1399179995059967,
+      "learning_rate": 8.81377785964266e-08,
+      "loss": 0.0063,
+      "step": 50525
+    },
+    {
+      "epoch": 9.14270211611503,
+      "grad_norm": 0.05243779718875885,
+      "learning_rate": 8.767728863510775e-08,
+      "loss": 0.002,
+      "step": 50550
+    },
+    {
+      "epoch": 9.14722372942666,
+      "grad_norm": 0.03155532851815224,
+      "learning_rate": 8.72167986737889e-08,
+      "loss": 0.0164,
+      "step": 50575
+    },
+    {
+      "epoch": 9.15174534273829,
+      "grad_norm": 0.025026287883520126,
+      "learning_rate": 8.675630871247007e-08,
+      "loss": 0.0037,
+      "step": 50600
+    },
+    {
+      "epoch": 9.156266956049919,
+      "grad_norm": 0.03365040570497513,
+      "learning_rate": 8.629581875115122e-08,
+      "loss": 0.0006,
+      "step": 50625
+    },
+    {
+      "epoch": 9.160788569361548,
+      "grad_norm": 0.01698467880487442,
+      "learning_rate": 8.583532878983237e-08,
+      "loss": 0.0075,
+      "step": 50650
+    },
+    {
+      "epoch": 9.165310182673178,
+      "grad_norm": 2.0226006507873535,
+      "learning_rate": 8.537483882851354e-08,
+      "loss": 0.0007,
+      "step": 50675
+    },
+    {
+      "epoch": 9.169831795984807,
+      "grad_norm": 0.2596571445465088,
+      "learning_rate": 8.49143488671947e-08,
+      "loss": 0.0019,
+      "step": 50700
+    },
+    {
+      "epoch": 9.174353409296437,
+      "grad_norm": 0.1478416472673416,
+      "learning_rate": 8.445385890587585e-08,
+      "loss": 0.0014,
+      "step": 50725
+    },
+    {
+      "epoch": 9.178875022608066,
+      "grad_norm": 0.015776338055729866,
+      "learning_rate": 8.399336894455701e-08,
+      "loss": 0.0152,
+      "step": 50750
+    },
+    {
+      "epoch": 9.183396635919696,
+      "grad_norm": 0.011738612316548824,
+      "learning_rate": 8.353287898323815e-08,
+      "loss": 0.0007,
+      "step": 50775
+    },
+    {
+      "epoch": 9.187918249231325,
+      "grad_norm": 0.031315308064222336,
+      "learning_rate": 8.307238902191932e-08,
+      "loss": 0.0013,
+      "step": 50800
+    },
+    {
+      "epoch": 9.192439862542955,
+      "grad_norm": 0.19211354851722717,
+      "learning_rate": 8.261189906060049e-08,
+      "loss": 0.001,
+      "step": 50825
+    },
+    {
+      "epoch": 9.196961475854584,
+      "grad_norm": 0.03229336068034172,
+      "learning_rate": 8.215140909928163e-08,
+      "loss": 0.0097,
+      "step": 50850
+    },
+    {
+      "epoch": 9.201483089166215,
+      "grad_norm": 0.010991367511451244,
+      "learning_rate": 8.169091913796279e-08,
+      "loss": 0.0149,
+      "step": 50875
+    },
+    {
+      "epoch": 9.206004702477845,
+      "grad_norm": 0.014507956802845001,
+      "learning_rate": 8.123042917664394e-08,
+      "loss": 0.0074,
+      "step": 50900
+    },
+    {
+      "epoch": 9.210526315789474,
+      "grad_norm": 2.2018537521362305,
+      "learning_rate": 8.07699392153251e-08,
+      "loss": 0.0245,
+      "step": 50925
+    },
+    {
+      "epoch": 9.215047929101104,
+      "grad_norm": 0.4151630699634552,
+      "learning_rate": 8.030944925400626e-08,
+      "loss": 0.0402,
+      "step": 50950
+    },
+    {
+      "epoch": 9.219569542412733,
+      "grad_norm": 0.041080426424741745,
+      "learning_rate": 7.98489592926874e-08,
+      "loss": 0.0219,
+      "step": 50975
+    },
+    {
+      "epoch": 9.224091155724363,
+      "grad_norm": 0.026268433779478073,
+      "learning_rate": 7.938846933136857e-08,
+      "loss": 0.0091,
+      "step": 51000
+    },
+    {
+      "epoch": 9.228612769035992,
+      "grad_norm": 0.04076811671257019,
+      "learning_rate": 7.892797937004974e-08,
+      "loss": 0.0012,
+      "step": 51025
+    },
+    {
+      "epoch": 9.233134382347622,
+      "grad_norm": 0.011797059327363968,
+      "learning_rate": 7.846748940873088e-08,
+      "loss": 0.005,
+      "step": 51050
+    },
+    {
+      "epoch": 9.237655995659251,
+      "grad_norm": 2.813450574874878,
+      "learning_rate": 7.800699944741204e-08,
+      "loss": 0.001,
+      "step": 51075
+    },
+    {
+      "epoch": 9.24217760897088,
+      "grad_norm": 0.11257357895374298,
+      "learning_rate": 7.754650948609321e-08,
+      "loss": 0.0089,
+      "step": 51100
+    },
+    {
+      "epoch": 9.24669922228251,
+      "grad_norm": 0.01619679108262062,
+      "learning_rate": 7.708601952477435e-08,
+      "loss": 0.0009,
+      "step": 51125
+    },
+    {
+      "epoch": 9.25122083559414,
+      "grad_norm": 0.0227675624191761,
+      "learning_rate": 7.662552956345551e-08,
+      "loss": 0.001,
+      "step": 51150
+    },
+    {
+      "epoch": 9.25574244890577,
+      "grad_norm": 19.843429565429688,
+      "learning_rate": 7.616503960213668e-08,
+      "loss": 0.0012,
+      "step": 51175
+    },
+    {
+      "epoch": 9.260264062217399,
+      "grad_norm": 0.17110028862953186,
+      "learning_rate": 7.570454964081782e-08,
+      "loss": 0.0048,
+      "step": 51200
+    },
+    {
+      "epoch": 9.264785675529028,
+      "grad_norm": 0.034038007259368896,
+      "learning_rate": 7.524405967949899e-08,
+      "loss": 0.0061,
+      "step": 51225
+    },
+    {
+      "epoch": 9.269307288840658,
+      "grad_norm": 0.8703758716583252,
+      "learning_rate": 7.478356971818014e-08,
+      "loss": 0.0041,
+      "step": 51250
+    },
+    {
+      "epoch": 9.273828902152289,
+      "grad_norm": 0.019711392000317574,
+      "learning_rate": 7.432307975686129e-08,
+      "loss": 0.0105,
+      "step": 51275
+    },
+    {
+      "epoch": 9.278350515463918,
+      "grad_norm": 12.372844696044922,
+      "learning_rate": 7.386258979554246e-08,
+      "loss": 0.0258,
+      "step": 51300
+    },
+    {
+      "epoch": 9.282872128775548,
+      "grad_norm": 0.03570927679538727,
+      "learning_rate": 7.340209983422361e-08,
+      "loss": 0.027,
+      "step": 51325
+    },
+    {
+      "epoch": 9.287393742087177,
+      "grad_norm": 0.2612115144729614,
+      "learning_rate": 7.294160987290476e-08,
+      "loss": 0.0166,
+      "step": 51350
+    },
+    {
+      "epoch": 9.291915355398807,
+      "grad_norm": 0.10630948096513748,
+      "learning_rate": 7.248111991158593e-08,
+      "loss": 0.0233,
+      "step": 51375
+    },
+    {
+      "epoch": 9.296436968710436,
+      "grad_norm": 0.08389817923307419,
+      "learning_rate": 7.202062995026708e-08,
+      "loss": 0.0012,
+      "step": 51400
+    },
+    {
+      "epoch": 9.300958582022066,
+      "grad_norm": 0.012098311446607113,
+      "learning_rate": 7.156013998894824e-08,
+      "loss": 0.0049,
+      "step": 51425
+    },
+    {
+      "epoch": 9.305480195333695,
+      "grad_norm": 1.7195242643356323,
+      "learning_rate": 7.109965002762939e-08,
+      "loss": 0.0013,
+      "step": 51450
+    },
+    {
+      "epoch": 9.310001808645325,
+      "grad_norm": 0.005529410671442747,
+      "learning_rate": 7.063916006631056e-08,
+      "loss": 0.0005,
+      "step": 51475
+    },
+    {
+      "epoch": 9.314523421956954,
+      "grad_norm": 14.70275592803955,
+      "learning_rate": 7.017867010499171e-08,
+      "loss": 0.0074,
+      "step": 51500
+    },
+    {
+      "epoch": 9.319045035268584,
+      "grad_norm": 0.03697911649942398,
+      "learning_rate": 6.971818014367286e-08,
+      "loss": 0.0019,
+      "step": 51525
+    },
+    {
+      "epoch": 9.323566648580213,
+      "grad_norm": 0.023678451776504517,
+      "learning_rate": 6.927610978080678e-08,
+      "loss": 0.0067,
+      "step": 51550
+    },
+    {
+      "epoch": 9.328088261891843,
+      "grad_norm": 0.056655462831258774,
+      "learning_rate": 6.881561981948793e-08,
+      "loss": 0.0029,
+      "step": 51575
+    },
+    {
+      "epoch": 9.332609875203472,
+      "grad_norm": 0.2097851186990738,
+      "learning_rate": 6.835512985816909e-08,
+      "loss": 0.0002,
+      "step": 51600
+    },
+    {
+      "epoch": 9.337131488515102,
+      "grad_norm": 0.030322756618261337,
+      "learning_rate": 6.789463989685025e-08,
+      "loss": 0.02,
+      "step": 51625
+    },
+    {
+      "epoch": 9.341653101826731,
+      "grad_norm": 0.05204546079039574,
+      "learning_rate": 6.74341499355314e-08,
+      "loss": 0.0002,
+      "step": 51650
+    },
+    {
+      "epoch": 9.34617471513836,
+      "grad_norm": 5.798641204833984,
+      "learning_rate": 6.697365997421256e-08,
+      "loss": 0.0012,
+      "step": 51675
+    },
+    {
+      "epoch": 9.350696328449992,
+      "grad_norm": 1.0529894828796387,
+      "learning_rate": 6.651317001289372e-08,
+      "loss": 0.016,
+      "step": 51700
+    },
+    {
+      "epoch": 9.355217941761621,
+      "grad_norm": 0.03712335228919983,
+      "learning_rate": 6.605268005157486e-08,
+      "loss": 0.0015,
+      "step": 51725
+    },
+    {
+      "epoch": 9.35973955507325,
+      "grad_norm": 0.20770879089832306,
+      "learning_rate": 6.559219009025603e-08,
+      "loss": 0.0245,
+      "step": 51750
+    },
+    {
+      "epoch": 9.36426116838488,
+      "grad_norm": 0.034741051495075226,
+      "learning_rate": 6.51317001289372e-08,
+      "loss": 0.0384,
+      "step": 51775
+    },
+    {
+      "epoch": 9.36878278169651,
+      "grad_norm": 0.06246571242809296,
+      "learning_rate": 6.467121016761834e-08,
+      "loss": 0.0029,
+      "step": 51800
+    },
+    {
+      "epoch": 9.37330439500814,
+      "grad_norm": 0.019802218303084373,
+      "learning_rate": 6.42107202062995e-08,
+      "loss": 0.0017,
+      "step": 51825
+    },
+    {
+      "epoch": 9.377826008319769,
+      "grad_norm": 0.035850297659635544,
+      "learning_rate": 6.375023024498067e-08,
+      "loss": 0.0024,
+      "step": 51850
+    },
+    {
+      "epoch": 9.382347621631398,
+      "grad_norm": 0.01637556403875351,
+      "learning_rate": 6.328974028366181e-08,
+      "loss": 0.0005,
+      "step": 51875
+    },
+    {
+      "epoch": 9.386869234943028,
+      "grad_norm": 0.04414265602827072,
+      "learning_rate": 6.282925032234297e-08,
+      "loss": 0.0038,
+      "step": 51900
+    },
+    {
+      "epoch": 9.391390848254657,
+      "grad_norm": 0.048500921577215195,
+      "learning_rate": 6.236876036102413e-08,
+      "loss": 0.0088,
+      "step": 51925
+    },
+    {
+      "epoch": 9.395912461566287,
+      "grad_norm": 0.22167189419269562,
+      "learning_rate": 6.190827039970528e-08,
+      "loss": 0.0033,
+      "step": 51950
+    },
+    {
+      "epoch": 9.400434074877916,
+      "grad_norm": 0.02152320370078087,
+      "learning_rate": 6.144778043838645e-08,
+      "loss": 0.0014,
+      "step": 51975
+    },
+    {
+      "epoch": 9.404955688189546,
+      "grad_norm": 0.016991982236504555,
+      "learning_rate": 6.09872904770676e-08,
+      "loss": 0.007,
+      "step": 52000
+    },
+    {
+      "epoch": 9.409477301501175,
+      "grad_norm": 0.01895890012383461,
+      "learning_rate": 6.052680051574875e-08,
+      "loss": 0.0049,
+      "step": 52025
+    },
+    {
+      "epoch": 9.413998914812804,
+      "grad_norm": 0.02496323734521866,
+      "learning_rate": 6.006631055442992e-08,
+      "loss": 0.0107,
+      "step": 52050
+    },
+    {
+      "epoch": 9.418520528124434,
+      "grad_norm": 5.717367649078369,
+      "learning_rate": 5.960582059311107e-08,
+      "loss": 0.0118,
+      "step": 52075
+    },
+    {
+      "epoch": 9.423042141436065,
+      "grad_norm": 0.02132461778819561,
+      "learning_rate": 5.9145330631792224e-08,
+      "loss": 0.008,
+      "step": 52100
+    },
+    {
+      "epoch": 9.427563754747695,
+      "grad_norm": 0.018231956288218498,
+      "learning_rate": 5.8684840670473384e-08,
+      "loss": 0.0376,
+      "step": 52125
+    },
+    {
+      "epoch": 9.432085368059324,
+      "grad_norm": 19.91530418395996,
+      "learning_rate": 5.822435070915454e-08,
+      "loss": 0.0246,
+      "step": 52150
+    },
+    {
+      "epoch": 9.436606981370954,
+      "grad_norm": 0.037468716502189636,
+      "learning_rate": 5.7763860747835697e-08,
+      "loss": 0.0306,
+      "step": 52175
+    },
+    {
+      "epoch": 9.441128594682583,
+      "grad_norm": 0.013554728589951992,
+      "learning_rate": 5.730337078651685e-08,
+      "loss": 0.0054,
+      "step": 52200
+    },
+    {
+      "epoch": 9.445650207994213,
+      "grad_norm": 0.023764220997691154,
+      "learning_rate": 5.684288082519801e-08,
+      "loss": 0.006,
+      "step": 52225
+    },
+    {
+      "epoch": 9.450171821305842,
+      "grad_norm": 0.040070127695798874,
+      "learning_rate": 5.638239086387916e-08,
+      "loss": 0.0008,
+      "step": 52250
+    },
+    {
+      "epoch": 9.454693434617472,
+      "grad_norm": 0.10571596771478653,
+      "learning_rate": 5.592190090256032e-08,
+      "loss": 0.0013,
+      "step": 52275
+    },
+    {
+      "epoch": 9.459215047929101,
+      "grad_norm": 0.14270947873592377,
+      "learning_rate": 5.546141094124148e-08,
+      "loss": 0.0129,
+      "step": 52300
+    },
+    {
+      "epoch": 9.46373666124073,
+      "grad_norm": 0.04533557966351509,
+      "learning_rate": 5.5000920979922634e-08,
+      "loss": 0.0006,
+      "step": 52325
+    },
+    {
+      "epoch": 9.46825827455236,
+      "grad_norm": 0.06353598833084106,
+      "learning_rate": 5.454043101860379e-08,
+      "loss": 0.0005,
+      "step": 52350
+    },
+    {
+      "epoch": 9.47277988786399,
+      "grad_norm": 0.016834860667586327,
+      "learning_rate": 5.4079941057284954e-08,
+      "loss": 0.0031,
+      "step": 52375
+    },
+    {
+      "epoch": 9.477301501175619,
+      "grad_norm": 0.012295857071876526,
+      "learning_rate": 5.3619451095966107e-08,
+      "loss": 0.0003,
+      "step": 52400
+    },
+    {
+      "epoch": 9.481823114487248,
+      "grad_norm": 10.514667510986328,
+      "learning_rate": 5.315896113464726e-08,
+      "loss": 0.008,
+      "step": 52425
+    },
+    {
+      "epoch": 9.486344727798878,
+      "grad_norm": 0.813252329826355,
+      "learning_rate": 5.269847117332842e-08,
+      "loss": 0.0014,
+      "step": 52450
+    },
+    {
+      "epoch": 9.490866341110507,
+      "grad_norm": 2.42742657661438,
+      "learning_rate": 5.223798121200958e-08,
+      "loss": 0.0026,
+      "step": 52475
+    },
+    {
+      "epoch": 9.495387954422139,
+      "grad_norm": 0.03881525248289108,
+      "learning_rate": 5.177749125069073e-08,
+      "loss": 0.0107,
+      "step": 52500
+    },
+    {
+      "epoch": 9.499909567733768,
+      "grad_norm": 0.09639015793800354,
+      "learning_rate": 5.1317001289371885e-08,
+      "loss": 0.0052,
+      "step": 52525
+    },
+    {
+      "epoch": 9.504431181045398,
+      "grad_norm": 0.16553708910942078,
+      "learning_rate": 5.085651132805305e-08,
+      "loss": 0.0188,
+      "step": 52550
+    },
+    {
+      "epoch": 9.508952794357027,
+      "grad_norm": 9.378378868103027,
+      "learning_rate": 5.0396021366734204e-08,
+      "loss": 0.0339,
+      "step": 52575
+    },
+    {
+      "epoch": 9.513474407668657,
+      "grad_norm": 0.2166759967803955,
+      "learning_rate": 4.993553140541536e-08,
+      "loss": 0.0012,
+      "step": 52600
+    },
+    {
+      "epoch": 9.517996020980286,
+      "grad_norm": 0.026047270745038986,
+      "learning_rate": 4.9475041444096517e-08,
+      "loss": 0.0091,
+      "step": 52625
+    },
+    {
+      "epoch": 9.522517634291916,
+      "grad_norm": 0.03240982070565224,
+      "learning_rate": 4.9014551482777676e-08,
+      "loss": 0.0085,
+      "step": 52650
+    },
+    {
+      "epoch": 9.527039247603545,
+      "grad_norm": 0.07469449937343597,
+      "learning_rate": 4.855406152145883e-08,
+      "loss": 0.0016,
+      "step": 52675
+    },
+    {
+      "epoch": 9.531560860915175,
+      "grad_norm": 0.034668173640966415,
+      "learning_rate": 4.809357156013999e-08,
+      "loss": 0.0008,
+      "step": 52700
+    },
+    {
+      "epoch": 9.536082474226804,
+      "grad_norm": 0.00523386849090457,
+      "learning_rate": 4.763308159882114e-08,
+      "loss": 0.0074,
+      "step": 52725
+    },
+    {
+      "epoch": 9.540604087538433,
+      "grad_norm": 0.018630068749189377,
+      "learning_rate": 4.71725916375023e-08,
+      "loss": 0.0108,
+      "step": 52750
+    },
+    {
+      "epoch": 9.545125700850063,
+      "grad_norm": 0.02157723344862461,
+      "learning_rate": 4.671210167618346e-08,
+      "loss": 0.0078,
+      "step": 52775
+    },
+    {
+      "epoch": 9.549647314161692,
+      "grad_norm": 0.016346458345651627,
+      "learning_rate": 4.6251611714864614e-08,
+      "loss": 0.0123,
+      "step": 52800
+    },
+    {
+      "epoch": 9.554168927473322,
+      "grad_norm": 0.027240611612796783,
+      "learning_rate": 4.579112175354577e-08,
+      "loss": 0.0002,
+      "step": 52825
+    },
+    {
+      "epoch": 9.558690540784951,
+      "grad_norm": 1.3178123235702515,
+      "learning_rate": 4.533063179222693e-08,
+      "loss": 0.0038,
+      "step": 52850
+    },
+    {
+      "epoch": 9.56321215409658,
+      "grad_norm": 0.025551458820700645,
+      "learning_rate": 4.4870141830908086e-08,
+      "loss": 0.011,
+      "step": 52875
+    },
+    {
+      "epoch": 9.56773376740821,
+      "grad_norm": 0.06759845465421677,
+      "learning_rate": 4.440965186958924e-08,
+      "loss": 0.0013,
+      "step": 52900
+    },
+    {
+      "epoch": 9.572255380719842,
+      "grad_norm": 0.01742335967719555,
+      "learning_rate": 4.394916190827039e-08,
+      "loss": 0.0124,
+      "step": 52925
+    },
+    {
+      "epoch": 9.576776994031471,
+      "grad_norm": 0.09457490593194962,
+      "learning_rate": 4.348867194695156e-08,
+      "loss": 0.0146,
+      "step": 52950
+    },
+    {
+      "epoch": 9.5812986073431,
+      "grad_norm": 0.11490760743618011,
+      "learning_rate": 4.302818198563271e-08,
+      "loss": 0.0145,
+      "step": 52975
+    },
+    {
+      "epoch": 9.58582022065473,
+      "grad_norm": 0.17125943303108215,
+      "learning_rate": 4.2567692024313865e-08,
+      "loss": 0.0047,
+      "step": 53000
+    },
+    {
+      "epoch": 9.59034183396636,
+      "grad_norm": 0.08460250496864319,
+      "learning_rate": 4.2107202062995024e-08,
+      "loss": 0.0045,
+      "step": 53025
+    },
+    {
+      "epoch": 9.594863447277989,
+      "grad_norm": 0.5364235639572144,
+      "learning_rate": 4.1646712101676184e-08,
+      "loss": 0.0014,
+      "step": 53050
+    },
+    {
+      "epoch": 9.599385060589618,
+      "grad_norm": 0.2898434102535248,
+      "learning_rate": 4.118622214035734e-08,
+      "loss": 0.0034,
+      "step": 53075
+    },
+    {
+      "epoch": 9.603906673901248,
+      "grad_norm": 0.003405811497941613,
+      "learning_rate": 4.0725732179038496e-08,
+      "loss": 0.0064,
+      "step": 53100
+    },
+    {
+      "epoch": 9.608428287212877,
+      "grad_norm": 0.03517955541610718,
+      "learning_rate": 4.0265242217719656e-08,
+      "loss": 0.005,
+      "step": 53125
+    },
+    {
+      "epoch": 9.612949900524507,
+      "grad_norm": 0.005290856584906578,
+      "learning_rate": 3.980475225640081e-08,
+      "loss": 0.006,
+      "step": 53150
+    },
+    {
+      "epoch": 9.617471513836136,
+      "grad_norm": 0.026814907789230347,
+      "learning_rate": 3.934426229508197e-08,
+      "loss": 0.0026,
+      "step": 53175
+    },
+    {
+      "epoch": 9.621993127147766,
+      "grad_norm": 0.03042653575539589,
+      "learning_rate": 3.888377233376312e-08,
+      "loss": 0.0061,
+      "step": 53200
+    },
+    {
+      "epoch": 9.626514740459395,
+      "grad_norm": 0.010026361793279648,
+      "learning_rate": 3.842328237244428e-08,
+      "loss": 0.0002,
+      "step": 53225
+    },
+    {
+      "epoch": 9.631036353771025,
+      "grad_norm": 0.01914265938103199,
+      "learning_rate": 3.7962792411125434e-08,
+      "loss": 0.005,
+      "step": 53250
+    },
+    {
+      "epoch": 9.635557967082654,
+      "grad_norm": 0.05811558663845062,
+      "learning_rate": 3.7502302449806594e-08,
+      "loss": 0.0126,
+      "step": 53275
+    },
+    {
+      "epoch": 9.640079580394286,
+      "grad_norm": 0.013714855536818504,
+      "learning_rate": 3.704181248848775e-08,
+      "loss": 0.0215,
+      "step": 53300
+    },
+    {
+      "epoch": 9.644601193705915,
+      "grad_norm": 0.0346146784722805,
+      "learning_rate": 3.6581322527168906e-08,
+      "loss": 0.0144,
+      "step": 53325
+    },
+    {
+      "epoch": 9.649122807017545,
+      "grad_norm": 0.0485786534845829,
+      "learning_rate": 3.6120832565850066e-08,
+      "loss": 0.015,
+      "step": 53350
+    },
+    {
+      "epoch": 9.653644420329174,
+      "grad_norm": 0.06555884331464767,
+      "learning_rate": 3.566034260453122e-08,
+      "loss": 0.0078,
+      "step": 53375
+    },
+    {
+      "epoch": 9.658166033640804,
+      "grad_norm": 0.03746599331498146,
+      "learning_rate": 3.519985264321237e-08,
+      "loss": 0.0021,
+      "step": 53400
+    },
+    {
+      "epoch": 9.662687646952433,
+      "grad_norm": 0.04957371950149536,
+      "learning_rate": 3.473936268189354e-08,
+      "loss": 0.009,
+      "step": 53425
+    },
+    {
+      "epoch": 9.667209260264062,
+      "grad_norm": 0.020930081605911255,
+      "learning_rate": 3.427887272057469e-08,
+      "loss": 0.0011,
+      "step": 53450
+    },
+    {
+      "epoch": 9.671730873575692,
+      "grad_norm": 0.014865943230688572,
+      "learning_rate": 3.3818382759255844e-08,
+      "loss": 0.0005,
+      "step": 53475
+    },
+    {
+      "epoch": 9.676252486887321,
+      "grad_norm": 0.025175156071782112,
+      "learning_rate": 3.3357892797937004e-08,
+      "loss": 0.0012,
+      "step": 53500
+    },
+    {
+      "epoch": 9.680774100198951,
+      "grad_norm": 0.03397619351744652,
+      "learning_rate": 3.2897402836618163e-08,
+      "loss": 0.0011,
+      "step": 53525
+    },
+    {
+      "epoch": 9.68529571351058,
+      "grad_norm": 0.02394242398440838,
+      "learning_rate": 3.2436912875299316e-08,
+      "loss": 0.0036,
+      "step": 53550
+    },
+    {
+      "epoch": 9.68981732682221,
+      "grad_norm": 0.08645796030759811,
+      "learning_rate": 3.197642291398047e-08,
+      "loss": 0.0005,
+      "step": 53575
+    },
+    {
+      "epoch": 9.69433894013384,
+      "grad_norm": 0.08837512135505676,
+      "learning_rate": 3.151593295266163e-08,
+      "loss": 0.0034,
+      "step": 53600
+    },
+    {
+      "epoch": 9.698860553445469,
+      "grad_norm": 0.0059136999770998955,
+      "learning_rate": 3.105544299134279e-08,
+      "loss": 0.0028,
+      "step": 53625
+    },
+    {
+      "epoch": 9.703382166757098,
+      "grad_norm": 0.06296961009502411,
+      "learning_rate": 3.059495303002394e-08,
+      "loss": 0.0057,
+      "step": 53650
+    },
+    {
+      "epoch": 9.707903780068728,
+      "grad_norm": 24.09531593322754,
+      "learning_rate": 3.01344630687051e-08,
+      "loss": 0.0048,
+      "step": 53675
+    },
+    {
+      "epoch": 9.712425393380357,
+      "grad_norm": 0.07932830601930618,
+      "learning_rate": 2.9673973107386258e-08,
+      "loss": 0.0354,
+      "step": 53700
+    },
+    {
+      "epoch": 9.716947006691989,
+      "grad_norm": 0.6279693841934204,
+      "learning_rate": 2.9213483146067417e-08,
+      "loss": 0.0024,
+      "step": 53725
+    },
+    {
+      "epoch": 9.721468620003618,
+      "grad_norm": 0.018382834270596504,
+      "learning_rate": 2.875299318474857e-08,
+      "loss": 0.0033,
+      "step": 53750
+    },
+    {
+      "epoch": 9.725990233315247,
+      "grad_norm": 0.13161396980285645,
+      "learning_rate": 2.829250322342973e-08,
+      "loss": 0.0269,
+      "step": 53775
+    },
+    {
+      "epoch": 9.730511846626877,
+      "grad_norm": 0.08527792245149612,
+      "learning_rate": 2.7832013262110883e-08,
+      "loss": 0.0098,
+      "step": 53800
+    },
+    {
+      "epoch": 9.735033459938506,
+      "grad_norm": 0.020729778334498405,
+      "learning_rate": 2.7371523300792042e-08,
+      "loss": 0.0041,
+      "step": 53825
+    },
+    {
+      "epoch": 9.739555073250136,
+      "grad_norm": 0.19172513484954834,
+      "learning_rate": 2.69110333394732e-08,
+      "loss": 0.0041,
+      "step": 53850
+    },
+    {
+      "epoch": 9.744076686561765,
+      "grad_norm": 0.03183312341570854,
+      "learning_rate": 2.6450543378154355e-08,
+      "loss": 0.0012,
+      "step": 53875
+    },
+    {
+      "epoch": 9.748598299873395,
+      "grad_norm": 5.034526824951172,
+      "learning_rate": 2.599005341683551e-08,
+      "loss": 0.0018,
+      "step": 53900
+    },
+    {
+      "epoch": 9.753119913185024,
+      "grad_norm": 0.0769072026014328,
+      "learning_rate": 2.552956345551667e-08,
+      "loss": 0.0005,
+      "step": 53925
+    },
+    {
+      "epoch": 9.757641526496654,
+      "grad_norm": 0.032027099281549454,
+      "learning_rate": 2.5069073494197824e-08,
+      "loss": 0.0056,
+      "step": 53950
+    },
+    {
+      "epoch": 9.762163139808283,
+      "grad_norm": 0.08156726509332657,
+      "learning_rate": 2.4608583532878984e-08,
+      "loss": 0.0068,
+      "step": 53975
+    },
+    {
+      "epoch": 9.766684753119913,
+      "grad_norm": 0.010565202683210373,
+      "learning_rate": 2.4148093571560137e-08,
+      "loss": 0.0032,
+      "step": 54000
+    },
+    {
+      "epoch": 9.771206366431542,
+      "grad_norm": 0.027527930215001106,
+      "learning_rate": 2.3687603610241296e-08,
+      "loss": 0.0003,
+      "step": 54025
+    },
+    {
+      "epoch": 9.775727979743172,
+      "grad_norm": 0.07169647514820099,
+      "learning_rate": 2.3227113648922452e-08,
+      "loss": 0.0066,
+      "step": 54050
+    },
+    {
+      "epoch": 9.780249593054801,
+      "grad_norm": 0.0017488193698227406,
+      "learning_rate": 2.276662368760361e-08,
+      "loss": 0.0004,
+      "step": 54075
+    },
+    {
+      "epoch": 9.78477120636643,
+      "grad_norm": 0.011047018691897392,
+      "learning_rate": 2.2306133726284765e-08,
+      "loss": 0.0077,
+      "step": 54100
+    },
+    {
+      "epoch": 9.78929281967806,
+      "grad_norm": 0.05196432024240494,
+      "learning_rate": 2.1845643764965925e-08,
+      "loss": 0.0067,
+      "step": 54125
+    },
+    {
+      "epoch": 9.793814432989691,
+      "grad_norm": 0.19424718618392944,
+      "learning_rate": 2.1385153803647078e-08,
+      "loss": 0.0115,
+      "step": 54150
+    },
+    {
+      "epoch": 9.798336046301321,
+      "grad_norm": 0.016021044924855232,
+      "learning_rate": 2.0924663842328237e-08,
+      "loss": 0.0218,
+      "step": 54175
+    },
+    {
+      "epoch": 9.80285765961295,
+      "grad_norm": 0.031037848442792892,
+      "learning_rate": 2.0482593479462148e-08,
+      "loss": 0.0081,
+      "step": 54200
+    },
+    {
+      "epoch": 9.80737927292458,
+      "grad_norm": 5.485635757446289,
+      "learning_rate": 2.0022103518143304e-08,
+      "loss": 0.0037,
+      "step": 54225
+    },
+    {
+      "epoch": 9.81190088623621,
+      "grad_norm": 0.02747327648103237,
+      "learning_rate": 1.956161355682446e-08,
+      "loss": 0.0044,
+      "step": 54250
+    },
+    {
+      "epoch": 9.816422499547839,
+      "grad_norm": 0.1059911772608757,
+      "learning_rate": 1.9101123595505617e-08,
+      "loss": 0.0038,
+      "step": 54275
+    },
+    {
+      "epoch": 9.820944112859468,
+      "grad_norm": 0.010986040346324444,
+      "learning_rate": 1.8640633634186776e-08,
+      "loss": 0.0066,
+      "step": 54300
+    },
+    {
+      "epoch": 9.825465726171098,
+      "grad_norm": 0.17033414542675018,
+      "learning_rate": 1.818014367286793e-08,
+      "loss": 0.0047,
+      "step": 54325
+    },
+    {
+      "epoch": 9.829987339482727,
+      "grad_norm": 0.007194284815341234,
+      "learning_rate": 1.771965371154909e-08,
+      "loss": 0.0036,
+      "step": 54350
+    },
+    {
+      "epoch": 9.834508952794357,
+      "grad_norm": 0.02418128214776516,
+      "learning_rate": 1.7259163750230242e-08,
+      "loss": 0.002,
+      "step": 54375
+    },
+    {
+      "epoch": 9.839030566105986,
+      "grad_norm": 0.11744951456785202,
+      "learning_rate": 1.67986737889114e-08,
+      "loss": 0.0021,
+      "step": 54400
+    },
+    {
+      "epoch": 9.843552179417616,
+      "grad_norm": 0.0121499327942729,
+      "learning_rate": 1.6338183827592558e-08,
+      "loss": 0.0079,
+      "step": 54425
+    },
+    {
+      "epoch": 9.848073792729245,
+      "grad_norm": 0.014060701243579388,
+      "learning_rate": 1.5877693866273714e-08,
+      "loss": 0.001,
+      "step": 54450
+    },
+    {
+      "epoch": 9.852595406040875,
+      "grad_norm": 2.7887675762176514,
+      "learning_rate": 1.541720390495487e-08,
+      "loss": 0.0038,
+      "step": 54475
+    },
+    {
+      "epoch": 9.857117019352504,
+      "grad_norm": 0.006395564880222082,
+      "learning_rate": 1.4956713943636027e-08,
+      "loss": 0.029,
+      "step": 54500
+    },
+    {
+      "epoch": 9.861638632664135,
+      "grad_norm": 0.06236935779452324,
+      "learning_rate": 1.4496223982317185e-08,
+      "loss": 0.0107,
+      "step": 54525
+    },
+    {
+      "epoch": 9.866160245975765,
+      "grad_norm": 4.1836981773376465,
+      "learning_rate": 1.4035734020998342e-08,
+      "loss": 0.0343,
+      "step": 54550
+    },
+    {
+      "epoch": 9.870681859287394,
+      "grad_norm": 0.0794193297624588,
+      "learning_rate": 1.3575244059679499e-08,
+      "loss": 0.0146,
+      "step": 54575
+    },
+    {
+      "epoch": 9.875203472599024,
+      "grad_norm": 0.023560110479593277,
+      "learning_rate": 1.3114754098360655e-08,
+      "loss": 0.0052,
+      "step": 54600
+    },
+    {
+      "epoch": 9.879725085910653,
+      "grad_norm": 0.2117091715335846,
+      "learning_rate": 1.2654264137041811e-08,
+      "loss": 0.005,
+      "step": 54625
+    },
+    {
+      "epoch": 9.884246699222283,
+      "grad_norm": 0.0047377352602779865,
+      "learning_rate": 1.2193774175722968e-08,
+      "loss": 0.0005,
+      "step": 54650
+    },
+    {
+      "epoch": 9.888768312533912,
+      "grad_norm": 0.25400134921073914,
+      "learning_rate": 1.1733284214404126e-08,
+      "loss": 0.008,
+      "step": 54675
+    },
+    {
+      "epoch": 9.893289925845542,
+      "grad_norm": 0.09486464411020279,
+      "learning_rate": 1.1272794253085282e-08,
+      "loss": 0.0044,
+      "step": 54700
+    },
+    {
+      "epoch": 9.897811539157171,
+      "grad_norm": 0.02684643305838108,
+      "learning_rate": 1.0812304291766438e-08,
+      "loss": 0.0069,
+      "step": 54725
+    },
+    {
+      "epoch": 9.9023331524688,
+      "grad_norm": 0.025187574326992035,
+      "learning_rate": 1.0351814330447595e-08,
+      "loss": 0.0007,
+      "step": 54750
+    },
+    {
+      "epoch": 9.90685476578043,
+      "grad_norm": 0.034975674003362656,
+      "learning_rate": 9.891324369128753e-09,
+      "loss": 0.0071,
+      "step": 54775
+    },
+    {
+      "epoch": 9.91137637909206,
+      "grad_norm": 15.95773983001709,
+      "learning_rate": 9.430834407809909e-09,
+      "loss": 0.011,
+      "step": 54800
+    },
+    {
+      "epoch": 9.91589799240369,
+      "grad_norm": 0.026726465672254562,
+      "learning_rate": 8.970344446491065e-09,
+      "loss": 0.0006,
+      "step": 54825
+    },
+    {
+      "epoch": 9.920419605715319,
+      "grad_norm": 0.020367203280329704,
+      "learning_rate": 8.509854485172221e-09,
+      "loss": 0.0035,
+      "step": 54850
+    },
+    {
+      "epoch": 9.924941219026948,
+      "grad_norm": 0.12988033890724182,
+      "learning_rate": 8.049364523853381e-09,
+      "loss": 0.005,
+      "step": 54875
+    },
+    {
+      "epoch": 9.929462832338578,
+      "grad_norm": 0.06096798926591873,
+      "learning_rate": 7.588874562534537e-09,
+      "loss": 0.004,
+      "step": 54900
+    },
+    {
+      "epoch": 9.933984445650207,
+      "grad_norm": 0.9536722898483276,
+      "learning_rate": 7.128384601215693e-09,
+      "loss": 0.0036,
+      "step": 54925
+    },
+    {
+      "epoch": 9.938506058961838,
+      "grad_norm": 17.520708084106445,
+      "learning_rate": 6.66789463989685e-09,
+      "loss": 0.022,
+      "step": 54950
+    },
+    {
+      "epoch": 9.943027672273468,
+      "grad_norm": 0.010756449773907661,
+      "learning_rate": 6.207404678578006e-09,
+      "loss": 0.0402,
+      "step": 54975
+    },
+    {
+      "epoch": 9.947549285585097,
+      "grad_norm": 0.015861673280596733,
+      "learning_rate": 5.746914717259163e-09,
+      "loss": 0.0017,
+      "step": 55000
+    },
+    {
+      "epoch": 9.952070898896727,
+      "grad_norm": 0.008379822596907616,
+      "learning_rate": 5.2864247559403205e-09,
+      "loss": 0.0052,
+      "step": 55025
+    },
+    {
+      "epoch": 9.956592512208356,
+      "grad_norm": 0.15586745738983154,
+      "learning_rate": 4.825934794621478e-09,
+      "loss": 0.0054,
+      "step": 55050
+    },
+    {
+      "epoch": 9.961114125519986,
+      "grad_norm": 0.022011611610651016,
+      "learning_rate": 4.365444833302634e-09,
+      "loss": 0.0012,
+      "step": 55075
+    },
+    {
+      "epoch": 9.965635738831615,
+      "grad_norm": 0.012549543753266335,
+      "learning_rate": 3.90495487198379e-09,
+      "loss": 0.0007,
+      "step": 55100
+    },
+    {
+      "epoch": 9.970157352143245,
+      "grad_norm": 0.8270652890205383,
+      "learning_rate": 3.4444649106649474e-09,
+      "loss": 0.0063,
+      "step": 55125
+    },
+    {
+      "epoch": 9.974678965454874,
+      "grad_norm": 0.0325373113155365,
+      "learning_rate": 2.983974949346104e-09,
+      "loss": 0.0008,
+      "step": 55150
+    },
+    {
+      "epoch": 9.979200578766504,
+      "grad_norm": 0.015854543074965477,
+      "learning_rate": 2.523484988027261e-09,
+      "loss": 0.0065,
+      "step": 55175
+    },
+    {
+      "epoch": 9.983722192078133,
+      "grad_norm": 16.109289169311523,
+      "learning_rate": 2.0629950267084176e-09,
+      "loss": 0.0122,
+      "step": 55200
+    },
+    {
+      "epoch": 9.988243805389763,
+      "grad_norm": 3.6053996086120605,
+      "learning_rate": 1.6025050653895745e-09,
+      "loss": 0.0018,
+      "step": 55225
+    },
+    {
+      "epoch": 9.992765418701392,
+      "grad_norm": 0.009253941476345062,
+      "learning_rate": 1.1420151040707312e-09,
+      "loss": 0.0008,
+      "step": 55250
+    },
+    {
+      "epoch": 9.997287032013022,
+      "grad_norm": 11.093884468078613,
+      "learning_rate": 6.815251427518879e-10,
+      "loss": 0.0159,
+      "step": 55275
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.3697284162044525,
+      "eval_runtime": 8323.367,
+      "eval_samples_per_second": 1.141,
+      "eval_steps_per_second": 0.143,
+      "eval_wer": 0.10309096732863549,
+      "step": 55290
+    },
+    {
+      "epoch": 10.0,
+      "step": 55290,
+      "total_flos": 7.517848352823706e+20,
+      "train_loss": 0.018776766294569685,
+      "train_runtime": 360719.5114,
+      "train_samples_per_second": 0.613,
+      "train_steps_per_second": 0.153
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 55290,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.517848352823706e+20,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}